Social learning in multimedia

Social Learning in Multimedia Final Report January 2000 EC Targeted Socio-Economic Research Project: 4141 PL 951003 Robin Williams, Roger Slack, James Stewart Research Centre for Social Sciences The University of Edinburgh Contact: Dr Robin Williams Director Research Centre for Social Sciences/ Technology Studies Unit University of Edinburgh High School Yards Edinburgh EH1 1LZ Tel: +131-650-6387 Fax: +131-650-6399 email: R.Williams@ed.ac.uk CONTENTS Abstract ........................................................................................................................ii I Executive summary ...................................................................................................... 1 1 Introduction and Background to the Study .................................................................................... 1 2 Objectives of the SLIM project..................................................................................................... 2 3 Conduct of Study .......................................................................................................................... 3 3a National/International Studies ...................................................................................................... 3 4 Findings ....................................................................................................................................... 6 5 Conclusions and Policy Implications........................................................................................... 12 Figure ES 1 Table of Deliverables ................................................................................................ 20 II Background and objectives of the project................................................................... 22 Chapter 1: Introduction.............................................................................................................22 1.1 1.2 1.3 The context of the study ......................................................................................................... 22 Debates about the Information Society.................................................................................... 23 Coupling innovation and society: Social Learning in Multimedia ........................................... 28 Chapter 2: The Intellectual Context and Concerns of the project.............................................31 2.1 2.2 2.3 2.4 III The Evolution of Technology Policy....................................................................................... 31 The Social Shaping of Technology ......................................................................................... 38 The Social Shaping of Information and Communication Technologies ................................... 50 A Model of Innovation in Multimedia .................................................................................... 58 Scientific description of the project results and methodology ................................ 67 Chapter 3: The Scope and Methods of the Study ......................................................................67 3.1 3.2 Goals of the SLIM Project ...................................................................................................... 67 Research Design and Conduct of the Studies .......................................................................... 69 Chapter 4: National Settings for Multimedia Appropriation ....................................................77 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 Introduction ........................................................................................................................... 77 Methodology and conduct of national studies ......................................................................... 78 Main Multimedia Actors and Constituencies .......................................................................... 79 The Demographic, Geographical and Cultural Context .......................................................... 82 The Policy Context................................................................................................................. 83 The Review of Multimedia Pilots and Social Experiments ...................................................... 88 How the National Context shaped understandings of Multimedia ........................................... 91 Conclusions............................................................................................................................ 93 Chapter 5: What do we mean by Social Learning? ...................................................................97 Chapter 6: The Process of Social Learning .............................................................................109 6.1 6.2 the space for social learning: strategic options for design: appropriation...............................109 The organisation of social learning........................................................................................118 Chapter 7: The Substance of Social Learning .........................................................................131 7.1 7.2 design and the representation of the user ...............................................................................131 appropriation/domestication ..................................................................................................149 Chapter 8: The conduct and management of multimedia experiments ....................................165 8.1 8.2 8.3 IV Conclusions and policy implications .................................................................... 175 9.1 9.2 9.3 9.4 V Mode of Experimentation and Mode of Control in Multimedia Experiments .........................165 The involvement of final users .............................................................................................171 the role of the state ................................................................................................................173 Challenges arising from the current context and dynamics of innovation of multimedia.............175 Exploitation and Transferability of Results of Multimedia Experiments.................................181 How can public policy support social learning? ..........................................................................191 Post Script.............................................................................................................................195 Dissemination........................................................................................................... 199 i VI References ............................................................................................................ 204 Abstract Social Learning and the Process of Innovation in Multimedia This study showed that the eventual uses and utility of multimedia products are often far removed from supplier presumptions. Social learning is therefore crucial to how generic Information and Communication Technology (ICT) capabilities are applied and used in particular settings. In creating new multimedia products and services, diverse players are forced to collaborate: suppliers of ICTs and complementary products, media specialists and users. Certain actors (intermediaries) play a key role in maintaining such collaboration and knowledge flows. The importance of social learning is reflected in the proliferation of multimedia experiments: pilots, feasibility studies and trials, which provide a forum for resolving the uncertainties and differences surrounding the development of new multimedia products. Multimedia projects remain inherently experimental. However the importance of this innovative effort, and the knowledge it throws up, has often been overlooked. The study highlighted the various options for organising social learning, from user-centred design, to evolutionary models in which technical and market development go hand in hand, and laissez-faire approaches in which users configure standard commodified technical components to their particular purposes. Multimedia is thus an ‘unfinished’ technology, which evolves, and acquires its meanings in its implementation and use (innofusion). Non-specialist ‘users’ play an active role in fitting these offerings to their purposes, making them useful and imparting significance (domestication). Conclusions and Policy Implications The key policy challenge posed by the SLIM study surrounds the need for a 'double shift' in the focus of technology policy from Research and Technological Development of ICTs, towards: • The appropriation activities (innofusion and domestication) of intermediate and final users, • The development and appropriation of cultural and information content. Public support for the appropriation of multimedia should include provision for a creative effort in implementing and using technologies, and for the dissemination of appropriation experiences to other appropriators and to future technology supply. Decision-makers, in seeking to demonstrate the wider exploitation of public-funded projects, tend to look towards the development of novel technological artefacts. This is unhelpful. First this promise of wider commercial exploitation is rarely fulfilled (especially in the short term). Second it may discourage experimentation around usage. Finally, it may divert attention from the important non-material outcomes of a multimedia experiment: knowledge of potential users and markets; developing relationships with collaborators. The social learning perspective draws attention to the transferability of results and how best to utilise the experiences gained in experiments. The lessons learnt may be contingent and difficult to communicate and generalise. It may not be helpful to search for best practice exemplars: attribution of success or failure is often contested and uninformative; there are many valuable lessons in projects formally defined as ‘failures’. Knowledge about change processes provides a more reliable basis for transferability than correlations between specific factors and outcomes. A key question however concerns ii whether the players involved in an experiment are motivated to apply the experience gained more broadly. Public support provides crucial resources – but needs to be carefully configured to avoid unhelpful outcomes (e.g. where funding favours launching new projects over exploiting existing products and building markets). iii I Executive summary Social Learning in Multimedia Targeted Socio-Economic Research programme European Commission Fourth Framework Programme. Contract 4141 PL 951003 May 1996 – January 1999 The SLIM Research Consortium was led by the Research Centre for Social Sciences/ Technology Studies Unit at the University of Edinburgh (UK). The other SLIM research centres were: Cellule Interfacultaire Technology Assessment (CITA), Faculte Universitaire Namur de la Paix (FUNdP) (BE); Department of Technology and Social Sciences, Technical University of Denmark (DK); The Telecommunications Research Group, University of Bremen (D); Communications, Technology & Culture Research Centre (COMTEC), Dublin City University (IRL); Faculteit der Cultuurwetenschappen, Maastricht University (NL); Senter for Technologi og Samfunn, Norwegian University of Science and Technology (NO); Institute for Research on the Built Environment (IREC-EPFL), Ecole Polytechnique Federale de Lausanne (CH). 1 Introduction and Background to the Study The installation of ‘Information Superhighways’, coupled with advances in processing power and usability of information technology, have led to widespread expectations of the rapid adoption of a new cluster of technologies under the rubric of multimedia. Applications based upon these technologies are, moreover, expected to be widely diffused in many areas of working and social life, and to have profound social and economic implications. In short, these technologies are expected to underpin the transition to an information society. However, beyond this global vision, there is little certainty about the kinds of applications that can be expected to emerge. How can we assess the prospects and societal implications of these new technologies? At present expectations revolve largely around visions emerging from technology practitioners, pundits and suppliers, driven by promises of what will be technically-achievable. However history shows us that what eventually emerges falls far short of these technically-driven visions, in terms of both the rate of change and its economic and social benefits. It also shows us that in some instances change can be far more rapid and far reaching than expected, the fax and the world-wide-web being cases in point. Pressing research and policy questions? Given these expectations, many groups are interested to develop a better understanding of the prospects for multimedia? Industrial and technical players have long been keen to identify what will be the killer application that will bring multimedia into the home. Public players may ask broader questions, regarding what will be the social and economic and technological implications of the particular technological paths chosen? 1 How can governments best promote technological development and the benefits anticipated in terms of wealth creation and the quality of life? Traditional ‘linear’ models, based on promoting technology supply, have been shown to be often ineffective. Laissez-faire strategies may side-step the problem of selecting technologies; however there are very real risks of market failure – for example where beneficial technologies do not achieve critical mass and thus fail to become commercially viable, or where the needs of economically marginal groups get overlooked. How should Europe respond to the global emergence of these technologies? Will these be globalised in their supply – with the development and supply of many of the key technologies dominated by USA? Will these be globalising in their implementation, use and social implications – in terms of cutting across and eroding national and regional cultural differences? Socio-economic research into the development, implementation and use of existing and emerging technologies has an important contribution here. Origins and approach of the SLIM study The Social Learning in Multimedia (SLIM) research project emerged from discussions amongst leading European research centres concerned with the social shaping of technology – and in particular the social shaping of Information and Communications Technologies (ICT). This perspective seek to understand the prospects and social implications of new technologies by detailed analysis of the processes of technological change and of the choices this involves. Distinctively, this project sought to address in tandem the implementation and consumption of multimedia as well as its design/ development. Drawing insights from the 'social shaping of technology' perspective and other recent advances in technology studies, the SLIM research consortium sought an interdisciplinary understanding of the process of technological change – and one that was open also to insights from cultural and media studies, which was judged to be of particular importance to analysing the development and uptake of multimedia. The study highlights the role of social learning - the inventive efforts of a wide array of players including intermediate and final users as well as technology suppliers and promoters - and the interactions between them, in which generic technical capabilities are fitted to evolving markets and the needs of different groups of ‘users’. In the practical efforts of these players to apply and use technology (learning by doing) and communicate and apply their experiences more broadly (learning by interacting), the design, uses and meanings of current and emerging technologies are articulated. These detailed processes of experimentation and negotiation are critical to the future evolution and success of the multimedia technologies that are expected to underpin the transition to an information society.. The profound uncertainties facing multimedia developers and promoters about how these new ICT applications will unfold are reflected in a proliferation of initiatives ranging from exemplary 'social experiments' informed by socio-political goals to more conventional technology pilots, feasibility studies and commercial trials. These 'multimedia experiments' provided the main focus for the study. 2 Objectives of the SLIM project I. To improve general theoretical understanding of the relationship between technology and society in the case of dynamic and complex technologies such as multimedia, Drawing upon a review of work on the social shaping of technology and of information and communications technologies, a set of concepts and tools were developed for analysing social 2 learning. (Deliverable 1a, 1b) II. To examine how different national/regional settings shape strategies for the appropriation of multimedia: A set of comparative national studies were undertaken, together with selected international studies, to explore how features of the local multimedia constituency and the broader market and policy context influenced the way that opportunities for multimedia are conceived and pursued. (Deliverable 2a) Reviews were conducted of the range of initiatives across Europe in multimedia applications and services, including commercial pilots and social experiments, in order to chart the range of conceptions of multimedia applications and the different ways of organising social learning about multimedia. (Deliverable 2b) III. To examine processes of social learning through integrated studies of the development, implementation/ consumption and use of new multimedia-based products and services Sets of detailed case-studies were developed across three settings seen as being crucial in the overall emergence of multimedia products and services: Education, (Deliverable 3a) Cultural industries, (Deliverable 3b) Public services, (Deliverable 3c) To analyse features of these sectors/ areas of activity and further to consider particular features of such social learning through cross-cutting studies of social learning in the home (Deliverable 3d) and in the organisation (Deliverable 3e) IV. To consider the implications for public policy at regional national and European level. To examine how social learning and innovation in multimedia might best be focused on meeting existing and emerging user requirements, through social experiments, commercial trials and other means, and to explore public policy options for the appropriation of multimedia, and how they might best promote innovation and social learning. (Deliverable 4) Deliverables were produced in relation to each of the specific objectives and commitments agreed in the Technical Annex. A Table of formal Deliverables is appended to this executive summary, showing further availability of these products (either on the web or in publications). 3 Conduct of Study 3a National/International Studies The first empirical research undertaken in the SLIM project revolved around a series of national studies in the eight countries represented in the SLIM consortium. The Belgian centre also addressed developments in France, while Edinburgh conducted a study of multimedia in Japan. These brief studies sought to capture what general concepts and understandings of multimedia were emerging, regarding how the technology infrastructure was to be configured, the kinds of application that would run upon this; the conception of who would be the users and how the technology would be used (and thus its social implications). The studies sought in particular to establish whether there was a dominant image of multimedia within a country (or whether a number of competing conceptions prevailed), and what this was and, finally, whether, and how these images could be related to the particular national etc context In some countries, our surveys were able to identify relatively clear national images of multimedia and conceptions that could be related to the relative strengths of these apparently counterposed concepts. Notably, in Germany, digital TV was placed centre stage, which may be related to the 3 dominant role of Deutsche Telekom and two large media conglomerates - the Kirch-Group and Bertelsmann. In contrast, in Norway, where apart from Telenor, there are no powerful national champions that could shape national perceptions of multimedia around their particular offerings, the 'internet model' prevails in a context in which many people already have access to the web from their homes. However, in most countries, a range of models of multimedia and its uses seemed to co-exist, ranging from established CD-ROM technologies, to the emerging technologies of the internet and world-wide-web. We can draw some conclusions from these complex national patterns. The studies reveal international differences and similarities in the perceptions of emerging multimedia technologies by a range of actors in different national settings. The world-wide web is showing exponential growth rates across all countries surveyed (though there are important differences on penetration and growth rates). However there is no sense of convergence around a single global model. What is most striking is the diversity of these responses to an ostensibly global technology, both in terms of images of multimedia technology and approaches to multimedia policy. There are some clear differences in the way that multimedia has been conceived in different countries - both regarding the extent to which dominant conceptions of multimedia products and services have emerged, and in the nature of those conceptions. These differences can be related to the diverse social, historical institutional and technological settings for the development and use of multimedia within different nations and regions, and the arrays of players involved - confirming the relevance of the social shaping perspective to understanding these complex patterns. These observations serve to underline the shortcomings of accounts which see technology supply as a universalising force which cuts across and homogenises different social settings. There were, of course, some largely common technological developments across countries - the core multimedia technologies are after all largely available at a global level. However, the anticipated convergence around global technological models has not taken place. Instead, multimedia technologies emerge as complex assemblages of global and local elements as generic ICT capabilities are appropriated and incorporated to particular purposes within particular social settings. Though there is some evidence of rhetorical convergence at the level of public policy discourses, this has only been partial. Some elements of technology policy recur: conceptions of the capability of the new multimedia systems that will emerge; the economic and social benefits that will accrue from their adoption; and the competitive dangers of being 'left out'. Common patterns can also be detected in the areas and activities in which public-funding multimedia initiatives and experiments have been set up (for example in education and public information). However, there are also continued differences in historical national policy styles and traditions - notably between interventionist and more laissez-faire models of the role of the state. There is some evidence that the broader preoccupations of particular nations may be taken up and projected within multimedia and information society policies. 3b The Integrated Studies of Multimedia Development and Use The integrated studies of multimedia development and use comprised the main empirical stage focus of SLIM research (phase 3). Each SLIM research Centre (except Bremen, which was only contracted to undertake the national studies) carried out a detailed investigation of a set of multimedia experiments, drawn from three sectors/ areas of human activity held to be of particular interest and significance. Case study selection, writing up and analysis in a particular area was coordinated by a lead site, as summarised below. Education lead site: The University of Maastricht 4 Education is one of the most advanced areas of multimedia application today, especially in higher education, where a strong technological infrastructure and expertise in multimedia technologies/ techniques, is matched by understanding of the education process and the substantive knowledge needed for the content of programmes. Conditions in other areas of education - such as primary and secondary schools, industrial training and education/ ’edutainment ’ in the home may be rather different in terms of access to such financial and technical resources and pedagogic challenges. A particular interest was with the interaction between multimedia design and choice of pedagogy. The education study explored how the adoption of multimedia differs between different sectors and sites of learning. Cultural industries/entertainment lead site: Dublin City University This stream focused on cultural products and services that are directed at the final consumer, individual households and citizens. Developments in these industries constitute one of the biggest commercial markets for multimedia, as well as having direct social significance. Studies in this domain drew attention to the importance of the cultural content of multimedia as well as its more narrowly ‘technical’ features, and made a special contribution to its conceptualisation. The research examined the particular conceptions and definitions of the 'multimedia product' and 'user' (and markets) which inform the development of these innovative products and services in the relevant national and regional contexts in an era of increasing globalisation of cultural and social relations. Public serviceslead site :Namur FUNdP Though initially conceived as addressing both public services and private services, it became clear that this would involve too broad a focus for developing effective insights. Comparisons between public and private services were in any case possible across the other integrated studies (for example many of the cultural products were developed by commercial organisations). A tighter focus was therefore sought on public administration and information systems. The opportunity arose for a more detailed focus upon the important European development of ‘Digital Cities’ including the original Amsterdam Digital City. Cross-Cutting Studies Cross-cutting studies were also undertaken, pulling together themes and findings about particular sites of multimedia use and issues arising there, viz.: 'the organisation' and 'the home'. These primarily drew upon fieldwork and research undertaken for the integrated studies, supplemented by selected additional studies in the case of organisational adoption. The Organisation: lead site: Lausanne EPFL This cross-cutting study addressed the process of implementation and appropriation of Multimedia configurations in organisational settings, bringing together findings from relevant integrated studies and a limited amount specific additional fieldwork in firms that have been in the forefront of adopting multimedia. It explored the range of organisational applications and the extent to which the uptake of multimedia varies between different organisational activities and settings. A large number of multimedia products are becoming available for use at work (e.g. desk-top videoconferencing, CD-I, CSCW tools). Organisations have started experimenting with this bewildering array of offerings to determine which are appropriate for different kinds of organisational functions (such as training, co-operative telework, external communication). This poses new problems for the management of technological change. An interesting feature is the combination of standard tools and components and their customisation to meet the particular the technical and organisational setting. The development of these specific organisational configurations of multimedia is of particular interest as an arena of social learning. 5 The Home: lead site : Trondheim, NUST This activity was intended as a study of incorporation of multimedia-based products and services within the household and their uptake and use by family members. However, in practice, none of the case-studies came to have the home as their main locus of inquiry, and only a few contain explicit information about the multimedia situation in homes. It was necessary to modify the ambitions of this part of the project – addressing the policy discourses that surround the use of multimedia in the home; conceptions of the home informing multimedia projects and current adoption. 4 Findings 4a The Process of Innovation with Multimedia the Experimental character of Multimedia Applications The influential Bangemann Report (European Commission COM (96) 607 1996) adopted a markedly laissez-faire view, that multimedia technology is already available more or less 'off the shelf', and 'offers attractive and cost-efficient solutions to meet consumer needs'. In this case, the key remaining task is to ensure the wider diffusion of these technologies and the growth of multimedia markets. In contrast our study shows that multimedia offerings are far from being 'finished products' in a number of ways: i) First the technologies are still unstable. They are still evolving rapidly and often prove rather difficult to deploy. ii) More profoundly their applications and uses are, with only a very few exceptions, still a long way away from becoming established and widely accepted. There is still a need for inventive activity - by suppliers, by producers of content and by users - to work out ways in which these technologies may be made relevant and useful to potential consumers. These observations lead us to move away from conventional technology policy approaches that conceive innovation as limited to initial research & development and followed by a separate stage of technology diffusion. Instead a rather different kind of activity may be needed that recognises the important innovative effort that follows initial product design and development and focuses upon the local appropriation of generic multimedia capabilities. The contemporary development of Multimedia-based products and services remains profoundly experimental. Government and private sector players feel compelled to become involved and gain experience with the technology. However, they face deep uncertainties - in particular about the behaviour of other players (e.g. suppliers of complementary products) and about the responses of 'final users' to their offerings. In this context one of the key ways in which a technology may be carried forward has been through demonstration projects, feasibility studies, pilots and commercial trials (as well as formal social experiments). A plethora of such experiments have been launched in most developed (and many developing) countries worldwide, often as part of 'Information Society' initiatives. In the SLIM project we refer to these different types of initiatives collectively by the term multimedia experiment, since there are important commonalities between them - particularly in relation to social learning. Multimedia projects were experimental even where they were not intended to be so. Where projects were set up with rather limited goals, their outcomes tended to be rather more open and to offer longer-term opportunities for learning (even with projects that got into difficulties and failed to meet their objectives) in ways that were not anticipated at the outset. The experimental nature of these developments was not however necessarily recognised or properly catered for. 6 4b Social Learning The SLIM project tried to capture the processes of social learning involved in the processes of deploying generic ICT capabilities into operating multimedia applications and in getting these offerings to be meaningful and prove useful. This involves two key related processes: Innofusion - innovation does not stop when artefacts emerge from technology supply but continues (in the so-called 'diffusion stage') as they are implemented and used. Considerable innovative effort may be needed to get technologies to work and prove useful in particular user circumstances. Domestication - refers to the creative ways in which final users incorporate artefacts within their local practices, purposes and culture. It is through innofusion and domestication that novel technologies are adapted to local contexts, incorporated within particular technical configurations and practices of technology use (and, through these processes, acquire their meaning and social significance). These are two closely related facets of the way in which the more or less generic technical capabilities, arising from technology supply, are appropriated - coupled to current and emerging social needs. The appropriation perspective being developed in this study emphasises the active role of technology users and other local players, in interaction with supply-side players, in establishing new uses of technology and in creating new markets. The social learning perspective first highlights the importance of practical local activity and knowledge – of learning by doing – both in developing multimedia systems and in developing usages of multimedia. It second draws attention to the processes of learning by interacting, whereby such locally acquired knowledge may be transferred and applied in other contexts. However, to be intelligible and useful in other contexts such knowledge cannot simply be transported – it must also be translated, combined with other knowledges, and transformed. This kind of learning by interacting is central to the operation of ‘the learning economy’ of supplier-user linkages in developing new products. Another key type of learning by interacting, particularly from a policy perspective, involves learning by regulating. This relates to the attempts by those involved with technologies to create the broader circumstances needed for their technology to become established and operate. In addition to formal state regulatory activities, this includes the attempts by private and public players to create technological infrastructures and standards that may be needed for a technology to run, and more generally to enlist other players around particular visions, technology policies and strategies. 4c Strategic Options for Innovation in Multimedia – the Space for Social Learning The SLIM study charted the various ways in which multimedia experiments / innovations are carried out in terms of which actors were involved at which stages and around which aspects of the configuration of a system. This defines the space for social learning (see Figure 6.3 of the report). One possible model for social learning which has received much attention involves the concept of ‘usercentred design’ in which knowledge about the many kinds of potential user and their requirements is captured and fed into the initial design. There were however rather few instances of this kind of prior user involvement. User-centred design is by no means the only, or the best way to adapt technologies to user purposes and contexts. The evolutionary model The SLIM study instead revealed a range of other avenues and modes of social learning. One important model, exemplified by the Amsterdam Digital City (DDS), is the Evolutionary model, in which there is a permanent process of experimentation (encompassing both innofusion and domestication) as the project develops and becomes progressively entrenched. Rather than design 7 preceding an implementation and appropriation effort, both activities were conducted in parallel on a more or less continuous basis. In DDS over time a technology enthusiast subculture opened the project to others – notably media and cultural content providers and, latterly non-specialist users who became information providers as well as consumers. In short, the constituency of players involved broadened in the course of building the market for its products. The project never stabilised; technological development never stopped. Instead the project has undergone a process of permanent evolution in which innofusion and domestication are combined. Design and implementation in multimedia experiments necessitates the involvement of wider arrays of players. This strategy is carried further in the ‘evolutionary development model’: developing the technology, and building the market go hand in hand. In the evolutionary model these are continuing activities and there is no clear boundary between technology development and diffusion. The 'pick and mix' or laissez-faire model Multimedia technologies are increasingly taking the form of configurational technologies - deploying an array of standard and customised element on a 'pick and mix' basis to meet particular local purposes and exigencies. The majority of multimedia experiments identified by SLIM involved applications configured from largely standard technological components - typically today the internet/ multimedia capable personal computer and the world-wide web (but also, for example, the CD-ROM). Although some of our longer-running projects (eg DDS) started off with more specialised text-based communication facilities, they were mainly forced to migrate to the web in the light of its growing popularity – due to the wider array of services and sites thereby available, and the growing preference of users for its attractive and readily understood user interface. The power and flexibility of this platform, and attempts to make it more transparent for non-specialist users opens up the possibility for final users with relatively modest technical skills and resources to configure systems for themselves. This ‘pick and mix’ model can be seen as a rather interesting kind of user-led model of innovation. Although the users involved in these specific cases have virtually no direct involvement in the innovation of the generic component technologies - their links to technology supply are almost entirely indirect and mediated through an impersonal market - they have considerable control over the final configuration and deployment of their system. This is a very different interpretation of user involvement than in classic 'social experiments' with multimedia, in which non-expert users are involved as actors in initial system development. However when we recognise the active and creative nature of innofusion and technology consumption/ domestication – and particularly when we address these processes over multiple overlapping product cycles, rather than limit our view to a snapshot view of a single product-cycle - then we see this pick and mix strategy as an important way in which a broader range of players (beyond technical specialists and decision-makers) can contribute to technological innovation. Overall we find that social learning around multimedia innovation and domestication takes place in many different settings - and can be ordered and organised in very different ways. This mapping exercise illustrates the diversity and richness of social learning and does not suggest that one way of carrying out such learning is 'better' than another.. 4d The Organisation of Social Learning – Mode of Experimentation versus Control An important empirical variable across the range of cases studies by SLIM, concerns their location on a spectrum between what we have described as the Mode of Experimentation and the Mode of Control. This relates to the level of open-ness of the project regarding both participation in the project (and the scope for extending participation) and its goals. In the mode of control, typically, 8 there may be an attempt to define the goals and expected outcomes of a project in advance and rather precisely, and also to prescribe who will participate in development decisions. In the mode of experimentation, there is scope for a more open learning process, with greater open-ness to other participants and agendas. The benefit of the mode of experimentation is a greater willingness to change direction and approach in the light of experience; the 'down-side' may be uncertainty and loss of predictability and control over the direction and outcomes of a project. The modes of experimentation or control involve a number of different, albeit related, dimensions. Particular experiments may be differently located on these dimensions on a spectrum between experimentation and control. These elements are summarised in the table below. Mode of Experimentation open to change process oriented minimum of boundaries and constraints exploratory experimentation open style of management Relevant dimension Mode of Control semi-open to change Product oriented explicit boundaries and constraints --- mode of practice --purposeful experimentation --- style of management --hierarchical style of management Figure 1: Dimensions of mode of experimentation vis-a-vis mode of control ---dominant attitude ---- goal/objective ----- environment --- Control is of course very much a question about who is seeking to exercise control and over what issues. In distinguishing the mode of experimentation and the mode of control we highlight differences in control and open-ness along two key dimensions: the range of players involved the extent of their involvement in the various stages of project development In other words, in the mode of control, control over the project is narrowly restricted (perhaps to a narrow technical or managerial group), and broader arrays of users are doubly excluded - both by not being party to the experiment and because the key design decisions have been taken prior to the roll-out and use of the system. This exclusion and preclusion is less clear in the mode of experimentation. In some of our cases the adoption of the Mode of Control seems to reflect particular contingencies for the Multimedia experiment, notably in the case of projects with a stronger technical development focus, where the mode of control corresponded also to the imposition of a narrow technical division of labour. The development of some of the cultural content products would also seem to resemble the mode of control insofar as cultural production is carried out by professional designers. However there are many dimensions of control. There is no simple link between particular exigencies of a project and the mode of control. For example, rather similar types of projects could differ sharply in this respect. There appears to be no simple relationship between the selection of a particular mode (of experimentation or design) and the success of a project. Our concern with social learning would, in principle, suggest that there might be important learning advantages from a more experimental approach. On the other hand such broader learning may incur potential costs and uncertainties. Our emphasis on experimentation should not obscure the importance of some degree of control and co-ordination in multimedia projects. Indeed, the lack of a central player able to co-ordinate or control these experimental efforts could also prove problematic. The overall lesson from these discussions is that what is required is an appropriate combination of experimentation and control. 9 Both experimentation and control are thus important. However there is a sense that technically focussed projects have often veered unduly towards the mode of control, and that the necessary experimentation has had to be ‘smuggled in’. Technological objectives have taken first place, and only after they have been realised has space emerged for experimentation about usages. The presumption that the technology would provide a solution per se meant that there was a failure to provide sufficient time and a safe context for users to ‘play around’ with the technology. Thus one of the important strategic choices in building and conducting a multimedia experiment concerns the dimensions and timing of control. 4e The key role of intermediaries Multimedia applications require diverse resources and different kinds of knowledge: about the technology infrastructure, about graphics and presentation of information, about the user context. These resources are virtually never all available within a single organisation. This is because the viability of a product (eg a new delivery system) may depend upon the availability of programmes (content) to run upon it, as well as the ability of different technology offerings to work together. The incompleteness of information available to individual players, as well as economic advantages from sharing costs and reducing uncertainties generates pressures for collaboration, even between competing players. Multimedia developments typically involve more or less formalised coalitions of players rather than single organisations. There must be flows of knowledge and ideas between diverse players involved in supply and use. The development and use of multimedia products and services thus almost always involves an interchange between players with widely differing institutional locations, expertise, cultures and commitments (including ‘domain specialists’ eg in the media or commercial services, as well as technology suppliers). Social learning thus starts with the ways in which players with complementary capabilities and domains of experience and expertise, work together in resolving the issues surrounding the development and appropriation of new technology-based products and services (learning by doing). However for such localised learning to contribute to wider social learning, what has been learnt must be communicated and applied more widely (i.e. learning by interacting). The central issue for this study thus concerns how to organise this kind of collaboration. What are the mechanisms for acquiring, generating, enhancing, and exchanging knowledge? This study has highlighted the role of the formal and informal networks linking together constellations of players involved in technology development and use for such knowledge flows. Certain actors, whom we describe as intermediaries, played a crucial role in mobilising these networks and supporting these knowledge flows. Intermediation is frequently shared amongst several players - though some may play a more central role than others. Intermediating roles emerge in the course of a multimedia experiment rather than following formal and pre-determined structures and roles. Moreover there may be different kinds of intermediation taking place at different levels; different kinds of intermediaries may be nested within each other, and roles may change during the lifecycle of a project. Particular individuals often emerged as intermediaries in the course of multimedia projects - though this role was not always formally recognised, and often emerged in the interstices of organisational structures. This was particularly apparent in the education studies where, although their importance was not recognised at the outset, teachers were crucial as intermediary users and intermediary producers of multimedia, and could be the driving force of multimedia innofusion particularly where they have space to manoeuvre. An ability to mobilise knowledge and resources within and outside their organisation/department seems crucial for successful intermediation - particularly crucial in the early 'pre-project' stages, where no formal commitment of resources had been made. The role of intermediary requires an ability to 10 cross different spaces - between different organisations and different departments within organisations - and between different knowledge communities (especially between technical specialists and non-technical communities of providers from service and cultural sectors and users). Appropriation Intermediaries Many multimedia experiments and trials revolve around the role of certain appropriation intermediaries as key points of interface between potential users and new multimedia products and services. They may play a particular role in configuring ICT technologies and systems towards particular potential user constituencies. This seems to involve a domestication effort, to make a technology relevant and attractive to individual consumers. For example the success of Cybercafes in offering a point of access for users to new supplier offerings seems to revolve around the intermediating role of cybercafe managers in promoting and channelling the appropriation process. This reminds us that even fairly well established technologies such as computer games are still not yet ‘transparent’ for the general user in that they are difficult to set up and use, and their utility/ value for the user is not yet established. Appropriation intermediaries may thus play a crucial role in bridging the gap between 'generic' supplier offerings and consumers' perceptions of what they want to buy/ use, which may not mesh together. Cybercafe managers provide a commercially viable service not only by sharing the purchase costs and risks between multiple consumers but also, partly, by transferring products from one context to a more accessible one, thereby adding relevance and value for the customer. 4f) What motivates social learning? How it is shaped by its local context Put simply, social learning takes place where it is in the interests of players to collaborate and exchange information. Intermediation and participation in fora for social learning involves costs and commitments. Constellations survive because the parties are motivated to take part and remain part: the individuals or organisations involved make the necessary investments because they expect some benefit. Foremost are the benefits of co-production and sharing of knowledge about multimedia development and appropriation. It may be in the mutual self-interest of supply-side players to collaborate in building new markets. However the situation of final users may be rather different to that of suppliers and intermediate users. Constellations do not necessarily emerge spontaneously. Key intermediaries may need to steer the process - and motivate the inclusion of whatever players and resources are needed. Here the studies pointed to the important role of public support in providing the resources where these are lacking. Social Learning is a multi-level game The participants in multimedia experiments were often involved in a complex and multi-level game. As well as the formally stated shared goals of a projects, particular participants had certain more or less covert goals, which were not necessarily shared. Furthermore, the various players involved tended to have a wide range of commitments and interests in a project at a number of different levels. For example a technology supplier might have several goals: to forestall and warnoff competitors, to align expectations of potential customers, to signal competences and establish a reputation as a future player in this and related markets) which could be more or less closely geared to what they might expect to learn from taking part in an experiment. Social learning may also be taking place at a number of levels around different issues. For example the launch of a multimedia pilot may be motivated by a concern to address the technical and operational problems in getting an infrastructure to work as well as a desire to try-out particular applications or popularise certain service concepts. Learning about complementary players and 11 about how to work with them in developing and maintaining new multimedia services may be equally important. It is essential to stress the complexity of goals and interests against which particular projects may be developed. One consequence is that assessing whether a project can be deemed successful becomes rather a complex and subtle matter. Important issues arise on the conclusion of a project. If a project is wound-up some players may be forced to move to other areas where there is little prospect of applying their experience. Will the players involved continue to have an institutional context that provides the resources and motive to carry these ideas forwards subsequently. For example, problems may arise, particularly in relation to public-funded experiments, if the constellation and some of the key actors involved in the project exist only for the duration of the experiment. In assessing social learning in a multimedia experiment we also need to take into account different timescales and contexts. To assess the significance of a multimedia experiment we must address the experimentation and social learning about the design and use of multimedia products that may take place within the project. However we must also examine how these may form part of a broader set of learning processes around multimedia products and services over a longer-timescale and within a broader setting. 5 Conclusions and Policy Implications 5a) Shifting away from a technology supply perspective The key policy challenge posed by the SLIM study surrounds the need, we identified, for a 'double shift' in the focus of technology promotion efforts, away from their rather exclusive concern with promoting the development and supply of core technologies including: The appropriation activities of intermediate and final users; and, associated with this, The need to support the development (and appropriation) of cultural and information content. The SLIM project points, instead, to the need for public support to be geared towards multimedia innofusion and appropriation as well as, and perhaps as an adjunct to, RTD projects. Moreover, support for multimedia uptake (which is currently largely cast in terms of 'awareness raising' - as if there existed a finished set of multimedia capabilities which potential industrial or private users merely need to be aware of) needs to recognise the experimental nature of appropriation. The implication would be that uptake initiatives should include provision for a creative effort and for dissemination of appropriation experiences to other appropriators and to future technology supply. Second, attempts to shift away from the linear, technology-driven model of innovation must engage with the sharp disjuncture between the cycles of innofusion and the cycles of appropriation: a) Roll-out and commercial launch typically takes an order of magnitude more resources than initial technical development. However resources for commercialisation have been less-readily available than for technology development. b) The process of domestication/ appropriation typically occurs over a much longer time scale than technology innovation. Technological dynamism and ever-shortening product cycles mean that the ICT-supply industry considers 2 - 5 year old offerings as obsolete. In contrast it still takes well over a decade for consumer ICTs such as the CD to establish their market. Since different parts of the innovation system operate with different dynamics, tensions may exist between innovation/ innofusion of technologies (for example the technology infrastructure/ delivery 12 system) and the promotion of a stable and homogeneous market - particularly for information and cultural content. Innofusion processes will tend to lead to a patchwork of more or less unique configurations. This presents particular problems where an information infrastructure is emerging through innofusion. High levels of dynamism and turbulence around technological trajectories and standards will tend to encourage consumers to defer purchasing decisions. Moreover, the outcome of extremely dispersed and fragmented decisions about the technology platform will be an extremely heterogeneous technology infrastructure comprising many specific local configurations. This constitutes a very uneven and fragmented market for complementary products. We see this already in the problem presented to information service/ application providers of supporting a plethora of standards and platform capabilities. This affects the usability of the system as well as posing particular problems for design. 5b) Transferability Transferability is a central concern in discussions of social learning. The concept of social learning seeks to analyse not only local learning experiences but the ways in which they may be communicated more widely and 'translated' to make them applicable in other contexts. Public and private funders of multimedia experiments have the commendable desire to ensure that the projects they are supporting do lead to useful outcomes, and can be more generally applied. The pursuit such demonstrable research outcomes seems, paradoxically, to have often had the consequence of promoting a search for concrete 'deliverables', and in particular channelling expectations towards the idea that the transferability of project outcomes will be through the wide dissemination and use of artefacts. The focus on concrete deliverables has been unhelpful, we argue for a number of reasons: it may discourage the kind of experimentation about usage that we have identified as key to the domestication of technologies. the complexity and uncertainties surrounding commercialisation and appropriation of technology mean that this form of exploitation is often unsuccessful, at least in the short term. production of a material artefact is but one of a number of potential outcomes from a research project (or even a commercial trial), and is not necessarily the most significant transferable element. Equally important, potentially, are non-material outcomes. These include the development of specific social relationships amongst diverse players involved in technology development and use, and knowledges of the processes of developing these relationships and technologies. Though the nonmaterial outcomes of a multimedia experiment may be far more significant than the material artefacts generated, there may be particular difficulties in capturing and conveying them. Many of the lessons are extremely contingent - and therefore may be hard to communicate and generalise. However, the vision of developing a new multimedia artefact holds out the prospect of the wider dissemination of the product or service - offering an apparently clear and self-evident metaphor for the exploitation process, and one that contains within itself the promise of social benefits and economic gain. However, an examination of multimedia experiments shows that such a linear exploitation trajectory is more the exception than the rule. The costs of commercial development are enormous, even in those cases where technical innovation has been successful in creating a new artefact. The costs of roll-out and commercial launch of new, mass market platforms are probably an order of magnitude higher than the technical development costs. Such investments are only likely to be supportable by the very largest global players. One consequence of the prevalence of the linear model, and the focusing of public support for the development of artefacts, has been the relatively limited availability of funding for exploitation. RTD support tends to be finite - and there is a danger that, once a project comes to the end of the funding 'conveyor belt', there may not be resources to carry forwards the development and commercialisation of the artefacts (and, no less importantly, to maintain the expertise and 13 relationships developed amongst the collaborators). The danger is that developers may be drawn by the availability of RTD support into the further technical development of an artefact, or a whollynew development project, simply by the need to find follow-on funding - and away from the commercialisation and societal appropriation of existing offerings. Novel modes of support (with larger levels and longer timescales of support) may be needed for the commercial launch and wider uptake of multimedia artefacts. This has implications for the way in which public support is provided - for example RTD projects need to have more elaborate 'exit strategies' when public support ceases. The focus on artefacts has had a number of other unhelpful consequences from the point of view of social learning around multimedia products and services in terms of the way in which multimedia experiments are operate. There seems to be a general tendency for funders to seek to fund novel elements - and a tendency to see such novelty primarily in terms of technical novelty. In a context of limited time and resources, the emphasis on technical development may divert attention away from technology domestication. However some of the most valuable outcomes of a multimedia experiment or trial are non-material, including, notably, its contribution to the development of specific social relationships and the acquisition of new areas and combinations of knowledge. The circumstances for developing and exploiting these may be very different than those for material artefacts. There are significant problems about the transferability of these kinds of knowledge: Where such knowledge is tacit and difficult to formalise and communicate more broadly. Where such knowledge is highly contingent to particular circumstances and stages of innovation; raising questions about whether and how it may be possible to extrapolate and apply this knowledge in other circumstances. 5c) Success and failure of a Multimedia Experiment Social learning shows that attribution of success or failure is often contested across projects - not least because different parties are likely to have different commitments to and perspectives upon a project, and may draw different kinds of benefits and lessons. Formal definitions of success and failure are uninformative and may speak more to local political contingencies. Our study showed how dominant actors were able to impose their criteria and notions of success and failure. However, failure was largely denied and neglected while success was announced and exploited. It is only under very particular circumstances that sponsoring groups find it in their interests to define a project as a failure. The search for 'success factors' and to identify 'best practice' can be misleading as well as potentially uninformative. Questions arise about how far we can reliably extrapolate from ‘success criteria’, templates and exemplars based upon established technology products and services. Are these metaphors reliable in different settings? Have the rules of the game changed over time? We found that it was not possible to formulate a simple account of 'best practice' either regarding the model of development, the goals of the project or the management of the technical changes. The factors that seemed to underpin particular successes were often highly contingent to the particular circumstances of the case in a way that impedes extrapolation. Knowledge about change processes seems to provide a more reliable basis for transferability than correlations between specific factors and outcomes. The outcomes of multimedia are often rather mixed and the assessment of success may thus be rather complex. From the point of view of social learning, what is more important is not whether a project is assessed as successful, but whether it provides for flows of information and alignments of players. From this perspective, it is not necessary for a Multimedia experiment to proceed to widespread use for it to make a helpful contribution to social learning. Some of the most important learning may take place in relation to projects that failed to meet the objectives set at the outset or that did not proceed to roll out. 14 A general consequence of social learning - and an important feature of multimedia experiments - is that it is not possible to fully prefigure the outcomes of a project at the outset. The conditions for promoting social learning may indeed be in conflict with the desire of project funders to seek to establish project outcomes in advance. We see this, for example in the tendency of many multimedia projects to be 'verification experiments' - geared towards confirming the value of multimedia products in meeting preconceived success criteria - rather than more open 'diversification experiments'. We must bear in mind indirect and unanticipated outcomes that may result over long timeframes and in other settings. Benefit can be drawn not only from experiments that succeeded in meeting their goals but also from those that apparently failed. The utility of a multimedia experiment cannot be determined solely in relation to how the experiment is conducted, but has to be assessed against its broader context and history of socio-technical change, and the specific challenges addressed. Let us examine a case where an experiment yields new artefacts and practices. Other broader and more diffuse forms of social learning may take place - for example as the system is rolled out. The utility and value of these direct outcomes will ultimately be verified by whether they are carried forwards and built upon or alternatively rejected and refuted. In other words, the value of an experiment is related to the extent to which the intense, focused and organised forms of social learning it affords adequately emulate the more diffuse 'organic' forms of social learning that may take place subsequently and across the full array of groups that may be involved. 5d) How can public policy support social learning? The social learning perspective poses a number of challenges regarding the policy process. On the one hand it criticises work that attempts to draw mechanistic policy conclusions, for example, around the linear model of technological change, and highlights the uncertainties that surround the development of policy. At the same time it opens up new policy opportunities. However the critical scrutiny that the social learning perspective encourages also problematises the process of policymaking. In particular it encourages a shift away from making particular substantive policy recommendations towards an approach which seeks to foster certain policymaking processes. A key issue here concerns the extent to which public policy can support social learning (and in particular the knowledge flows needed both to transfer social learning achievements and to apply them more widely). However flagging the importance of social learning does not resolve the policy problem. Social learning processes are far from straightforward. Although in theory all experiences may yield lessons that are in some sense useful, this does not imply that all social learning outcomes are equivalent. The concern by project funders to be seen to have supported 'successful' projects can have negative consequences by favouring less experimental approaches, and suppressing potentially valuable experiences from projects which appeared to be 'failures'. Social learning is a contradictory and 'lumpy' process. 'Unhelpful learning outcomes' are possible” for example there is a danger of inappropriate mimicry of external examples rather than adaptation to local circumstances, where developers and promoters become too closely committed to particular concepts of technology which prove neither feasible nor appropriate. Another question that must be addressed concerns the allocation of public resources. We are interested in the consequences of state resourcing both in relation to the conduct of an experiment and its further exploitation. Despite the laissez-faire rhetorics of public policies for the information society/ information superhighways, the state seems to play a key role across a wide range of areas of multimedia development, both in relation to areas of potential 'market failure' and for public information and services. Given the high costs and deep uncertainties surrounding multimedia developments and markets, private players may not find it in their short-term interest to embark upon developments, particularly where there are uncertainties about whether a critical mass will be attained to make a project viable. State resources - knowledge and co-ordination efforts as well as financial support - are crucial, and underpin the bringing into being of many of the multimedia 15 experiments. Furthermore such public support often provided the main resources deployed by the intermediaries that were at the heart of multimedia constellations, in building and maintaining experiments. 5e) Unhelpful Learning Outcomes Public support may have 'unhelpful' learning outcomes (for example the 'suppression of public failure' which may impede 'learning from failure'). Similarly, the survival contingencies of particular projects that are dependent upon external income may have unintended and undesired consequences on the behaviour of particular actors or groups. Learning to be fundable: For example, one challenge for many projects, especially those depending upon state-subsidies, concerns the need to learn to be fundable. To enlist the support of external funders it may be necessary to cast a project proposal in terms that match the presuppositions of potential funders. This may be reflected in the adoption of unrealistic or inappropriate goals in order to bring funders on board. Sometimes the result could be that sponsors' goals (driven perhaps by their particular visions of the benefits of multimedia) took precedence over user needs. Once a project has been established, how can it continue to secure the resources it needed? It proved relatively easy for many of the public-funded projects studied to get money for new project development. They found it much harder to obtain long-term funding, e.g. to establish an ongoing service or continued support for an existing project (particularly since an increasing proportion of central state funding is allocated, often very-competitively, on a project by project basis, in the context of attempts to improve the effectiveness of state support). Selectors look for novelty and additionality from new services - which may force those already involved in particular experiments and services to continually recast their offerings according to the exigencies of funds being currently disbursed to obtain further funding. Where project actors have to continually reapply to do new and novel things to gain renewed funding, there are dangers of a drift in goals. Instead of projects developing cumulative achievements, new goals are adopted that undercut fulfilment of goals from previous rounds of funding. There is also a danger of 'project fatigue' where those involved appear to lose a sense of direction and mission. Learning to be viable Conversely, as a project moves towards completion, there may be a need to learn to be viable. This has historically been a problem for many state-funded demonstrator projects. One criticism that can be made of guaranteed state funding for an experiment is that it may divert participants' attention from the need to assess and build the future market needed to ensure the wider uptake and perhaps commercially viable of a new product and service. This had certainly been an important failing of certain historical social experiments. Some of our cases achieved viability by building an effective market of 'users'. However this remained the exception. Other cases are still dependent upon state support - and there would be questions about their future viability were this to be withdrawn. 5f) Modes of public support to promote social learning? Social learning may be said to reach fruition when a technology becomes disseminated and embedded within particular social settings. However, current modes of public support for multimedia may not cater well for such evolution. In particular, there is a lack of fit between the relatively short-term discrete funding models that are typical of much public support (under the project-based mode of support) and the longer-term requirements of multimedia experiments 16 driven by the dynamics and timescales of social learning. Public support for this 'post-project' stage proved difficult to obtain. Many multimedia experiments experienced difficulties in obtaining the kinds of longer-term funding needed to support the exploitation of technologies they had developed and in particular for the appropriation, entrenchment and commercialisation of their offerings. The prevalent model of finite, project based funding may give little scope for a project to become established, bed down, and provide the basis for more far-reaching societal learning process. One implication might be that current regimes of project-based funding might need to be reformed specifically to cater for these long-term requirements - for example by encouraging longer-term multi-phase funding - either from the outset or through extension awards as a development project comes to an end. In many countries there are more or less covert modes of public support for commercialisation. However these tend to be separate from research support. The social learning perspective calls into question this traditional separation between (imputedly pre-competitive) research and development and a competitive commercialisation phase (which by implication does not involve research). Cultural products may be a special case here: Their high up-front development costs may create disincentives to producing content geared towards minority cultures and small markets, raising the risk of market failure. There could be a synergy here between policies geared towards technology appropriation and those concerned with the preservation of cultural diversity in the face of the globalisation of the media economy How best to integrate public and commercial funding? On the other hand provision of long-term public funding is not without problems as it may divert the attention of participants away from the need to become viable by building a user base. There are dangers of a culture of dependency and more significantly through a failure to address the market for products seriously, a tendency towards an approach dominated by technology supply. The question arises of whether different funding strategies can be devised that balance between public and private support, and between long-term security and the necessity for short term survival - to offset such potential problems and encourage more effective social learning? Here we need to look at the integration of public and private support. As already noted, many multimedia projects were hybrids, involving various private and public sector players. We have already highlighted the need for systems of resourcing to provide incentives for social learning, in terms of the exploitation of experiences gained within a project. We should consider here both 'direct' exploitation - for example in terms of the entrenchment/ commercialisation of a new multimedia product and service - and indirect exploitation in further trials and experiments. These two may be in tension, with the former perhaps requiring that existing project constituencies to remain broadly intact and grow. For the latter this is not necessary. Closure of a project could of course be a vehicle for dispersing staff and disseminating their embodied knowledge more broadly. However, for this to happen, it is necessary that the players continue to be involved in the broad field, in roles where this experience can be applied. Either way the key question is whether the intermediaries involved continue to exist albeit in a different setting, with the necessary incentives and context to apply their experiences more broadly. 5g) What has been achieved; what remains to be done When the research questions and perspectives of the SLIM project were being elaborated, prevailing knowledge and opinion stressed the need to acquire more detailed knowledge of the setting and purposes of different kinds of 'users', and build this knowledge into the initial design of multimedia offerings. With hindsight we can see that this was an idealised and simplified schema. For example, designers do not possess clear representations of the user and the user's setting and instead work with rather implicit and poorly specified models based on incomplete and often 17 unreliable sources of information. We found further that ‘user-centred design’ was by no means the only way by which users could contribute to improving the design and usability of multimedia artefacts. There were a variety of social learning strategies including evolutionary and laissez-faire models in which users contributed directly to system configuration in the innofusion/ appropriation of artefacts. Multimedia offerings are not ‘solutions’ to social need, but are elements which ‘users’ selectively appropriate and configure with other artefacts, meanings and practices, around their evolving purposes. In this sense multimedia offerings are always to some degree ‘unfinished’. In short, though the SLIM project showed that social learning is important - it revealed a rather different empirical reality from that presumed at the outset. For example, the initial presumption that multimedia experiments would provide a context in which suppliers, users and other players could meet and negotiate their requirements was undermined by the scarcity of such exemplary ‘social experiments’ in which users were involved at the outset. On the other hand all multimedia developments were in some way experimental, involving various forms of collaboration and learning within and between organisations. The diverse forms of multimedia initiatives confirmed both our sense that social learning was important, and also that there was considerable uncertainty about how such learning should be organised. Learning did not imply a smooth and progressive accumulation and sharing of knowledge; we found that the institutional arrangements were in most cases fragmented and subject to deep fissures and rifts. Activities tended to be clustered around either technology development or more frequently its configuration and appropriation - and there was a marked absence of constellations bringing together developers and final users. In this context we identified the importance of intermediaries as key players in social learning, linking together actors from different sites and serving as conduits for the exchange of knowledge and other resources. The important contribution of intermediaries was, however, often overlooked. Public policy needs to pay more attention to intermediaries as a resource for innofusion and the domestication of multimedia. Many contemporary multimedia developments are configured from combinations of more or less standard technology components. Development of novel technical artefacts may be less important than this local appropriation effort. There is a consequent need to shift effort away from technology development, narrowly conceived, towards an appropriation effort, focused upon innofusion and domestication. The development of multimedia content is also resource intensive and requires support. Widespread criticism of linear, supply-driven concepts of innovation has heralded more integrated approaches which give more attention to commercialisation efforts and building markets. It does not follow that policies can be developed that will secure better supplier-user coupling and a harmonious integration of technology development and diffusion. In contrast this study has revealed the tensions and contradictions surrounding multimedia development, for example between the different dynamics of the development of component technologies and of the appropriation of applications. Whilst an integrated approach to innovation policy is indeed needed, this does not imply the adoption of a monolithic model that couples these tightly together. Instead loose coupling may be needed between some policy elements within an overall framework. For example, strategies may need to acknowledge the differing dynamics of the development of core technologies, delivery systems/ platforms, applications and the information and cultural content they hold, and thus allow greater autonomy between policy interventions across these fields. We cannot yet reliably predict whether user-centred design, evolutionary development or laissez-faire models will be more effective in particular cases! Finally we have addressed the relationship between the local and the global. Technology innovation has traditionally been seen in global terms - and European technology policy has often been couched in terms of combating and competing with global competitors from the USA and Japan. In contrast the SLIM study has demonstrated the enduring importance of the local - particularly in relation to certain kinds of cultural content concerned with local identities. Rather than the dominance of the global over the local, the SLIM study points to the interweaving of local and global elements - and the continual reworking of the relationship between them. In this context there are enormous 18 opportunities for players at local, regional, national and European levels to contribute to the social learning processes that will enable us to genuinely domesticate global ICT capabilities and put them to the services of the people of Europe, and to bring about the benefits promised by the Information Society rhetorics of wealth creation, knowledge sharing, enhancing services, strengthening community and enriching culture. 19 Figure ES 1 Table of Deliverables Deliverable Deliverable 1a State of the Art Review Deliverable 1b State of the Art Review Deliverable 2a National and International Studies Reference Williams, R. (1997) The Social Shaping of Information and Communications Technologies. pp. 299-338. In Kubicek, H., Dutton, W. H. and Williams, R. (Eds.) The Social Shaping of Information Superhighways: European and American Roads to the Information Society Frankfurt and New York: Campus Verlag and St Martin’s Press. http:/ / www.ed.ac.uk/ ~rcss/ SLIM/ public/ phase1/ SSICT.html Sørensen, Knut H. (1996) ‘Learning technology, constructing culture. Sociotechnical change as social learning’ STS working paper no 18/ 96, University of Trondheim: Centre for technology and society. http:/ / www.ed.ac.uk/ ~rcss/ SLIM/ public/ phase1/ knut.html Williams, R. and R. Slack (Eds.) Europe Appropriates Multimedia: A study of the National Uptake of Multimedia in Eight European Countries and Japan, Senter for Teknologi og Samfunn (Centre for Technology and Society), Report No. 42 Norwegian University of Science and Technology, Trondheim: Norway. ISSN 0802 3581 42 Kubicek, H., Beckert, B., Williams, R. and Stewart, J. (forthcoming) The Social Shaping of Multimedia: an international pespective Deliverable 2b International Review of Multimedia Experiments Deliverable 3a Education Deliverable 3b Cultural industries Deliverable 3c Public Services Deliverable 3d The Home Lobet-Maris, C. (forthcoming) A Comparison of the United States NII Policy and EU Multimedia Policy Jaeger, B. Slack, R. and Williams, R. (forthcoming) ‘Europe Experiments with Multimedia: An Overview of Social Experiments and Trials’ The Information Society Special issue on Multimedia. http:/ / www.ed.ac.uk/ ~rcss/ SLIM/ private/ phase2/ SocExp.pdf Lieshout, M. van, Egyedi, T. M. and Bijker, W. E. (Eds.) Social Learning in Multimedia: Education (Provisional title) Aldershot: Ashgate. http:/ / www.ed.ac.uk/ ~rcss/ SLIM/ private/ I-studies/ mvl/ mvl.html Preston P. The Same Old Story? Multimedia as ‘cultural content’ and emergent new media form http:/ / www.ed.ac.uk/ ~rcss/ SLIM/ private/ I-studies/ PP/ PP.html C. Lobet-Maris and B. van Bastelaer (Eds.) Social Learning Regarding Multimedia Developments at a Local Level: The Case of Digital Cities Namur: CITA, University of Namur. http:/ / www.ed.ac.uk/ ~rcss/ SLIM/ private/ I-studies/ BvB/ digitalcities.html Sørensen, Knut H. Homeward bound? Multimedia in everyday life Deliverable 3e The organisation http:/ / www.ed.ac.uk/ ~rcss/ SLIM/ private/ crosscutting/ KHS/ homestudy.PDF Rossel, P. (1998) Cross-cutting perspectives on the social learning of multimedia in the organisation. Deliverable 4 Final report http:/ / www.ed.ac.uk/ ~rcss/ SLIM/ private/ crosscutting/ PR/ PRorgsum.html Williams, Robin, with Roger Slack and James Stewart Social Learning in Multimedia: Final Report to European Commission, DGXII, Edinburgh, RCSS, 1999. R. Williams 'Public choices and social learning: the new multimedia technologies 20 in Europe' The Information Society Special Edition on Multimedia, guest editor R. Williams forthcoming 2000. 21 II Background and objectives of the project Chapter 1 Introduction The installation of ‘Information Superhighways’, coupled with advances in processing power and usability of information technology, have led to widespread expectations of the rapid adoption of a new cluster of technologies under the rubric of multimedia. Applications based upon these technologies are, moreover, expected to be widely diffused in many areas of working and social life, and to have profound social and economic implications. In short, these technologies are expected to underpin the transition to an information society. However, beyond this global vision, there is little certainty about the kinds of applications that can be expected to emerge. How can we assess the prospects and societal implications of these new technologies? At present expectations revolve largely around visions emerging from technology practitioners, pundits and suppliers, driven by promises of what will be technically-achievable. However history shows us that what eventually emerges falls far short of these technically-driven visions, in terms of both the rate of change and its economic and social benefits (Bruce 1988, Dutton 1995). It also shows us that in some instances change can be far more rapid and far reaching than expected, the fax and the world-wide-web being cases in point (Coopersmith 1993, Oakley 1990). Socio-economic research into the development, implementation and use of existing and emerging technologies has an important contribution here. A major European study has been undertaken of Social Learning in Multimedia, funded under the Targetted Socio-Economic Research programme of the European Commission Fourth Framework Programme. This report summarises some of the main findings. This eight country study analysed the development, implementation and consumption of the multimedia technologies that are expected to underpin the transition to an information society. The study draws insights from the 'social shaping of technology' perspective and other recent advances in technology studies (MacKenzie and Wajcman 1985, Williams and Edge 1996). 1.1 The context of the study There has been much recent debate about the social and economic implications of new technologies, and about the public policies (and business strategies) needed to promote beneficial technological advance. 'Linear models' which focus narrowly on generating novel technology artefacts and knowledge are no longer seen as sufficient to achieve successful innovation and thus the creation of new wealth and jobs, let alone broader social goals such as a better quality of life. The technological 22 systems of modern society are becoming more complex and elaborate in terms of both their 'technical' operation and their societal use. And as technologies become more pervasive and central to society, we come to expect more of them. These factors, coupled with the growing dynamism and pace of social and technical change, add to the difficulties of technology planning and management. Some have suggested that we are finding our way towards a 'third phase' of Research and Innovation Policies in which the key challenge is to couple innovation and society combining social objectives with priorities for the dynamics of innovation (Caracostas and Muldur 1998). Perhaps the foremost example of such endeavours is the European Commission's Fifth Framework Programme (EC5FP), which seeks to recast its earlier approach to Research and Technology Development (RTD) within a distinctive 'integrated, problem-solving approach'.1 Many of these points apply with special force in relation to information and communications technologies (ICT). Earlier technology policies that focused upon promoting the supply of new technologies had not delivered all the economic and social benefits that had been promised. For example, Europe's balance of trade deficit in ICTs was growing. There was moreover a gulf between people's everyday experience of ICTs and the confident predictions and compelling visions, articulated by technology suppliers and pundits, that ICT would transform our lives for the better (Forester 1998, Dutton 1996). In this context, 'meeting the needs of the user' became a slogan for the IT industry - taken up for example by ESPRIT the European Strategic Programme for Research on Information Technology under the European Commission's Fourth Framework Programme. There was, however, little understanding of how to match emerging technical potentialities to user need. This was compounded on the one hand by the dynamism of technology supply at a global level (involving an unprecedented pace of technological innovation coupled with the extreme turbulence of technologies and markets). On the other hand were the deep uncertainties surrounding the application and use of these technologies: a challenge magnified by the fact that advanced ICTs were becoming generic in their application across all industrial sectors and, with the development of multimedia, were expected to become ever more pervasive in their potential adoption in work and in many areas of everyday life. 1.2 Debates about the Information Society: the Multimedia Revolution and the emerging Information Superhighways It has been widely presumed that digital data processing systems that could handle text, graphics, sound and video signals would be easier to use and more engaging for consumers. The emerging 'Information Superhighways' would bring these 1In particular, the European Commission has laid out its aims for the Fifth Framework Programme as being to make research not only 'more efficient' but also 'increasingly directed towards meeting basic social and economic needs by bringing about the changes which each individual citizen desires' (European Commission, 1996 COM(96) 332). The EC5FP, for example integrates research and technology development within targeted, policy oriented Thematic Programmes encompassing the entire spectrum of activities and disciplines (social as well as natural science) needed to achieve these economic and social objectives. 23 multimedia products and services into every home and workplace. By the mid-1990s, public policy and debate focused upon the implications of these changes - with new policy frameworks emerging from the USA, Japan, Europe and its member states (Kubicek et al. 1997). The influential Bangemann Report (High-Level Group on the Information Society 1994) argued that Europe was becoming an information society - already in the throes of a new industrial revolution 'as significant and far-reaching as those of the past'. The Bangemann Group saw ICTs as leading to new jobs and wealth creation, coupled with 'a significantly higher quality of life and a wider choice of services and entertainment'. Bangemann forms part of a broader current of discourse about the imputed digital or multimedia revolution and the shape of the emerging information society. Though particular accounts differ in some respects they also broadly exhibit a number of common elements. The Social Learning in Multimedia (SLIM) project represents one of the most large scale and in-depth studies to date of empirical developments in the creation and adoption of new ICTs. We are thus able to subject to critical scrutiny some of the widespread presumptions that have been made about this imputed multimedia revolution. SLIM has yielded a mass of empirical findings and has developed important analytical concepts which offer an extremely rich account of the process of technological innovation with multimedia and its social implications. These are reported in the following chapters. By way of an introduction we can pick out the ways in which SLIM findings engage with some of the most prevalent presumptions about the nature and implications of these changes. a) the multimedia revolution? The Bangemann Report argued that ICTs posed a 'revolutionary challenge to decision makers': a revolution that would generate 'uncertainty, discontinuity and opportunity' (High-Level Group on the Information Society 1994). The key technical change in question involves technologically-driven convergence - in terms of the coming together of core data transmission and processing technologies into an integrated digitised technology infrastructure able to handle different kinds of signal (data, voice, graphics, video); the resulting convergence of computing, communications and broadcasting; and finally the integration of different industries, services and realms of activity through these digitised media. This kind of presumption of a change in 'techno-economic paradigm' (Freeman et al 1982, Perez 1983) tends to be conceived as a change whose broad directions are in some ways already mapped out - and that the key policy questions surround how quickly the socio-economic system can adapt to the new exigencies.2 Such arguments have 2An illustration arises in the way that the European Commission's Rolling Action Plan ( European Commission COM (96) 607 27 Nov. 1996) addresses the influence of 'social acceptability' to 'the success of the information society' entirely in terms of measures to facilitate social adjustment (rather than, for example, to change the ways technologies may be utilised!). Thus it states The information society is bringing fundamental changes in the way we live and work in a period of high unemployment. This situation brings uncertainties and poses big challenges for workers, 24 been criticised as forms of 'technological determinism' for conceiving social change as being driven by changes in core technologies. Linked to these general criticisms, some more specific questions can be raised. Discourses of revolutionary change tend to embody a rather monolithic concept of technological development. As a consequence they tend to emphasise the global character of these changes. By implication, the specific technological paths and choices are not seen as problematic, as if the trajectories of technology were well established. Finally, multimedia is treated as if it already provided ready solutions to social need. We shall examine each of these presumptions in turn. b) technology as a global development These changes are portrayed as part of a set of global developments - technological capabilities are seen to have global applicability. Since a global market is presumed for new technology products and services, the question that remains is what share of this global trade Europe shall obtain. Under such a view, the main question about our future concerns which countries will 'get there first' and thereby benefit from new industries that will emerge. Thus, if 'Europe arrives late our suppliers of technologies and services will lack the commercial muscle to win a share of the enormous global opportunities which lie ahead.' (High-Level Group on the Information Society 1994). We would comment that this view of the global multimedia market - with the major economic rewards going overwhelmingly to the successful globally-dominant players - sets up an extremely pessimistic view of Europe's prospects given the current strength of the USA in ICT supply and in the internet. Such a view could be damaging to the extent that it understates the potential opportunities for European firms by concentrating on the most globalised markets (eg the core technologies underlying computer platforms and delivery systems) and downplaying the market for products and services (e.g. services, information and cultural content) which have a much stronger local dimension. This globalised and monolithic concept of multimedia technologies often seems to carry a further implication that the global availability of technological capabilities, and the increasing role of global players in technology and service supply will result in the progressive homogenisation of technical infrastructures, applications and service concepts - and through this, a homogenisation of societies as a whole. In Chapter 4 we pose the question of whether Multimedia is indeed an international medium that cuts across and erodes national differences or whether it is appropriated differently within particular local settings? As we shall see , one of the main findings from our review of national patterns of Multimedia uptake citizens, enterprises and public institutions. Therefore, to ensure social acceptance of the information society in Europe, it is essential to meet these challenges with public policies that provide opportunities to develop the employability of workers, support change in the organisation of enterprises, reinforce social cohesion and cultural diversity and enhance people's ability to participate in the information society. (European Commission COM (96) 607 27 Nov. 1996 ) 25 concerns the enormous diversity of models of multimedia between countries (and, for that matter, within countries), given their different histories, their differing industrial, geographic, technical and cultural contexts. Generic technological capabilities can be taken up differently in different settings. The local dimension proves even more significant in relation to multimedia applications and the services, information and cultural content that run upon them. This finding has important socio-economic implications. Though some elements of the Information Society may be globally available, others will remain localised. Local initiatives will continue to be key. Technology policy needs to encompass this - and some of the main recommendations emerging from this study concerns the importance of local appropriation efforts for multimedia services and content. This finding brings similar economic implications that some of the key market opportunities may be for local applications. We will suggest later that the scope for local markets is particularly marked in relation to cultural products with local symbolism - for example those concerned with local identities. c) multimedia trajectories and development paths The global and monolithic vision of the development of multimedia is often associated with a view that sees the particular development paths and choices for multimedia as in some way not problematic. In some accounts this reflects a presumption that there is a well-established development trajectory - perhaps reflecting imputed trends in core technologies (for example, ideas about the convergence of currently separate media and services [broadcasting, telephony, computing] into a single digitised medium). Other accounts see 'the market' as the driving force. The Bangemann Report falls in to the latter category. Its emphatically laissez-faire view insists that the market, rather than the state, will 'decide winners and losers'. The 'prime task of government is to safeguard competitive forces and ensure a strong and lasting political welcome for the information society, so that demand-pull can finance growth' (High-Level Group on the Information Society 1994). However, after over a decade of waiting, 'killer-application' that was expected to bring multimedia into every home has still not emerged (Dutton 1995, 1996). Some features have become established (for example, the exponential growth in the internet and world-wide web have consolidated their position as probably the dominant framework for on-line services. However digital TV and other platforms (notably mobile telephony) are still 'waiting in the wings'. A variety of technical configurations are possible (and we will suggest that the outcome is likely to be a web of overlapping, rather technologically heterogeneous, webs rather than the kind of orderly and seamless web suggested by the 'Information Superhighway' model). There is, moreover, little certainty over what kinds of service model will prevail. 26 d) multimedia offers already available solutions to social need - that merely need to be diffused into society The presumption underlying Bangemann's laissez-faire view is that multimedia technology is already available, more or less 'off the shelf', and that it 'offers attractive and cost-efficient solutions to meet consumer needs' (European Commission COM (96) 607 1996). In this case, the key remaining task is to ensure the wider diffusion of these technologies and the growth of multimedia markets. However our study shows that multimedia offerings are far from being 'finished products' in a number of ways: i) ii) iii) First the technologies are still unstable. They are still evolving rapidly and often prove rather difficult to deploy. More profoundly their applications and uses are, with only a very few exceptions, still a long way away from becoming established and widely accepted. There is still a need for inventive activity - by suppliers, by producers of content and by users - to work out ways in which these technologies may be made relevant and useful to potential consumers. This lack of stabilisation and closure of technology is compounded in the case of ICTs by the fact that these are designed to be flexible technologies - able to be reconfigured and applied in ways not anticipated by the designers. They increasingly take the form of generic platforms and media that can be used with a range of applications, services and types of information and cultural content. In this sense multimedia technologies may never become 'finished products', demonstrating product stabilisation and maturation of markets in the way that has been seen with other consumer goods. These observations lead us to move away from conventional technology policy approaches which conceive innovation as limited to initial research and development and followed by a separate stage of technology diffusion (for example through demonstrator projects and awareness campaigns). Instead a rather different kind of activity may be needed that recognises the important innovative effort that follows initial product design and development and focuses upon the local appropriation of generic multimedia capabilities. We return to this in the next section. e) experimental multimedia applications Despite these weaknesses, the Bangemann Report contains a range of important proposals. One of the features that has proved most important is its proposal to launch experimental applications of multimedia. This proposal is driven by recognition of the possibility of market failure - despite the report's overall emphasis on the role of market forces - where a technology cannot establish the critical mass of users to become financially viable. Thus heightened competition by itself will fail to "produce - or produce too slowly - the critical mass which has the power to drive investment in new networks and services". In order to "create a 27 virtuous circle of supply and demand" it calls for a significant number of market testing applications based on information networks and services can be launched across Europe to create critical mass. Initiatives taking the form of experimental applications are the most effective means of addressing the slow take-off of demand and supply. They have a demonstration function which would help to promote their wider use; they provide an early test bed for suppliers to fine-tune applications to customer requirements, and they can stimulate advanced users, still relatively few in number in Europe as compared to the US. (High-Level Group on the Information Society 1994). We are intrigued by Bangemann's insistence that these initiatives 'are not pilot projects in the traditional sense' and but are demonstrators: since their first objective is for market testing 'to test the value to the user, and the economic feasibility of the information systems' they therefore need to be launched in real commercial environments, preferably on a large scale to be truly effective. Despite these protestations, the support given to public and private experiments has proved to be one of the most important and effective features of the Bangemann report (an approach that has been extended to the G7 countries - and globally, for example through the Global Bangemann Challenge). The value of these initiative has not been as narrow demonstrators of a finished technology, but as a space for experimenting and learning, bringing together technology providers with other players (e.g. service providers and some kinds of users) in addressing not only technical and operational aspects but also, above all questions about usage and user responses. Initiatives designed as demonstrators provide important opportunities for broader and more open-ended learning. The SLIM project takes observation this further to suggest that all multimedia projects are in some ways experimental - not just those explicitly established as feasibility studies or social experiments. However, one of the main conclusions to emerge from this study is that the failure to recognise the experimental character of multimedia initiatives means that the lessons from these projects are often not sought or properly utilised. 1.3 Coupling innovation and society: Social Learning in Multimedia These observations point to the need for a more sophisticated and intricate understanding of the relationship between technology and society. In particular, new concepts are required that recognise the importance of this local innovation effort, and the active role of 'users' with their own knowledges and concerns. We have sought to conceptualise these as processes of social learning. We briefly summarise the main features of this perspective below. In Chapter 2 we review the body of research on which these concepts are based. We go on to develop and explore the concept of social learning in detail in chapter 5. 28 From studies of the social shaping of industrial technologies, we deploy the concept of innofusion (Fleck 1988) to show that innovation does not stop when artefacts emerge from technology supply but continues (in the so-called 'diffusion stage') as they are implemented and used. Considerable innovative effort may be needed to get technologies to work and prove useful in particular user circumstances. In addition from studies of the adoption of technologies in the household we advance the concept of domestication to refer to the creative ways in which final users incorporate artefacts within their local practices, purposes and culture. Domestication here carries the sense of taming something which was hitherto perhaps wild or alien, rather than referring specifically to the household. It is through innofusion and domestication that novel technologies are adapted to local contexts; incorporated within particular technical configurations and practices of technology use (and, through these processes, acquire their meaning and social significance). These are two closely related facets of the way in which the more or less generic technical capabilities, arising from technology supply, are appropriated (du Gay et al. 1997) - coupled to current and emerging social needs. The appropriation perspective being developed in this study emphasises the active role of technology users and other local players, in interaction with supply-side players, in establishing new uses of technology and in creating new markets. These kinds of issues of user acceptance and the establishment of new markets have come to be seen as central to the prospects of multimedia, in a context in which the pace and dynamism of technological innovation have reached an unprecedented level. Indeed leading ICT companies like Microsoft have come to see user responses as a crucial, albeit difficult to predict, factor shaping their success, indeed survival, in the turbulent market for information and communications technologies and services (Myervohld, 1998). We have developed the concept of ‘social learning’ to explicate the detailed mechanisms and processes involved. The social learning perspective first highlights the importance of practical local activity and knowledge – of learning by doing (Arrow 1962) - both in developing multimedia systems and in developing usages of multimedia. It second draws attention to the processes of learning by interacting (Sørensen 1996), whereby such locally-acquired knowledge may be transferred and applied in other contexts. However, to be intelligible and useful in other contexts such knowledge cannot simply be transported – it must also be translated, combined with other knowledges, and transformed (Latour 1986, Callon 1987). One of the important kinds of learning by interacting involves supplier-user linkages in developing new products. Their contribution to successful innovation has been highlighted by evolutionary economic theorists (for example Andersen & Lundvall 1988) through the concept of ‘the learning economy’. Social learning seeks to explore empirically and in detail the operation of the learning economy, to analyse how it is shaped by its particular socio-economic context and the strategies of the actors involved. We are thus interested to explore social learning as a process of negotiation, subject to conflicts of interest amongst players with rather different capabilities, commitments, cultures and contexts. 29 One important kind of learning by interacting, particularly from a policy perspective, involves learning by regulating. This relates to the attempts by those involved with technologies to create the broader circumstances needed for their technology to become established and operate. In addition to formal state regulatory activities, this includes the attempts by private and public players to create technological infrastructures and standards that may be needed for a technology to run, and more generally to enlist other players around particular visions, technology policies and strategies. This is one of the ways in which the social learning perspective seeks to enrich analysis of the social shaping of technology, and to explicate its policy implications. We will return to these questions in detail in Chapter 6. 30 Chapter 2 The Intellectual Context and Concerns of the project The Technology-Society Relationship This chapter explores our changing understanding of the relationship between technology and society. The first section addresses the evolution of technology policies. It starts by examining the dominant Post-War policy tradition - based upon a 'linear' model of the relationship between technical and social change - and the growing difficulties it encountered. A period of reassessment of policy frameworks has culminated in the emergence of more sophisticated, integrated policy approaches notably in the European Commission's Fifth Framework Programme. This rethinking of the question of technology policy has coincided and been mutually reinforced by important advances in understanding of the relationship between technology and society. Section 2 reviews these developments in the field of technology studies. Here an important area of interdisciplinary inquiry has thrown up a distinctive approach, which we describe under the banner of 'the social shaping of technology' (SST). The third section reviews recent research findings which bear upon the social shaping of Information and Communication Technologies. Finally we explore the understanding that emerges from this of the process of innovation in Multimedia – and of social learning in multimedia. 2.1 The Evolution of Technology Policy a) The failure of the orthodox 'linear model' of technology policy The success of science intensive industries in the decades after the second world war, and the enormous advances made possible by 'Big Science' during the war provided the rationale for the adoption across most developed states of a technology policy framework that looked to the promotion of advances in core sciences and technology to drive technical progress and economic growth. Under this perspective, technical progress was treated as largely synonymous with social progress. This view saw technological change as a driver of social change and embodied a further 'technological determinist' presumption that new technologies required particular sorts of social arrangements. The social policy issue was restricted to a very narrow one of how quickly and smoothly a society could adjust to meet the exigencies of new technology (and it was argued that the biggest risks would arise from a failure to ensure uptake of technology and rapid adaptation). The role of social science was limited in this view to monitoring and 31 predicting future technological trajectories, and in this way assessing the future 'socio-economic impacts' of new technologies.3 Over the last two decades, the marked shortcomings of this traditional perspective have been progressively revealed. Both of its key elements - the emphasis on technology push and its inadequate model of the socio-economic and other implications of technological change - have been called into question. Failure of Science/ Technology Push policies - uncertainties in technical innovation Traditional technology policies in the European Union and most OECD countries pursued economic growth and competitiveness through support for Research and Technical Development (RTD). They presumed a ‘linear model’ of innovation, involving a one-way flow of information, ideas and solutions from basic science, through applied RTD , to industrial production and the diffusion of finished artefacts through the market to consumers. The basic driver of technological innovation was seen as advances in underlying scientific and technological knowledge, arising perhaps through public funded research and development, and the resulting creation of new technological artefacts. However this 'science and technology push' model proved not to be very effective in delivering successful technological advances, let alone the economic and social benefits that were expected to accrue. There was an increasing number of examples in which technologies developed in the laboratory were not taken up on a commercial basis. In contrast to the confident promises of the engineers, it turned out that new technologies often failed and, even when implemented, did not yield the predicted improvements in wealth creation and the quality of life. Technological innovation proved to be a rather uncertain endeavour. As we see below, as well as the narrowly 'technical difficulties' encountered, it often proved rather difficult to fit new technological offerings to social need and build new markets. As a result, public technology policies based on such linear models are increasingly seen as unhelpful – not least because of their privileging of technological supply and their artificial division of innovation into separate phases. 4 Lack of attention to unanticipated and undesirable outcomes of technology 3 We can detect the intellectual heritage of this kind of perspective in the requirements in parts of the European 3rd and 4th Framework Programmes to conduct assessments of the economic and social impacts of technologies developed under their auspices. 4 Though the linear model is firmly out of fashion today we should note that linear presumptions about the process of innovation are remarkably resilient – perhaps in part because they appear offer the promise that would be enormously enticing to public and industrial decision-makers of a world in which the development, uptake and outcomes of new technologies can be planned and controlled together with rather straightforward mechanisms for intervening – through support for technological development (Tait and Williams 1999). 32 The traditional model, with its emphatically positive attitude towards new technologies, tended to downgrade consideration of 'socio-economic outcomes' of technological change. To the extent that negative outcomes were acknowledged, they were largely presumed to be modest, and were seen as a necessary, but acceptable corollary of the presumed benefits of technology. Since the 1970s, in particular, there has been increasing recognition that technologies could often bring unanticipated, and undesired consequences. The starting point for this was growing awareness of the environmental and other risks of new technological activities.5 The linear model also separated consideration of technology promotion from consideration of its costs and how they might be addressed. This was a consequence of its presumption that the costs and benefits of a technology were somehow inherent in the technology. In contrast, a growing body of critical discourse suggested that the particular socio-economic outcomes of a technology, its costs and benefits, were not an inevitable, technical matter but arose from the ways in which a technology was designed and implemented. For example environmental concerns - culminating with the recognition of global climate change - raised questions about how existing technologies might have been designed without taking into account social goals such as pollution avoidance, and suggested the possible of creating more environmentally sustainable technologies and industrial systems. This broadened the technology policy agenda, beyond a narrow concern solely with technology promotion to consider also choices in the design and application of technologies. It raised questions regarding how to assess in advance the prospects and implication of new technologies and select appropriate technological options accordingly. b) The search for an interactive technology policy framework Coupling technology supply to markets ? The need for more integrated approaches to technology policy (both public policy and commercial strategy) that link technology to markets was first signalled almost 30 years ago by Chris Freeman. Freeman (1974) stressed the importance of 'coupling' between technology supply and its user markets - linkages which are today analysed by evolutionary economists as the basis for the learning economy.6 For example an important turning point in attitudes to the new petrochemical technologies was associated with Rachel Carson's book, Silent Spring, which documented the damage to wild-life associated with the Post-War use of synthetic pesticides and other biologically active materials. Further challenges arose in the 1970s as a succession of health and environmental hazards testified to the possibole dangers associated with new technological activities (asbestos, PVC, radiation) dangers which science and technology had failed to predict (Faulkner, Fleck and Williams 1998). 6 Although focussing in this account upon the shortcomings of frameworks which emphasised 'science and technology push', we should also note their mirror image in those policy approaches which gave primary emphasis to the role of markets. Such approaches became prevalent in the 1980s, notably in the UK. Reacting against the failure of large-scale state sponsored technologies (for example, supersonic air-travel) it was suggested that the state was not very effective in 'picking winners'. Instead it proposed that such choices should be left to the market. However the evidence from this episode was that markets could fail in certain circumstances and that resort to short-term 5 33 However, linking new technologies to the 'market' of potential users has posed considerable challenges. This has been particularly notable in areas such as Information and Communications Technology (ICT) in which an accelerating rate of change in core technologies has been associated with enormous turbulence in product markets. The falling costs of computing and communications has opened up the possibility of new kinds of products and services. But what will these look like? There has been considerable uncertainty about which products will prove sufficiently attractive to consumers to be commercially viable - particularly given expectations that radical technological change would both allow very different ways of meeting user needs and create new markets. In a context where there is no existing customer base, knowledge of earlier products and markets may no longer be a reliable guide. Meeting 'user need' became a rallying call for industry and for government alike - for example under the later stages of ESPRIT (European Strategic Programme of Research in Information Technology) Programme within the European Commission's Fourth Framework Programme (EC4FP). These uncertainties were given greater salience by the potentially enormous costs of research and development of new knowledge intensive technological systems. In industries such as ICT, these uncertainties regarding future markets and competitive challenges combined. Technological changes have tended to cut across existing product markets and industrial sectors. The brief history of this sector has been characterised by the displacement of many established players by new market entrants. Correctly anticipating – and shaping – evolving markets and user needs are key to the commercial survival of firms in these sectors (Meyersdahl 1999). As we see later, these uncertainties regarding 'user acceptance' of new technologies have become more salient in recent years because technologies such as ICT are being taken up by an ever growing share of the population. Greater scale increases the potential development costs and the commercial benefits of a successful systems. It also increases the difficulties faced by suppliers in knowing their user market. Furthermore these systems are becoming part of everyday life - at work, in the community and in the home - in a much more direct way than hitherto. For example over the four decades of the application of computers in administration most workers had only indirect experience of computing. As technology users moved out from the domain of technical specialists, they became subject to a range of new and potentially differing demands, from managers and 'end-users' at work; from people using them for entertainment in the home. User demands have become more critical. It is no longer sufficient to generalise about 'the user'. Although some user needs are generic (for example regarding the physical and perceptual ergonomics of the computer interface), other needs may be more specific. ICT applications are increasingly bound up with social practice. ‘Industry-specific’ applications must support the competitive strategy and coordination processes of the organisation – they are ‘organisational technologies’, replete with information about the specific organisation and its sectoral etc context. ICT applications in market exigency still left unanswered problems of technology policy. The implication is that neither a 'science push' nor a 'market pull' model are adequate. Instead a realistic and useful policy framework needs to address the interaction between them. 34 everyday life are more subject to perhaps more diverse and challenging user expectations, not just functional requirements of productivity and cost-effectiveness, but also complex personal goals relating to individual and group identity and satisfaction. Questions regarding advances in core technical capacities are matched by questions about how these may be applied and used. These questions about how a technology is domesticated and socialised are not, of course, restricted to ICTs. A similar array of issues is coming onto the horizon, for example, with respect to the application of biotechnology into new medical treatments and in agriculture. From the Linear Model to 'Linear Plus' EC 4FP, and comparable national programmes, such as the 'Foresight Initiatives' launched in the 1990s by a range of developed economies, gave greater attention to questions of dissemination and application as well as the generation of new technological knowledge. Various measures were proposed to assist the transfer of technologies from public sector research laboratories and their commercial exploitation. These developments, though important, reflected only a partial rethinking of the traditional model. Rather than abandon the 'linear model of innovation' altogether they reflected an attempt to counteract the reasons for its ineffectiveness. For example, measures are proposed to assist the dissemination of new technological knowledge from the Research and Education sector to industry, ignoring the fact that much industrial innovation is rooted in knowledge generated within industry itself. Tait and Williams (1999) have described this as a 'linear plus' model to signify the way it seeks to get the innovation process to operate in an effectively linear manner. In certain circumstances, for example where a technological paradigm and an infrastructure of institutions and complementary technologies have already been established, innovation can proceed in an effectively linear manner, driven by the supply of new technology offerings (the development of new drugs is perhaps a case in point). And where the linear model is effective, it can be a rather efficient way of organising innovation. However, where product concepts are changing and diverge from existing societal and technical arrangements, effective innovation requires other kinds of adaptation and learning involving knowledge and other inputs from a broader range of players. The linear and ‘linear plus’ models both tend to downplay the importance of the array of nontechnical actors (eg players ‘downstream’ in the innovation process and a broader array of actors indirectly involved), their knowledge, experience and concerns. The challenge is thus to develop innovation models that deal adequately with the full range of relevant players. This will call for more attention to be paid to demandside players and the process of appropriation of new technological capabilities. However this is not to suggest a shift to the obverse of the ‘technology-push’ model: a ‘supply-pull’ model. Instead it is necessary to tackle the complex interactions between demand-side and supply-side actors, and with others who may not be directly involved but who may influence technology choice as regulators, policymakers and informally by shaping opinions. In the search for such an 35 integrated model, linear plus is a step in the right direction but one that may not yet go far enough. The search for an integrated technology policy framework Technology policies are being reassessed across the developed world. In particular, the European Community Fifth Framework Programme (EC5FP) represents a challenging rethink of previous technology policy frameworks. It signals a break from technology-push models of the past and instead proposes an integrated approach, encompassing technology promotion and exploitation and geared towards broad social and policy goals of competitive and sustainable growth etc. etc. The emphasis is upon interdisciplinary research, bringing together social sciences as well as natural science and engineering disciplines, in a targeted effort, guiding individual RTD and uptake projects and linking them into clusters around certain strategic technological targets.7 The integrated model proposed by EC5FP builds upon and complements developments in thinking about the process of technological innovation - the critique of the linear model and the espousal of an interactive model of innovation, as well as its recognition of the need for interdisciplinary approaches which bring together insights from social and policy sciences as well as economics and the disciplines of science and engineering. However, the challenge posed by EC5FP has, arguably, run ahead of theoretical understanding and our ability to make generalisations about policy approaches and solutions. Today we see developments on a range of fronts in terms of both technology policy frameworks and practices. A range of accounts seek to move away from a linear understanding of the innovation process (see for example Caracostas, Berkhout and Dits 1999). However this endeavour throws up a range of difficulties. If the relationship between technology and society is not unidirectional and deterministic, but involves a variety of more or less loose linkages at different levels, the setting for technology policy becomes extremely complex. Thus an interactive model, which starts by highlighting linkages between technology supply and user markets rapidly encounters questions regarding the different structures of markets, the diversity of intermediate and final ‘users’, the negotiated and unstable character of demand as needs evolve in the face of social and technical change. A growing body of empirical evidence points to this complexity and highlights the contingent nature of outcomes of technological innovation processes. For example, various writers have stressed the local and differentiated cultures of technology supply (Vincenti 1994, Hård 1994). Similar issues emerge about the local character of user communities and markets and the relationship between local markets and contexts of use and how they may be transformed into more global markets. Given the This is not to say that EC5FP has overcome all the failings of previous approaches. The legacy of earlier linear policy frameworks is represented, for example, in the very structure of the programme in the distinction it makes between research and take-up activities. As we note in the conclusions this separation overlooks the important innovative opportunities which may arise as new technologies are implemented and used. 7 36 importance of local contexts and contingencies, analysts have moved away from seeking to make strong generalisations about the role of particular institutional and policy settings in supporting successful innovation; there has been a shift away from a search for 'cook book' recipes for success towards a processual approach (Clark and Staunton 1989, Edge and Williams 1996). processual approaches to technological innovation - social shaping and social learning The ‘social shaping of technology’ perspective, and other processual approaches, would therefore be critical of over-generalised approaches to technology policy. SST accounts would tend to emphasise the need for detailed empirical investigation and would offer a critical interrogation of the validity of presumptions and generalisations from other technological contexts. However such approaches may seem to offer cold comfort for hard-pushed decision-makers in government and industry. The challenge that decision-makers would pose for SST surround its ability to go beyond exercising a solely cautionary and critical voice and offer instead some stronger generalisations. We see three sets of developments that offer some ways forwards in addressing this challenge. First is the growing body of empirical research that begin to offer a more adequate account of the process of technological innovation and how this may vary across a range of settings - in terms of differing technical domains; differing socio-economic contexts of use; differing national, cultural, economic and policy contexts. One of the aims of the SLIM project was to present a detailed enquiry into the character of the innovation process in relation to multimedia products and services that can provide an adequate account of the complexity surrounding both the technical elaboration and the social appropriation and use of new artefacts. The diversity thus revealed belies any simple correlation between context, process and outcomes of innovation - as we show in Chapter 4 in relation to our survey of the national settings for multimedia development and application. This complexity at first sight seems to belie attempts at generalisation. However these studies have also given rise to a rich conceptual framework which helps to map out points of regularity and similarity. SST has been extremely productive in its theoretical and conceptual contribution. Two broad strategies have been adopted to make these insights useful to non-academic audiences. On the one hand, various writers have sought to make these ideas available to broader audiences - for example through a ‘tool-box’ approach, whereby a range of frequently encountered processes and issues in the social shaping process are identified. These are offered not as ‘one size fits all’ generalisations, but as simplifications about innovation processes together with indications of the contexts in which they may be relevant – as tools for thinking and acting which can help guide decision-makers. See for example Molina’s ‘sociotechnical constituencies’ model, which identifies issues surrounding the internal alignment of groups of diverse players involved in technology development. 37 On the other hand has been an approach targetted towards the reflexive activities of the players involved in the innovation process itself. This work focuses upon the adaptive behaviours and learning opportunities available to the players involved. From a processual perspective, in a context that is complex and variable, an important question about the conduct and outcome of innovation process concerns the ability of individuals, groups or organisations (or, for that matter, an economy) to react creatively to particular exigencies and a changing context. These kind of considerations underpin the move in this research project from the social shaping of technology perspective to a ‘social learning’ perspective. This move also parallels a growing concern more generally with processes such as learning (Gibbons et al. 1994) and capability building (TASM 1999 articles).and with the growing salience of ideas of the ‘knowledge economy’ and the ‘learning organisation’ 2.2 The Social Shaping of Technology a) Introduction The term 'the social shaping of technology' (SST) draws attention to an important development in socio-economic research on technology: the move by scholars from a wide range of backgrounds to address the content of technology and the particular processes involved in innovation (MacKenzie & Wajcman 1985, Bijker and Law 1992, Williams and Edge 1996). Their common concern to analyse the social shaping or 'social construction' of technology, stood in contrast to traditional 'deterministic' approaches which took for granted the character and direction of technological advance and thus limited the scope of socio-economic inquiry to addressing its outcomes or 'impacts'. Instead, SST research highlights the range of social as well as technical factors - organisational, political, economic and cultural - which pattern the design and implementation of technology (and which thus also shape its social implications). This broad approach has been taken up, with differences of emphasis and intellectual tradition, across a number of European countries (Cronberg and Sørensen 1995, Williams and Edge 1996). SST has gained increasing recognition in recent years as a valuable research focus, and one which has a potentially important policy contribution. It broadens the technology policy agenda by drawing attention to the many choices inherent in every stage of technological change (and their possible implications) and offers a better understanding of the relationship between scientific excellence, technological innovation and economic and social well-being (European Science Foundation 1991). In this way it holds out the prospect of more effective promotion of technology as well as its assessment and control. SST approaches attempt to grasp the complexity of the socio-economic processes involved in technological innovation. In contrast to the presumption that technology supply would generate solutions that corresponded to user requirements, that could then be simply diffused to fulfill society's needs, SST has shown that identifying, let alone meeting, current and emerging demand for 38 technologies can be difficult. Indeed a key challenge for technological innovation concerns how to match new technical capacities to 'social need'. This is a potentially difficult problem because social needs, and the means by which they may be fulfilled, are not fixed entities, but evolve, partly in the face of new technical capabilities. These problems are particularly acute in relation to radical innovation (as opposed to the incremental enhancement of existing devices with wellestablished uses), since the potential uses and usefulness of an emerging technology are, at the outset, often not well understood by suppliers let alone by potential users. SST has contributed to the rather belated recognition of the importance of 'the user' in technological change over the last decade, as well as to the development of an 'interactive' model of innovation. In contrast to traditional 'linear' models which saw innovation as restricted to technology supply, the interactive model emphasises the innovative effort as technologies are implemented, consumed and used and highlights the interaction between technology supply and use. Consideration of (intermediate and final) users draws our attention to the range of players involved in innovation, with their different relationships to the technology and varying commitments in terms of past experience and expertise. These players including technical specialists and other groups from supplier organisations, policymakers, existing and potential users - may have widely differing understandings of technology and its utility. The character and utility of technologies are not defined by any one player but emerge through a complex process of negotiation between diverse players. The complexity of these interactions is one of the reasons why the development and application of technology involves deep uncertainties. Technological innovations often fail altogether; they usually develop far slower than suppliers and promoters predict and may follow rather different trajectories than was initially anticipated. Similar uncertainties surround the outcomes of innovation - which mean that it is extremely difficult to predict the 'socio-economic impacts' of technologies with any useful accuracy. One reason for this is that these outcomes are not inherent in the artefacts, but relate to the ways these are applied within particular social settings. So even where the design of artefacts embodies particular values or the objectives of particular actors, there is still scope for others involved to achieve some flexibility in their implementation and use. The implications of technology for society are therefore rather complex and frequently involve unintended and unpredicted outcomes. Given this situation, it is hardly surprising that attempts to promote or to control technology have not been very effective to date. In seeking to capture this complexity, SST has developed various sets of concepts which address first the dynamism and turbulence of technological innovation, highlighting the negotiability of technology (Cronberg 1992); the scope for particular groups and forces to shape technologies to their ends, the flexibility of technology (i.e. its open-ness to re-negotiation by local actors) and the possibility of different kinds of ('technological' and 'social') outcome. Second it seeks to explore the extent and manner in which technological choices may be foreclosed. SST points to processes of closure - the ways in which innovation may become stabilised (Pinch and Bijker 1984) and of entrenchment, through which particular technological 39 options become generalised. Entrenchment underpins the cumulative development of technological knowledge and increasing returns to established technological options as a result of earlier 'sunk' investment. Entrenchment is critical to the success of a technology. However it may, equally, result in 'lock-in' to particular options that may become outdated or have undesirable consequences. Such, entrenchment is countered by the dynamism, of both new technological capacities and of 'society', which may undermine and destabilise existing expectations and development paths. SST has begun to explore this complex and contradictory interplay surrounding technological innovation between entrenchment and dynamism. b) SST and the critique of technological determinism Mainstream 'technocratic' approaches to technology, inherited from PostEnlightenment traditions a concept of technological progress that did not seek to problematise the process of technological innovation. In its deterministic versions, technology was seen as having its own self-evident dynamic - reflecting a single (for example technical) rationality, or an economic imperative. Alternatively, instrumental views saw technology as readily available and capable of being rationally applied to meet social and economic needs - for example, the dominant neo-classical tradition of economic analysis, with its assumptions that technologies will emerge fluidly in response to market demands (Coombs et al 1987). Either way, technological progress could be taken for granted. These views of technological progress were not well equipped to deal with the experiences of technology, which have been seen as increasingly problematic since the 1970s, on at least two fronts. First is the experience of unintended and undesired consequences of technology (for example health and environmental hazards). Second, the growing pace and salience of technological change has drawn attention to the difficulty of achieving successful technological innovation - in terms of generating new technological knowledge and applying it; fitting it to existing and emerging demand. The latter concern was given impetus by the growing realisation that the traditional approach of supporting technological supply was not, by itself, sufficient to achieve technological advance, let alone its application to achieve improvements in economic performance and social well-being. These experiences prompted research into the relationship between technology and society - including SST research. In contrast to the certainties held out by images of social and technological progress, technological change was revealed as a highly uncertain process. A critique of this simplistic and unidimensional interpretation led to the development of a complex, 'interactive' model of technological change that highlighted the influence of a range of players, involved in the use of technology, and its broader setting as well as just technological supply. SST emerged through a critique of the 'technological determinism' inherent in such traditional views of technological progress. For example they criticised ideologies of 'technological imperative', prevalent the early 1980s, which suggested that 40 particular paths of technological change were inevitable (Edge 1988). This involved two presumptions: i) ii) that the nature of technologies and the direction of change were unproblematic or pre-determined (perhaps subject to an inner 'technical logic' or 'economic imperative'). that technology had necessary and determinate 'impacts' upon society: technological change thus produces social and organisational change . Instead, SST studies showed that technology does not develop according to an inner technical logic but is instead a social product, patterned by the conditions of its creation and use. A variety of technical options are available at every stage in both the generation and implementation of new technologies; which is selected cannot be reduced to simple 'technical' considerations, but is patterned by a range of broader 'social' factors. Central to SST is the concept that there are 'choices' (though not necessarily conscious choices) inherent in both the design of individual artefacts and systems. If technology does not emerge from the unfolding of a predetermined logic or a single determinant, then innovation is a 'garden of forking paths'. Different technological routes are available, potentially leading to different outcomes in terms of the form of technology: the content of technological artefacts and practices. Significantly, these choices could have differing implications for society and for particular social groups. The general thrust of SST is to problematise and open up for inquiry the character of technologies, as well as their social implications. We can analyse social and economic influences over the particular technological routes taken (and their consequences). SST thus goes beyond traditional approaches, that were merely concerned with assessing the 'social impacts' of technology, to examine what shapes the technology which is having these 'impacts', and the way in which these impacts are achieved (MacKenzie and Wajcman 1985). SST research could, it was hoped, identify opportunities to influence technological change and its social consequences, at an early stage - moments at which accountability and control could be exercised. SST broadens the policy agenda: no more mere tinkering at the margins of technology policy, seeking to grapple with its products but leaving unaltered the direction and goals of innovation. SST allowed people to get inside science and technology themselves (Latour 1986, 1988). It offered the prospect of moving beyond defensive and reactive responses to technology, towards a more pro-active role. c) the interactive model of innovation Innofusion and the critique of the linear model of innovation 41 Fleck (1988) highlighted the shortcomings of traditional 'linear models' which saw innovation as being restricted to the technology supply side - as if finished artefacts would emerge from the research and development laboratory as 'black-boxed' technical solutions, already corresponding to user needs, that could simply be diffused through the market to potential users. Instead he advanced the term `innofusion' to describe the "processes of technological design, trial and exploration, in which user needs and requirements are discovered and incorporated in the course of the struggle to get the technology to work in useful ways, at the point of application (1988:3). The concept of innofusion specifically counters prevalent models of diffusion, with its implicit view - inherited from the idea of transport of atoms in physics – of an essentially passive process, in which what is diffused is unchanged and all that was at question would be the rate and equilibrium point of diffusion (Rogers 1983). In contrast Rosenberg (1979) showed that, in technological diffusion: "The diffusion process is typically dependent on a stream of improvements in performance characteristics of an innovation, its progressive modification and adaptation to suit the specialized requirements of various sub-markets, and the availability and introduction of other complementary inputs which decisively affect the economic usefulness of an original innovation" (Rosenberg 1979:75). In contrast, the terminology of innofusion insists that innovation continues beyond the laboratory as technologies are diffused, implemented and used. Fleck's account focuses on the implementation of industrial technologies. Other work has focused upon the translation (Latour 1986, Callon 1987) and transformation of artefacts as they move from the laboratory to commercial production to widespread adoption. As we see in Chapter 6, homologous concepts have been advanced regarding the active and innovative characteristics of the processes of consumption and use of technologies in other settings - and in particular their domestication within the household (Silverstone and Hirsch 1992). The importance of innovation around the application and use of technologies is one of the reasons why the potential uses and utility of a technology may not be wellunderstood at the outset of a programme of innovation. The history of innovation is full of such examples (Williams 1997). This observation underlines the serendipity of the innovation process, in terms of our limited ability to anticipate the development of technological capacities and, more importantly, to pre-conceive future applications/ uses. Perhaps the biggest difficulties here surround the responses of the user and the evolution of 'social needs' . The future uses and utility of a new application may not at first be self-evident. Partly this is because of the difficulty of anticipating the outcome of the protracted learning processes involving suppliers and users alike as technologies are applied and used. Partly it arises because people may have widely differing attitudes to a particular technology. interpretive flexibility Thought we tend, with hindsight, to take for granted functional benefits of a new technology has often been presumed to be self-evident, historical and contemporary 42 studies of the development and uptake of new technologies have highlighted the diversity of perceptions of different groups of technologists and engineers (Vincenti, Hård 1994) and 'users' (Pinch and Bijker 1984) regarding technical benefits of different design decisions and the social significance and use of the technology. Perceptions are shaped by the particular circumstances of actors (their histories, experiences of technologies, interests and commitments knowledges, local cultures). The concept of 'interpretive flexibility' has been proposed (borrowed from the sociology of scientific knowledge) to capture the multiplicity of meanings that may be associated with a technology and the scope for the various groups involved to articulate their own definitions (Pinch and Bijker, 1984). This flexibility may be countered by processes of 'closure', in which particular design solutions and interpretations gain ascendancy amongst the array of 'Relevant Social Groups' involved (idem). This framework calls for a revision of traditional conceptions which saw the meanings and significance of technologies being built in to technological design and which consequently portrayed consumers and users as essentially passive. In contrast SST shows that, though designers may seek to inscribe particular intentions into material artefacts, this does not foreclose questions about the meanings and uses of a technology. The final consumer may have little opportunity to influence the design and development of artefacts such as domestic goods, other than the 'veto power' to adopt or not (Cockburn 1993). However, even in this setting it is important to acknowledge the scope for these actors to articulate their own representations of technologies and uses which may differ from those articulated by technology suppliers (Sørensen and Berg 1992, Akrich 1992, 1992a, Cockburn 1993). Consumers have shown remarkable inventiveness in sidestepping design presumptions about the use of artefacts, and adapting technologies to their own purposes. In this sense closure is never final and consumers and users are actors, able to exercise choice and inventiveness in the deployment and use of supplier offerings (Sørensen 1994). This is not to imply total flexibility regarding the use and outcomes of artefacts - some interpretations and some outcomes are more probable than others. However the significance of an artefact is not fixed in the laboratory, but is only finally resolved in its implementation and use, incorporated within the practices and culture The negotiability of technology SST in emphasising the scope for local actors to redefine and impose their own understandings thus suggests that it is unhelpful to talk of the socio-economic 'impacts’ of technology, as if these were somehow inherent in the artefact. Outcomes, arise through the interaction between artefact and its social setting of use. Rather than imposing particular uses and outcomes, artefacts offer a range of constraints and affordances in their use. And the fluidity and flexibility in use, 43 which may be 'designed into' many artefacts (especially ICTs) suggests that some artefacts may be associated with an extremely wide range of outcomes.8 SST in this way emphasises the negotiability of technologies - in the sense that artefacts typically emerge through a complex process of action and interaction between a wide range of players involved and affected - including the various groups of (intermediate as well as final) users, suppliers of complementary as well as competing products, promoters, consultants etc.. One corollary of this is that the development and application of new technologies typically requires the bringing together of different kinds of knowledge and other resources - which are often unevenly distributed across different organisations and institutions. Studies of the development of new technologies and technological systems have emphasised how they emerge through the creation of formal or informal alliances of players with the resources and technical expertise needed, united by certain broadly shared concepts or visions of as yet unrealised technologies. SST has begun to uncover some of the strategies for creating/ mobilising such 'sociotechnical systems' (Hughes 1983), 'constituencies' (Molina 1989) 'ensembles' (Bijker 1993, 1994) or 'actor-networks' (Law and Callon 1992), for example, 'enrolling' local players in a broader network (Law and Callon 1992) and 'aligning' their expectations around realisable objectives (Molina 1994). Though early SST research was primarily concerned with the initial development of new technologies and artefacts, and the socio-technical systems/ constituencies of players involved in the generation of new technologies, the interactive model highlights the need to look at whole 'circuit of technology' (Cockburn and FürstDilic 1994:3), from design and production through to consumption and use, to understand how technologies and the social implications are shaped. The entrenchment of technology Stabilisation and entrenchment are crucial to the success of a technology. First, the commercial viability of a new technology may rely on achieving a critical mass - or rather on convincing a critical mass of players to invest in the technology (Williams 1995, Graham et al 1996, Schneider 1991, 1998). Where technologies exhibit economies of scale and network externalities, uncertainties arise about whether enough people would be convinced to adopt a technology to recoup development 8This is not to suggest that outcomes are entirely a local achievement. Indeed some have argued the need to take on board forms of 'soft technological determinism - for example that the availability of certain technical facilities by changing the time, money and other forms of investment needed to achieve certain results changes the parameters on which paths of action are chosen. However, the patterning achieved by technologies is not a simple result of the functional characteristics of an artefact; its symbolic dimensions are also important. Societal outcomes of technological change are thus a complex product of the interaction between a range of players involved in the supply and use of a technology. They cannot be understood simply by snapshot assessments of the immediate settings, but must be examined as in their longer-term and broader context. These considerations suggest that conventional ideas about how to conduct socio-economic 'impact assessments' for new technologies are more or less meaningless. 44 costs; to bring down the price of a product and to make it useful to consumers (eg by establishing a market of content providers and/ or service users). This is particularly important where technologies demonstrate what economists describe as 'positive returns to sunk investment' (the gradual improvements in performance of technologies as technological knowledge is refined and developed) and 'positive network externalities'. The latter refers to technologies such as the telephone or Electronic Data Interchange (Graham et al 1996) the benefit of which to existing users is increased by the arrival of additional users. In these settings, markets may fail, at least in the short-term, to guarantee the uptake of more productive technologies. These considerations point to the benefits of 'closure' and reaching agreements about the technological paths to be followed. On the other hand such rigidity can be counter-productive. The results of earlier technological choices can constrain later technological decision-making (Rosenberg 1994). Path dependencies arise from the increasing returns on ‘sunk investment’ into particular technical standards, given the cumulative nature of technological advances. Such 'path dependencies' can result in 'lock-in' to established solutions and standards, even where these technologies are no longer seen as optimal (David 1975, Arthur 1989, Cowan 1992). Well-known examples of this are the QWERTY keyboard and railway gauges. Entrenching factors are not just economic, but include, importantly, shared perceptions and expectations of a technology. Alignment of perception is an important step in innovation. For example, engineers must project visions and align expectations of a technology and its capabilities in order to enroll the support of other players if they are to obtain the technical and human resources needed to create it (Molina 1994). This may involve establishing consensus around particular technological concepts and options. The extension of such consensus to include the range of suppliers, consumers of an artefact and other relevant social groups presages technological 'closure' and the stabilisation of technological artefacts (Pinch and Bijker 1984). However, alignments that are premature or embody particular presumptions or visions of a technology can focus attention too narrowly on particular technological paths in a way that can prove disastrous if circumstances and perceived user requirements change (an example is the recently abandoned $150 million investment in the EFTPoS UK pilot for Electronic Funds Transfer at Point of Sale [Howells and Hine 1994]).9 This is just one example of how inflexible development contexts may result in inflexible technological designs (Collingridge 1992). Entrenchment therefore represents something of a two-edged sword. 9 Many of the dilemmas around technology closure – between long-term and short-term, between local and universal – arise in relation to standard setting for ICTs. It is clearly desirable to avoid technological closure around incompatible systems, that would entrench incompatibilities and might result in 'stove piping' in telecommunications networks (Spacek 1997). Many have therefore argued the need instead for the development of public, de jure, open standards. However the search for universal, open standards is not without dangers - as is shown in the case of Electronic Data Interchange standards (Williams 1995). The development of public standards can be slow and result in unworkable, over-engineered solutions – as exemplified by the case of OSI. 45 However, closure is never final (Rosen 1993). The possibility remains of reversing earlier choices (Latour 1988). However the perceived costs of modifying an earlier decision (in their broadest terms including time, money, uncertainties and the effort of abandoning existing routines and knowledge investments and learning new approaches) may appear prohibitive. Irreversibility, is always a complex sociotechnical accomplishment, rather than a technical fact (Collingridge 1992, Callon 1991, 1993). Established facts and artefacts remain perennially vulnerable, in particular, to dynamism in the technological system and broader society. So we find that technological innovation is subject to two contradictory tendencies. On the one hand, we identify processes forces that will tend to stabilise technologies, by aligning expectations, and reducing the uncertainties and costs of established models. On the other hand, the dynamics of the development both of new technological opportunities and of user requirements - new problems thrown up by societal changes, and the articulation of new ways of linking those problems with technical possibilities - which may open up new application possibilities, and undermine existing solutions, reversing the trend to stabilisation (Brady et al. 1992). An important influence here concerns the dynamics of particular product markets. Economic pressures may favour established approaches, in a context of positive returns on past investment and economies of scale. Standardisation of technology has played an important role - yielding substantial economic benefits for various players: bigger markets and greater profits for suppliers and important price advantages for consumers who can share development costs (which may be very high with high-technology, knowledge intensive products). In this way, some technical artefacts may become stabilised and standardised. They may be made available to the user through the market as 'black-boxed' solutions, as 'commodities' with well-established attributes. However, competitive considerations may also counteract this. Once new markets have been established this will attract new entrants to compete with established suppliers, who in turn may seek to differentiate their offerings and retain their existing customer base through technological leadership. And users may seek to gain 'competitive advantage' by adopting these new offerings. So here again we find a complex interplay between standardisation and commodification - consolidating and undermining technological entrenchment over time. d) The policy concerns of SST SST was not just seeking a better understanding of technological innovation, it also has policy claims and objectives. SST has been strongly influenced by a concern with the political choices surrounding technology. It criticised ideas that science and technology stood outside or above society, instead viewing them as areas of social activity, subject to social forces and amenable to social analysis (Bijker 1993). In this way it opened up technological decision-making for critical scrutiny. SST, by rendering the social processes of innovation problematic, has opened up policy issues that had been obscured by technological determinism, and by related simplistic models. For example, as already noted, SST criticised established 'linear models' of innovation, and drew attention to the close and reciprocal interactions 46 between these stages, and the transformation of technologies between their initial conception and their eventual application. In particular SST contributed towards the development of public policies which emphasised the role of the user as well as the supplier, and the need for linkages between them (Fleck 1988). It has been claimed that SST could help to broaden technology policy agendas and make them more 'pro-active' (Williams and Edge 1996). In particular, the policy utility of conventional approaches, such as technology assessment (TA), is limited by their concern with the retrospective assessment of the costs and benefits of technologies already designed and developed (Kranakis 1988). As Collingridge (1980) pointed out, such ‘reactive’ attempts to control technology fall foul of a central dilemma: When a new technology is first introduced, its societal implications may not be understood with any accuracy. There will typically be too little evidence of potential risks to justify social intervention and control. The costs and benefits of a technology, and their distribution between different social groups, may only become reliably known when the technology has become matured and widely adopted. However, by this stage the technology may be deeply entrenched; many important choices will have been taken, and other options foreclosed. Reversing these choices will be expensive and may conflict with by now wellestablished social and economic interests (e.g. firms and specialists committed to the technology). The social shaping view, and in particular the model of 'constructive technology assessment' (CTA) being articulated notably in the Netherlands (Schot 1992, Rip et al. 1995), holds out the prospect of strategic intervention from the early stages of innovation - highlighting both the possible consequences of different technological routes, and the problems and opportunities of controlling them to meet societal goals. The other main way in which SST suggests a broadening of the policy agenda is by extending the range of locales and players seen as significant in shaping innovation and its technical and societal outcomes. The focus of traditional technology policy on technology supply produces a 'problem of diffusion', in terms of transforming technological R&D into marketable innovations; getting these innovations implemented and making them profitable (Sørensen 1996). In particular, the increased attention within SST to the related appropriation processes of implementation (and innofusion), consumption and use of artefacts draws attention to different actors and suggests that measures geared towards the technology supply chain may usefully be supplemented by an effort to support technology appropriation. In particular the concept of social learning puts the user and use at centre stage and integrates technology development with societal uptake - flagging how members of a society learn about technology offerings and how designers learn about users and users. When one acknowledges the need for creativity in order to be able to gainfully employ new technologies, one discovers the need to support and stimulate, but also to regulate, this creativity. However, this also demands a different understanding of and greater concern for what users do. SST further explores the way in which both technological innovation and appropriation may be shaped by the broader societal context. Successful 47 technologies are those which succeed in establishing a supportive socio-technical setting. Perhaps the most obvious illustration here is the car (Sørensen 1994a). SST draws our attention to the wider technological regime – how regimes become entrenched and may favour some artefacts over others, and how new regimes can become established (see for example Rip 1995, Kemp, Schot & Hoogma 1998, Rip and Schot 1999). e) Social shaping and social learning This study seeks to explore these issues in more detail under the banner of ‘social learning’. The social learning framework we are developing here is not wholly distinct from an SST perspective – but tries instead to carry forwards some aspects of SST and address earlier weaknesses. One key aspect concerns the importance to the cultural symbolic dimension. Early SST studies had been criticised for their emphasis on the initial design phase at the expense of the consumption of technology and for their neglect of the symbolic dimensions particularly around consumption (Mackay & Gillespie 1992). In developing the SLIM project we deliberately sought to address this by including in the research consortium researchers and research centres from cultural studies and semiotic perspectives whose work focussed upon the symbolic dimensions and the 'appropriation' of technology.10 The need to address the symbolic as well as functional dimensions of artefacts is particularly relevant with information and communication technologies: to the extent that they function as media that carry various kinds of information and cultural content. This first remind us of the need to not just restrict our inquiry to hardware and the material nature of things but to focus more explicitly on the symbolic dimension. Second, and more specifically, multimedia artefacts can be seen to have a ‘double symbolic’ dimension (Preston 1999): not only does the physical artefact have meaning; there is also the information and cultural content it conveys. The SLIM project sought to further enhance studies of the social shaping of technology by promoting a dialogue with other strands of research: notably related evolutionary models of technological development; and, work from organisational studies (for example on ‘organisational learning, Schon) and elsewhere which have emphasised the adaptive and reflexive capabilities of actors. In relation to this last point, our use of the terminology of social learning rather than SST also signals a focus not only upon analysing processes of innovation but also upon the possibilities for intervention – and in particular on the scope for adjusting innovation processes and their outcomes as a result of the reflexive activities either of the players involved or of technology researchers. In short our interest in social learning represents an attempt to broaden the intellectual base of SST and broaden and clarify opportunities for intervention. We Another area of work which has been largely overlooked by the more rigorously established social sciences is studies of consumption from a marketing perspective (see for example Mick and Fournier 1998). 10 48 sought to explore these concerns in particular domains of ‘social learning’ around Multimedia. At the outset of the SLIM study, social learning issues were conceived rather narrowly through a focus upon: i) representation-design When a new supplier offering is developed the designer/ developer must construct it around some kind of representation of the use and user of the artefact (though this may be largely implicit). Here the designer is trying to appropriate the future user. ii) domestication/appropriation the active and creative process in which ‘users’ selectively incorporate supplier offerings and adapt them to their specific needs, routines and practices . Here supplier hypotheses about the use and utility of their offerings are tested and the meanings and significance of an artefact are finally established. In the course of the study a rather broader and more elaborate view of social learning has emerged encompassing not only these immediate processes of ‘learning by doing’ and ‘learning by interacting’ involved in getting a technology to work and to be useful in particular settings, and also the activities of learning by regulating involved in creating the broader circumstances for a technology to be successful (Sørensen 1996). We return to this discussion in Chapter 6. The research project we are reporting explores these issues in relation to Multimedia and other advanced applications of Information and Communications Technologies. A growing body of research exists into the social shaping of ICTs which can throw light upon these matters. We briefly review this in the next section. 49 2.3 The Social Shaping of Information and Communication Technologies The installation of Information Superhighways, coupled with advances in processing power and usability of information technology, have led to widespread expectations of the rapid adoption of a new cluster of technologies under the rubric of multimedia (Kubicek et al. 1997, Slack, Stewart and Williams 1997). Applications based upon these technologies are, moreover, expected to be widely diffused in many areas of working and social life, and to have profound social and economic implications. In short, these technologies are expected to underpin the transition to an information society. However, beyond this global vision, there is little certainty about the kinds of applications that can be expected to emerge. How can we assess the prospects and societal implications of these new technologies? Experience with these emerging technologies is still rather limited. Initial applications may, anyway, be far from typical of future offerings. Much contemporary discussion is based upon visions of future offerings - which are predominantly technology driven visions and rhetorics, and informed by the perspectives of supplier and a technically-focussed community of practitioners and pundits. The past can provide us with important insights into the future (Bruce 1988, Dutton 1995). The lessons from earlier technologies suggest that these early visions may be deeply misleading (Williams 2000). A growing body of research has addressed the social, economic factors shaping the development and use of ICTs and its outcomes (Williams 1995, Dutton 1996, Clausen and Williams 1997, Kubicek et al. 1997, Slack Stewart and Williams 1999). As there is only space in this chapter for a selective review of this body of findings (see Williams 1997 for a more extensive account), we focus on two features which recur throughout the history of ICTs and in different technological settings (though they may be resolved differently in different periods and contexts),which have shaped the character and use of ICTs: first the complex interaction between ICT supply and use, and, second, the contribution of ICT users in stimulating innovations which may give rise to new core technologies, and even transforming our understanding of the significance of an artefact. Our analysis of these features forms the basis for the final section which presents a model for understanding the character of contemporary multimedia artefacts; the process of innovation and future prospects of multimedia. Let us start by considering why the promises of social transformation and ‘revolutionary change’ that would arise from the supply of new ICTs have not materialised. Though the universalising rhetorics of ICT suggested that the supply of new technological offerings would provide finished solutions that could by readily and widely applied to the full range of social activities and problems. ICT supply has certainly been extraordinarily dynamic in terms of the increasing power and falling costs of new products. However, the domain of societal application and use of ICTs has proved more complex and obdurate than many technologists had presumed. There are a number of reasons for this (Williams 1997). In brief: 50 Computing has its roots in formal and mathematical models and representations. Successes in early administrative applications of ‘electronic data processing’ to routine and simplified information processing activities (e.g. accounting) which could readily be described in mathematical terms, converted to algorithms and implemented in software encouraged technical specialists to underestimate the complexity of application areas, and the consequent difficulties of applying ICTs more broadly. It proved rather difficult to apply ICTs to human communication and decision-making processes involving complex judgments and interpretation in contexts that are typically characterised by ambiguity and uncertainty and that are inherently more difficult to describe in formal mathematical terms. The claims to universality of ICT, based upon formal mathematical techniques, came into conflict with the diversity and specificity of the social contexts of application and use of ICT occupied by diverse kinds of ‘users’ with their own practices, purposes and culture. Users are various and may have quite different perceptions of ICT artefacts and their utility. Matching supplier offerings to user need can thus prove problematic, particularly with novel technologies where there are few established models of the application and use. While technology driven views typically takes the utility of the artefact for granted - assuming that new functionalities offered will somehow match user requirements, users do not have a finite and determinate set of existing 'requirements'; requirements are constructed built upon earlier templates and evolve as artefacts and practices change..11 The logics inherent in new technologies did not drive social and organisational change. Indeed, a body of research into the industrial implementation of ICTs suggested a mutual shaping between technology and organisation, in which, in the short-term at least, organisational structures and practices seemed to be more powerful reshapers of technology than the other way around (Webster 1990, Williams 1995, Clausen and Williams 1997). Rather than the transformation of ‘society’ by ‘technology’ we thus find a complex interplay between the dynamics of technology supply and use. Following on from this, the application domain has been the site of considerable experimentation and innovation as users and others struggled to get these generic offerings to work and prove useful in actual contexts of use (Fleck 1988, 1988a). Indeed ‘users’ have made an important contribution to innovation in ICT – even in relation to core technologies which today are seen as the virtually exclusive preserve of specialised IT suppliers (Brady et al. 1992). Thus industrial users of automatic data processing machines, such as Prudential Assurance and Lyons, were heavily involved in the construction of the earliest commercial computers (Campbell-Kelly 1989) in the commercial application of robotics (Fleck 1998). ‘User-led’ innovations throw up ideas which may be taken up in future technology supply and ultimately become sedimented in core technologies. Industrial users were involved in the The relationship between suppliers and users may be particularly difficult given the uneven distribution between them of technical and other pertinent knowledges (for example of the application domain). This is reflected, for example in the 'difficulties in communication' frequently experienced between technical specialist suppliers and non-expert users. 11 51 creation of the first operating systems (Friedman 1989). This continues today - the most famous recent example being the development by CERN scientists of the ideas that underpinned the World Wide Web. In some instances (e.g. the telephone, the home computer) user innovationhas led to the transformation of our understandings of a technology; radically rethinking its uses and significance. a) The interplay between technology supply and use Perhaps the most important factor shaping the evolution of ICT and its uses has been the interplay between the dynamism of ICT supply (the increasing power and falling prices, particularly of standardised mass produced technologies) and the specificity of the social settings in which they are introduced. These tensions surrounding the entrenchment of technology are played out through various related strategies including battles around the creation of interoperability standards and platforms (particularly in relation to component technologies); through the emergence of ICT tools and media which are relatively independent of the activities being conducted; and through the development of complex ICT applications not as integrated ‘systems technologies’ but as ‘configurations’ (Fleck 1988a) of more or less standardised component technologies and tools. Knowledge intensive products exhibit peculiar economics, particularly where products can be mass-produced at low cost. The economics of certain 'globalised' ICT products such as microprocessors and computer operating systems involves huge R&D costs of launching new products, coupled with massive potential economies of scale if they are successful in the market. This combination brings great uncertainties: huge losses for those that fail and potentially enormous returns for those that prevail. As a result the production of many core technologies such as microprocessors is increasingly globalised. The market for a component technology is effectively defined by the installed base of complementary products (for example software that can run on a particular computer's operating system). The entrenchment and erosion of standards Our earlier discussion of the problem of entrenchment. highlighted particular uncertainties and difficulties in creating and for that matter in changing large scale technological systems based on network technologies like ICT with high requirements for interoperability. Returns to such investment and network externalities provide enormous incentives for the development of industry and de facto interoperability standards. ICT supply has been profoundly shaped by the commercial strategies of the supply side-players seeking to create markets and maximise their share of established markets, often involving battles around the creation or erosion of standards. Entrenched players have adopted competitive strategies of 'architectural technology', notably in the microprocessor industry, where some elements of a product remain constant through several different generations (Morris and Ferguson 1993). This allows producers to innovate their 52 products without abandoning their existing markets and provides some guarantee of compatibility over several product generations for consumers and producers of complementary products. Conversely, attempts to launch radical new microprocessor designs, such as the Japanese TRON project, have often been unsuccessful, given the widespread diffusion of Intel and Motorola's CISC (complex instruction set computing) architectures. However, new RISC (reduced instruction set computing) architectures have emerged, involving an 'open' alliance of suppliers and users, that have begun to challenge the market dominance of CISC. This commercial strategy is reflected in the very design of their products for example by offering open access to codes to enable interoperability, and incorporating the ability for users of existing CISC products to run their current software (Molina 1992, 1994). The prospect that established standards and approaches may be by-passed or come to be seen as outmoded and limiting creates risks for users and suppliers (including suppliers of competing and complementary products) of 'inappropriate' investment in technologies that later become obsolete. 12 The existence of multiple competing systems imposes additional costs and uncertainties on suppliers and users, culminating in standards wars (as demonstrated, for example, by various well known cases, such as that between Betamax and VHS systems for video recording). These experiences in turn have mobilised a further dynamic to agree a single standard (de Laat 1999). Following on from this, suppliers of new delivery systems and platforms may seek to collaborate closely with their intermediate users suppliers of complementary products (for example Sony's tie up with content providers in developing their Interactive CD product [Collinson 1993, Collinson and Molina 1998]). This kind of collaboration - in developing standards and building markets - becomes crucial to the likelihood of success of a technology. These two interlocking processes in the creation and entrenchment of new technologies (aligning co-producers and enrolling future users in both developing artefacts and building markets) set into train 'risk management' strategies that paradoxically compound the uncertainty and indeterminacy surrounding technological development. Though in theory a firm might seek to develop and launch new technologies on its own, escalating development costs, globalisation and pressures for standardisation of many ICT systems mean that technology development has an increasingly collective character - taking the form of a networking activity, involving inter-organisational linkages and modes of engagement between suppliers and 'users' or their proxies. Various forms of collaboration emerge as an attempt to share costs and risks - and reduce uncertainties by foreclosing options in advance - rather than incur the costs of fighting out standards wars in the marketplace.13 However, as we see later, these efforts, paradoxically, These issues are particularly acute in relation to the information infrastructure, and especially the delivery systems and end-user technology platforms in the workplace and especially the home, which are widely dispersed but together involve enormous sunk investments. 13 Pressures towards harmonisation should not be seen as a simple victory of cooperation over competition. In fact we see an extremely complex politics around contradictory tendencies around on the one hand pressures to align supply-side players around overall creation of markets and on 12 53 end up by imparting a higher degree of indeterminacy and uncertainty to technology development processes. Similar commercial strategies are being played out in the area of software - in terms of operating systems and applications software. In particular we see attempts to establish standardised applications (eg databases, or spreadsheets) as open platforms on which other applications could be mounted (Swann & Lamaison 1989, Swann 1990). Today much attention is directed towards the anticipated battle between proprietary Microsoft and 'open' Java operating systems and principles for the next generation of applications on the Internet. Software supply strategies and the user These interactions between global (technologies, markets) and local (user contexts) are often played out in relation to software. Software represents the critical layer in ICT systems - it forms the interface between the 'universal' calculating engine of the computer, and the wide range of social activities to which ICTs are applied. For ICT systems to be useful, they must, to some extent, model and replicate parts of social and organisational activity. Software is designed to achieve particular purposes; its design embodies particular values and social relationships. The various social groups involved in or affected by IT may have different objectives and priorities. Software is thus a potential site of conflict and controversy (Dunlop and Kling 1991, Quintas 1993). This is particularly evident in relation to the complex integrated software systems being developed to support the activities of large organisations. There has at the same time been an important recent shift in the design and supply of software involving the development of packaged software tools - such as spreadsheets - which can be adapted by the user to a wide range of purposes. Some of the most successful applications in communications technologies, such as electronic mail or latterly desk-top video-conferencing, are in the form of media, that offer little constraint on the content to be exchanged and that can thus be applied in a wide variety of contexts. So we find a dichotomy in software development strategies between on the one hand attempts to designing more of 'society' (and specific contexts of use) into the software and on the other, attempts to design 'society' (and specific contexts of use) out of the artefact. We can show how these conflicting pressures, between maximising the extent to which software matches particular social settings or maximising market size in the case of strategies for the supply the industrial software applications - displayed graphically in Figure 2.1 below (Procter and Williams 1996). This shows at the one extreme, custom software designed around the needs of a particular user (highly complex and with a market size of 1) and, at the other, discrete applications (e.g. spreadsheet or word processing software) that are supplied as cheap commodified solutions to an increasingly globalised market. The former is expensive, but offers the other competitive desires to dominate (and at times fragment market or lock customers in to proprietary standards). The case of Electronic Funds Transfer at Point of Sale in the UK provides an effective illustration (Howells and Hine 1993). 54 solutions that are well-matched to local needs; the latter offers very cheap solutions to particular tasks. The failure to recognise this trade-off underpin the critique we advance in Chapter 6 of the concept of ‘user-centred design’ as the solution to the problem of usability of software. Figure 2.1 Volume Variety Characteristics of Packaged Software Solutions from Procter and Williams 1996 The focus of new supply strategies has been between these extremes: for example attempts to adapt and ‘sell-on’ complex customised solutions to organisation with broadly similar settings and activities (e.g. credit-card processing) as niche-specific applications; or generic applications designed to be adapted by the user to meet a variety of particular circumstances (e.g. production control and scheduling systems) (Fincham et al 1995, Webster and Williams 1993). One other strategy is becoming increasingly influential in contemporary ICT applications: that of configurational technology solutions. With the emergence of inter-operability standards it becomes more easy to select a range of standard components (hardware and software - e.g. personal computers, network management software, database technologies) and to knit them together in conjunction with customised components into a particular configuration that matches the particular circumstances of use. Such configurational technologies offer a much cheaper way to meet the particular needs of a complex organisation than fully customised solutions. Complex ICT systems today are increasingly taking on these 'configurational' characteristics. Indeed, given the enormous cost advantages of standard hardware and software components, few developers would build a new system entirely from scratch. 55 b) transforming our understandings of a technology. Paradoxically, it is in the area of mass-market ICT products for use in the home and in everyday life that we have some of the clearest evidence about the centrality of user responses in shaping not only the differential success of new products, but the very conception of these products (for example the camera [Jenkins 1975] or the Video Cassette Recorder [Roosenbloom 1982]). Perhaps the most striking and wellresearch example is the telephone, which was originally conceived and promoted as a business communication tool for conveying price information to farmers, but which was re-invented by people in rural areas, particularly women, as a medium for social communication (Fischer 1992). The final consumer may have little opportunity to engage with the design and development of mass-produced goods. This is particularly marked in relation to ICT products for the home which remains a largely private sphere, with only weak, and predominantly indirect linkages between supplier and user (Silverstone and Hirsch 1992, Cockburn and Furst-Dilic 1994, Cawson et al. 1995). However, even in this setting it is important to acknowledge the scope for these actors to articulate their own representations of technologies and uses which may differ from those articulated by technology suppliers (Sørensen and Berg 1992, Akrich 1992, Cockburn 1993). Consumption is an active process, involving selective decisions to purchase the technology and appropriate it within the household - in terms of where it is located and how it is incorporated within family routines (Silverstone and Morley 1990). Although the designer may seek to prefigure the user - and thus implicitly to constrain the ways in which the product is used - the final user still retains flexibility in the meanings they attribute to technologies, and in choices about the artefact will be appropriated. This often involves innovation by the consumer - using technology in ways not anticipated by the designer (Berg 1994a). Designers and developers have had problems in addressing the needs of mass consumption products. Domestic ICT users (and refusers) are not an amorphous and homogeneous mass; their responses are differentiated by gender, generation and class, and shaped in the complex social dynamics (or 'moral economy') of the family (Silverstone and Morley 1990, Silverstone 1991). Given the paucity of links between designers and potential users, designers have often relied primarily upon their own experience and expertise; starting from their understanding of technological opportunities and imagining how these might be taken up in their own households - which may be far from typical (Cawson et al 1995). This may cause problems in the acceptability of ICT offerings. Perhaps the most obvious example today is the 'baroque' design of most contemporary video cassette recorders - which the vast majority of consumers find very difficult to use. Similarly, suppliers' lack understanding of 'the housewife' as a possible user, and of 'her' needs and have had little appeal to many customers (Berg and Aune 1994) For example ‘smart house’ technology demonstrators reflect technology-push rather than user-need; they have not really addressed the realities of domestic labour (Berg 1994b). This may be one reason why the adoption of domestic ICTs has often fallen far short of expectations (Thomas and Miles 1990). 56 The case of the home computer provides an illustration of how technologies are appropriated by domestic users (Silverstone and Hirsch 1992). The evolution of this technology became subject to a web of competing conceptions articulated by various players: government, suppliers, parents, children. Though initially promoted as a means of carrying out various 'useful' activities (word-processing, household accounts, and in particular as a support for educational programmes), this was largely subverted by consumers in the family. In particular boys, though possibly pressurising their parents to acquire a computer for educational reasons, were really interested in using them for playing computer games. Indeed their enormous interest in computer-games has shaped the evolution of home computers, leading to the creation of a specialised market for these products (Haddon 1992). The markedly uneven success of telematics products and services (such electronic mail, videotex and fax) points to some of the complexities surrounding the design and uptake of technologies (van Rijn and Williams 1988). The success of new telematics services does not simply reflect their functionality and price, but also the extent to which they are compatible with the skills, understandings and habitual practices of potential users (Miles 1990, Thomas & Miles 1990). The history of facsimile technology presents an extremely interesting case. Though conceived as early as 1843, with its first commercial launch in 1865, this and successive attempts to develop the technology did not meet with success apart from particular niche applications - for example its use by newspapers to transmit photographic etc. images to remote offices. Fax did not take-off until the 1980s, when its (largely unexpected) explosive growth was attributed to the success of Japanese suppliers in 'manufacturing a superior machine' that was cheap and designed to be easy-to-use (Coopersmith 1993). Only then was fax technology close enough to the requirements of everyday business and domestic users for them to be enrolled as consumers.14 Network externalities are particularly important here - and a technology may not be attractive to a potential user if sufficient numbers of other users cannot be convinced to sign up - as in the case of videophones (Dutton 1995). Similar factors may explain the uneven success of technologies such as videotex, which have been strongly promoted over the last two decades, but which in most countries dramatically failed to fulfill the predicted rates of uptake (Bruce 1988, Thomas & Miles 1990, Schneider 1991, Christoffersen and Bouwman 1992). The case of Minitel in France provides an important, and enlightening exception. Here France Telecom created circumstances favourable to the diffusion of the technology, and to social learning in innovating new services. This included making terminals available to large numbers of domestic consumers at low cost, and creating a framework for service suppliers to operate (including provision for charging for services through the customer's phone bill). As a result a huge range of information services could be made available many of which proved commercially viable. Significantly, the fax, and the other most widely spread telematics application, electronic mail, are both genuine media, making few presumptions about the kinds of communication they support. In this sense they parallel the tool-based approach which proved so successful in personal computing. 14 57 2.4 A Model of Innovation in Multimedia This brief review provides the intellectual starting point for the study of Social Learning in Multimedia. The processes identified shaping the development and uptake of ICTs are extremely relevant for understanding the innovation and future prospects for multimedia. We lay out below the broad model of innovation in multimedia which informed the SLIM project. a) The definition of Multimedia Multimedia refers literally and most generically to the facility to present information in a variety of media (for example, voice, and graphics as well as text) through a single integrated channel. Today it is used more specifically to refer to the expected convergence15 of information and communication and broadcasting technologies, involving the installation of high-speed broadband communications networks, coupled with advances in processing power and usability of information technology, enabling the storage, processing and transmission of large volumes of digitised information creates the scope to handle graphics (static and animated) and sound as well as text. Multimedia thus does not refer to a particular technology, but to a cluster of innovations. Their ability to make information technology systems easier and more engaging to use underpin expectations of the rapid adoption of a new cluster of technologies under the rubric of multimedia, and the further expectation that applications based upon these converging technologies will rapidly become widely adopted in many areas of working and social life, and have profound social and economic implications. Most definitions of Multimedia start from particular technical characteristics. Two features are salient - interactivity between the user and the product or other users (for networked products) and 'multimedia-ness' - the scope to present data through more vivid forms than simple text (i.e. voice, graphics, moving images). However, the key aspect of Multimedia does not reside in any particular technical feature, but in the increased control and choices it offers for processing information. One of the implications of the perspective being developed here is that what is key is not a technical definition of interactivity eg in terms of bits of data being exchanged – but the interactivity in use of a system – which may be achieved through a variety of different technical configurations (Curry 1999, Schmutzer 1999). There are a variety of alternative technical paths for delivering essentially the same kind of functionality. The SLIM study therefore adopted a broad definition of multimedia, particularly in selecting cases for study based on the opportunities they afforded for examining 15 We should however treat claims about convergence with some caution. There is doubt as to the extent to which industries and product markets will indeed converge. (Fransman 1996). Concepts of convergence have gained currency at several stages in the history of ICT - underpinning the imputed 'Information Technology Revolution' (combining telecommunications and computing [see Forester 1985]) and the subsequent emergence of Information and Communications Technology (Dutton 1995). 58 social learning about the application and use of multimedia rather than their conformance with particular technical criteria. 16 b) The dynamics of Multimedia development In chapter 1 we drew attention to the enormous uncertainties that surround the development of multimedia. Whilst continued progress can be expected in the performance of the generic technical components (e.g. increasing information processing power) there is much less agreement about how these will be configured into delivery systems and about which applications will succeed. As in the case of many other ICTs there is a tension between the promises held out rapidly developing technologies and what seems to be delivered (especially in the shortterm) and between the imputed universal applicability of new generic technical capabilities and the specificities that surround their adoption within various social settings. Multimedia as configurational technology The concept of configurational technology is also highly relevant to understanding these developments. In particular it provides a schema for understanding how contradictory requirements are reconciled - in particular the tensions between global technological development and its local appropriation, and how supplier offerings may be matched to local user requirements. Like other rapidly-changing complex technologies, multimedia is heterogeneous, combining offerings from a range of different suppliers to meet particular requirements, affording considerable flexibility in development, implementation and use. The concept of technological configurations can be applied at two levels: i) to highlight the way that a range of component technologies (fibroptic cable, microprocessors, software tools) are assembled into particular configurations in 16 Multimedia definitions and taxonomies have traditionally been constructed around particular technical characteristics – distinguishing for example between on-line (eg multimedia pc) and standalone (eg CD-ROM) media. However, in terms of social learning about multimedia applications and their appropriation, specific technical features may not be the central issue. Moreover, such distinctions may be rapidly eroded by rather modest technical development – for example many CD-ROMS today are designed as gateways and directories into the world-wide web; the ‘attachments’ facilities of email software allows email to be used to transmit graphic data. Internet standards have allowed telematic systems to rapidly upgrade from text-processing to sound and graphics. These considerations point to the need for flexibility in the definition adopted of multimedia. For these reasons the SLIM project adopted a rather broad definition of multimedia, arguing that it would be unhelpful to restrict the scope of the study around an exclusive technical definition of multimedia as systems that are both interactive and use multiple media. The selection of cases for study could, instead, encompass systems fulfilling only one of these features where they presented significant opportunities for social learning today. In order to address the key issue of how society is learning to use new technical capabilities, studies could, for example, include stand-alone systems such as multimedia capable personal computers (PCs) and Interactive CD ROM based systems, or networked systems (such as email or groupware systems) which might not yet offer multiple media. 59 terms of delivery systems (for example public kiosks, interactive TV or networked personal computers taking these systems into the home); ii) to show how these delivery systems are themselves configured to the purposes of the particular networked multimedia applications that run upon them (for example teleshopping, on-line newspapers, games, interactive TV services). This is summarised in figure 2.2 below (Collinson et al. 1996). APPLICATIONS Specific configurations services, applications and products, within particular sectors and contexts video-on demand tele-shopping CD-I education packages video-conferencing electronic cash-cards DELIVERY SYSTEMS Combinations of CD-ROMs, technologies for storage, PDAs display, delivery, Interactive-TV distribution / set top boxes The Internet and services COMPONENTS basic building blocks microprocessors, that can be combined to video standards enable product and ISDN bandwidth system development screen resolution software tools compression techniques definition Figure 2.2 examples The Three Layer Model of Innovation in Multimedia adapted from Collinson et al. 1996 The importance of this schema is in charting the complexity of the innovation process underpinning the emerging information infrastructure. In particular, it opens up for examination the level of autonomy between the different levels. It suggests that particular combinations of components can be put together to build a variety of different platforms and applications, with different social and technical 60 characteristics and implications. Conversely, particular components can be transferred into different socio-technical settings – as we see, for example with digital TV set top boxes being opened up as devices for accessing the Internet. In terms of the influence of the user and social learning processes on multimedia the key social shaping processes concern the development and implementation of applications and their attendant delivery systems. It is here that the most salient societal choices about the character of the information superhighways and the applications running upon them will be made. For example some kinds of delivery system will be closely dedicated to particular kinds of application (for example, an interactive television terminal) - while others will provide a more open platform on which a range of networked services can be mounted (for example a personal computer linked to the Internet).17 These choices which affect the flexibility of the delivery system could therefore have important social implications. These considerations may be particularly significant in relation to the main multimedia platform for the home – it was perceived that choices here could constrain the types of applications that the mass of users would readily be able to access, and the range of roles that these consumers will have. For example, will they mainly be passive consumers of informational products (on a broadcasting model) or will they be able to engage in more interactive uses, or will they even be information producers in their own right? This range of scenarios encompasses very different images of the character of the future 'Information Society'. The significance of these processes is heightened by 'path dependencies', which may be particularly pronounced with network technologies like multimedia. There has been a widespread expectation that particular multimedia applications would be the 'key drivers', bringing networked multimedia products into the home and workplace, and establishing models for other products and services, which may in turn constrain the kind of further applications that can be mounted upon them. The first widespread application will involve enormous sunk investments, by suppliers and users. This could be an important influence upon the shape of multimedia and its societal implications in the medium term, particularly if it leads to the widespread implementation of delivery systems and devices that are dedicated around the requirements of those specific applications (e.g. video-on-demand) rather than offering more flexible configurations (e.g. networked PCs). An interesting example is provided here by the case of Minitel in France. Though this was seen as exemplary in creating a market for on-line services, it has been bypassed by the emergence and global uptake of Internet-based standards. Difficulties in upgrading the dedicated Minitel terminals to Internet standards constitutes a barrier to access to the world-wide web by the mass market. We have already referred to the increasingly media-like characteristics of many ICT applications – in terms of their relative independence from the information and cultural content that is being communicated. As we see later one of the main findings from the SLIM project concerns the importance of content development in the appropriation of multimedia. 17 61 Paradoxically, an unplanned consequence of the selection of an industry-standard pc as the terminal for the German videotext system has allowed German users ready access to both delivery systems (Schneider 1999). Our understanding of the multimedia platform as configuration also highlights the scope and constraints upon its reconfiguration. One implication concerns the desirability of aligning wherever possible with emerging de facto global standards. Other public policy and technology strategy implications may follow. For example, reconciling the desire to benefit from using technologies before closure with the need to avoid entrenching incompatible solutions suggests the need for strategies geared towards longer term flexibility for example involving gateways between different systems and migration paths to allow subsequent upgrading of systems to be compatible with emerging standards (Spacek 1997). c) Social learning in multimedia Significant technical obstacles remain for example regarding the reliability and security of transactions across the Internet. The ‘local loop’ with its limited carrying capacity continues to be seen as a bottleneck to bringing broadband services into the home. However we can be rather confident about continued improvements in performance of many of the core technologies. The geometrical growth in computing power suggested by Moore’s law is being outstripped by the rate of installation of new telecommunications capabilities. A new generation of products and services are being envisaged around the development of 'multimedia' technologies and the 'information superhighways' that could bring digitised video and sound, as well as text messages, into the home (Collinson 1993, Cawson et al. 1995). A plethora of products are being launched for the workplace (Procter and Williams 1996). Central and local government are investing in developments particularly in relation to education. Huge markets are anticipated for new products that are interactive, easier to use, and more engaging. However, as this review has shown, there is little understanding of what the products that will eventually prevail will look like (Dutton et al. 1996, Kahin & Keller 1995). Perhaps the most profound uncertainties surround the social uptake and use of the new products and services. This is an area where we have very little direct experience or relevant knowledge. Indeed one of the key features of 'the multimedia revolution' to date is that it has primarily been driven by the perspectives of suppliers of equipment and services, informed largely by their expectations of what might be technically feasible. This may not be altogether surprising insofar as few exemplars of these future services have been created to date - so there is little possibility to assess their attractiveness to the user. However, despite the huge amount of energy devoted by suppliers, promoters and media popularisers of multimedia-based products and services, there is remarkably little evidence about the nature and strength of user demand. The long awaited 'killer application' that will herald the multimedia revolution still somehow eludes (Dutton et al 1996). Many of the first round services launched (e.g. interactive TV) are only on the margins of commercial viability. Some of the early commercial ventures and industrial alliances - driven by particular concepts of the new market 62 opportunities that were anticipated from technology convergence and which in turn fuelled wider expectations about the multimedia revolution – have already been quietly shelved, often following substantial losses. Most of the multimedia applications that have emerged to date are markedly conservative innovations - often simply on-line replicas of existing products and services (Collinson et al 1995). This represents a very rational strategy by suppliers seeking to minimise uncertainty, drawing on their existing expertise and links with existing customers. However, as we saw, the experience of earlier generations of ICT suggests that the main products and services that eventually prevail in the future information society may well be far removed from the embryonic offerings available today. It is extremely difficult for suppliers to assess what user requirements will emerge. Suppliers of new products and services are forced to operate in a context of only limited information and considerable uncertainty about 'the user' and 'their needs' and how they might be satisfied by multimedia. Moreover 'users' and 'user needs' are not pre-existing and static entities, but emerge and evolve through an interaction between supplier offerings and user responses. In some ways, the supplier can be said to 'configure’ or ‘construct the user' - or rather to seek so to do (Woolgar 1991, Akrich 1992a). The key questions about the future of multimedia thus concern the engagement between supplier and user. Successful applications will doubtless emerge. In the meantime public and commercial promoters have recognised the need to develop better understanding of existing and nascent requirements of potential consumers/ users of their offerings. They have even looked to social science for answers as well as launching a growing array of feasibility studies, pilot projects and commercial trials of multimedia and related technologies (for example interactive-TV) to provide this information. Social shaping of technology research can offer important insights here, as the preceding review has shown. However it also points to the need for continued scepticism. For example, we still do not know the answer to the most basic question regarding how best to match user requirements to the new technical possibilities. For example, we do not know whether the most effective approach will involve building more societal knowledge into the design of new applications - or alternatively will be achieved by designing and launching generic offerings, and letting final users learn how to adapt supplier offerings to their purposes. Furthermore, what opportunities exist between these two extremes of ‘user-centred design’ and ‘laissez-faire’ supply and user-led appropriation? How might these best be combined? If multimedia is to become widespread, we suggest that an important feature of its evolution will be the highly dispersed processes that we describe as social learning involving a range of social actors: suppliers, policymakers, promoters and intermediate and final users (Procter and Williams 1996). The thesis of the SLIM study is that society and its various publics can draw more general lessons from 63 studying the creative attempts of these players to resolve particular issues around multimedia development, implementation and use. d) the experimental character of contemporary multimedia developments When the SLIM research design was developed we anticipated an important focus upon specific ‘social experiments’ in multimedia - perhaps of the sort conducted in Scandinavia in the 1980s (Cronberg et al 1991). Such social experiments sought to offer exemplary solutions for particular social groups, articulating values and concerns that might be overlooked by conventional commercial developments (for example the needs of economically marginal and socially excluded groups) and based on the involvement of users. One finding from this project is that this particular of social experiments are today rather rare. However, the contemporary development of Multimedia-based products and services remains profoundly experimental (Jaeger, Slack and Williams, 2000,forthcoming). On the one hand, government and private sector players feel compelled to become involved and gain experience with the technology; on the other hand they face deep uncertainties - in particular about the responses of 'final users' to their offerings, as well as the behaviour of other players - such as collaborators and suppliers of complementary products. In this context one of the key ways in which a technology may be carried forward has been through demonstration projects, feasibility studies, pilots and commercial trials (as well as formal social experiments). A plethora of such experiments, feasibility studies and pilots have been launched in most developed (and many developing) countries worldwide, often as part of 'Information Society' initiatives, together with a host of commercial initiatives and trials. There are important commonalities between these different types of initiatives particularly in relation to social learning. In the SLIM project we refer to these collectively by the term multimedia experiment. The other reason for using the term multimedia experiment is that multimedia projects were experimental even where they was not intended to be so. Where projects were set up with rather limited goals, their outcomes tended to be rather more open and to offer longer-term opportunities for learning (even with projects that got into difficulties and failed to meet their objectives) in ways that were not anticipated at the outset. The experimental nature of these developments was not however necessarily recognised or properly catered for.18 The lack of prior attention to social learning possibilities was particularly notable in projects that were launched with the primary objective of developing technical capacities, where there was often very little prior discussion of technology application and use. Of course, even the most 18 Similarly the potential for innoffusion was often not recognised (Fleck 1988, Fleck et al 1990). 64 'technical' of trials needed to involve some kind of social application and concept of use. Indeed such trials were often conducted with 'real life' applications. However issues around application and use were not always given proper attention. Although experimentation inevitably took place in the course of these trials these were not necessarily planned for and specified in advance. As we see in Chapter 4, this was particularly noticable around technology infrastructure developments, such as the Metropolitan Area Networks, which created broad-band transmission capabilities around which a swathe of multimedia application initiatives arose. In this context, experimentation and social learning around the application and use of multimedia are thus smuggled in - as a more or less implicit and covert outcome rather than a central goal of a technically oriented trial. This lack of attention and resources has potentially serious consequences. For example where projects were only given short-term support, resources were not necessarily available for the longer-term exploitation of experiences (as we see in the concluding chapter). This is not to accept that multimedia experiments can ever be narrowly 'technical'; these infrastructure initiatives need some kinds of application for testing to take place. Even though formal trials may be couched in terms of a technical feasibility study, this is likely to be the starting point for a more extended informal process of social learning around the development and use of applications that run upon the infrastructure (Jaeger, Slack and Williams forthcoming, 2000). Though one cannot prevent social learning, in many such de facto trials and experiments, it seems that the lessons are not always being systematically sought or communicated. As a result, it seems, some lessons are lost and have to be learnt time and time again. Despite the powerful rhetorics of the impending Information Society and the multimedia revolution (and the substantial proportions of GNP being invested in ICTs generally) it is important to remember that multimedia projects - i.e. those involving more innovative technologies or uses of technologies - were often rather marginal to the core concerns and activities of many of the key commercial and government players involved in creating new multimedia services.19 As a result, the levels of investments were often modest, and it often proved difficult for the proponents to obtain the levels of resources they felt necessary. We can see the resort to this experimental approach as representing a trade-off between competing concerns - to avoid the risk of being 'left out' of a technology that was deemed to represent the future, while reducing the risks of wasting significant sums on inappropriate investments. Whilst the lack of resources and commitments did present problems for the players, it also did provide a space for groups within these organisations to experiment with these new technical capabilities. We see the proliferation of such multimedia experiments and trials as an indicator both of the importance of social learning in such a novel technological arena and a pointer to the growing recognition of such social learning by many of the players involved. The diversity of these kinds of multimedia experiment is of particular interest. Beyond these we note the variety of ways in which multimedia is being developed and implemented more generally. Multimedia experiments can be A similar observation could also be made about the commercial and individual consumers of multimedia. Multimedia was of course extremely central to ICT suppliers. 19 65 expected to provide a more elaborate and formalised context for social learning. There is however, little understanding about the actual processes of social learning taking place in such experiments and more broadly. The SLIM project emphasised that we do not know how best to organise such learning processes. We were not committed at the outset to particular approaches such as social experiments specifically, or even formalised multimedia experiments in general but were instead open, for example to the idea that conventional models of technology supply through the market might suffice. The intellectual challenge that this project sought to address was thus to conduct an open-minded exploration about how multimedia innovation proceeded in different contexts, and the social learning opportunities they afforded. . 66 III Scientific description of the project results and methodology Chapter 3 The Scope and Methods of the Study 3.1 Goals of the SLIM Project The Social Learning in Multimedia (SLIM) research project emerged from discussions amongst an informal network involving some of the leading European research centres in the area of social shaping of technology – and in particular the social shaping of Information and Communications Technologies.20 The consortium sought an interdisciplinary understanding of the process of technological change – and one that was open also to insights from cultural and media studies, which was judged to be of particular importance to analysing the development and uptake of multimedia. a) General Aims of the SLIM project: The general aims of the SLIM project were briefly summarised as: i) ii) To improve our empirical knowledge and our theoretical understanding of the development and implementation/ consumption of multimedia products and services through cross-national and multi-disciplinary research; to explore the social and policy implications of these research findings. These were pursued through a set of specific objectives. b) Specific Objectives of the SLIM project: 1. To improve general theoretical understanding of the relationship between technology and society in the case of dynamic and complex technologies such as multimedia, and develop concepts and tools for analysing social learning 2. To examine how different national/ regional settings shape strategies for the appropriation of multimedia: To undertake a set of comparative national studies and selected international studies, in order to explore how features of the local multimedia constituency and the broader market and policy context influence the way that opportunities for multimedia are conceived and pursued. a) Many of these centres had been previously active in the COST A4 European research collaboration on ‘the social shaping of technology’. 20 67 b) To review the range of initiatives across Europe in multimedia applications and services, including commercial pilots and social experiments, in order to chart the range of conceptions of multimedia applications and the different ways of organising social learning about multimedia. 3. To examine processes of social learning in multimedia, focussing upon the ways in which current and potential requirements of different groups of users are represented and matched with technical possibilities To explore this through a set of integrated studies of the development, implementation/ consumption and use of new multimedia-based products and services. To develop sets of detailed case-studies across three settings seen as being crucial in the overall emergence of multimedia products and services: education, cultural industries, public services, To analyse features of these sectors/ areas of activity and further to consider particular features of such social learning in the home and in the organisation 4. To consider the implications for public policy at regional national and European level. Specifically: to examine how social learning and innovation in multimedia might best be focused on meeting existing and emerging user requirements, through social experiments, commercial trials and other means, and to explore public policy options for the appropriation of multimedia, and how they might best promote innovation and social learning. To disseminate these research findings within the participating countries and at the European level and discuss their implications with academic, industry and policy figures. 68 3.2 Research Design and Conduct of the Studies a) Phase 1: Preparatory Studies In the first Phase of SLIM research sought to create a knowledge base and analytical foundations for the research project. Reviews were conducted of the state of the art of research in three areas: i) ii) the social shaping of technology and its policy implications Social shaping/ social learning in multimedia and other information and communication technologies These are summarised in Chapter 2. iii) processes of 'social learning' in new technologies - including its component concepts of domestication, innofusion and appropriation of technologies. This is explored in Chapter 5. Three short reports were produced and widely disseminated through the SLIM web pages and formal publications (Williams 1997). b) Phase 2: National and International Studies Cross-national comparison and international research has an important contribution to understanding the social shaping of multimedia and the scope for social learning. Specifically it allows us: to examine diversity in the way multimedia has been conceived and in the systems of innovation in different national and regional settings: to explore the influence of particular socio-economic settings in shaping these to explore the interaction between global and local factors. i) ii) iii) Each centre therefore undertook a national study as part of a cross-national comparison focusing on how the national/ regional setting shaped the strategies for the development and appropriation of multimedia. Selected ‘international research’ was undertaken to provide a setting for the national studies. A review was undertaken of social experiments and trials in multimedia in each nation, which formed the basis for developing the Integrated Studies of Development and Use. Methodology for national studies The national studies sought to provide a broad coverage of a set of extremely complex and rapidly changing developments - and moreover undertake this in a 69 short period and with a very limited level of resources for each country. We were not therefore able to undertake a large-scale survey, ab initio, that could be based upon a common methodology. Instead it was necessary to rely, where these were available, upon existing national statistics, published material and the wealth of 'grey literature' in this field for each country, supplemented by selected interviews with key players. This inevitably entailed a number of problems with the interpretation of material - and necessitates particular care in relation to making detailed comparison . Where possible, resort was made to international (and especially European) databases and statistics to provide comparitors and indicators. However our work with these shows that there may be continuing problems with the use of such data, particularly in such a rapidly changing field (where for example statistics on the use of internet need to show the month for which the data pertains rather than the year so great is the rate of growth in some countries), and where definitions are not sufficiently well-established to provide benchmarks for comparisons. Reliance on diverse (and extremely variable) secondary sources meant that it was not possible to adopt a wholly standardised methodology for the national studies. For example, the countries covered in the review differ greatly in their size and the number of initiatives going on. In the larger economies, it was necessary to have a more selective focus than in the smaller ones. Thus, the review of multimedia experiments for the larger countries encompassed what was judged to be the most significant and interesting national cases, together with an array of more modest cases in the immediate locality, whilst in the smaller economies it was feasible to attempt a more inclusive review. The national studies covered Belgium, Denmark, France, Germany, Ireland, Japan, the Netherlands, Norway, Switzerland, the UK, Each SLIM research centre undertook a study of its own national context. In addition, France was studied by the Belgium SLIM Centre, and Japan was studied by the UK SLIM Centre. To aid in the conduct of the national studies and presentation of their findings guidelines were agreed amongst the SLIM research teams regarding the methodology, content and format of the national studies. The national studies have been published separately as Williams and Slack (1999). Each national study comprised: 1) 2) 3) 4) a one page summary of the highlights of the national study a factual description of the national context an analysis of the key players shaping the development of multimedia comprising: a map of the main players, an overview of their strategies, and a review of each of the main sectors involved. a review of social experiments and commercial trials. The national studies sought to capture how the national/ regional setting and the array of major actors patterned the dominant conceptions of multimedia and the 70 perceived ‘rules of the game’ for multimedia innovation. The descriptive studies and their analysis were 'brought to life' by interrogating the available sources regarding a variety of research hypotheses about how the national/ regional setting might shape the appropriation of multimedia. These hypotheses served as a sensitisation tool for the researchers in conducting their national study, and guided the comparative analysis of findings. These hypotheses can be grouped around three areas: 1) 2) 3) the influence of the character and orientation of the main multimedia actors and 'constituencies' in the nation the influence of the demographic, geographical and cultural context the influence of the policy context international studies Though not able to investigate global multimedia developments per se the study explore how such developments, such as international standard setting and the activities of international suppliers, impinge upon multimedia development/ appropriation strategies within Europe. Thus the national studies and the 'Integrated Studies of Development and Use' both addressed the national/ local adoption of products developed elsewhere. In addition a limited amount of specific research was undertaken of international developments. Insights were drawn from a related study at the Technical University of Denmark of the role of different interests and players in the global development of multimedia standards and their implications for European multimedia industries. FUNDP - Namur University studied EC policies for multimedia, and compared them with those developed in the USA. Edinburgh University conducted a review of Japanese approaches to multimedia to establish its distinctive features and differences/ commonalities with European approaches. Review of multimedia experiments For the review of social experiments and trials, a more inclusive approach was adopted, with a more rigorous methodological foundation, supplementing published sources such as official documents with interviews with key players. These reviews were not only useful in their own right, but also provided the initial sample from which to draw the set of more detailed 'Integrated Studies' of the development and uptake of multimedia products and services in 3 areas: Education, Public Administration and Cultural Products under Phase 3 of the SLIM research project. c) Phase 3: Integrated Studies of Multimedia Development and Use The Integrated Studies 71 The main part of the SLIM research effort was devoted to ‘integrated studies’ of multimedia development and use. This attempt to study both the development and implementation/ use of Multimedia was a distinctive feature of the SLIM project and represents perhaps its most ambitious goal. Earlier studies of everyday technologies had typically failed to do this. With few exceptions (notably du Gay et al’s 1997 study of the Walkman) studies have ended up looking at EITHER development OR consumption/ use of particular artefacts, even where they set out to do otherwise (see for example Thomas & Miles 1990, Cockburn & Furst-Dilic 1994). The reasons for this are not difficult to find. The most immediate factor is the long time-scales for the development and adoption of new artefacts, which exceed the duration of most externally-funded socio-economic research projects. In addition development activities may be separated in geographical and institutional terms. The SLIM study, with rather limited overall resources and duration, faced particular challenges in this respect.21 As well as the studies undertaken with EC funding it was also possible to adapt case findings from research within SLIM centres funded by other bodies. In this way an array of extensive investigations were achieved, exploring the processes of multimedia development and use. The studies allowed varying insights into appropriation processes in different settings. Some studies were of projects that remained mainly at the development stage and did not provide an opportunity to address appropriation. An array of cases was developed in relation to three areas of activity,22 chosen because of their special significance as probable 'drivers' of multimedia and their socio-economic importance. They were : education; cultural industries/ entertainment; public services. At the outset of the study we conceived these studies as first addressing how the supplier conceives new multimedia applications, and thereby 'constructs the user'; then assessing the extent to which these constructs match the expectations of actual and emerging users. By examining the consumption and use of new multimedia offerings, we sought to explicate processes of user innovation, by which the user fits multimedia products to their own purposes highlighting the extent to which this simply replicates supplier concepts or extends them in unanticipated ways. In practice the cases revealed a rather more complex and inchoate situation than might be presumed from this initial schema. For example, as we shall see, the tacit expectation that design would be informed by a more or less explicit ‘representation’ of the user was by no means fulfilled. More generally, the separation between an initial design and subsequent implementation/ consumption phase was not always clear cut. To aid comparison between cases which differed to a greater or lesser degree in scale, timeframes and their socio-technical setting, a common analytical frame was used. This started by mapping the ‘transformation terrain’ of the project, including the various actors directly involved in the 21In particular the funding model negotiated by the EC only allowed 18 months for the 'integrated' and 'cross-cutting' studies of multimedia development and use that constituted Phase 3 of the SLIM project. 22 This focus on areas of human activity, rather than existing industrial sectors, was sought because of the potential of multimedia to cut across existing institutional boundaries and locales. 72 multimedia project (in relation to design, implementation and use), and the array of actors and institutions constituting the broader context for the development. 23 Relevant cases were developed in each area by the SLIM Research Centres.24 One Centre was the ‘lead site’ for each area, taking responsibility for coordinating the selection of cases, their analysis and writing up the substantive findings for that sector. Findings from each integrated study are summarised in the Annex. Education lead site: The University of Limburg Education is one of the most advanced areas of multimedia application today, especially in higher education, where a strong technological infrastructure and expertise in multimedia technologies/ techniques, is matched by understanding of the education process and the substantive knowledge needed for the content of programmes. Conditions in other areas of education - such as primary and secondary schools, industrial training and education/ ’edutainment ’ in the home may be rather different in terms of access to such financial and technical resources and pedagogic challenges. A particular interest was with the interaction between multimedia design and choice of pedagogy. The education study explored how the adoption of multimedia differs between different sectors and sites of learning. Cultural industries/ entertainment lead site: Dublin City University This element of the SLIM project will focus on cultural products and services which are directed at the final consumer, individual households and citizens. Developments in these industries will constitute one of the biggest commercial markets for multimedia, as well as having direct social significance. Studies in this domain drew attention to the importance of the cultural content of multimedia as well as its more narrowly ‘technical’ features, and made a special contribution to its conceptualisation. This research will examine the particular conceptions and definitions of the 'multimedia product' and 'user' (and markets) which inform the development of these innovative products and services in the relevant national and regional contexts in an era of increasing globalisation of cultural and social relations. Public services lead site :Namur FUNdP This part of the study was initially conceived as addressing both public services and private services (for example retailing/ shopping). However it became clear that this would involve too broad a focus for developing effective insights. Comparisons Charting this transformation terrain in diagrammatic terms proved an effective tool in communicating the features of case and drawing attention to differences and similarities that might otherwise have been overlooked. 24 Except Bremen University which was only funded to contribute to Phase 2 of the study. 23 73 between public and private services were in any case possible across the other integrated studies (for example many of the cultural products were developed by commercial organisations). A tighter focus was therefore sought on public administration and information systems. The opportunity arose for a more detailed focus upon the important European development of ‘Digital Cities’ including the original Amsterdam Digital City (DDS). Cross-Cutting Studies Cross-cutting studies were also undertaken, pulling together themes and findings about particular sites of multimedia use and issues arising there, viz.: 'the organisation' and 'the home'. These primarily drew upon fieldwork and research undertaken for the integrated studies, be supplemented by selected additional studies in the case of organisational adoption. The Organisation: lead site: Lausanne EPFL This cross-cutting study addressed the process of implementation and appropriation of Multimedia configurations in organisational settings, bringing together findings from relevant integrated studies and a limited amount specific additional fieldwork in firms that have been in the forefront of adopting multimedia. It explored the range of organisational applications and the extent to which the uptake of multimedia varies between different organisational activities and settings. A large number of multimedia products are becoming available for use at work (e.g. desk-top videoconferencing, CD-I, CSCW tools). Organisations have started experimenting with this bewildering array of offerings to determine which are appropriate for different kinds of organisational functions (such as training, cooperative telework, external communication). This poses new problems for the management of technological change. An interesting feature is the combination of standard tools and components and their customisation to meet the particular the technical and organisational setting. The development of these specific organisational configurations of multimedia is of particular interest as an arena of social learning. The Home: lead site : Trondheim (NUST) This activity was intended as a study of incorporation of multimedia-based products and services within the household and their uptake and use by family members. However, in practice, none of the case-studies came to have the home as their main locus of inquiry, and only a few contain explicit information about the multimedia situation in homes. It was necessary to modify the ambitions of this part of the project – addressing the policy discourses that surround the use of 74 multimedia in the home; conceptions of the home informing multimedia projects and current adoption. d) Phase 4: Analysis and Dissemination The SLIM research project had ambitious analytical goals – of developing a better conceptual understanding of the process of innovation, and particularly social learning with Multimedia to improve theoretical understanding our ability to generalise about these phenomena and to intervene in policy and practice. This involved a substantial effort of analysing research findings. Analysis started within the Phase 3 studies through a detailed comparison of individual cases to identify common features in particular settings (Education, Culture/ Entertainment, Public Services, The Organisation, The Home). Our initial hope that the summative reports from these activities could simply be further distilled to produce over-arching insights into multimedia innovation turned out to be rather optimistic. The SLIM project had been extraordinarily productive – and this hierarchical mode of analysis failed to do justice to the richness of our findings. Instead it was necessary to conduct a detailed examination of the full array of casefindings coupled with a reworking of the various conceptual discussions and debates throughout the course of the project. We were consequently forced to request an extension of the project. In tandem with we sought to draw out the policy implications of this research, particularly in relation to two related issues: a) how innovation in multimedia might best be focused upon meeting existing and emerging social needs (specifically, but not exclusively, addressing best practice in building social experiments in multimedia and how to involve 'end-users' in multimedia development). b) b)how public policy intervention might best promote social learning and the appropriation of Multimedia (which might involve various means, ranging from direct interventions such as social experiments, through to the regulation of markets.) In addition, specific substantive policy recommendations were sought from the individual research elements (National and International Studies, Education, Culture/ Entertainment, Public Services, The Organisation, The Home). A major effort was conducted to discuss these detailed findings with relevant policy and practitioner communities and disseminate them through a range of national and international fora. Early findings from the study were discussed with interested officials from DGXII and DGXIII at a workshop (Brussels, 22-3 January 1999). Presentations of SLIM research findings were made to the European SocioEconomic Research Conference (Brussels, 28-30 April 1999). A presentation on Lessons from the SLIM Project formed the keynote for a workshop on “ SocioEconomic Research on Multimedia Content and Tools” (Luxembourg, 24.3.99) 75 organised under the EC 5FP Programme on Information Society Technologies to explore the scope for Socio-Economic Research within their Key Action on Multimedia Content and Tools. SLIM papers formed the backbone of a specialist stream at the ‘4S’ Conference: of the Society for Social Studies of Science and Technology, San Diego 27 – 31 October 1999). A conference on The Social Shaping of Multimedia was jointly organised by the SLIM project and the COST A4 action on ‘Social Shaping of Technology’ (Edinburgh 27 – 9 June 1997) and has given rise to an edited collection (Slack, Stewart and Williams 1999). A special edition of a journal The Information Society will appear in 2000 (Williams 2000 forthcoming). All the integrated studies have yielded substantial monographs, which are being revised for publication though academic fora, either as a special journal edition or as edited collections. Early dissemination of results has been made possible through publication in the series of working papers run by many of the SLIM Centres as well as through an extensive set of pages and sites on the world wide web. This work is now becoming widely published in more formal channels. The full list of publications in the appended Dissemination Strategy testifies to both the volume and the quality of research findings. This document describes in detail the extensive dissemination effort undertaken for this project. 76 Chapter 4 National Settings for Multimedia Appropriation 4.1 Introduction One of the key questions relating to the societal appropriation of multimedia concerns whether Multimedia is an international medium that cuts across and erodes national differences or whether it is appropriated differently within particular national techno-economic and cultural contexts? Modern technologies are often conceived as providing universally applicable techniques and solutions that can be applied in a wide range of social settings; the use of technologies is seen is a simple extension of their powerful technical functionality (often taken as self-evident). Such technological determinist discourses have tended to portray ICT and multimedia as a vehicle for global trends and indeed have proposed some kind of societal convergence. On the other hand, more sociologically-informed accounts have tended to emphasise the contingencies and specificities that surround the application and consumption of technology - and in particular the enduring importance of local cultures and actors. Empirical research (and common sense) would suggest that global and local are both important, and that what is at issue is the interplay and balance between the global and the local in multimedia development and use (Castells, 1996). Here the social learning perspective suggests that the uses and outcomes of technologies are not simply inscribed in the design of artefacts, but depend equally upon the way these are selectively taken up, adapted and incorporated within the practices of individuals and groups. And it is in these local consumption processes that the meanings and significance of technologies are ultimately expressed and achieved. As a result, intermediate and final consumers are part of the equation as well as technology promoters and suppliers. This is particularly marked in relation to technologies such as multimedia, which are not restricted to particular industrial or occupational settings, but which are seen as becoming widely adopted across all areas of work and everyday life. Some of the elements of multimedia are indeed designed and produced at a global level. Certain players have achieved global reach - for example microprocessor manufacturers, who must pursue enormous economies of scale from global production to cover the escalating costs of developing new products. However the development and implementation of ICT delivery systems and applications must inevitably include a local element. The importance of the local may vary, for example, between simply adopting and configuring standard commodified products to the wholly customised production of products and services for particular national or regional users or niche markets. There may be significant differences between different parts of the multimedia field. The core hardware and 77 software technologies at the heart of the information infrastructure may of course largely conform to global technical standards; their selection may be strongly shaped by widely established 'technical' criteria of price and performance. On the other hand, the applications and content that run upon this infrastructure are more liable to be specific to particular local cultures, contexts and requirements. Important differences can be noted in this respect between sectors and application domains. For example one may expect to find more international standardisation in relation to on-line banking or credit card services than for cultural products; the former two are perhaps more subject to common functional requirements and there are strong pressures and templates for international harmonisation. The balance between local and global in the societal appropriation of multimedia is thus likely to vary. This has important economic, social and policy implications. For example, the respective contribution of local and global players in multimedia provision bears critically upon the opportunities for national and regional economies to become involved. This balance of contributions is neither predetermined nor uniform between countries and sectors, but will depend upon which players are active, their resources, the relationships between them (and their relationship with global players) and the particular context (especially the market size, but also government policies etc.). National or regional settings present more or less distinctive terrains for the appropriation of multimedia constituted, shaping the strategies by which local players (and global players acting locally) may seek to develop and implement multimedia, This raises a key research question for the SLIM study concerning how national and regional settings shape the ways in which opportunities for multimedia are conceived and pursued. This is the topic which lies at the heart of this part of the SLIM research project. To explore these questions, the SLIM research network embarked upon a crossnational investigation through a series of national studies geared towards comparative insight. These sought to understand the national setting for the appropriation of multimedia, how this might be patterned by structural (e.g. demographic) and institutional factors and how this, in turn, might influence the strategies and visions of the main players involved. These national studies also included a review of social experiments and trials in multimedia, conducted as a prelude to a series of detailed case-studies of the development and use of multimedia that formed the main empirical focus of the SLIM project. 4.2 Methodology and conduct of national studies A series of ‘national studies’ were undertaken which sought to explore how the national / regional setting and the constituency of leading players in the multimedia arena shaped ‘the rules of the game’ for multimedia innovation and the prevailing conceptions and visions of multimedia. Detailed empirical reviews were undertaken encompassing developments in Belgium, Denmark, France, Germany, Ireland, Japan, the Netherlands, Norway, Switzerland, the UK, These have been published as Williams and Slack (1999). The methodology adopted was described in Chapter 3. The national studies sought to provide a broad coverage of a set of 78 extremely complex and rapidly changing developments despite a very limited level of research time and resources for each country. It was necessary to rely, where these were available, upon existing national statistics, published material and the wealth of 'grey literature' in this field. This inevitably entailed a number of problems with the interpretation of material - and necessitates particular care in relation to making detailed comparisons. The descriptive studies and their analysis were 'brought to life' by interrogating the available information sources in the light of a variety of research hypotheses about how the national/ regional setting might shape the appropriation of multimedia. These hypotheses served as a sensitisation tool for the researchers in conducting their national study, and guided the comparative analysis of findings. These hypotheses can be grouped around three areas: 1) 2) 3) the influence of the character and orientation of the main multimedia actors and 'constituencies' in the nation the influence of the demographic, geographical and cultural context the influence of the policy context We shall examine each of these in turn below in the three sub-sections that follow. They are however closely related. Thus, as we see in Section 4.3, public policy and regulation influence the salience of national players - notably the number and role of telecommunications providers, while Section 4.5 highlights the influence of national industrial players on public policy. 4.3 Main Multimedia Actors and Constituencies Each SLIM research centre undertook a study of how national/ regional settings shaped the prevailing ways in which multimedia was conceived, and the strategies for the appropriation of Multimedia of the main actors involved. An important empirical focus of the study was thus upon the structure and operation of the multimedia constituencies in their national/ regional settings. This addressed the identity and strategies of the key players involved in multimedia provision, the relationships that existed between them and with external players. It sought to explore how the structure and orientation of these national multimedia constituencies shaped the ways in which opportunities for multimedia are conceived and the overall pattern of Multimedia developments. We were interested here in assessing how the strategies of these various players and the interactions between them set the 'rules of the game' as well as the attendant models and visions of multimedia and its uses - in this way perhaps constituting a particular 'technology regime' (Rip 1995). The starting point was an exercise to map the multimedia constituency, by identifying the main active players - who constituted the 'key drivers' within the nation/ region. These players potentially included research organisations and relevant public policy bodies, telecommunication service providers, and the range 79 of firms supplying multimedia and complementary products and services 'intermediate users' such as media and business information companies. 25 The study sought to address the relationships between the main players (e.g. ownership, collaboration) and their relationship with global players. It examined the economic strength and technological competencies of these players, and other factors bearing upon their influence (an important question here concerned telecommunication service providers; whether they were in public or private ownership, and the level of competition in telecommunications provision - see below). Research sought to establish which concepts of multimedia these key players promote, and how this was shaping overall conceptions of Multimedia. We were particularly interested to examine the concepts they had adopted of the market for multimedia products and services - particularly regarding who they see as their user and what uses they envisage. An important variable is the extent to which there are 'national champions' in multimedia provision in particular nations - and in this case, whether and how that player influenced the conceptions of multimedia that prevailed? Such a 'champion' might be a major ICT company or the national Telecommunications provider. Only some countries have major ICT suppliers (Siemens in Germany; Philips in the Netherlands, Thomson in France); all countries have powerful established telecommunications companies - Telcos - though their relative strength and roles may vary. Other players, in particular from existing media industries could also be significant (notably in Germany and the UK). The national and indeed global strategies of these players could be a significant influence over multimedia initiatives, public policies and wider perceptions of technology within a nation - as already noted in relation to the salience given to Digital TV in Germany. Our concern to address how the industrial base and expertise within a nation (including the presence or absence of a national champion) shape the way multimedia is pursued required some assessment of the respective strengths of the main players or groups of players - such as telecoms network providers, the TV/ Video producers, cable companies, the press etc. - as well as the relationships that existed between them. Finally the studies sought to address how the multimedia constituency is changing with the emergence of new actors and the evolution of existing actors, together with shifts in their respective roles and power. 25Although some large industrial multimedia users (e.g. in the finance sector) did appear as key drivers in the national studies, this was not the case with 'final users' consuming and using multimedia in their working and everyday lives. As we see in the Integrated Studies, these actors, widely dispersed across different workplaces, households and communities, were of crucial importance to the uptake and evolution of multimedia products and services. Despite this they were not directly represented in the national public debates and interactions shaping the uptake of multimedia. This raises an interesting question about who represents the final consumer/ user particularly in a context that these technologies are promoted in terms of the substantial benefits they will bring to members of society. 80 Regulation and Competition in Telecommunications Established national telecommunications providers are some of the largest technology players. One crucial feature in relation to the operation of national multimedia constituencies and policy formation (see below) appeared to be the level of competition in telecommunications (and Value Added Network Service [VANS]) provision. There is a global trend, underpinned by the policies of the World Trade Organisation and the European Union towards 'liberalisation' - and in particular the privatisation of Telcos and the introduction of competition in telecommunications provision. However there were marked differences across the European countries studied by SLIM. On the one hand were countries like the UK and Norway, which had made early and rapid moves towards liberalisation, and at the other extreme France, Germany and Ireland in which the process is still under-way. The extent of liberalisation in telecommunications services - the level of competition: whether there was a single dominant player or multiple telecoms providers, together with the regulatory restrictions that might apply to the activities of players - could have important consequences for the development of networked multimedia trials and services. This was particularly an issue in the aftermath of liberalisation in which the incumbent national PTT would typically start from a monopoly position. In this context, public policy and regulation might be driven by the need to offset tendencies towards monopoly and the exclusion of potential entrants by placing restrictions on dominant players. One example can be provided by the UK, in which industrial competition policy constrained the pricing policies of BT, the former public monopoly provider, and protected new entrants to the telephony market, while also preventing it from providing local video-services to give time for the fledgling cable franchises to become established (Slack & Stewart 1999).26 Where there was a single dominant supplier (i.e. a national monopoly or effective monopoly), that player would be virtually a necessary partner in any multimedia pilot or initiative. This would also affect the parameters for participation in trials. In a monopoly or near monopoly context a dominant Telco may be motivated to take part simply to help stimulate uptake of telecommunications services overall thereby building its own market. Such firms were often in a position to make substantial investments (for example Telecom Eirean is a significant player in Ireland [Kerr and Preston 1999] and has recently funded a 'cabled town' initiative). In addition to their large size, their monopoly situation allows them to deploy considerable resources, whilst minimising risk - for example by channeling user responses and shaping market development. In a more competitive context, Telcos may need to give more attention to whether, when and how they will gain a return on investments in such third-party trials, and their attempts to prefigure and stimulate market developments may encounter more immediate challenges. For example France Telecom which in the past has been a major player in large-scale ICT developments such as Minitel has become more cautious since Equally, policies might be geared to preventing technical barriers to free trade - for example by requiring the provision of gateways and interoperability between competing networks (see Spacek 1997). 26 81 telecommunications privatisation and deregulation (Vedel 1997, Lobet-Maris with van Bastelaer and Cammaerts 1999). Though the extent of progress towards liberalisation of telecommunications was an important shaping factors across a range of countries examined, it would be unhelpful to see this as a uniform process leading to a single set of outcomes. There is certainly a shift (and a deliberate shift) towards plurality in telecommunications provision. However, developments were complex - and particular historical contingencies continued to shape national provision in particular ways. For example in the Netherlands, for historical reasons, local cable companies have found themselves in a key position in multimedia developments (van Lente 1999). 4.4 The Demographic, Geographical and Cultural Context Second the research sought to identify how the national appropriation of Multimedia is shaped by its demographic, geographical and cultural context. These constitute the 'structural' physical and socio-political terrain, providing the setting for other considerations such as public policy and the structure and orientation of multimedia constituencies. This part of the study started by examining particular structural features of the national/ regional context including, inter-alia, nation size, geographic and demographic features and linguistic groupings. For example population size and the size of linguistic groups have a very immediate bearing upon market size - and thus the likely return on investment from developing products for a particular market. We were interested to explore how these factors may have patterned the 'rules of the game' in particular regions. Let us explore these points in more detail. At the most immediate level, a question arises about how the density and distribution of population affect the use of telematics and other technologies for remote communication. It seems plausible to suggest a relationship between the high level of uptake of cellular telephony in Norway (or for that matter in Finland), the emphasis in that country upon communication and internal linkages in public discourse and policy and the physical size of the country and physical dispersion of its population. However the structural context does not per se impose particular views. Conversely we also find high rates of mobile phone adoption in small, densely populated countries such as Belgium. Various factors are clearly at play here, and much depends upon how these structural features were construed within social and policy discourses. The size of linguistic communities provides an interesting illustration of this kind of interplay. One area of interest for the study concerned how countries (or language communities) might react to the dominance of US/ English language offerings in the field of multimedia. There seemed to be some interesting differences between countries and regions. In some countries there have been strategies to oppose the dominance of US culture and English language products, as part of a broader 82 concern to maintain national cultural integrity.27 Conversely there was less evidence of this in certain smaller countries such as Denmark and The Netherlands in which English is already widely spoken. (This may in part be explained by the fact that - for reasons of work, or from the need to follow imported, sub-titled films - many people in these smaller countries have already found it sufficiently attractive/ expedient to make the investment needed to learn the English language; in contrast larger economies and linguistic groups it may be more cost-effective to support local cultural production or the translation, rather than just sub-titling, of films, and fewer people may be drawn to learn English). The importance of language group and cultural identities is not simply a factor of population size, of course, but relates crucially to matters of history and political and cultural sensibilities. For example there was more vocal evidence of support for Dutch language programming within the Dutch-speaking communities of Belgium groups which are concerned to maintain their identity in competition with other linguistic (Francophonic and German) communities - than within the Netherlands itself (van Bastelaer, Lobet-Maris & Pierson 1999). Similarly the most salient initiatives in French language programming were noted in Quebec (van Bastelaer with Lobet-Maris 1999). The importance of cultural and linguistic considerations thus appears to be part of a broader political picture. 4.5 The Policy Context Many countries have developed policies and initiatives about the imputed emergence of national Information Infrastructures/ Information Superhighways, and the multimedia services that would run upon them. These policies are often linked to discourses about an imputed current transition towards an Information Society. The latest wave of initiatives across Europe and the developed countries seems to have been stimulated by the high-profile launch of the US National Information Infrastructure (NII) initiative in 1993 (Kahin & Wilson 1997), followed in 1994 by the Bangemann Report (the European Union action plan 'Europe's Way to the Information Society') (Schneider 1997). Governments have sought to keep ahead of the field - or at least not be left out of these technological advances, and the social and commercial benefits expected to flow from them through a range of related measures: i) ii) iii) by creating a supportive policy and regulatory context; by directly or indirectly ensuring the development of the technical infrastructure (e.g. the Information Superhighways); and, by supporting initiatives and experiments in the application of ICTs. In their attempts to match or outdo each other, we find some processes of convergence - arising perhaps through mimicry of policies developed in other 27Paradoxically France's long-standing policy of combating Anglophonic domination in cultural products has not been applied strongly in relation to multimedia products and services - partly perhaps because, through her earlier Minitel initiative, France already has a well-established information market (van Bastelaer with Lobet-Maris 1999). 83 countries and through alignment of views. Many Information Society policies thus seem to be united in their resort to arguments about the competitive necessity of being at the forefront of multimedia developments, and the dangers of being left behind in the technological competition (van Lente 1999). Another feature has been the articulation of 'technological utopian' views of the presumed societal benefits of these changes (West 1996). On the other hand, differences in emphasis and style remain (e.g. around long-standing differences in national policy styles and contexts [Lobet-Maris with Bastelaer and Cammaerts 1999]) - for example in the approach to the role of the state and the balance between state and market in overall provision. As we see below, a spectrum of approaches can be found across the European Union with the UK at one end emphasising laissez-faire approaches (Stewart & Slack 1999) and a range of countries attributing a greater role and legitimacy for the state. To address these, the national studies sought to capture the national policy framework concerning Multimedia. In addressing this, we were particularly interested to identify whether it was underpinned by a particular conception of multimedia, its use and social implications (for example in relation to the dichotomy, discussed below, that was posed between the internet and digital TV). In assessing policy frameworks, we sought to identify points of similarity with, and influence by, external reference points and to assess whether governments had adopted these policies wholesale or more selectively. For example we were interested to assess the influence of the EU Bangemann Report and the activities of the G7 group of countries, which played an active role, for example, in promoting and coordinating multimedia experiments), This part of the studies also sought to find some evidence of the consequences of these policies. Did public policy merely shape national rhetorics surrounding multimedia, or does it also affect patterns of investment in new multimedia infrastructures and service and even the social appropriation of multimedia. A review was undertaken of policy initiatives to promote multimedia creation and adoption. This examined the goals of these initiatives, the diverse concerns being pursued around multimedia, and the balance between these. This showed that multimedia and information society initiatives were geared towards a range of different goals: between economic development goals - for example promoting competitiveness and job creation and societal goals - for example regulations and policies geared to countering potential negative effects on privacy or social exclusion; fulfilling broader social objectives e.g. for public service and the quality of life or simply the need for democratic control over multimedia development. Different parts of the administration - for example, Education, Trade and Industry, local government - tended to have different priorities and concerns. As a result gulfs and contradictions could be noted between the policy perspectives of various public bodies, though the overarching responsibility for policy tended to be located with ministries concerned with trade and industry (Lobet-Maris with van Bastelaer and Cammaerts 1999). We were particularly concerned to examine how this affected the 84 conduct and goals of social experiments and public trials with multimedia (see section 4.6). The review also sought to characterise the means deployed by the state, which varied across a spectrum, and seemed to be changing. At one extreme was the dirigiste model of the state acting as a director of change (selecting and developing particular technologies; providing infrastructures; running projects [and the state is a major potential user of multimedia in its many service and administrative activities]). Such a model has deep roots - in interventionist approaches to science and technology, and economic development more generally, and has, for example, been a major feature of the telecommunications technology policy in most European countries until recent decade. This tradition has been particularly marked in France, where central-direction over strategic and high-profile technology projects has been given the title "High Tech Colbertism" - in reference to the successful industrialisation strategy carried out by its 18th Century premier (Cohen 1992). However even here a shift from dirigisme to pragmatism has been noted, linked to recent failures of large high-technology public policy initiatives, such as the 'cable plan', and the more widespread shift towards liberal approaches (Vedel 1997, LobetMaris with Bastelaer and Cammaerts 1999).28 In Japan, the state plays a no-less central role, organising choices in innovation and directing the activities of private ICT firms (Collinson 1999). Denmark also stresses the role of the public sector, both as a driver for multimedia development and, rather distinctively, as a vehicle for addressing broader social objectives about how multimedia will be deployed. Public sector intervention is seen as necessary in Denmark to offset potential negative effects of technology and ensure that it will be geared to particular social objectives which might otherwise be overlooked and also to ensure the involvement of citizens in decision-making over technology (Jaeger & Hansen 1999). However a range of less direct models and methods of intervention exist - and extend, at the other end of the spectrum, to the a situation where the state merely acts as a coordinator or catalyst of change - for example by facilitating and organising linkages and flows of knowledge between players from industry or technology institutes. In many countries, public intervention around multimedia technologies has veered towards the latter form, involving indirect methods, such as training and awareness initiatives, and providing resources (e.g. grants) to help other public or private bodies to carry out projects (Lobet-Maris with Bastelaer and Cammaerts 1999). Such a model would seem well-suited to a setting in which technological decision-making cannot be centralised within the administration. This may arise because choices in technology evolution depend upon the interaction between a range of players - some of which who may operate largely outside the realm of the state's powers (for example global suppliers) or upon the behaviours of widely dispersed consumers which are difficult to predict and organise. As Lobet Maris et al (1999) have noted, this is explicitly recognised in the Dutch 'Information Superhighway' policies are based upon recognition of the limited role a government NB the analysis of the French situation presented by van Bastelaar and Lobet-Maris in Williams and Slack (1999) preceded the actions and initiatives taken by the Jospin government. 28 85 can play in shaping the Information Society.29 Similarly, Brosveet and Sørensen (1999) have pointed to the shift in the model of state activity in Norway from planned development and growth of technological capabilities towards a more opportunistic model drawing on global developments, which they characterise as 'Fishing not farming'. Whereas 'farming' describes 1980s strategies of nations to grow technological capacities in core technological fields, 'fishing' implies a selective strategy, drawing on offerings found in that global market. These developments have to be seen in the light of the increasing influence of neo-liberal approaches and the globalisation of technology development. The language of 'de-regulation' should not, however, lead us to confuse this shift with the departure of the state from the field, nor a withering away in the role of the state.30 Indeed one of the key findings of this study has been of the central role of public administration (at local, national and international levels) in multimedia experiments and initiatives (Jaeger, Slack & Williams 2000). Within the European Union, the European Commission has become a major sponsor of trials and projects. Its role is particularly salient in countries such as Ireland with a modest indigenous development base (Kerr & Preston 1999). In carrying out comparisons, an important difference between different European states is in the structure of state activities - particularly regarding the significance of regional and local state structures and in the relationships/ allocation of responsibilities between local government, regional structures and national/ federal government functions. These differences in governance structures may be significant (and in particular in the extent to which there is a significant regional dimension to state activities), especially since technology development involves both national elements (e.g. the development of information infrastructures) and local elements (e.g. particular applications - such as, for example digital cities and community information systems). The review therefore sought to capture the involvement of the different levels of government. We were interested to address how the relative salience of different levels of state activity, and the allocation of functions between them might influence the effectiveness of state intervention. For example, might the dilution of responsibilities between different bodies, the lack of funds or power of key bodies; or the strength of laissez-faire ideologies, give rise to weaknesses or gulfs in the state's activity? There is a sense that the Federal governance systems in Belgium and Switzerland may have particular problems in developing and implementing technology policies for multimedia - for example where responsibility was delegated to regional structures, and overall national 29Thus the Dutch 1994 National Action Plan states that the government 'does not intend to produce a blueprint for the exact design and scope of the information superhighway (...) Only society, in all its diversity, can continually decide which way developments will go' Information Superhighway - From Metaphor to Action. Action programme, Part I (1.2. The government’s role), Dutch National Action Plan (1994), http:/ / info.minez.nl/ docs/ nap-en.htm 30Indeed one of the features of liberalisation of telecommunications and the privatisation of national telecommunications service monopoly providers has been a move from a mode of control based upon direct political/ administrative control towards a model in which private firms are subject to often extremely elaborate regulatory requirements (Mansell 1993). 86 coordination or vision were largely lacking (though the lack of a national technology policy in this area is not perceived as problematic within Switzerland) (van Bastelaer, Lobet-Maris & Pierson 1999, Rossel 1999). Finally, and in part following on from this last observation, we note that the longstanding differences that could be noted between states in their policy concerns and traditions, did not always seem to derive from the specificities of multimedia in that country. Not only were existing policy styles reflected within multimedia policies, particular national concerns could end up being projected into discussions of multimedia policy. For example concerns about the effectiveness of the Swiss and Belgian Federal systems were re-visited in their Information Society debates. Similarly we note the way in which concerns about intercommunal relations resurfaced in Belgium, In this sense, we could see multimedia serving as a kind of 'Rorschach Test' - a recipient of long-standing policy concerns. The relationship between the State and Industry Any assessment of the significance and impact of state intervention must address the reciprocal relationship between the state and industry (and other technology players such as research and education institutes). On the one hand the state is an important actor in establishing the terrain and setting the 'rules of the game' for multimedia developments (as already illustrated by the important example, already discussed, of the extent of competition in telecommunications). One the other hand, the policies adopted by the state, and the measures available to it for exerting influence, were shaped by the size and structure of the multimedia constituencies in these countries. The existence of powerful national champions able to influence the terms of state policy was an important factor. This factor however interacts with the established mode of operation of the state - particularly regarding whether it adopts an interventionist or more laissez-faire approach, and the balance it adopts between economic and social policy objectives - to create a range of possible situations (Lobet-Maris 1997, Lobet-Maris with Bastelaer and Cammaerts 1999). Not all of the possible outcomes that could be generated by such a schema were identified in our sample of countries. We could distinguish four particular situations amongst the nations studied: first are countries where there were dominant national champions (e.g. Deutsche Telekom and Siemens in Germany) which were at the centre of discussions of public policy. In the case of Germany this was coupled with an interventionist model to create what could be described as a corporatist framework; second, we find countries in where there is no clear pattern of dominance amongst the large and more pluralistic array of technology and media companies. In the case of the UK this is coupled with a laissez-faire policy tradition, to create what could be described as a highly competitive framework for multimedia innovation. In the case of Japan an array of competing corporations (more oligopolistic rather than pluralist) interact with a state that is highly interventionist in technology development strategies in what could be described as a collective development model; finally we find countries like Belgium, in which the weak development of local industrial players means that a state with an interventionist tradition may be 87 drawn into a wider role as a 'proxy' for the private sector (Lobet-Maris with Bastelaer and Cammaerts 1999). 4.6 The Review of Multimedia Pilots and Social Experiments An important part of the national studies was a review of the main multimedia experiments and trials in that nation/ region. This has been published and analysed elsewhere (Jaeger Slack and Williams 2000). This article reviews some of the specific features of these experiments. Here we will simply make a number of observations about the distribution of the sample of experiments before going on to draw out some overall observations about the character of Multimedia Pilots and Social Experiments as a whole today. We have categorised all the pilots and experiments from each national review in terms of the area and type of activity they involved. The results of this exercise are shown in Table 4.1. Table 4.1 The Distribution of Multimedia Pilots and Social Experiments31 Categories BE CH D DK IRE NL NO UK Total % 1. Public Admin.32 6 2 4 8 0 6 4 2 32 18 2. Health33 3 2 0 0 0 2 0 0 7 4 3. Culture and Media34 3 4 5 3 7 2 2 2 28 16 4. Education35 6 3 6 6 6 4 6 5 42 21 5. Economic Development36 3 4 8 5 2 2 0 4 28 15 6. Infrastructure37 4 5 7 5 0 1 3 3 28 16 7. E-Commerce38 0 5 4 5 0 2 2 1 19 11 31This table is adapted from Jaeger, Slack and Williams 1999, and particular thanks are due to Roger Slack for compiling this information. 32 Citizen information, administration of government and communication with representatives. 33 Health information and promotion. 34 Includes both cultural heritage and experiments with, for example, interactive television. 35 Electronic classrooms, CD-ROM’s and the like, connections between schools and other education establishments. 36 Includes national, regional, local projects. 37 MAN’s and the like. 38 Teleshopping, tourism and the like. 88 Total 25 25 34 32 15 19 17 17 184 101 18 of the 146 separate projects covered in the survey were ‘hybrid’ projects, concerned with more than one area of activity. Most of these (12/ 18) were technology infrastructure projects (typically Metropolitan Area Networks), which also involved trials and initiatives in specific activities such as Public Information, Education and Economic Development. To simplify comparison, these compound hybrid projects have been disaggregated, and each element has been counted separately, bringing the sample size up to 184. The distribution of these multimedia pilots and experiments between different areas of activity is shown in Table 4.1. This reveals that the vast majority of projects fall within a small number of finite areas. Projects geared towards Education, Public Administration, Culture/ Media, Technology Infrastructure provision, local Economic Development and Commerce accounted for 97% of our sample. The overall distribution is largely in line with that found by the European Survey of the Information Society projects.39 The largest single group, comprising 42 experiments and pilots, and accounting for over one fifth of the total (21%), were concerned with Education. These included ‘electronic classrooms’, connections between schools and other establishments, and CD-ROMs, and were mainly, but not exclusively, public sector initiatives. These were motivated by a range of goals - a view that education was per se ‘a good thing’, and that technology could improve the quality of education and reduce the costs of its provisions. Further there was a view that these initiatives would promote skill development in general and IT literacy in particular and that these initiatives could contribute to broader policy goals by, for example, helping to combat social exclusion or stimulating local economic development. The next largest group, of 32 projects, accounting for nearly a fifth of the projects surveyed (18%), was Public Administration - comprising mainly public information systems, together with some systems more narrowly targeted on the administration of government (i.e. internal communication within the administration and communication with representatives). The public information systems diverged in their organisation and goals along a spectrum between two poles: at one end were citizen information systems, where new ICTs were seen as an efficient means of information provision for citizens and which seemed often to have been conceived as simply another means of dissemination in electronic form of versions of paper leaflets and notices currently circulated by the administration. At the other end was the more bottom-up development of electronic communities, in which ICT was seen as providing, through electronic means, close linkages and exchanges of information within a community (particularly in relation to socially fragmented or physically dispersed communities).40 (http:/ / www.ispo.cec.be/ esis/ December 1997). Public information systems are evolving and it is not clear whether the distinction between these two types will be maintained. For example, citizen information systems that were established on a top-down basis may subsequently form the nucleus of broader and more pluralistic community 39 40 89 One related area, where we might have expected to find a significant number of projects, is that of Health care. Indeed it was somewhat surprising that the survey only yielded information (and often only schematic information) on seven healthoriented projects (4% of the total) that were underway in Belgium and the Netherlands. We are aware, from other sources, of a number Health-related multimedia initiatives that have been established in Europe. However these are not necessarily captured within this survey (for example where work is still primarily developmental it may not involve any public trials). The low level of health projects might thus be an artefact of the timing of the survey or an error from small sample sizes. This, interestingly, was the one area in which our findings differed from the European Survey of Information Society projects, which showed a somewhat higher proportion (16% of the total) of health-related projects. 41 The other significant groups of projects were concerned with: Culture and Media - 28 projects (16% of total) including both ‘cultural heritage’ projects, such as electronic museums (notably in Ireland), and experiments with, for instance, interactive television (notably in Germany and the UK). The former were typically publicly-funded and the latter privately. Economic Development - for example attempts to use multimedia to promote regional economic development (‘smart regions’). There were 28 such projects (16% of the total), spanning local, regional and national levels. Electronic Commerce - a smaller group of 19 projects (11% of the total) were concerned with Teleshopping and electronic marketing of Tourism facilities etc. When comparing the profile of experiments in the various countries surveyed, some areas of broad similarity can be noted - in particular, the across the board attention to Education. More generally the coverage of initiatives was rather similar between countries in terms of the range of topics and the relative attention to each. Differences can however be noted between countries, though the small numbers involved necessitate considerable caution in how to interpret them. For example, as already noted, the significant representation of Ireland amongst the Culture and Media projects (7 out of 28 such projects), which can be linked to the importance of tourism and cultural activities to the Irish economy. However it may also in part reflect the familiarity of the Irish SLIM research team with projects in this area, and their effectiveness in locating a more complete sample of cultural than other kinds kind of project, as well as the absolute frequency of such projects and the relative lack of other initiatives in Ireland at the time of the survey.42 information systems, and vice versa. These issues are explored in more detail in the reports of the Integrated Study on Digital Cities (Lobet and van Bastelaar 1999). 41 (http:/ / www.ispo.cec.be/ esis/ December 1997) 42 Since the survey was undertaken, the Irish PTT has announced a number of relatively large-scale trials with advanced telematic services, including a 'wired city' initiative. 90 There was some evidence, in this survey of projects, of differences between countries in the relative level of attention given to economic goals (local economic development and commerce), which were particularly noticeable in the Danish, German and Swiss samples, and public goals (public administration and health), in which Belgium, Denmark and Netherlands all figured most prominently (followed to a lesser extent by Germany and Norway). Taken together this suggests a divergence in the relative emphasis in multimedia projects between public and economic goals. Although most countries were engaged with both sets of goals there was some indication of a trade-off between the two. 4.7 How the National Context shaped understandings of Multimedia The national studies had a number of particular concerns and foci. At the most general level the studies were concerned to look at general concepts and understandings of multimedia that were emerging - and in particular to establish whether there was a dominant image of multimedia within a country (or whether a number of competing conceptions prevailed), and what this was in terms of the concepts of how the technology infrastructure was to be configured, the kinds of application that would run upon them; the conception of who would be the users and how the technology would be used - and thus its social implications? When this study was being launched in 1996, a point of sharp debate in this area concerned the metaphor and organising principles that were to be applied in relation to the delivery system/ platform for mass market multimedia applications particular into the home - where a dichotomy had been suggested between views of multimedia as an extension of the world-wide web (i.e. based on a networked, multimedia-capable personal computer) or whether it was seen as an extension of the set-top box for digital TV. These contrasting technical models - world-wide-web versus interactive TV - were associated, more or less implicitly, with contrasting visions (or more precisely suites of visions) of the use and the user of multimedia. These different sets of views were associated with the influence and involvement of different sectoral actors; with different concepts of the uses and utility of multimedia; and, with sharply contrasting moral visions of the information society. Thus the set-top box, with its limited technical interactivity (e.g. if operated by a TV remote control device rather than a keyboard and with its limited 'back channel') seemed to place the user in a passive consumer role - ordering pizza from your armchair - in contrast to the socially valorised activities of communication and information search through the personal computer linked up to the internet. Today this dichotomy seems less clear than when the SLIM research project was being developed. Fewer people believe that a single platform will emerge for multimedia in the house. Instead we will continue to have a range of different devices for particular purposes (as we already do, for example, with consumer electronic, where TV, radio and Personal Stereo continue as distinct products despite the technical possibility of integrating them. Moreover, the dichotomy 91 between these routes has become blurred as key players in each market have sought to open up their offerings to the alternative option. Thus, for example, interactive TV is being proposed as a point of access, rather than an alternative, to the web (Stewart 1999 ITV paper). In some countries, our surveys were able to identify relatively clear national images of multimedia - and conceptions that could be related to the relative strengths of these apparently counterposed concepts. Notably, in Germany, digital TV was placed centre stage - a development that may be related to the way the German multimedia landscape is dominated by two large media conglomerates - the KirchGroup and Bertelsmann - and Deutsche Telekom (Kubicek, Schmid & Beckert 1999). In Norway in contrast, the 'internet model' prevails in a context in which many people already have access to the web from their homes (Brosveet & Sørensen 1999). Norway is rather open to the technological visions emerging from the USA and, apart from Telenor, there are not powerful national champions that could shape national perceptions of multimedia around their particular offerings. However, in most countries, a range of models of multimedia and its uses seemed to co-exist, ranging from established CD-ROM technologies, to the emerging technologies of the internet and world-wide-web. Our studies yielded some rather interesting differences in emphasis of national images on these two technologies between, for example, Denmark, in which multimedia has been primarily understood in terms of CD ROM and a country like the Netherlands in which CD technology is today taken for granted 43. In Ireland both CD ROM and internet figure while in Belgium and France, attention is mainly focused upon the internet. In the UK and Netherlands we see markedly pluralistic conceptions of multimedia. It could be misleading to seek to draw too much inference from these accounts of national image - which may be influenced by recent public relations activities and initiatives (for example, in Switzerland, the idea of multimedia is associated with the recent high profile opening of the new Olympic Museum). In most cases the prevailing conceptions of multimedia are poorly bounded and defined, and rather heterogeneous. We would seek to draw rather different conclusions from these complex and confused national patterns in terms of the perception of multimedia. First the predicted convergence of information and communication technologies has not really materialised. The world-wide web is showing epidemic growth rates across all countries surveyed (though there are important differences on penetration and growth rates). However, beyond this, technology pundits are still awaiting the 'killer application' that will bring multimedia into the home. The search to identify national models/ images would imply some level of resolution of the lines of development of multimedia technologies and uses around relatively clear models (which could be global models or specific national ideal Perhaps because it was here that CD technology was developed by Philips jointly with Sony in Japan). 43 92 types). In some circumstances, of course, particular national players may be strong enough, or associations amongst players may have sufficient coherence and influence, to shape overarching views. Germany provides one such instance. France has provided another, for example in the case of Minitel, though there has been some rethinking of the desirability and feasibility of such strategic statesponsored initiatives following the difficulties with the French Cable Plan (LobetMaris with van Bastelaer and Cammaerts 1999). However, in most contexts multimedia refers to an extremely diverse, dynamic and turbulent set of developments. It involves a wide range of players, involved in content provision and application domains as well as diverse technology supply. In such a situation it seems implausible that specific models will come to prevail. This could, of course, be taken to reflect that fact that we are a transitional situation in a context in which 'technological closure' has not yet taken place 4.8 Conclusions A substantial set of empirical findings has been amassed through the national studies of multimedia uptake. There are extensive opportunities for more detailed cross-national comparative analysis of these findings. Some of this work has been published elsewhere (Jaeger, Slack & Williams 2000, forthcoming, Kubicek et al forthcoming). Some immediate conclusions from this inquiry can however be briefly stated. First is that there are some clear differences in the way that multimedia has been conceived in different countries - both regarding the extent to which dominant conceptions of multimedia products and services have emerged, and in the nature of those conceptions. Second, these differences can be related to the diverse social, historical institutional and technological settings for the development and use of multimedia within different nations and regions, and the arrays of players involved. There is not, however a simple relationship between particular factors and actors and the pattern of multimedia developments. This comparative review highlights the complexity of interactions, and the way that particular influences may be mediated by other historical factors. This could be illustrated by the way that the influence of the size of linguistic groups depended upon how this factor might be taken up within political concerns. In this content, it seems rather difficult - and indeed unhelpful to attempt to produce a simple characterisation and explanation of national patterns of technological change in multimedia. These observations serve however to underline the shortcomings of accounts which see technology supply as a universalising force which cuts across and gradually homogenises different social settings. There were, of course, some largely common technological developments across countries - the core multimedia technologies are after all largely available at a global level. However, the anticipated convergence 93 around global technological models has not taken place. Instead, multimedia technologies emerge as complex assemblages of global and local elements as generic ICT capabilities are appropriated and incorporated to particular purposes within particular social settings. One important 'differentiator' arises through path dependencies (primarily arising as a result of 'sunk investment' particularly by final consumers). This was illustrated in the case of France, in which the earlier success of Teletel/ Minitel - a large installed base of terminals and an established information market - constitutes a key factor shaping future multimedia developments, for example holding back the uptake of PCs and the internet (van Bastelaer with LobetMaris 1999). The different outcomes of Minitel in France and Germany (with the latter able to migrate more readily to the internet as an unintended consequence of the choice of an industry standard PCs, rather than a dedicated terminal, as the domestic Minitel terminal [Schneider 1999]) highlights the potential influence of particular contingencies. Third though there does seem to have been some evidence of rhetorical convergence at the level of public policy discourses, this has only been partial. Some elements of technology policy recur: conceptions of the capability of the new multimedia systems that will emerge; the economic and social benefits that will accrue from their adoption; and the competitive dangers of being 'left out'. Common patterns can also be detected in the areas and activities in which public-funded multimedia initiatives and experiments have been set up (for example in education and public information) (Jaeger, Slack & Williams 2000). However, examination also reveals continued differences in historical national policy styles and traditions - notably regarding the role of the state, between interventionist and more laissez-faire models. There is some evidence that the broader preoccupations of particular nations may be taken up and projected within multimedia and information society policies. These points reinforce the relevance of the social shaping of technology perspective to understanding how multimedia technologies may be appropriated differently in particular national and regional settings. We have tried to map out these factors here. More detailed comparative analysis will be needed to capture the dynamics of this process - and the way in which regional and national settings shape technological innovation and policymaking. The national studies, presented in Williams and Slack (1999), provide an important resource for researchers and decision-makers. They draw attention to both differences and similarities in the perceptions of emerging multimedia technologies by a range of actors in different national settings. What is most striking is the diversity of these responses to an ostensibly global technology, both in terms of images of multimedia technology and approaches to multimedia policy. Cross-national research has a particular value here, as it allows us to open up for critical examination precisely those things which are taken for granted within particular societies and cultures. By drawing attention to the range of possible approaches in public policies and technology development strategies for multimedia we hope to contribute to more informed discussion and debate. The other goal of this phase of 94 the SLIM project was to provide a foundation for the main empirical stage of SLIM research (phase III, the integrated studies of multimedia development and use) by mapping and analysing the influence of different national settings. It is to this part of the research that we now turn. 95 Chapter 5 What do we mean by Social Learning? a) What is Social Learning? What do we mean by Social Learning? The chapter lays out our findings about the processes of social learning. In the first part we define social learning and explore how, in the course of this research project, we have developed our conceptualisation of social learning to achieve a more detailed and nuanced understanding of the process of technological innovation. We next seek to map out some of the different ways in which innovation may take place to identify strategic options in relation to social learning. The final section looks at the how social learning is conducted and organised. Our definition of social learning at the outset of the study focused upon two elements – to which a third was added i) the ways in which technologies are designed and developed (particularly regarding the ways in which the expected users and uses are represented in design); ii) how these offerings are taken up within a society through both the implementation/ consumption and of particular artefacts (highlighting two closely related social learning processes of innofusion and domestication); and, finally iii) the development of the broader context of regulations, policies, expectations and institutions that can be said to set the 'rules of the game' for multimedia innovation and constitute the 'technology regime' (Rip 1995). We use the concept of social learning to provide tools for addressing the broad spaces in which technological artefacts and systems are technically and culturally appropriated and made part of (or rejected by) particular social systems and settings (Rip et al 1995). Fleck (1988) argued that the 'implementation arena' was a laboratory for the further innovation of industrial artefacts emerging from technology supply. The concept of social learning takes this further and suggests that we should see wider ‘Society’ as a laboratory for learning about the generic technical capabilities (Herbold 1995). Put thus way, the term ‘social learning’ sounds eminently laudable; however we do not imply that the processes are smooth, homogeneous and consensual. We express caution regarding the interpretation of the term learning. Social learning is not for us a psychological construct. We are seeking to understand processes of sociotechnical change, not as a narrowly cognitive process, but as a process of negotiation and interaction between different players and thus subject to conflicts, and differences of power and interest. (For a more detailed discussion see Sørensen 97 [1997], on whose exegesis, this section draws heavily). Social learning, thus conceived, seeks to provide an inclusive analytical framework. For example, as outlined above, it encompasses cycles of development and consumption of technology as well as the bringing into being of the broader circumstances needed for the technology. We explore below the intellectual debt to the ideas emerging from economic history and evolutionary economics of 'learning by doing' and of 'supplier-user interaction'. However, social learning is much more than this. As Sørensen (1996) puts it: Social learning can be characterised as a combined act of discovery and analysis, of understanding and giving meaning, and of tinkering and the development of routines. In order to make an artefact work, it has to be placed, spatially, temporally, and conceptually. It has to be fitted into the existing, heterogeneous networks of machines, systems, routines, and culture. Thus, to analyse social learning we have to go beyond the narrow instrumental understandings of economists and address issues of meaning and identity as well (Sørensen 1994). The importance of this symbolic effort - to make technology meaningful - is of obvious relevance when we come to look at the domestication and cultural appropriation of novel technologies. Similarly, social learning is not restricted to the operation of the learning economy, but extends into the efforts of various players associated with multimedia to regulate and provide order in and reshape the broader context. We have characterised this as learning by regulating. It involves both of the main types of social learning we have identified: both learning by doing in the 'trial and error' process of developing their own strategies and orientations around technology and learning by interacting in their efforts to align and regulate the behaviour of other players. We are seeking to develop an analytical perspective that pulls together elements from approaches with have hitherto been not well integrated (for example technology development and consumption; cultural and material aspects). We are, moreover trying to do this in a context in which the particular terms have been used in often rather inconsistent ways. We therefore start our analysis by exploring the origins and definition of these different elements. b) Origins of the concept of Social Learning: learning by doing; learning by interacting learning by doing Economists and economic historians became interested in social learning through studies of productivity that showed continuous improvements over long periods of time without any investments in new technology. Arrow (1962) calls this phenomenon 'learning by doing'. The idea is that workers, individually as well as 98 collectively, develop more efficient ways of employing machinery through their experience from usage. This kind of 'learning curve' effect is well-known. Though often taken to be limited to the initial "introduction" of new equipment, it is now recognised that learning by doing may improve the efficiency of production over a very long period of time. A related phenomenon is 'learning by using'. Rosenberg (1982) suggests this concept to describe the process through which a user gains familiarity with a given piece of technology and develops skills in making use of it. These concepts point to the fact that the properties of a technology (its affordances and limitations) may not be immediately apparent, but are discovered and learnt through experience, often in relation to particular productive processes and activities. It highlights the development of knowledge that is contingent and often tacit (Williams, Faulkner and Fleck 1998). Such learning by doing provides a potentially very important source of information on the effective use of a technology and thus about how to make process innovations. Similarly, by giving suppliers access to what users have learnt about their products and what deficiencies and potentialities they have discovered, it could provide invaluable information for product innovation. This underlines the importance of the linkages between users and producers. To innovate successfully, producers may depend critically on information from users, and vice versa. This is the basis of the idea of learning-by-interacting and the learning economy (Andersen & Lundvall 1988). learning by interacting - the learning economy The idea of learning by interacting has emerged importantly amongst evolutionary approaches. Evolutionary models of development of technology have a long history in technology studies (see, e.g. Gilfillan 1970). They sought to understand the complex interaction between technology supply and use by applying evolutionary metaphors of the generation of variation and selection (see for example, Schneider 2000 forthcoming). Freeman’s classic work (Freeman 1994) highlighted the importance of supplier-user coupling for successful innovation. Some stability in inter-firm relations, as well as a supportive institutional environment, may be needed to provide the necessary preconditions to maintain such linkages in the supply chain. Evolutionary accounts have analysed these linkages as constituting a learning economy. In seeking to understand differences in the regional and sectoral performance of economies, Andersen & Lundvall (1988) emphasised the institutionalisation of such linkages through the development of channels of communication, shared codes of conduct, and conceptualisations between suppliers and users. The operation of the learning economy is shaped by the systemic institutional features and setting of a given regional or national economy, including public policy. The concept of the learning economy calls into question simplistic beliefs in the market as a mechanism of communication as it demands a greater stability of 99 economic relations and more developed patterns of communication than those held up by idealised market forces alone (Sørensen 1996). There may also be need for new institutional arrangements in which governments - locally, nationally, and even supranationally - may play a crucial role. OECD countries are beginning to take up the policy challenge of developing instruments to facilitate such links and provide better institutions and instruments for communication. Supplier-user communication can be relatively straightforward in relation to industrial technologies, where the high value of transactions may motivate collaboration and there are direct links between supplier and consumer, and where both supplier and consumer share relatively high levels of technical skills as well as presumptions about the values associated with technology. Supplier-user relations, conversely, can be expected to be more problematic in the case of mass-produced consumer goods, where there is little scope for direct engagement between suppliers and the dispersed mass of anonymous users who are only indirectly linked to the supplier through the market. To return to Hirschman's (1970) frame of reference, the consumer has - in particular in mass markets - only a choice between exit and loyalty, and thus very limited possibilities of 'voice' - communication with producers/ designers. Particularly in relation to novel products where there is no established customer-base (where indeed the customers do not even exist yet!), the supplier may have little understanding of the user and the use setting. The supplier is thus forced to play a major role in prefiguring, or indeed constructing, the customer and the market, and must find some way of modeling user requirements. Fortunately, many companies recognise this and make efforts to reproduce userproducer relations, at least on a semi-permanent basis for example, by conducting market research on 'proxy users' (e.g. panels of potential customers) or through discussions with intermediate users (eg retailers of the goods and services being produced). It may also be necessary to enroll a range of other actors, such as competing suppliers and suppliers of complementary products (Collinson 1993). This remains a rather difficult and uncertain process for many suppliers particularly in relation to domestic technologies consumed in the private sphere of the household (Silverstone 1991). There may be an important role here for public policy. Evolutionary economists have been concerned to characterise and explain the differential performance and outcomes of these national or regional innovation systems (eg Andersen & Lundvall 1988). What is more interesting for our purposes, however, is to understand the process of innovation: how learning by interacting takes place in relation to particular technologies and in relation to particular groups of actors. Scholars within the field of Social Shaping of Technology thus sought to draw different insights from these same coupling processes – through detailed studies of technological innovation. Thus Fleck’s (1988) concept of innofusion explored the ways in which artefacts emerging from technology supply were transformed and further innovated as they were implemented and in the struggle to get them to work and be productive under actual circumstances of use. He highlighted the possibility 100 of feedback from implementation to future technological supply – a possibility which depends critically upon kinds of linkages and channels of communication between supplier and user which make up the learning economy. The concept of innofusion was developed to account for innovation in industrial technologies. It provides an important tool for analysing social learning processes in the adaptation and transformation in use of technical artefacts. The concept of innofusion does not per se, provide an adequate framework for analysing then ways in which artefacts become incorporated within practices and acquire their meanings and significance. To address these dimensions we have turned to work from cultural studies and related perspectives into the consumption of technologies in everyday life. c) Developing the Social Learning framework: appropriation : domestication and learning by regulating the appropriation framework: the domestication of technology We are seeking to develop our framework to analyse social learning processes around the application, consumption and use of technologies. Our aim is to be able to analyse in greater detail the nature of "learning by using". What is the scope of such learning, what is achieved, how does it affect the outcomes of technology and how may such insights be helpful to the design of better technology? Here we draw from anthropological or ethnographic perspectives the concept of 'appropriation of technology' (Pacey 1983, du Guy at al. 1997). The concept of appropriation reminds us that the adoption, consumption and use of artefacts is an active process (in contrast to the passivity implied by traditional terminologies such as diffusion); it highlights the degrees of freedom remaining to users and other players outside technology supply in exercising choices around the selection and local deployment of artefacts and it emphasises the symbolic as well as instrumental aspects of these processes.44 In particular, we emphasise that technology has to become encultured (or embedded) in order to function (Silverstone & Hirsch 1992, McCracken 1988, Lie & Sørensen 1996). Similarly, media and consumption studies focus on the symbolic aspects of 44 In some senses the concept of appropriation is a relative term, that can be applied at different levels, including, for example, how a country reacts to the availability of a set of global technologies. At the level of particular artefacts, we can use the term appropriation to explore both: i) the ways in which generic ICT capabilities are applied and incorporated into final multimedia products and services intended for particular social settings (the home, the workplace, particular regions and nations etc.) and ii) the related processes by which these products and services are incorporated into particular uses in these settings. Where we are seeking to refer specifically to the former we may use the term 'intermediate appropriation', insofar as it concerns what industrial economists would describe as intermediate consumers. The latter, we could describe as 'final appropriation', though this is perhaps the default meaning when the term appropriation used in cultural and technology studies. 101 technology and the need to analyse how artefacts acquire their meanings. For example, McCracken (1988) draws attention to the many rituals which people employ to give meaning to material objects. Without meaning, artefacts remain artificial and alien. Work has to be performed to integrate a technology into a culture. This is important for our understanding above all in that it emphasises that social learning involves the creation of meaning as well as practical efforts to make technology work. The interpretation of an innovation is of great importance to its eventual success or failure. The reinterpretation of the camera from a professional to an amateur artefact (Jenkins 1975) or of the VCR from a tool for broadcasting companies to a household appliance (Roosenbloom 1982), extended in a dramatic fashion the market for these products. The future role of multimedia may depend on the public understanding of this technology in terms of moral acceptability (is it harmful to children?, is it mainly "toys for the boys"?, etc.). 45 Domestication We were tempted to make ‘appropriation’ the central concept in our analysis of social learning. However there is no general theory of appropriation, and the term appropriation has been used in rather different ways by authors from various disciplinary backgrounds and purposes.46 No dominant definition or methodology for analysing it has yet emerged – or seems in prospect. We therefore turn instead to the concept of domestication to provide a more well-defined approach to the study of appropriation of technology.47 The concept of domestication emerged from a series of important studies which sought to understand the appropriation of artefacts in the specific social setting of the home . Silverstone, Hirsch & Morley (1992) use this term to describe how 45 Meaning is projected internally as well as externally. The objects we use, should be compatible with our personal identity as well as with the perception of self that we want to convey to an outside world. This is one of the reasons why some technologies are met with greater enthusiasm from some groups of people than other. Reactions towards new technologies that are often labeled "resistance", may on closer examination prove to be rooted in this dynamic of meaning: The common interpretation of the innovation is not in accordance with our self identity. 46 For example it is used both to refer to specific processes of incorporation of artefacts within particular social settings and to the ways in which generic technological capabilities are incorporated into products and cultures (eg du Guy et al. 1997). Thus cultural studies' analyses of appropriation tend to focus rather singularly upon the symbolic aspects (see for example du Gay et al. 1977), while work from technology studies retains a slightly instrumentalist bias upon the concrete use of the artefact (MacKay and Gillespie 1992)). 47 The domestication concept provides some interesting analytic elements which feel able to build upon. At the same time it has less ‘intellectual baggage’ than the more established terminology. For ease of exegesis we adopt a simple schema here which sees innofusion and domestication as both being important processes in the appropriation of technological artefacts. However usages of the terms domestication and appropriation are various, geared towards different purposes and contexts around which they are deployed – and we are not arguing for or expecting a harmonisation of usage. 102 artefacts are integrated into what they call "the moral economy of the household". They highlight the interaction between economic and symbolic transactions within the household and between the household and the outside world. Moral economies are negotiated spaces. To domesticate an artefact is to negotiate its meaning and practice. Lie and Sørensen (1998 or 1996) define domestication as a kind of taming process: "Metaphorically speaking, we tame the technologies that surround us in everyday life." Domestication shows the need to "tame" the facts and artefacts that are taken from a "wild" outside world. Thus, domestication is a way of theorising what the cultural appropriation of technology is all about.48 Basically, domestication is necessary both to make artefacts work and to make sense. Both action and meaning are important. Artefacts have to be: * acquired, either bought or in some other way made accessible. * placed which means that they are situated in a physical, symbolic and mental space. * interpreted to be given meaning within the household or a similar local context of identity as well as symbolic value to the outside world. * integrated into social practices of action (Silverstone, Hirsch & Morley 1992, Lie & Sørensen 1996, Sørensen, Aune and Hatling, 1996). Strategies of domestication thus involve the following three main dimensions: (1) practical, (2) symbolic, and (3) cognitive. In the practical dimension, domestication brings forward a pattern of usage. How shall the artefact be employed? What are the practical implications of a given set of knowledge? Symbolic efforts are about production of meaning and the relationship between meaning, identity, and the public presentation of self. The cognitive work is related to learning about the artefact or the intellectual appropriation of new knowledge. A second important feature of the domestication is that it reminds us of some of the trivial challenges that have to be solved in order to make things work. In particular, it points to the necessity of constructing routines - including working out the tasks that have to be performed and the skills necessary to do these tasks and an appropriate division of labour - for the smooth and successful implementation and operation of an artefact (Suchman 1987, see also Nelson & Winter 1982).This does not mean that these routines are inscribed in the artefact, just waiting to be discovered. People construct different routines and different meanings from the same artefact. Of course, these processes of construction are contingent, and may be 48 We are moving the concept of domestication away from its initial association with domestic technologies in the home – it is about taming rather than the household as an arena. Indeed we can see domestication being carried out not just by individuals or households but also by institutions and other collectives, even nations. 103 affected by class, ethnicity, gender, and age. Studies (for example of the ostensibly straightforward task of making use of a PC [Aune 1996, Sørensen, Aune and Hatling 1996]) have emphasised the diversity of learning strategies as well as formal training, noting the importance of practical activities of experimentation pointing to the identity between using and learning, and tinkering where you learn in a pragmatic manner what you need to know in order to perform necessary tasks. Learning by trial-and-error is of general importance, but in addition users get information and knowledge from each other and through a host of different channels. The learning economy observable on the macro level has its parallel in local learning communities. However, social learning is not just the acquisition of skills and knowledge. When we emphasise domestication, we highlight the construction and reconstruction of culture as old and new combinations of artefacts, skills, knowledge, and social relations. This means that users are no longer seen as passive consumers, fed by industry. In line with the idea of "active audiences" from media studies (e.g. Silverstone 1994), users are perceived as struggling with the socio-material relationships that they encounter in their everyday life. Domestication is a way of conceptualising this struggle. To sum up, we use the term domestication specifically to refer to the way in which technical capabilities are explored, meanings attributed and practices developed as artefacts are integrated within local social settings. Appropriation and domestication processes are thus crucial to the successful uptake of new multimedia products - it is here that presumptions about the use and utility embedded in a designed artefact are tested, in which user acceptance or rejection of new offerings takes place and in which the uses and significance of technologies are finally determined. Technology regimes and learning by regulating So far we have focused upon the activities of actors directly involved in technology supply and consumption. However, social learning is not restricted to this 'learning economy', but also encompasses the activity of public policymakers, as well as promoters and other players in civil society in setting the 'rules of the game'. We have drawn here upon the concept of ‘technology regimes’ developed by Rip and co-workers(Rip 1995, Rip and Schot 1999) to capture the visions and sets of rules that help to shape the ways in which new technologies are developed and applied . These rules thus provide a broader framework and resources for local design/ appropriation activities. Finally this 'learning by regulating' (Sørensen 1996) also includes the attempts by policymakers to find mechanisms of influence and control that are appropriate to this evolving sociotechnical terrain. We can distinguish (at least) two generic types of learning by regulating: i) first (and conceptually precedent) is the creation of mobilising metaphors and visions which inspire the development of new technological systems; ii) second is closer to the more conventional use of the term regulation in the 104 sense of setting the rules of the game, including the activities of public regulators, (including legislators, promoters and other policymakers) and other players (including pundits, and the activities of major industry and technology players). Mobilising metaphors for technology regimes Over the last decade we have seen the articulation of powerful metaphors in the field of ICT - for example of The Information Society; Information Superhighways/ Infrastructure. These serve to organise, give direction, mobilise resources and set norms for programmes of technological change. As we saw in Chapter 4, public policies provide an important impetus in these developments. For example Information Society policies have been adopted by all the industrialised and many industrialising countries (Dutton 1996, West 1996, Williams 1999). There is however a close interaction between public and private players - with technology suppliers, severally or in combination, projecting their visions of the particular technological routes to be adopted and the kinds of economic and social benefits that can be expected. In a context of technological dynamism and uncertainty such visions hold out the prospect of simplifying the choices for intermediate and final users and others by aligning the strategies of a range of suppliers of complementary products. On the other hand an important feature of these metaphors seems to be that they are generic or generalisable enough to be applicable in a range of situations, irrespective of the diversity of contexts, and their scope for flexibility in interpretation. Their ability to work in this way seems to be underpinned by richness and heterogeneity of these metaphors - they articulate generic elements which many different actors can 'buy-in' to, offering lots of ‘hooks’ onto which local actors can project meaning. In this sense mobilising visions and metaphors do not resolve uncertainty, but provide a meeting place for different communities (there are important inputs here from parallel concepts from interactional sociology and the sociology of science and technology such as 'boundary objects' (Star & Griesemer 1989) and the recently propounded concept of 'trading spaces' [Gallison 1997). We can thus see the deployment of metaphors as operating at a number of levels in shaping expectations around multimedia: i) ii) iii) at the level of particular concepts of technology application and use (and we explore this in the next chapter, in relation to the promulgation of the idea of Digital Cities as an exemplar of public information systems; In the attempts by commercial players and promoters to align expectations of others in the market about the future configuration of ICTs. We have advanced the concept of ‘poles of attraction’ to explore this. As this is a novel concept that arose in the course of the SLIM study we describe it briefly below. Overarching views and policy pronouncements about the general character and social implications of ICT. We have already noted the way in which the 105 idea of an Information Society has been embraced by public policy in very different countries ranging from the USA to Malaysia. Poles of attraction: uncertainty and entrenchment in technology development In Chapter 2 we examined the process of entrenchment, and the attendant problems and uncertainties that beset attempts to create new, or change existing, complex technological systems such as ICT. These are particularly acute for what have been described as 'network technologies', whose functioning and utility depend not on the individual artefact but on how the artefact is integrated into a wider network of inter-operating elements (for example in relation to new 'delivery systems' and platforms), and which are subject to marked path dependencies and scale effects. Though in some circumstances a firm might seek to develop and launch new technologies on its own, escalating development costs, globalisation and pressures for standardisation of many ICT systems mean that technology development has an increasingly collective character. Various forms of collaboration emerge as an attempt to share costs and risks - and reduce uncertainties by foreclosing options in advance - rather than incur the costs of fighting out standards wars in the marketplace. These dilemmas had stimulated suppliers to seek to collaborate with other players (competitors, suppliers of complementary products, intermediate and final consumers) in developing standards and building markets. Since the commercial viability of a new technology may rely on convincing a critical mass of players to invest in it, such alignments becomes crucial to the prospects of success of a technology. Technology development increasingly takes the form of a networking activity, involving inter-organisational linkages and modes of engagement between various suppliers and 'users' or their proxies. Firms take part in an ever widening array of public and semi-public fora (Jakobs and Williams 1999). Paradoxically these attempts to align co-producers and enroll future users in building markets and entrenching new technologies set into train 'risk management' strategies that compound the uncertainty and indeterminacy surrounding technological development. These efforts of various players to manoeuvre to maximise their prospects, hedge their bets and foreclose technology choices end up by imparting a higher degree of indeterminacy and apparent fluidity to development processes. Indeed, some of the key decisions affecting the future prospects of a technology may be taken in the 'virtual space' of standards-setting committees and of industry fora, where key players seek to align expectations around their particular offerings. The term "poles of attraction" (Stewart forthcoming) was coined to describe the way in which metaphors and visions of future technological systems and services may be advanced to try and align the behaviour of other players (attract customers, enlist collaborators, deter competitors). Suppliers and other interested players may propose particular conceptions of future ICTs - to moot possible support: to orient and win commitments from potential suppliers of complementary products; to inform customer expectations, to ward off competitors and more generally to test 106 out and shape ideas about technological futures. Recent examples include the espousal of the 'network computer' as a solution to platform harmonisation problems in distributed business computing, and a challenge to an established pole in the shape of the IBM clone personal computer and the MicroSoft Windows environment. Other salient current instances are provided by the concept of Information Superhighways (Kubicek et al. 1997) which, it is argued, has been overtaken by (changing and sometimes contested) models of the Internet (Thomas & Wyatt 1999). The concept of poles of attraction draws attention to the way in which options may gain support and momentum - by aligning expectations and winning commitments and investments in particular technological routes – and may become entrenched and materialised in technologies and institutions.. Equally commitments and alignments can very rapidly evaporate. Proffered poles of attraction may fail to win commitments. Even where some momentum has been established this may be reversed and the option may lose support and fall back into the fermenting brew of emerging technologies. In this way the future of ICT is being to some extent fought out in advance in a virtual space. Poles of attraction, thus conceived, are complex entities. We can see close links between this concept and the idea discussed above of technological metaphors - not least in their operation as boundary objects. Indeed it is their multiple overlapping meanings that make poles of attraction effective as a tool for orienting thinking across diverse constituencies. To succeed a proposal must offer something for various different groups. For example when proposing a new technology it is necessary to articulate ideas not just about proposed technological functions, but also about how technologies will be used and who will use them. Learning by regulating The emergence of new media with apparently rather different socio-technical characteristics to its predecessors potentially throws up an array of issues for the state. The issues include (but are not restricted to) explicit types of regulation about legal and policy issues, such as: intellectual property and its enforcement (i.e. through charging systems); privacy and data security (for example the current controversy about encryption); decency (for example in relation to pornography), fraud and the problems of policing cyber-society. In relation to multimedia in particular, the thesis of convergence has immediate policy implications, as it suggests the coming together of the fields of computing, telecommunications and broadcasting which have been subject to very different regulatory regimes and traditions. The state is also involved as a provider or guarantor of the infrastructures that may be critical to the operation of a technology. Just as the widespread uptake of the car rested upon the establishment of a host of different institutions to regulate and promote this technology, today the emerging Information Society is seen to require the establishment and maintenance of an Information Infrastructure (Kahin & Wilson 1997). 107 Public policy has played an important role. As we saw in Chapter 4, there have been important attempts to learn from other countries. A fear of being left behind in technological and economic competition has been reflected by powerful rhetorics of technological imperative. Mimicry of initiatives elsewhere, mutual reinforcement and efforts to harmonise policy and regulatory requirements have all tended to encourage a convergence of public perspectives. However at the same time the practical concerns of policymakers have also stimulated them to match policy to particular local conditions, traditions and concerns. Private actors also make efforts of regulation, for example by developing and trying to impose standards, producing infrastructure, and - more generally - by trying to enrol political constituencies as well as customers and clients to their visions and designs. In short, regulation is not just located in government institutions. We see efforts to impose order and set the rules of the game across a range of bodies and in different activities. Thus private players in promoting visions of technology futures seek to establish, and put their offerings at the heart of, emerging technology regimes. Thus today we can see suppliers simultaneously competing to pursue their particular views and interests while at the same time acting in concert to establish ideas of the imperative of the widespread adoption ICTs as we move to an Information Society. There is equally an interplay between the state's regulatory and technology promotion activities. Regulation, and in particular the formation of technological regimes, is not restricted to a 'top-level' political domain, but is permeates and ultimately becomes rooted in the dispersed activities of technology design, implementation and use. 108 Chapter 6 The Process of Social Learning 6.1) the space for social learning: strategic options for design: appropriation a) the linear model and mapping the space for social learning In explicating the different sites and processes of social learning we have begun to define something about the broader space for 'social learning'. Our empirical studies have shown that the processes of innovation (in its broadest sense to include the full cycle from design to implementation and use) vary significantly from case to case in terms of the range of players involved, in different types of activity over time. We can represent this graphically. It is in this sense that we use the term space, while at the same time using it as a metaphor for the scope for choice. We can then map the strategic options for innovation as ‘paths’ through this space. We can conveniently start by examining an ideotypical linear, technology-driven model of innovation. Figure 6.1 provides a graphic representation of such a linear model, which sees technological development as limited to a constituency of design/ RTD specialists and leading to stable artefacts, which is followed by separate phase of diffusion of these artefacts through the market to satisfy user demand. The criticisms of this model have already been well-rehearsed in Chapter 2. The obvious criticisms are that these stages are not separate; that technology supply does not yield finished artefacts already configured to existing need, but that there is an iteration between supply and use and that uptake involves an active appropriation rather than the passive model conveyed by the term diffusion. A range of actors is thus involved. This is why we insist that Figure 6.1 represents an ideal type (that is flawed and unrealistic), rather than an empirical instance of innovation. Another weakness of the linear model is that it tends towards a static, 'snap-shot ' view of innovation. A dynamic model is needed which sees technological innovation as an interactive process involving repeated cycles from technology design and development and the implementation and appropriation of artefacts. We will try to capture this in later representations. Figure 6.1 is, however, useful here as it lays out something of the socio-technical space in which social learning in the generation and appropriation of multimedia artefacts may take place. It thus provides dimensions against which the diversity of strategies and options for technology design: appropriation can be displayed. We will briefly discuss some of these strategic options. 109 Designers Intermediate Users Proxy Users Final Users DESIGN /DEVT. diffusion USE Figure 6.1 the linear model of innovation b) the user centred model – its origins; critique of the ethnographic fallacy One of the main initial responses to the perceived failure of ICT offerings to match the culture and requirements of 'final users', particularly in relation to workplace applications, has been the pursuit of user-centred design. Failings of existing systems were attributed to the shortcomings of dominant ‘technocratic’ design approaches: the difficulties experienced computer scientists and engineers in capturing user requirements; their narrow functionalist understandings of the tasks being automated and their lack of understanding the culture and practices of users. Initiatives were launched to develop richer understandings of the context and purposes of the user and build them into technology design. New design methodologies and models were proposed. Often user-centred design involved the deployment of social scientists alongside technology developers to study user contexts or to bring user representatives into the design process directly. Some interesting work has been done (see for example, Ehn 1988, Bødker & Greenbaum 1992, Green et al 1993). This model is represented in Figure 6.2. However this kind of project seems to have had only modest influence over system design overall, and there have been problems - for example in relation to the uptake and wider applicability of models that emerged, and, more profoundly, in their failure to generate distinctly different models of artefact. A number of the multimedia experiments we explored – particularly those with broader ambitions to be exemplars and fulfill certain social ambitions rather than narrow pragmatic goals - have tended to resemble this kind of project. 110 Whilst the shift towards user-centred design represents a significant and positive development, we need to avoid the pitfalls of what we have termed the ethnographic fallacy: the presumption that the primary solution to meeting user needs is to build ever more extensive knowledge of the user's context and purposes into technology design. This problem arises because the emphasis on the complexity, diversity and thus specificity of ‘user requirements and contexts’ (and the consequent importance of local knowledge about the user) is taken up within an essentially linear, design-centred model of innovation to emphasise the need for artefacts to be designed around the largely unique culture and practices of particular users. By seeing computer artefacts as largely fixed in their properties, and thus privileging prior design (Procter and Williams 1999) the key question becomes one of building ever more extensive amounts of knowledge about the context culture and purposes of users into the designed system. However, recognition of the complexity and diversity of user settings does not necessarily imply that technological design should or will be entirely shaped around the detailed needs of particular users. For example, the cost and other benefits of reusing software 'code' generates a trade-off in software acquisition strategies between the increased utility to the user and higher cost of solutions custom-built around their particular requirements and cheaper generic solutions which may match their requirements less exactly (Fincham et al 1994, Brady et al. 1992). Users may choose to adapt to the constraints of cheaper packages software.49 These trade-offs yield a range of technology supply strategies, as we saw in Chapter 2 – especially Figure 2.1). Thus the rapid spread of packaged software and tools reminds us that the possible price (per unit functionality) advantages of mass-produced standard solutions may outweigh the costs to particular users of adapting systems or adapting their activities to system constraints and affordances. The attractiveness of standardised offerings is further increased by possibility of combining them with customised elements into configurational solutions. This is further assisted by conscious attempts to design such component technologies to be readily linked together and customised. The configuration and customisation of cheap, generic component technologies has proved a remarkably effective way of acquiring ICTs – and one that has had far more impact than user-centred design. What is critical in terms of this current discussion is that the development of configurational technology is a model in which technology design and implementation are closely coupled, with a strong resemblance to the evolutionary development models outlined in d) below. 49 Users may find additional benefits from adapting to standard offerings – for example regarding the greater availability of skills to maintain and use a package. 111 Designers Intermediate Users Proxy Users Final Users DESIGN /DEVT. user input to design USE Figure 6.2 User- centred design c) innofusion In the innofusion model, the boundary between technology development and implementation/ use becomes eroded. Technology applications emerge through an iterative process involving various kinds of feedback from the implementation and appropriation of artefacts to technology design and development. The innovation concept draws our attention both to the struggle to get things to work in a particular context, which may transform the artefact and understandings of its utility, and the possibility of feedback to future technological supply. This is shown in Figure 6.3. Through this process, 'users', in their efforts to apply and use technology offerings, become actors in technology design - directly contributing to the reconfiguration of their own systems and potentially providing resources for future technology supply. The latter case of course depends upon the existence of channels to relay implementation experiences to suppliers and users of technology elsewhere. Fleck's innofusion model was articulated in relation to industrial automation, where supplier and user are large organisations with substantial financial and technical resources, and where there is a contract and direct contacts between engineers and managers of both organisations. Even in these circumstances, there was often a failure to collect and utilise implementation experiences (Fleck 1988, Webster and Williams 1993). The problems of supplier-user links are obviously far greater in relation to mass-produced consumer goods distributed through the impersonal market. In Figure 6.3 we show innofusion and domestication as separate cycles. Their 112 location indicates the different centre of gravity of these processes, respectively, around the technical elaboration of an artefact, particularly in the implementation phase, and around the elaboration of practices and meanings in consumption50 and use of an artefact. However we should not see these as being different and necessarily separate kinds of activity. Both of these kinds of social learning activity take place, and to a large degree simultaneously in the course of the appropriation of an artefact.51 The distinction then is not primarily in terms of the kinds of activity – but in terms of its centre of gravity and in the kinds of outcome that are being analysed. The distinction is, however, worth making in that it does highlight differences in social learning potential between different settings. It reminds us that different groups of actors may be drawing different things from an innovation process (e.g. suppliers may be enhancing their offerings; domestic consumers, individually and collectively may be learning whether and how an artefact may be relevant to their purposes). The SLIM case-studies provided ample evidence of innofusion. These are discussed in the next Chapter. Many of the multimedia experiments involved users in trialing new multimedia offerings. Multimedia experiments often go beyond user-centred design by involving token or proxy users in more or less natural settings - giving them space to experiment and develop practices and usages around trial applications. This was particularly notable in relation to the applications of multimedia in education (van Lieshout et al 1999) and within organisations (Rossell 1999a). Innofusion was less in evidence, however, in relation to the 'cultural content' products for more mass markets (Preston 1999), where there was a bigger gulf between the development and appropriating constituencies. 50 Strictly speaking, consumption of course starts prior to implementation, though continuing into the stage of normal use of an artefact – insofar as the purchaser is enlisted to acquire an artefact: there is thus some prior engagement between supplier promise and user expectations 51 We considered (on the grounds of simplicity in presentation) the idea of defining innofusion as the technical appropriation and domestication as the cultural appropriation of an artefact. However we explicitly reject such a formulation as it implies a separation between ‘technical’ and ‘cultural’ that is impossible to make in practice and unhelpful for analysis in that it cuts against our emphasis on the socio-technical character of many of these processes. The boundaries between innofusion and domestication are by no means clear cut. 113 Designers Intermediate Users Proxy Users Fina DESIGN /DEVT. innofusion USE domestication Figure 6.3 Innofusion and the domestication of artefacts52 d) the evolutionary model We can use this way of mapping the space for social learning to explore the various ways in which multimedia experiments were conducted – and the different locales for innovation and social learning in terms of which actors were involved at which stages and around which aspects of the configuration of a system. One important model, perhaps best exemplified by the Amsterdam Digital City (DDS), is the Evolutionary model, in which there is a permanent process of experimentation (encompassing both innofusion and domestication) as the project was developed and progressively became entrenched. Rather than design preceding an implementation and appropriation effort both activities were conducted in parallel on a more or less continuous basis. An important part of the evolution of DDS and the development of content for it was the way that the technology enthusiast sub-culture which first developed the concept were willing to open their project to other constituencies: initially groups involved in media and cultural 52 NB in these diagrams, locations on the two axes are indicative rather than exact. Thus no implication is intended in the horizontal dimension in terms of the fact or sequence of involvement of particular players such as intermediate or proxy users in every case. Nor is there a presumption that there will always be some kind of trial use between development and use. However there is an implication that projects which do roll-out to widespread use will move from the top left to the bottom right of the picture. These figures then seek to convey something of the different routes that may be taken in this process. 114 products and furthermore, especially latterly, including non-specialised users as information producers as well as consumers. The constituency of players involved broadened in the course of building the market for its products. The project never stabilised; technological development never stopped. Instead the project has undergone a process of permanent evolution in which innofusion and domestication are combined (van Lieshout). Figure 6.4 shows both the involvement of wider ranges of players in design and implementation in multimedia experiments, and the extension of this strategy in the ‘evolutionary development model’, in which developing the technology, and building the market go hand in hand. In the evolutionary model these are continuing activities and there is no clear boundary between technology development and diffusion. In the case of DDS we saw a progressive broadening of the socio-technical constituency of involved players; barriers between technology developers and users were eroded, and the boundaries of the project expanded. Amsterdam Digital City DDS DDS was somewhat exceptional among the Digital City initiatives we examined as it was the earliest such development, and has established a substantial and self-sustaining information system. An interesting feature of the DDS case is the way in which it has managed to simultaneously involve both expert and nonexperts - further technical innovation of the system has proceeded hand in hand with the appropriation of DDS by non-experts (both as users of the system and producers of information pages). For example the interface for DDS went through three distinct technical stages, (see The Interface Design for Amsterdam Digital City – chap. 7) What was perhaps more important about DDS (which distinguishes it from many other Digital City initiatives which have not achieved the same momentum and market size) was the way in which project development was opened up to a range of groups. DDS has been successful in enrolling a critical mass and sustaining this 'market' of information producers and users thereby ensuring the viability and survival of the experiment. Why has DDS become entrenched more effectively than the hundreds of Digital Cities that followed it? Our study was not able to identify a single cause for this success; instead what seemed to be important concerned the process of development - and in particular the open-ness of the experiment in its conduct; the heterogeneity of enrollment of diverse actors that was achieved by a 'cascade of activities that together created the network in which the digital city became grounded' (van Lieshout 1998). In this process domestication and innofusion take place in tandem. A project and concept of a community information system that started off amongst a sub-culture of politically-motivated technology enthusiasts has been opened up to range of different constituencies and cultures (notably to actors and groups from media/cultural product circles, and to non-specialists in the community. (van Lieshout 1998) 115 Designers Intermediate Users Proxy Users Final Users DESIGN /DEVT. Multimedia Experiment TRIAL USE Evolutionary system building Pick and Mix laissez-faire USE Figure 6.4 The Evolutionary Model and the Pick and Mix Model e) the 'pick and mix' or laissez-faire model Figure 6.4 also shows the ‘pick and mix’ laissez-faire model of technology appropriation. As noted in Section 2.4, multimedia technologies are increasingly taking the form of configurational technologies - deploying an array of standard and customised element on a 'pick and mix' basis to meet particular local purposes and exigencies. This has been particularly marked with the emergence of the multimedia and internet capable PC as today the basic platform. The majority of multimedia experiments identified by SLIM involved the development of applications from largely standard technological components typically today the internet capable personal computer and the world-wide web (but also, for example, the CD-ROM). Although some of our longer-running projects (DDS and the Scottish Community Information Service) started off with more limited, specialised text-based communication facilities, they were mainly forced to migrate to the web in the light of its growing popularity - both in relation to the wider array of services and sites thereby available, and the growing expectations of users for its attractive and readily understood user interface. The power and flexibility of this platform, and attempts to make it more transparent for non-specialist users opens up the possibility for final users who have relatively modest technical skills and resources to configure systems for themselves. 116 This can be seen as a rather interesting kind of user-led model of innovation. Although the users involved in these specific cases have virtually no direct involvement in the innovation of the generic component technologies - their links to technology supply are almost entirely indirect and mediated through an impersonal market - they have considerable control about the final configuration and deployment of their system. Lobet (1999) therefore describes this as a 'laissez-faire' strategy. This is, of course, a very different interpretation of the concept of user-led than in a classic 'social experiment' with multimedia, in which non-expert users are involved as actors in system development (Jaeger & Qvortrup 1991, Jaeger, Slack & Williams 2000). However when we recognise the active and creative nature of innofusion and technology consumption/ domestication – and particularly when we address these processes over multiple overlapping product cycles, rather than limit our view to a snapshot view of a single product-cycle - then we see this pick and mix strategy, shown in figure 6.4, as one of a wide range of ways in which a broader range of players (beyond technical specialists and decision-makers) can contribute to technological innovation. Overall we find that social learning around multimedia innovation and domestication takes place in many different settings - and this overview has pointed to the very different ways in which this can be ordered and organised. The motive of this mapping exercise has been to illustrate the diversity and richness of social learning rather than to suggest that one way of carrying out such learning is 'better' than another. In what way, and on what basis, could we assess the strengths and weaknesses of these different models? In terms of whether a project proceeds to roll-out and becomes viable, perhaps the key question in assessing these development strategies concerns how well these attempts to organise social learning within particular experimental settings prefigure, match and are sustained by the wider, 'organic' social learning processes that will take place when a new product or service is launched. However, the social learning perspective seeks a broader approach to consideration of innovation strategy. Thus, for example, a key question, from the point of view of the viability of a new system, concerns whether users can be enrolled quickly enough and in sufficient numbers. This is not to reduce the concern with social learning to questions of market acceptance: as we argue in chapter 9, the rejection of a product could provide a valuable lesson. Our mapping exercise is motivated precisely by a concern to demonstrate the variety of ways in which lessons may be drawn. For example the gulf we have noted between technology design and use could be addressed in a number of ways; one interpretation might focus on exploring how social learning processes could help identify ways in which products could be more closely configured to the expressed needs of existing (e.g. experimental) users. However it could equally suggest the need for the design of a product to be more generic to make it appeal to a wider range of potential users while seeking to give them more opportunity to intervene around its final configuration and use (i.e. design for configurability/ further social learning). The social learning perspective does not favour particular approaches - to the contrary it 117 draws attention to alternative strategies. The test, then, is whether social learning in experimental settings will be reinforced by wider de facto social learning following launch. Can this organised social learning make the learning curve for new technologies more effective? Of course what actors mean by effectiveness may depend upon their role and commitments - for some it may be to shorten and intensify learning; for others it may be to broaden the range of concerns or include new actors. We return to this point in Chapter 9 where we explore the policy implications of this study and present some conclusions regarding transferability and the exploitation of multimedia experiments. But now we turn to examine the conduct and management of multimedia projects. 6.2) The organisation of social learning a) The key role of intermediaries and constellations The development and use of multimedia products and services always involves an interchange between players separated in time and space, and, most significantly, in terms of their institutional location, expertise and commitments. One important separation, that we have already examined, may clearly be that between the developers and users of multimedia artefacts. However with multimedia products we may also find important differences between technical specialists (eg in computing and telecommunications) concerned with creating the delivery system and producers of complementary products: their 'intermediate users/ consumers' i.e. the providers of content and services. Thus in order to supplement their services we found the Irish broadcaster RTE collaborating with Real Audio, and Compuflex buying into other technology companies. The intermediate users will often include what ICT specialists might describe as ‘domain specialists’. In the SLIM study these included, teachers, TV and printed media specialists, museum curators, public administrators. These groups also have their own institutions, cultures and traditions. Social learning thus starts with the ways in which players with complementary capabilities and domains of experience and expertise, work together in resolving the issues surrounding the development and appropriation of new technology-based products and services. This practical exercise in working together and solving problems (learning by doing) is a crucial component of social learning. The experience of being involved in such consortia, and learning about other players, may be as important a benefit motivating actors to take part in a multimedia project as the formal goals and expected outcomes of the project (Curry 1999, Nicoll 1999). Perhaps the central issue for this study thus concerns how to organise this kind of collaboration. This draws our attention most immediately to the mechanisms for acquiring, generating, enhancing, and exchanging knowledge. This study, and others, have in particular highlighted the role of the formal and informal networks linking together constellations of players involved in technology development and use for such knowledge flows. However for such localised learning to contribute to wider social learning, what has been learnt must be communicated and applied 118 more widely (i.e. learning by interacting). This brings us to another question about how best to communicate and apply more broadly the experiences of multimedia trials. Many of the substantive outputs from a trial may be highly specific to particular players and settings, and hard to generalise. The most important lessons may be about the process of building technologies and institutions - again raising questions about how to apply particular experiences more broadly. However, the lessons may be hard to formalise in a way that allows their broader communication. We return to these issues in Chapter 9. The final element of social learning as explicated above concerned the attempts by players to influence their intellectual and policy environment and the broader technological regime (learning by regulating). Collaboration and linkages are thus key to social learning. The question we turn to now concerns how to organise these kinds of collaboration and exchange. We first suggest that Multimedia applications typically emerge through the activities of socio-technical constellations involving an often dispersed network of diverse players who may be more or less directly linked together. Secondly we point to the crucial role of intermediaries in maintaining these networks, both in linking these disparate players together and by applying elsewhere lessons learnt in particular projects and aligning and ordering views more generally. In analysing social learning with multimedia, we deliberately chose the metaphor of socio-technical constellations as it more adequately conveys the idea that there may be gulfs between players with some remaining rather peripheral and looselyconnected, in preference to the more established terminology of socio-technical constituencies (Molina 1989), systems (Hughes 1983) and actor-networks (Latour 1986, Callon 1991), which conveys a sense of closely-linked and coherently organised arrays of players with direct relationships with and commitments to particular facts and artefacts. This is not to suggest that these terms may not be more appropriate for analysing other technology development settings, such as RTD projects, but simply to point out the heterogeneity and loose coupling in many of the contemporary multimedia experiments and initiatives that we encountered. Further, our empirical studies point to two particular types of socio-technical constellation: * in application development - in the learning economy - involving suppliers and intermediate users * in appropriation - concerned with technology consumption and use. The great majority of the cases we studied were of this sort, and involved only a modest technical development effort, as they were often based upon largely standard technological offerings). These groupings typically emerge around the activities of one or more intermediaries. Let us explore these crucial observations in more detail. 119 The importance of intermediaries The importance of intermediaries arises from the fact we have already noted that technology applications do not arise unproblematically from the emergence of supplier offerings already articulated towards the expressed and nascent requirements of current and future 'users' and that the impersonal 'pure market' is a rather poor communication device.53 It is particularly significant in relation to emerging multimedia products and services. Multimedia applications require diverse resources and different kinds of knowledge: about the technology infrastructure, about graphics and presentation of information, about the user context.54 These resources are virtually never all available within a single organisation. In consequence, as our research has shown, multimedia experiments cut across organisational boundaries. Multimedia developments typically involve more or less formalised coalitions of players rather than single organisations. There must be flows of knowledge and ideas between diverse players involved in supply and use. In certain circumstances this interaction may be mediated primarily through the impersonal market - through the rejection or uptake of the artefact by the user. One example from the SLIM research was the Girls ROM product (a CD-ROM given away with a Norwegian girls' magazine). In this case we saw how the developers configured the product around an existing technology platform and an established market niche to increase the probable acceptance of the product. Moreover, as we see later, though this development was driven by a particular small firm, it involved collaboration with other firms in publishing and advertising. In most cases, however, we have found a more active intermediation process - with certain players (intermediaries) acting to bridge the various domains of development and use. The use of the term intermediary should not obscure the fact that intermediation is frequently shared amongst several players - though some may play a more central role than others. Intermediating roles emerge in the course of a multimedia experiment rather than following formal and pre-determined structures and roles. Moreover there may be different kinds of intermediation taking place at different levels; different kinds of intermediaries may be nested within each other, and roles 53The concept of intermediary, of course, appears in industrial economics, in studies of the development of network technologies such as inter-organisational IT networks (Williams 1995, Graham et al. 1996) and in cultural studies (in relation to cultural intermediaries - whose role is bound up with the circulation of ideas). The SLIM study seeks to draw on the different dimensions of intermediation thereby elucidated. 54 And, as the study of cultural products concluded, in many of these cases, ‘success’ also depends critically upon various kinds of knowledge and know-hows related to content authoring, design and editing and marketing functions rather than the purely technical skills/ competencies (Preston 1999). 120 may change during the lifecycle of a project (Rossell 1999a). For example in the CableSchool project, while Cable Co. was the initiator and key intermediary in project development, within the schools, individual teachers were key intermediaries in the local appropriation of these technologies. Telepoly Telepoly is a broadband real-time teleteaching system, set up by the two Swiss Federal Institutes of Technology, in Zurich and Lausanne, together with the Swiss Centre for Scientific Computing, to distribute higher education curricula components in a multiplexed way. The case shows how originally educational aspects are neglected in favour of realising a technologically stable infrastructure. Only after the ATM-network was established did educational perspectives get a chance. Meanwhile, the case study shows the importance of a proper adaptation of teaching strategies to the requirements of a distance teaching environment. Different strategies are possible – successful examples include one lecturer who placed teaching materials on the web, and another who adapted a more demonstrative teaching style suitable for broadcast. Only teachers who are willing and able to adapt their teaching to the constraints and affordances of the new medium appear to be successful in their teaching. (Rossell 1999a) Particular individuals often emerged as intermediaries in the course of multimedia projects - though this role was not always formally recognised, and often emerged in the interstices of organisational structures. This was particularly apparent in the education studies where, across a number of cases it was clear that ‘teachers are crucial as intermediary users and intermediary producers of multimedia, and could be the driving force of multimedia innofusion particularly where they have space to manoeuvre’ (Bijker, Egyedi and Lieshout 1999). However their importance was not recognised at the outset, as Telepoly, Bornholm and the CableSchool cases show. Multimedia intermediaries can be seen, in some ways, as another incarnation of roles such as 'project champion' and the 'project manager' roles which have come to be seen as central to the successful management of technological change. However what seemed to be critical to their success in intermediation is their ability to mobilise knowledge and resources within and outside their organisation/ department - a role that was particularly crucial in the early 'preproject' stages, where no formal commitment of resources had been made (for example, in UK Bank, the Research Division had to enlist sponsors from the Bank's business departments to gain legitimacy and resources to set up a videoconferencing trial by pulling together potential users and external players at a time when there was no budget and no authority to proceed [Williams Procter and Cashin 1999]). What is crucial to this intermediation role then is this ability to cross different spaces - between different organisations and different departments within organisations - and between different knowledge communities (especially that between the technical specialists and non-technical communities of providers from service and cultural sectors and users). 121 Intermediaries and Constellations in Application Development. Let us consider first the development of new multimedia applications. These often, indeed typically, emerge at the interface between organisations with very different commitments (goals, expertise, networks). This is because the viability of a product (eg a new delivery system) may depend upon the availability of programmes (content) to run upon it, as well as the ability of different technology offerings to work together (Collinson 1993, Nicoll 1999). Pressures for collaboration are rooted in the incompleteness of information available to individual players, as well as economic considerations (cost sharing; uncertainty reduction). When we examine the array of multimedia initiatives studied by the SLIM research Centres we find that most of these are dealing with largely standardised technology components and systems. In particular, the internet and the world-wide web became established as the dominant vehicle for multimedia applications in the period of the SLIM study. In most of these cases, system development was primarily a process of selecting and combining established technologies and configuring them to particular local needs. Some customisation is needed in building such configurational technologies. However, in most cases there was only a modest technical development effort. 55 An important observation regarding the overall array of cases studied concerns the rather marked absence of constellations directly bringing together final-users as well as supply-side players.56 We explore this further in the next section. Appropriation Constellations and Intermediaries A number of the multimedia experiments and trials revolve around the role of certain appropriation intermediaries as key points of interface between potential users and new multimedia products and services. They may play a particular role in configuring ICT component technologies and systems towards particular potential user constituencies. This seems to involve mainly a domestication effort, to make a technology relevant and attractive to individual consumers. The SLIM studies provide a number of examples, particularly in relation to cultural products including Nerve Centre and the cybercafes. The cybercafe phenomenon is of interest - despite the rhetorics of the Information Society which emphasise individual ownership and private use of ICTs (eg in the home), cybercafes have been formed, and seem set to continue, as collective and public spaces for different groups to gain access to information society technologies (Stewart 1999). There are different kinds of cybercafes – and even cyberpubs – offering different kinds of 55 In part this is only to be expected insofar as there a wide array of multimedia projects – and developments of new delivery systems and platforms would tend to be large-scale projects and therefore relatively infrequent. However 56 Interestingly, as we see in the next Chapter, Nerve Centre operated to some extent as an intermediary both ‘upstream’ in the supply chain (in configuring generic ICT offerings into particular cultural products) and ‘downstream’ in subsequently supporting the appropriation of their Virtual Museum offerings. 122 atmosphere and geared towards different markets (Preston 1999). Cybercafes seem to play a key role here in acting as a point of access for users to new supplier offerings, and cybercafe managers are revealed as important intermediaries, promoting and channeling the appropriation process. They highlight the point that even fairly well-established multimedia technologies such as computer games still are not yet ‘transparent’ for the general user in two senses: they are difficult to set up and get to work, and their utility/ value for the user is not yet established. This kind of appropriation intermediary may thus play a crucial role in bridging the gap between 'generic' supplier offerings and consumers' perceptions of what they want to buy/ use, which may not mesh together. Thus the cybercafe managers provide a commercially viable service not only by sharing the purchase costs and risks between multiple consumers but also, partly, by transferring products from one context to a more accessible one, increasing their perceived relevance and value to the customer.57 Thus, cybercafe managers can be seen to act as important gatekeepers and brokers in the appropriation of multimedia in entertainment and other everyday activities. In seeking to maintain the attractiveness of their cafes and bring in customers they creatively deploy knowledges of both supplier offerings and user expectations and culture. They review products coming on to the market, select them and install them locally. For the customer this reduces 'search costs' (as well as reducing the risks and uncertainty of purchases). In addition by installing it and going through the learning curve of set-up and use the manager helps reduce the barriers to access for the consumer. SLIM research also identified occasions where the potential contribution of certain appropriation intermediaries was not adequately recognised in advance, - as exemplified, notably, by the failure to involve teachers in relation to several of the educational multimedia experiments (van Lieshout et al 1999). These observations have a potentially important implication for industry strategy and public policy might surround the contribution of appropriation intermediaries towards innofusion goals, and the feedback of appropriation experiences to future supply. For example, in Japan, Sony utilised feedback from consumer electronic stores to gain insights into their newly launched CD-I products (Collinson 1993). It might be of interest to identify other effective, but low cost channels for such feedback in relation to multimedia offerings (which might merit appropriate public policy support if the circumstances did not favour voluntary participation). b) what motivates social learning? how it is shaped by its local context 57 We can see appropriation intermediaries as playing an important role in reducing the costs of acquisition which may impede the uptake, and thus even threaten the commercial viability of a new product. Thus technology promoters may need to sell a product cheap to get a sufficient number of users to 'buy-in' to make a viable market - as demonstrated in a number of cases such as French Minitel. They may need to overcome 'non-cost barriers', for example to provide for (or subsidise provision of) the knowledge needed to adopt and use a product - which may otherwise deter potential users. 123 Having identified the importance of intermediaries and collaborative networks to the development and appropriation of multimedia, the question arises of what factors promote this kind of exchange, and how such social learning is influenced by the local setting? Put simply, social learning takes place where it is in the interests of players to collaborate and exchange information. Learning in this sense involves linking players across socio-technical space (i.e. from different industrial cultures, or between supplier and user communities). Intermediation and participation in fora for social learning involves them in costs and commitments (as well as opportunity costs from foreclosing other opportunities) Constellations survive because the parties are motivated to take part and remain part: the make the necessary investments because the individuals or organisations involved expect of some benefit. Foremost are the benefits of co-production and sharing of knowledge about multimedia development and appropriation. Such collaboration may bring expectation of economic gain (directly or indirectly, eg through reputation building58) or similarly a reduction in costs and risks (eg from 'closure' of technological controversies; through alignment of views and enhanced social learning). It may be in the mutual self-interest of players in supply and consumption (especially intermediate users) to collaborate in building new markets. However the situation of final users may be rather different to that of suppliers and intermediate users. There may be areas and players who do not experience sufficient incentives, but whose involvement may be critical (again particularly in relation to final users). Constellations do not necessarily emerge spontaneously. Key intermediaries may need to steer the process - and motivate the inclusion of whatever players and resources are needed. Here the studies pointed to the important role of public support in providing the resources where these are lacking This has policy implications that we return to below.. Our study emphasised the particularities and localisation of social learning processes. Social learning was shaped by its local context and differed between players. In analysing the different patterns that emerged, our study highlighted the importance of the specific 'translation terrain' - the immediate array of players with their historical and contingent concerns and capabilities, each trying to map out a strategy in interaction with other players and in the light of their broader social, economic and cultural setting. First, although all actors in some sense had access to the same (notionally universally available) technical resources, they made different constructions of these opportunities in the light of their particular local setting and their commitments and concerns. Second, in contrast to expectations of a progressive alignment of multimedia around certain global models, the SLIM studies emphasise the specificity of local translation terrains. This was demonstrated with particular clarity in relation to the Digital City experiments 58 As we saw in our studies of Cable School (Slack 1998) and the Language course tele-education project (Mourik 1998), and other cases such as the UK Videotron trial (Curry 1999), positioning in relation to future markets and reputation building may be important elements behind a decision to take part in an initiative- to demonstrate an organisation's membership of an advanced technology culture and to ensure that it figures in "know-who" - knowledge about who is who in the field 124 (Lobet and van Bastelaar forthcoming). The emergence of Amsterdam Digital City (DDS) has provided a very powerful exemplar of successful development. However the Digital City projects that have emerged in its wake have been many and various. Quite different constructions have been made of the goals and methods of development of these initiatives. c) Social Learning is a multi-level game Following on from these observations, another matter that became apparent in the course of in-depth investigations of the conduct of particular multimedia experiments was that the participants were often involved in a complex and multilevel game. As well as the formally stated shared goals of a projects, particular participants had certain more or less covert goals, which were not necessarily shared.59 Furthermore, the various players involved tended to have a wide range of commitments and interests in a project at a number of different levels. For example a technology supplier might have several goals: to forestall and warn-off competitors, to align expectations of potential customers, to signal competences and establish a reputation as a future player in this and related markets) which could be more or less closely geared to what they might expect to learn from taking part in an experiment. For example the Cable Company promoting the UK CableSchool project was seeking to advertise its competences as a supplier of computer networks to the local education authorities in the run-up to a new information network acquisition. Cable School This service was developed to connect a number of British schools to the internet through the use of cable modems. CableCo, the supplier of the modems worked with the schools to develop both infrastructure and content. The service can be had a number of specific constituencies beyond the schools being served: it was regarded by CableCo as a showcase if its capability to offer a cable-based infrastructure for local councils as well as a trial of cable modems for this and other services. The orginal concept was that techers in different schools would collaborate in producing and sharing web-based teaching worksheets. The development of Cable School stressed the technical at the expense of the social – providing infrastructure as opposed to time and space for teachers to meet. This had implications for trust and commitment of teachers – who lacked the confidence and motivation to submit their worksheets to the system. However the teachers used the infrastructure to develop the service in unanticipated ways. Teachers, finding that there was not a great deal of content supplied within the Cable School Project innovated, used the web connection to 59 Given the dangers of premature alignment of views about a technology (see for example Howells and Hine 1993) it could be argued that development constellations need to contain a wide variety of different groups, with their own goals and criteria for assessing success. Though the diversity of commitments and perspectives may present problems for a consortium - see for example Nicoll (2000 forthcoming), it should not necessarily be seen as a weakness. 125 cache potentially useful we pages for later use within the school. The infrastructure was used to communicate – particularly via email within schools as well as between schools. in a manner unanticipated by the originators of the service. We have described the continued use of the infrastructure despite the lack of content as ‘the social learning of independence’. The project has implications for current ICT policy in education in that it appears that the stress on infrastructure at the expense of social dimensions is being replicated. (Slack 1999) The Language Course was similarly developed largely as a marketing and Public Relations exercise by an IT supplier (Mourik 1998). The Norwegian PTT, Telenor, supported and shaped FRIHUS 2000, an IT highway service in Fredrikstad, as a way of ensuring that it was considered as a player in emerging local information systems, particularly in the development of telecottage and telecommuting centres Social learning may also be taking place at a number of levels around different issues. For example the launch of a multimedia pilot may be motivated by a concern to address the technical and operational problems in getting an infrastructure to work as well as a desire to try-out particular applications or popularise certain service concepts. Learning about complementary players and about how to work with them in developing and maintaining new multimedia services may be equally important (Curry 1999, Nicoll 2000 forthcoming). It is essential to stress the complexity of goals and interests against which particular projects may be developed. One consequence is that assessing whether a project can be deemed successful becomes rather a complex and subtle matter. As we see below, many of the lessons may be covert - and indeed may remain tacit to those involved particularly in relation to the strategies of building constellations/ experiments. This also flags the role of intermediaries of carrying and applying these local experiences more widely. An important issue may thus arise on the conclusion of a project. If a project is wound-up some players may be forced to move to other areas where there is little prospect of applying their experience. Will the players involved continue to have an institutional context that provides the resources and motive to carry these ideas forwards subsequently. For example, problems may arise, particularly in relation to public-funded experiments, if the constellation and some of the key actors involved in the project exist only for the duration of the experiment. This has important policy implications that we discuss in Chapter 9. In assessing social learning in a multimedia experiment we also need to take into account different timescales and contexts. To assess the significance of a multimedia experiment we must address the experimentation and social learning about the design and use of multimedia products that may take place within the project. However we must also examine how these may form part of a broader set of learning processes around multimedia products and services over a longertimescale and within a broader setting. d) the reflexive intermediary - social learning and reflexivity 126 This leads us to consider further the broader role of intermediaries in steering a multimedia experiment, and their attempts to negotiate and navigate in a complex and changing context and deal with the different exigencies that may arise. All the projects we examined evolved substantially in the course of their lifetime. For example, the key actor in Craigmillar Community Information Service (CCIS) was highly successful in obtaining continuing public support for his project, by redefining the project in relation to new funding opportunities, new technologies, new potential users of the system - for example, joining up with other information services to increase the attractiveness and user base, upgrading the system to web standards (Slack, 1999). New challenges were encountered and new activities undertaken, not necessarily recognised at the outset, Amsterdam Digital City (DDS) has been through three distinct launches, in the face of changes in technology and concepts of the service and how it is organised. Similarly, after successfully developing the prototype of their Virtual Museum of Colm Cille as a CD-ROM and web-site, Derry Nerve Centre embarked upon a range of activities to support the appropriation of this product in local communities and in particular schools. These points bring us back to the central and role played by the individual and organisational intermediaries in identifying and developing a response to these new challenges. Craigmillar Community Information Service CCIS was developed in the early 1990s to connect community groups in a deprived area of Edinburgh. The project has evolved over time, shaping by changing technologies; changing concepts of community information services and how they might be used; and the opportunities for further funding. CCIS has learned to be fundable in an uncertain environment where the priorities for funding change regularly. It has also learned to exploit technologies in an effort to enrol new user constituencies. Thus CCIS tailors itself to the current environment in order to ensure its survival. The changes in CCIS’s articulations have led to critical comments from those involved in the project and in the wider community. These critiques turn on the question of how far CCIS fulfils its original remit, and how far it has become entranced by the potential of technology over and above the basic aims. The project shows how different user constituencies may have very different conceptions of the success and failure of the system depending on where they are sited within the discursive construction of CCIS as a service. Those involved in the project ab origine suggest that CCIS has become too outward looking – pointing to the development of web-based interfaces over the ‘First Class’ system introduced to connect community-based groups at the outset. The manager of the project counters these criticisms by presenting access to the Internet as a civil rights issue – what he calls ‘cyber rights’. He suggests that it is important that as many people as possible become enfranchised through connection to the web and actively prosecutes this through the establishment of groups such as ‘Keyboard Kids’ and ‘Cyber Grannies’. The uptake of the service among the original community groups reached initial targets; however there was a need to 127 develop further user constituencies in order for the project to remain viable and secure further funding. However, this expansion of the service is seen by some as moving on and away from the traditional community base and role of CCIS. (Slack 1999) In considering the changing role and strategies of intermediaries it is therefore important to analyse them as reflexive actors, responding creatively to novel and changing circumstances Indeed the role of the intermediary is one of providing a focus of reflexivity in the social learning process. This was to some degree recognised by the intermediaries themselves, who were often concerned to theorise about the world in which they operated, how it was changing, the challenges this posed and how they responded to the consequent difficulties and uncertainties. This kind of reflexive activity is thus key to our concept of social learning. The social learning perspective flags the importance of knowledge acquired through practice. It suggests that all players have opportunities for reflection and insight into multimedia - including everyday citizens in their daily decisions about whether and how to engage with a technology. However certain players may have a particularly advantageous viewpoint which offers particular opportunities and incentives for reflection and for developing a more systematic understanding - for example technology project leaders, managers and intermediaries who have a broad span of control and action. Intermediaries are then of special interest in terms of their what they do and the knowledge and experience they acquire. We conclude this chapter with a brief aside on the relationship between (reflexive) researchers and (reflexive) practitioners. As we noted in Chapter 2, the idea of social learning highlights two kinds of reflexive processes: both: i) the ways in which the players themselves are changing their strategies in the light of particular experiences (which are articulated into new models and programmes of multimedia innovation) and ii) the possibility that social science research can help us understand how better to organise this learning. These considerations suggest a range of social learning settings and potentialities. One focus of social learning then is with socio-economic researchers, who have access to particular skills, methods and theories and the time and other resources needed to develop a span of overview to develop and refine systematic understanding of social learning processes and outcomes. However researchers do not have exclusive insights. When we come to consider the policy implications of the social learning perspective we find that one of the key issues concerns how to extrapolate from particular multimedia experiments and contexts of learning to other settings. And this is, of course, an important challenge for social scientists as much as for practitioners. 128 129 Chapter 7 The Substance of Social Learning Chapter 6 explored the process of social learning. In this chapter we begin to explore the substance of social learning - i.e. what was learnt and how - in these various multimedia experiments. For ease of reference we divide this into two sections, addressing, respectively, social learning in the design process (and how the user was represented in this process), and social learning in the implementation and use of artefacts (and how the artefacts and its uses were reconceived in innofusion and domestication). However, it will be immediately apparent that the implicit linearity in this schema is artificial and not very helpful. Since multimedia is not developed as a series of isolated experiments but emerges through repeated cycles of innovation, many of the experiences cut across the boundary between design and use. 7.1 design and the representation of the user a) why representing the user and use is problematic In designing and developing an artefact some model is needed of the anticipated user and the ways in which the artefact will be used. These representations of the user/ use may be more or less specific. Designers do not simply develop an artefact - they develop some concept of the use context and lifestyle (van Lieshout 1999). The main current of social shaping and constructivist studies has typically conceived design as shaped by the values and context in which technologies are developed. Implicit in many of these accounts was a rather ’politicised’ view of technological design as being rather richly informed by particular concepts of the user and use. Thus Akrich (1992) and Latour (1992) claim that we may interpret the endeavours of designers as efforts to inscribe certain preferred programmes of action by users in the design of a given artefact or technological system. Designers visualise a script of preferred reactions to the artefact, and they try to shape the technology in order to make these reactions as mandatory as possible (Sørensen 1997). In a similar vein, Woolgar (1991) describes software design as seeking to 'configure' users. Technology studies has pointed to the flexibility of interpretation of technology (Pinch and Bijker 1984). For example the above writers, from a discourse theoretic background, see technology as a 'text' that is capable of different readings. However there is a strong sense that the technology offers a preferred reading. Many existing studies have thus tended to presume that there is a comprehensive representation of user and context (for example in relation to their skills, their identities (eg in terms of gender [Cockburn & Furst-Dilic 1994] or other social features such as race and class) and the activities that may be seen as appropriate. Researchers starting from this perspective have then tended to look for the problems that may arise where that representation is restrictive or out of line 131 with the actual users (for example highlighting the problems where engineers have relied on their personal experiences and presumptions to articulate a rather unrepresentative model of the user). The SLIM study sought to make rather fewer presumptions. We took the view that artefacts embody something of a hypothesis about the user (Lobet and van Bastelaar 1999). In this sense, multimedia experiments and trials provide an opportunity to test these hypotheses. The SLIM investigations however have shown that in the design and development process for multimedia products and services, these hypotheses about the user and use often remain implicit and under-specified. The presumptions made about the user are typically unstated and often poorly elaborated. This finding presents something of a challenge to some of the presumptions that underpinned our initial conceptions of social learning.60 We return to this point below. Despite this, Development/ Design needs to prefigure a number of elements about the context, purposes and activities of the user - we seek to capture these within the term representation. Representation relates to a number of different elements. Nicoll's (2000, forthcoming) contextual usability model conceives the usability of technology as a complex of interdependent elements within a particular context, including usefulness, [the development of] usage patterns, and the particular social and cognitive exigencies of situated use. Following on from this we suggest that representation encompasses: The Technical Configuration of the system, Content, Usage, Uses, 'Rules' (formal and informal) about proper usage/ users. In principle, the hypotheses embodied in the design/ representation are attempts to prefigure these very dimensions of the eventual use of the artefact. However, it is difficult, indeed impossible to prefigure these reliably - hence the importance of social learning, both in testing the design hypotheses and for feedback from appropriation to future technology design/ representation. This is shown schematically in Figure 7.1. 60Social learning was initially presumed to be important because it was necessary to adapt technology offerings to the requirements and context of specific users and uses - to build extensive knowledge about specific user contexts and purposes into multimedia offerings (in this sense, in drafting the SLIM research proposal, we were also victims of the ‘ethnographic fallacy’) . However for many kinds of application, generic interface designs (eg the Multimedia capable PC with an industry standard browser to allow use of now widely-adopted web conventions for accessing information) seem to be largely adequate! As we see below, what needs to be more specific relates largely to the Multimedia content (i.e. the content and presentation of material and services). 132 testing design hypotheses Representation Appropriation ---------------------> Tech configuration, Content, Usage, Uses, Rules Tech configuration, Content, Usage, <------------------ Uses, Rules --feedback to future supply Figure 7.1 Schematic diagram of User Representation and Appropriation Following on from this we ask what intellectual resources do designers and developers have for building a representation of the user? Figure 7.2 shows some of the sources of ideas and information that designers/ developers may deploy. indirect evidence about users direct involvement of user - little market info on existing users - demand/market for representation other products of user - competitors - user panels - market research - trials - innofusion constructions of the user - visions of technology - fictions/myths? about the user - expert proxies eg experience of engineers/intermediaries Figure 7.2 Resources for building representation of the user Figure 7.2 illustrates a number of points. First. in a context in which information about potential users is typically incomplete or of uncertain reliability, players may be obliged to 'knit together' different kinds of knowledge from diverse sources and with different degrees of gearing to 'actual' users. Second, there may be relatively little empirically grounded information about existing users. In the absence of direct knowledge of users, expert constructions of the user will be the main resort. Constructions of the user, created by 'experts' (eg 133 engineers or intermediaries) may be derived from their own personal experience and culture (Nicoll forthcoming) or may be more firmly rooted in experiences in this or other product markets. Expert constructions may be informed not only by rigorous evidence and pertinent experience but also by visions of technology and fictions (myths?) or stereotypes about the user which may turn out to be more or less close to actual users. Developers do not work in a vacuum - but may be influenced here by the behaviour of peers and competitors - which may be reflected in clustering of supplier offerings or the mutual reinforcement of supplier visions and presumptions. These kinds of alignment of expert views, and consequent mission-blindness, have been identified as the root of a number of high-profile and expensive failures of ICT systems. The potential weaknesses of the latter forms of evidence are clear. However we take this point further. Our third argument is that all the forms of information about future users carry their own uncertainties and difficulties. For example the most systematic empirical information available about user choices and preferences (as revealed, for example, in aggregate form through market behaviour) may exist in relation to established products. However in where products are changing, expert views are liable to be rooted in prior experiences in other related markets. The question arises as to how far one can extrapolate from such information. The problem perceived in relation to 'radical innovations' is that knowledge about the user and use of existing applications is seen as not providing a reliable guide to the novel application. 61 Empirically grounded information about users of a new product may be sought through direct involvement of users in panels, through market research surveys and trials. However, various difficulties arise regarding the interpretation of such direct information about users. For example user panels need to be introduced to new technologies and given some training in their use - their selection and training however mean that they are in some ways no longer independent and representative of wider publics. Sørensen (1999 section 2.3) describes this as 'simulated social learning' involving as it does "people that are supposed to act as if they were users, but under artificial, laboratory-like circumstances." Initiatives for 'user involvement' in the design/ use of prototypes raised similar problems about representativeness. How should appropriate 'proxy users' be selected? Is their behaviour in the laboratory a good basis for understanding behaviour in everyday life? Would these responses provide a secure foundation for anticipating the larger cohort of real users that the supplier wanted to attract? These factors tended to mandate in favour of live trials - which would allow experimentation around the acceptance and utility of a product in naturalistic settings (Nicoll 2000, forthcoming). As an aside we would suggest here that the defining characteristic of a radical innovation is the extent to which it cuts across existing institutional forms, in contrast to the earlier literature which tended to see 'radicalness' simply in terms of the technical properties of a new technology (see for example, Freeman 1974). 61 134 These observations underpin the potential importance of social learning in the innofusion and domestication of multimedia - in that it can provide rich sources of more direct and reliable information about 'actual' user responses to supplier offerings. It raises questions, which we return to below, about how such appropriation experiences can be fed back to generate more robust user representations for future design. b) on Genres and Interactivity: Learning New Media Models An important finding across the SLIM investigations was that the most important social learning processes were not primarily concerned with the technical artefact but concerned the content and use of multimedia.62 For example the Danish BlackOut case study found little of interest in relation to the technology (in this case CDROM which is relatively stabilised). Learning instead focused upon 'the way to tell a story... indicating a potential change in... cultural genres’ (Hansen, 1998). Inn relation to the latter, we can begin to pick out the different elements around which design and other social learning processes are taking place. Though different writers have used different terminologies, and there may be overlaps and interactions between these elements, we can differentiate the form of the medium from its specific cultural content, and distinguish, within the former, between the tools used to obtain access and they way in which that content is organised. Thus, in other words, we can distinguish: i) ii) iii) tools and navigational devices (e.g. buttons, search facilities, scroll bars) scripts, narrative architectures and devices (e.g. relating to how stories are told/ how a user may enter, make a journey through and leave a programme) and maps/ navigational structures and coordinates (i.e. showing where a user is within such a journey) 63 specific information and cultural content In terms of learning new genres, we are particularly concerned with ii) scripts and narrative architectures [and i) tools and navigational devices insofar as it impinges upon ii)]. We explore below how the Amsterdam Digital City (DDS) deployed a geographical metaphor of a city (with squares and houses) to present and organise the spatial layout of its various ‘pages’. The Virtual Museum of Colm Cille used a strikingly similar spatial layout presented through the metaphor of a museum with intersecting rooms and display walls. In this way we find a process of borrowing from areas of social life which are well-understood (or, to be more specific, where understandings are well established and broadly shared by many sections of society) – and in particular, as we see below, from other established media. There is however a tension between things being new and things remaining the same. 62This is not to suggest that 'the technical' was unimportant - indeed as we see below, technical problems in configuring a system continue to represent a major barrier to multimedia uptake and use. Moreover, particular delivery systems have their own affordances and technical limitations. 63We are intrigued to note how terms have been applied from other media forms which may be textual [scripts] or graphical [maps]. 135 Much has been made, in discussions of multimedia technology, of the rather different properties of this new medium compared to established physical media (eg printed paper, vinyl records) and its affordances in terms of new flexibilities in the way information is combined and accessed. For example, one strand of technical enthusiast culture has emphasised the ability of electronic media to radically transform the relationship between producer and consumer of cultural products; of artist and audience (Qvortrup 1999). Some have called for an end to the conventional fixed and linear narrative structures of physical media. The cultural product is transformed from finished object to a resource for reworking by its consumer. The Black-Out study explored these themes. It suggested that many multimedia products involve rather limited types of innovation/ social learning. The first generation of products were driven by the desire of the commercial computer market 'to sell the media without any real regard for content or creative forms of interactivity'; the second generation of CD-ROM product was used by cultural industries to present their traditional offerings without development of any new media content. As a result, Hansen (1996) argues, the multimedia products we have seen to date offer little more than re-cycling of old media content in new formats (eg CD-ROM etc.) with little real innovation in terms of media forms or 'interactivity'. Only a minority of products have fully exploited the potential interaction between media and content (Hansen 1996, Preston 1999). Hansen (1998) sees the ’Black-Out’ case as an important exception in this respects. Here the change is not only in the form of presentation, but is also reflected in the content of the story and structure of use. Black Out Black Out is a CD-ROM product developed in Denmark. The starting point for Black Out was on the content development side, with an author who wanted to tell a story in a multimedia form. His aim was to create a hybrid product combining the qualities of literary and computer game genres. Here multimedia producers integrate interactivity into the story. They have used multimedia to break the linear model of a story on traditional eg paper-based media, designed to be read sequentially from the start to the end. The original idea was that users should each be able to create their own story. Though has is not been possible in Black Out, the users are able to vary elements of the story. No two users get the same story, when they play. They each get their own version with their own details. This creates a space for experiences of the story which reflects the personality and choices of the user. (Hansen, 1998) Interactivity 136 Interactivity has been seen as one of the key features of multimedia. Most such accounts view interactivity in technical terms as the property of a medium or delivery system. Such a media-centred view has been criticised (Rafaeli, 1988, quoted in Hansen et al. 1996). For example, it fails to distinguish between routinised machine-machine interactions (such as Electronic Data Interchange) and computer mediated interpersonal communication. Schmutzer (1999) suggests that interactivity of a medium should instead be seen in terms of two dimensions of user control: i) control over the content of communication - in terms of the extent to which the user can influence the content exchanged in a communication process; and, ii) control over the flow of communication - in terms what influence the user has over which contents are exchanged when and in which order. In practice what is important for the appropriation of multimedia does not rest upon some technical definition of interactivity - for example in terms of the amounts of information exchanged between user and application and the size of the 'back channel' from users to system. Instead, the SLIM study sees interactivity primarily in terms of how the user interacts with the system - and engages with the content supplied. What is at issue is the extent of interactivity in use and the users' experience, rather than a narrow technical definition of interactivity. In this sense systems which are largely off-line can be highly interactive - for example some of the interactive television products that have already been launched on the basis of an extension of viewdata technology which has no electronic ’back channel’ (Curry 1999).64 Some have tried to draw distinctions between different degrees of interactivity distinguishing ’quasi-interactive’ systems, like that above in which interaction is constrained, and fully ’interactive’ ones which have the characteristics of human communication (Rafaeli op. Cit., van Lieshout et al 1999). Whilst this does point to an important dimension of multimedia, at this stage it seems premature to prejudge what its importance will be. Our emphasis on interactivity-in-use does not imply that the technical dimension of the delivery system is unimportant. The Nerve Centre study developed cultural products for distribution on off-line media (viz. CD-ROM) and an on-line media, the Internet. Kerr (1998a: 48) noted, in particular, that 'the distribution platform influenced the complexity of the content and the degree of interactivity between the final users and the media and/ or with the producers'. The CD-ROM was designed as a more immersive game like space in which one explored and interacted with the media directly via the keyboard and mouse'; in contrast, the web site 'was viewed within the guiding context of a browser window’. e-mail facilities were built in to In the Videotron case the interactivity is between the user and a complex array of responses articulated by the designer. However a lot can be done within a system that only allows a one-way transmission of data. For example, complex software allows quizzes and betting - with the user ringing up to claim their prize. 64 137 allow a degree of interactivity with the designers. She concludes that 'the form and content were influenced greatly by the technical limitations of the distribution media' (idem). Elsewhere however the boundaries between on-line and off-line media are becoming blurred - with widely adopted products like the Encarta CD-ROM encyclopaedia acting as a point of access to web-sites. Hansen (1996) reminds us that it may take some time before we can achieve a full understanding of the affordances of new technical capabilities, particularly if these differ significantly from existing media forms. On the other hand, some elements of multimedia operation have very rapidly become globally communicated and accepted (the ideas of windows, buttons, the world-wide-web metaphor). The point about such models and metaphors is that they are not only a resource for design, but also help to contribute to appropriation. We should remember that the imputed 'multimedia revolution' (or transition to an information society) is still in its infancy. In particular, we are still in the early stages in the process of developing new genres for multimedia products - in the sense of sets of protocols and ideas about product and the reader that can unite and guide the activities of reader and developer alike. This refers to devices and narrative structures that inform how to write narratives/ products and how to read them. The history of cinema shows us how these types of convention emerged and evolved very gradually (McBride 1998b). With hindsight many of today's multimedia offerings will doubtless seem as quaint to the next generation as the first movies seem to us today. c) The resort to Best Practice Exemplars and Metaphors of Design/Use Given these consideration, successful examples of multimedia adoption could become an important resource for designers and developers, as a source both of ideas about how to apply ICTs and of evidence of its acceptance and uptake. However the resort to 'best practice exemplars' still presents problems, regarding both the selection of appropriate exemplars and the diagnosis and attribution of success. How successful was a recipe? What were the reasons for its success? What elements can be transferred to a new setting? Will they prove successful there? Insights from the Digital City and Community Information cases Our studies of Digital Cities and Community Information Systems, by focusing upon a rather similar set of applications give us very interesting insights into these processes (Lobet and van Bastelaar 1999). For example the Craigmillar Community Information Service was explicitly initiated by a desire that Scotland should launch a project to emulate the Manchester Host - one of the earliest community information systems; the idea of using IT to integrate communities drew explicitly on the ’electronic village halls’ in the 138 Manchester project - although the ambitions and scope of the Craigmillar project were considerably scaled down in terms of geographical coverage and scope (Slack, 1999). The later Frihus 2000 feasibility study report, reviewed a number of international exemplars, and selected the Gemsis 2000 project in Salford as presenting a more realistic solution for a town such as Fredrikstad in Norway. The SLIM study concluded that Gemsis 2000 represented a questionable choice as model for the development of a future urban IT highway in Fredrikstad, which was at the periphery of on-going activities. Further one could suggest that external perceptions of the success of Gemsis 2000 might well have been much more favourable than that of players directly involved! This points to a more general risk of external observers confusing the expectations and objectives of a project with actual achievements. In the case of Digital Cities the diffusion of these exemplars was organised. Thus the early experiments with Manchester host provided one of the exemplars that informed the formation of the Telecities network (an off-shoot of an EC supported network of major European cities). The Amsterdam Digital City (DDS) provided a later, and highly convincing, exemplar of such systems - an integrated view that encompassed ideas about the design and use of information systems. In particular the analogy between the arrangement of web pages and the layout of a city into squares and houses etc. provided a compelling and readily understood metaphor about the design of the user interface and how the user would navigate through the information space. DDS has spawned a huge array of Digital City initiatives - a testament to the success of the Telecities initiative as a vehicle for disseminating these ideas, and to the attractiveness of the DDS design/ use metaphor. Once the concept of a Digital City has emerged and come to be seen as effective in one place, there are obvious opportunities to its wider adoption. What is striking about this array of Digital City studies, however, is the diversity of outcomes it reveals from projects that apparently drew inspiration from the same sources: especially DDS. For example the Digital Metropolis Antwerp (DMA), has a very similar structure to the Amsterdam DDS project due to the collaboration between the two cities in its development. However DMA utilises some rather different design metaphors such as a 'bridge' (between the administration and the inhabitants) and 'quarters' which indicate specific themes (sport, culture, education, ...) rather than specific places or streets. DMA is more closely integrated with the local authority and geared towards integrating the citizen with one-line public services. So what was apparently a twin sister project becomes transformed in a different translation terrain. The possibilities for diverse interpretation of a ‘best practice’ exemplar should not be seen as a problem. To the contrary, some space for experimentation was necessary for successful appropriation of a new technology. This was amply 139 demonstrated by the Educational Multimedia cases, as we see below when discussing unanticipated uses of technologies. The last two paragraphs illustrate some of the key feature of successful metaphors that we discussed in Chapter 6. What makes metaphors such as ’The Digital City’ effective is precisely their ability to provide meaning to diverse groups. And this seems to involve not only their immediacy and rhetorical compulsion but also their open-ness to local reinterpretation and redefinition. Thus the metaphor of the digital city has succeeded not merely because it provides a convincing description of what a new system might involve - but also offers a rather open and comprehensible frame that can readily be grasped, for example, both by designers and users of information pages and can cater for a variety of activities. Successful metaphors are not those which simply impose a uniform outcome, but allow a range of players to own and appropriate them, and redefine them according to their own particular circumstances. As such they are complex entities, nests of concepts (eg about user, technology, uses) rather than single ideas, which provide 'hooks' that may engage the interest of diverse audiences. d) Difficulties in representing the user Why then does it seem to be so difficult to develop robust models of the user at the outset? We have already pointed to one of the key problems perceived with radical innovations - that the user does not yet exist (i.e. that we are uncertain about the applicability to a new setting and new product of the knowledges we posses about users and users of existing products). These uncertainties are amplified because needs for a particular new product or service are not fixed and pre-determined but emerge through a dialogue between current user requirements and emerging technical opportunities in the course of the domestication of a technology. We know that needs will evolve. This links in to another uncertainty about the identity of the future user - which will only emerge through this engagement. Issues thus arise about how to target the future market - about how to define and put boundaries around particular classes of user. Providers may make presumptions about their key markets - and the priority attached to particular market segments - which may be called into question. This was exemplified in the SLIM study of 'The Den on the Net' (Kerr 1998c cited in Preston 1999), the set of web-sites developed by RTE, the Irish national TV and radio broadcaster, to 'add-value' to its traditional media offerings. The subsequent experience emphasised the importance of the ’diaspora’ market comprising Irish national abroad and emigrants (in particular the USA and other new commonwealth nations provided a potentially enormous market of people already using the web and interested in 'Irish culture’. The existence and importance of this particular market could have been anticipated, however this is by no means always the case. Indeed the very idea of cyberspace emphasises the profound uncertainties about the identity of the user. This was, for example, one of the key dilemmas facing the 140 digital city initiatives and community information systems. On the one hand there were 'grounded users' - those located within a particular physical community (or other interest group) whose characteristics were well known. However there were also visitors from outside the physical community, whose identities remained uncertain. Issues arose about how or whether to cater for them. Should they have the same status and access as grounded users? Different approaches have been adopted. For example, DDS in Amsterdam distinguishes between 'residents' and the presumed casual access of 'tourists'. Another point is that it is unhelpful to talk about 'the user' since users are extremely heterogeneous. Even where there is direct information about the actual users of a new product there may be uncertainty about how reliable it is as a basis for extrapolation. For example, we already know that early adopters may be technology enthusiasts with very different expectations and requirements than later adopters (Rogers 1983, Norman 1988). Technologies designed around the presumptions of technology enthusiasts in design and early adopters may exclude the rest of the market. Issues thus arise about how to segment the intended market about how to cater for various classes of user. An important factor concerns the level of competence available to the multimedia provider. Levels of in-house expertise in marketing (and associated issues such as product design) were extremely variable. Smaller providers were often lacking in this respect. In contrast some of the larger and more established players possessed relatively high and extensive knowledge about their current customers (for example media companies such as RTE in Ireland know a great deal about their established market and were able to apply this in designing their web-based offerings). The heterogeneity of users - and the difficulties in specifying likely users groups and their characteristics - underpinned another of the key problems in developing design representations. As designers tried to cater for ever wider arrays of groups users, concrete conceptions of the user derived from particular individual users may become more diluted, and the representation of the user becomes more abstract. Instead of differentiating between many various sorts of users - each of which might have their own peculiarities - designers trying to interpret user characteristics seek to incorporate user requirements wholesale into their design in an attempt to serve them all at once. In consequence, instead of working around a particular representation of ‘the’ user, no user at all is conjured up. For example the users were neglected in the design of the user interface for the first two generations of the Amsterdam Digital City (DDS). In this context, technology driven visions tended to serve as a proxy for representations derived from knowledge of actual user constituencies; user and use seemed often to be extrapolated from technical potential (Lobet and van Bastelaar 1999). Of course the initiative for some projects was precisely such a technically-driven vision. The Interface Design for Amsterdam Digital City (DDS) The first, text based version of Amsterdam Digital City, DDS1.0, entailed a menu-structured interface based on Freeport software. Facilities were created that 141 enabled users to browse through the information pages, and that enabled them to communicate with each other, but these (the menus, the underlying software) were not fine-tuned to peculiarities of different groups of users. The second interface, DDS2.0, was, in words of the designers themselves, the result of a technically driven approach, as they sought to be amongst the first organisations experimenting with graphical interfaces. Only with the start of the design of DDS3.0 did users became an explicit part of the design process, in terms of a largely shared presumption that the interface should be accessible and understandable to all users, irrespective of age, gender or education. During the design process, one group of users was explicitly taken into account. Interestingly enough, this was not the group of inexperienced beginners, but the opposite group: the so-called ’power-users’ with experience with a UNIX-based environment, who were offered entrance to the shell structure of the interface (van Lieshout 1999). However technically-driven visions were a frequent source of difficulty. This was very clear in relation to the educational multimedia projects, in which the presumption that multimedia would provide a technical fix - eg as a means for teachers to obtain web-based lessons 'off the peg' led to a failure to consider pedagogic models and concepts of the social context and processes of education and integrate these with the educational multimedia applications (Slack 1999). Overall the multimedia products being developed under SLIM were often far removed from the visions of radical change espoused by the technology pundits and forecasters. Whilst technology commentators and suppliers have often articulated science fiction visions of radically new ways of doing things, with new technologies that are here or just around the corner, the actual offerings put onto the market have been rather more modest in scope. and have tended to be little more than incremental developments of existing products and services. When faced with a conflict between the attractions and uncertainties of the rhetorics of radical change, developers seemed most likely to resort to their existing knowledge, routines and market links. As a result many multimedia applications were modest extensions to existing practices. This should not be surprising - wise and realistic investors will seek early returns and avoid uncertainties by building upon their knowledge of and links to their existing markets. This is the case for most of the projects studied by SLIM, which represent only rather modest innovations. We did however also encounter some attempts to restructure existing markets. The case of multimedia highlights one of the difficulties in trying build or transform mass markets - regarding the need to relate to the established understandings and requirements of existing consumers. There was thus an ongoing issue about whether the design of multimedia offerings should be constructed around understandings of existing groups of 'users' and market segments or around visions of how these user markets could be transformed. We explore this in detail below in considering the gender construction of multimedia artefacts, where the SLIM investigations afforded particular insight. We then go on to examine two contrasting views of the design process that underlay these two extremes. 142 e) building on existing users/markets or transforming them? the gender construction of artefacts An insightful illustration of the ways in which technologies have tended to be designed around existing socio-economic and cultural parameters, and of developments that may serve to transform these, can be obtained by examining the gender construction of multimedia artefacts. SLIM studies confirmed a theme in much contemporary literature on gender and technology that computer-based technologies have traditionally tended to be constructed as predominantly masculine artefacts and that their domestication has generally tended to follow clearly gendered patterns. As in other technical spheres, computer competencies and enthusiasms are heavily gendered as they are deeply embedded within, and shaped by existing social and cultural conditioning. As one case study report puts it: "males are the keepers of the knowledge of the machine, they control the practical use of it, and through its placing and signification, it is symbolically given meaning as their 'toy'"(Spilker, 1998). Multimedia content styles might also be deemed to reflect the user base at present which tends to be male, urban, well-paid, college educated, 25 - 35 years old etc. (Preston 1999). However our studies also emphasised the dynamism of these situations. They add weight to recent discussions of the mutual shaping and reshaping of gender relations and technology (Berg & Aune 1994). Most of the initiatives we encountered tended to operate within rather gender stereotyped parameters; we further we found that males have tended to predominate in the production teams involved in the construction of multimedia products (Preston 1999). However the set of SLIM studies also includes some attempts to move outside these parameters. In particular certain innovative initiatives have sought to exploit the potentially large but undeveloped women’s market for multimedia products (it may not be a coincidence that these were both in Norway). Perhaps the most striking example was Girls-ROM (Spilker, 1998). We could interpret this as a socially radical project, which in the long term could be seen as an attempt to transform the gender identification of the technology and to create a new market. It involved an attempt to open up an established medium/ delivery system (PCs with CD-ROM readers) to a new market - of young women. Despite this demonstrably radical goal, what is interesting about the project is the rather pragmatic approach it involved. GirlsROM, at least in the short-term, was built upon fairly conventional gendered understandings of this target market - and extrapolated from existing cultural product markets (eg Girls magazine) and included for example somewhat stereotypically gendered features such as a 'secret diary', and deployed them within a new delivery system.65 This incremental model of change (in terms of working within established parameters both of an existing Similarly the emerging subculture of young women computer enthusiasts identified in our ‘Spice Girls’ study emphasised their gender normality. 65 143 market segment [Girls' magazines] and of an established delivery systems [CDROM PC]) offered a more realistic and realisable pathway to the wider appropriation of multimedia products than might have been possible with a 'BigBang' approach - a head-on attempt to achieve rapid and profound realignments of both technology and markets. The GirlsROM project has had longer term effects: though driven by a small media/ technology based firm, the project involved a network of actors, including a large magazine publisher and advertising interests. Since then, the latter has launched its own larger-scale HomeNet project in collaboration with other large Norwegian publishers. This too seeks to address women users as a specific market segment within a broader service that is oriented to different (gender, age, etc.) categories of Norwegian home users. The study seems to indicate a rapidly growing awareness and ‘social learning’ on the part of the mainstream media corporations and advertisers that it is in their interests to construct the widest possible audience/ user base. This is prompting them to expand the range and forms of services which address specific gender group identities and thus address related aspects of ‘exclusion’ - at least from the more commercial or consumerist multimedia content products (Spilker, 1998). The Internet in particular has opened up more imaginative possibilities for technology. The diversification of computing applications has increased the potential relevance of these technologies to different groups in society (see for example the Craigmillar Community Information System, which fostered a number of specific user groups to extend the size and range of users of its services within the local community, including a 'cybergrannies' group. As in the case of "women's" technologies these groups of users/ uses were built around stereotyped constructions of user needs (in this case of older-users' needs as a way of recording their memories). There is some evidence that we are on the threshold of a sea change in the gender identification of ICTs. The growth of the Internet and the stabilisation of certain tools and functions (especially email which continues to be the most significant application for most users) has consolidated the role of the computer as a device for communication and sociability, and thus a potentially more convivial device for women, and as a useful object rather than a technical object for men to play with. This is also evinced by the emergence of sub-cultures of computer enthusiast young women explored in the SLIM study ‘From Spice Girls to Cyber Girls’ (Nordli 1999). f) Modes of Design and User Representation As well as comparing social learning processes in individual case-studies (severally and within integrated study areas) we have sought to make comparisons between the different integrated study areas. This has drawn attention to some important differences (alongside a number of similarities) between Cultural/ Content products and our other integrated study areas (Education and Digital Cities) in terms of the manner in which design (and in particular user representation) was approached. To capture these we found it useful to draw a distinction between two possible modes of design focusing respectively on accountability and creativity: 144 i) Accountable Design: The first is rooted in computer system design traditions and stresses the Accountability of the designer to the user. It is associated with an inclusive approach to design, targeted towards the full range of different user constituencies. In this model, good design is design which meets user specifications - a model which perhaps reaches its zenith in Structured Systems Design Methodologies. Its ultimate objective is a system that fully meets the users' perceived needs. This has been the dominant discourse around ICT design. ii) Creative Design: The second is rooted in models of artistic activity and sees design in terms of Creativity, and valorises Authorship. In this mode the designer is given leave, based on internalised representations of users, to construct new concepts of use - to reconstruct the user - within the bounds of what particular users can be convinced is acceptable/ attractive. Indeed creativity may be seen in terms of transforming existing genres. The ideal type is a system which enchants the user. The creative design mode would seem to be more typical of many kinds of cultural products - eg the Nerve Centre or games design. In contrast the accountability model is more representative of the mainstream of systems engineering and functional IT applications within organisations. 66 g) How much design/development? Following on from these considerations, questions arise about how far to prefigure the user? How far should multimedia development attempt to build concepts of the user and uses into an application, as opposed to leaving these questions relatively unconstrained? The supplier needs to convince potential users that their offerings will meet their particular purposes and will be sufficiently valuable to justify their acquisition. More dedicated applications may have the advantage of increasing the utility of the application to particular users (where these match particular user requirements) and making these inscribed utilities and their advantages more apparent to potential future users. On the other hand dedicated applications run a greater risk: i) of failing accurately to capture the current requirements of target users; ii) of excluding other potential users whose circumstances and requirements are somewhat different; iii) of foreclosing (rather than enabling) the active role of users in reinventing the technology and its uses in the course of its appropriation. There may be an important temporal element here. In the short-term it may be more feasible to enroll potential customers to buy-in to applications which are closer to their immediate perceptions of their current needs and how technology might fulfill them. However, if this incorrectly anticipates the kinds of use that will ultimately However, making such a dichotomy immediately draws our attention to similarities between cultural and other products - for example, Hollywood films are now put through increasingly formalised market testing before release; conversely we find attempts to change the teaching methods of computer system designers to enhance creativity. 66 145 prevail, it may deter potential users and constrain their ability to realise their objectives in future. It is very easy for designers to overlook apparently obvious features of the users and their contexts. This was amply illustrated by the design of the Language Class videoconferencing system (Mourik, 1998). Although the designers knew the users would be children they forgot one of the most obvious changes needed; the height of the computer screen. The camera mounted on the computer showed only the top of the head of the children (whose height had to be raised with pillows and books on the chair). As we see later, the design also overlooked the potential for playful use of the system by these children, as well as the possibility that the children or other members of the family might reprogramme the computer for other uses. There is thus a continuing dilemma surrounding the design and development of multimedia and other products about how far to prefigure the user in systems development and build specific user representations into designed artefacts. There is a danger that facilities that are too abstract run the risk of failing to engage and convince potential customers. Building applications around specific concepts of user and use may increase their perceived value to particular users. However the above considerations suggest that there are risks in seeking to foreclose user choice, and in trying to move too far beyond current practices and models. These factors may mandate in favour of adopting more generic design approaches. This may involves active strategies to build upon successful specific applications, but 'decontextualise' them - to design out from the artefact reference to its specific contexts of origination and use which may limit its future use and market - or more precisely to 're-design' and re-present the artefact to make it more generic and open it up to broader markets. These choices in modes of design have an obvious linkage with the strategic choices we identified in Chapter 6 regarding the organisation of social learning perspective. We find a range of different strategies for addressing this design dilemma, between at one extreme the prior design model - e.g. user-centred design, which seeks to build up a comprehensive knowledge of the user setting and purposes into the designed artefact by studying or involving particular users in artefact development, and at the other extreme the laissez faire model, in which the user configures standard generic components to their particular needs. Although many elements of design are divorced from users in the latter model (i.e. the design of component technologies), when we look at the whole product life-cycle we argued that this represents a rather clearly 'user-led' strategy in the scope afforded to intermediate and final users to actively select and configure elements to their particular purposes. This model has also been the basis for important successes - indeed the most dynamic field of contemporary technology supply and technology appropriation the world-wide-web/ Internet largely corresponds to this model. An important feature here seems to be the existence of a sub-culture that shares a view of the usage of technology - particularly where this view turns out to be one that can be generalised to other groups. The success of technologies such as the world-wide 146 web 67 suggests that a particularly flexible supply strategy may well be to keep supplier offerings generic and provide for the active role of local appropriation players in configuring and adapting them to their purposes. In practice it seems that multimedia artefacts rarely embody a single 'script', but instead are built around a multiplicity of overlapping possible scripts (indeed the interpretive flexibility of multimedia artefacts is rooted in this. h) Technical design, development, configuration The SLIM research proposal was developed at a period in which the nature of the delivery system/ technical platform that would bring multimedia into the home and the community was still open. Whilst the internet-capable personal computer presented a powerful exemplar, with very high rates of adoption in countries such as the USA, in other countries, for example Germany and to a lesser extent the UK the digital TV was seen as the basis for commercial multimedia services in the home. This focused attention on the battle for dominance between these two - with their very different characteristics in terms of the user interface and the capacity and potential uses of the 'back-channel'. Linked to this view was an expectation that the design of the delivery system/ domestic platform would be built around particular kinds of application/ service and thus favour some kinds of multimedia use over others (indeed clear moral stereotypes were articulated, for example between the liberatory uses of the web and the commercial consumption mode of use of web TV (as captured by visions of ordering pizza with your TV remote control). Such a vision has become less convincing, partly because of the continuing success and development of Internet and the TCP/ IP standards on which it is based. The idea of head-on confrontation between digital TV and the multimedia capable PC has given way as players from both constituencies have sought to open up their offerings to other markets (a model of competition with coexistence rather than a mutually exclusive battle to the death). The huge amount of information available on the Internet forces digital TV players to present their offerings as a point of access to the web, rather than an alternative source - even MicroSoft was forced to abandon its plan to develop alternative information systems. Partly as a by-product of the hegemony of the internet, the internet standards and the multimedia capable PC have become the de facto standard for the delivery system. As a consequence the development of multimedia applications has, in general, involved only modest levels of technical development activity in relation to the technology platform. This was certainly the case with the multimedia experiments investigated by SLIM. Most of the multimedia applications surveyed have been configured from largely standard technological offerings - and indeed have tended to involve well-established applications of technologies that are already widely adopted, rather than particularly novel applications of 'high-end' 67This pertains for example in relation to the raft of products around the world-wide web. It also seems to apply, for example, to DDS - where a group of technologists shared a particular agenda and vision of how IT might be used to underpin community information exchange. There may be grounds for suggesting that such technology sub-cultures could be a more effective basis for system design as creativity than as accountability. 147 technologies. Multimedia experiments and trials typically involved a personal computer with a CD-ROM reader or a network connection (mainly and increasingly the latter), combined with an assortment of packaged software components, often based on the emerging global de facto standards, particularly with the increasing dominance of MicroSoft Windows operating systems. The important local innovative effort has been to knit together these different technical and cultural elements into specific multimedia applications. This has to been seen as reflecting the practical exigencies of trials. There are two main reasons. First is the need to economise on costly innovation and learning. The availability of generic multimedia packages with facilities for easy customisation provides application developers with the option of minimising technological development costs and leaving shut certain technical 'black boxes'. For example UK Bank developed its videoconferencing system by combining existing market solutions - configuring them together and customising the interface (Procter, Williams and Cashin 1999).68 There was some indication of a trade off in the number of fronts on which innovation was pursued between innovation in core technologies and in applications. Projects with a stronger technical development focus were often those in which less attention was given to technology appropriation (not least because of the prevalence of linear sequential design approaches in which technical development tended to be conducted prior to implementation trials). Conversely, where established technologies were used, time and energy could be conserved for experimentation around the development of content and usages). A concern to minimise the amount of technical innovation that experimenters have to embark upon also underpinned the second factor. This was the desire to ensure that systems are compatible with the existing technical infrastructures and practices of, potential user communities, and the industry standards that are emerging here. There is little point in developing a sophisticated application if it cuts out may potential customers by its presumptions about what technical facilities and skills they possess. UK Bank found this out to its cost when it developed an on-line banking system that could only be accessed by customers with a particular and relatively recent version of a browser - and one that moreover required the user to have a relatively recent and powerful personal computer. Some of the longer running projects we studied emerged prior to the global hegemony of the internet. The project developers have found themselves more or less compelled by user expectations (over and above any enthusiasm they might have had for such a development) to upgrade their offerings from proprietary systems or low-end generic communication systems such as bulletin boards to web standards. This is exemplified in three of our community information/ Digital City The backbone was provided by a commercial Computer-Supported Cooperative Work package which was configured to allow staff to communicate by telephone and videoconferencing while jointly inspecting and manipulating information held on other, industry standards, spreadsheet and word-processing software. The Bank's developers configured these elements together, customised the layout and switched off certain functions that were not seen as necessary. 68 148 case studies. The cases in point are the Amsterdam Digital City (DDS), the Copenhagen Base and the Scottish Community Information System. We thus find that the technical infrastructure of delivery systems for multimedia are developed around global internet standards and cheap standard commodified solutions. There were exceptions. The kiosk developed for accessing the Antwerp Digital City (DMA) was deliberately designed to regulate use (Steenhoudt, 1997c). The keyboard is too low for the average user and it is installed at a too acute angle. This design improves access for the people in a wheelchair and makes it extremely uncomfortable for extended use by others. By inhibiting people from using the information kiosk for too long, especially for surfing on Internet, more people can make use of the services (van Bastelaar & Lobet 1999)! Even here, however, the component technologies, albeit not the final physical configuration, were broadly standard More generally however we found multimedia offerings being aligned to existing delivery systems. This is not surprising. Creating a new distribution channel is extremely expensive, and can only be attempted by the largest players and even then at great risk. Whilst some may see benefits from going it alone, the overwhelming majority will seek to offer their products into markets constituted by existing delivery systems. This is particularly relevant for the producers of cultural products. They can maximise their markets by opening up their offerings for different media channels. Thus the Virtual Museum of Colm Cille ported its products from multimedia displays in the museum to a CD-ROM product and to a web-site. 7.2) appropriation/domestication a) Issues in investigating appropriation and domestication Appropriation refer to highly dispersed processes, occurring often over extended periods and across an array of social spaces and involving an important informal element (this is particularly an issue in relation to domestication; innofusion processes in contrast are more focused spatially and temporally at the point of implementation). It thus presents particular difficulties for systematic research, which we should perhaps consider before we go on to address the specific findings of our study. One established research method involves intensive longitudinal ethnographic studies. However, these have, in consequence typically been limited in focus, for example to a very small sample of selected families (Silverstone 1991, Silverstone & Hirsch 1992)). As we noted in discussing the SLIM research design in Chapter 3, attempts to conduct in-depth study both the development, and appropriation of multimedia face considerable practical and methodological difficulties, not least because the time needed to go from product development to its use can far exceed the typical durations of externally-funded research projects. The SLIM research centres adopted various strategies in confronting this dilemma, including utilising prior 149 research within and outside the network and combining contemporary with historical analysis. The result was a rich set of integrated studies which allow varying insights into appropriation and domestication processes in different settings. Some studies were of projects which remained mainly at the stage of project development and did not provide an opportunity to address appropriation. It is important to bear in mind these differential opportunities in considering the SLIM findings. Four of the studies in which we are able to examine final appropriation (from our array of 30 investigations69) comprise cases in which the final appropriation is taking place within the same 'configuring constituency' that developed the multimedia application. The application was both put together and used within an organised and bounded array of players (hence our use of the term constituency rather than constellation here, explicitly to flag this fact ) - whether this was a single organisation deploying multimedia for its purposes (Edipresse, Scottish Bank70), or a coalition of organisations (Cable School). The Language Course case also involved proxy external users (teachers and students), bought in a rather tokenistic manner so that the pilot could be tested and demonstrated at a forthcoming trade exhibition (Mourik, 1998). Where appropriation is within a single organisation or close alliance of organisations, appropriation processes can be expected to arise more quickly - as the linkages between players involved are direct and appropriation is more readily subject to purposive management and incentives - than for example in mass consumption of products, diffused through the market (where links will be indirect, and where the consumer may be exhorted to take part but cannot really be directed). In contrast, mass-consumer products and services tend to have a particularly long development cycles. If we wish to investigate multimedia development, in the lifetime of a short-term study, projects are likely still to be at the developmental stage, and not to have 'rolled-out' for wider use - thereby offering relatively little scope for studying final appropriation. The SLIM research programme investigated a number of products that were geared towards such mass markets and domestic consumption. The time needed to set up and conduct studies is not the only constraint in exploring appropriation in such settings. The other, as discussed by Silverstone (1991), is the private and closed character of the domestic space. These barriers, of time, space and openness posed particular problems of access to understanding local appropriation (barriers which of course apply to technology developers as well as to SLIM researchers). For example, there was little scope for examining local appropriation processes in the Norwegian Girls-ROM case - in which a CD was given away in a magazine (especially since the project remained something of a one-off). The final appropriation process here was rather ephemeral. 694 cases by each of 7 SLIM Centres plus Edipresse and UK Bank - though the latter two both contain more than one case each. 70In the case of Edipresse and UK Bank some applications also involved relatively standardised information services (typically based on the world-wide-web) for a wider array of customers. 150 For these various reasons, the SLIM project was not able to address the consumption/ appropriation of multimedia in the home to the extent that was anticipated in the original research design (partly as a result of methodological and resource constraints and but mainly because the empirical reality was rather different to that anticipated - in particular we did not see the expected roll-out of multimedia products to widespread domestic use). However this did not prevent us from examining appropriation. Some of the multimedia applications were developed for use in and by commercial organisations; most of the educational multimedia developments were geared around their experimental use within educational establishments. Many of the products developed for mass markets (eg in edutainments) were not exclusively for consumption by the individual in the privacy of his/ her home. For example, cybercafes emerged as important sites for the collective and public consumption of various multimedia products. Here, and in other places, appropriation was actively organised. For instance, following development of the prototype 'Virtual Museum of Colm Cille' multimedia display, the developers, Nerve Centre, embarked upon an explicit appropriation phase as they sought to encourage the uptake of their products. Virtual Museum of St. Colm Cille The project was established within the ’Nerve Centre’ project based in Derry City. The virtual museum exemplifies the process of learning by doing in that the ’Nerve Centre’ had substantial experience in ’traditional’ media such as film and photography, but this was one of the first projects undertaken using multimedia. Further, the development of local content about the Saint’s life was undertaken with an eye to global markets and representations - dialogue between the local and the global both in terms of modes of representation and technology was integral to the development of the project. In other words local content, takenfor-granted representational conventions and technologies together with established practices had to be replaced by a praxis-based social learning in the transition to multimedia. The content itself was developed by and for young people, but other actors were enrolled in the project by the Nerve Center in order to gain funds and to ’ratify’ the project through the investment of cultural capital by older citizens who had little idea of the potentialities of multimedia. One concern was with the representation of a local area or culture to the world through the use of multimedia. The expectation was that younger citizens would become increasingly aware of their local culture (and the value of it) through placing it into a global context. Further, we note what Kerr (1998) calls ’social learning through Trojan horses’: through an attempt to enfranchise older citizens in the uses of ICTs by the use of local content using new media technologies expressing community memory and history. (Kerr 1998) 151 b) The appropriation of multimedia content Most discourses about multimedia focus upon the development of advanced multimedia artefacts. In contrast, our study of Social Learning in Multimedia highlights three key features: i) ii) iii) the importance of appropriation as a crucial stage of innovation and learning in the overall 'circuit of technology' from design to use (Cockburn & FurstDilic 1994); and, the importance of domestication and the incorporation of artefacts in local practice and culture; and, following on from this, the importance of multimedia content rather than artefactual design In relation to the former, some of the key potential difficulties and uncertainties surrounding the future of multimedia are associated with technology appropriation rather than development. The appropriation process is highly unpredictable and rather difficult to coordinate involving as it does, large numbers of players who may not be committed to the technology, but who may need to be enticed to use it (particularly with products geared to a mass of private consumers who cannot be directed or managed in a 'command and control' model). How, for a start, can potential customers be encouraged to see these 'strange' offerings as potentially relevant and useful to them? If it's difficult to get potential customers to commit time and resources to investigate whether these relatively unknown products are worthwhile, how can suppliers motivate them to make the more substantial investments of time and resources in gaining access to the technology and learning how to use it? Following on from this, the appropriation of multimedia can be extremely slow - it tends to operate on far longer timeframes than initial technology design and development – and rather obdurate. The difficulties associated with appropriation were brought home sharply in our study of the multinational marketing efforts of Compuflex. Compuflex found that simply translating its US products into European languages was not sufficient for it to be able to market these products successfully in Europe. Other kinds of translation in underlying concepts of use would also be needed to make them attractive. Even an organisation with the global financial and technical resources of Compuflex Corporation was unable to impose its offerings onto the market. In analysing the domestication of multimedia products and services we emphasise the complexity of the process involved in the development of meanings and practices of use around the technology, involving diverse users and intermediaries as well as designers and developers. We have already examined under the heading design/ representation that efforts of designers to articulate concepts and means of use and convey these to users. We now address the role of users in selectively taking up (or rejecting), adapting and reinventing these concepts. The interaction between supplier efforts and appropriation has complex and contradictory outcomes. Whilst noting some areas of dynamism in relation to the 152 uptake of multimedia products, developments remain very uneven - reflecting the intricacies of the social learning process. In this respect, the SLIM studies highlight the continued importance of 'the local', despite visions of a global convergence of cultural forms around offerings from multimedia supply. This was reflected in a complex set of outcomes in the cases examined between local and global elements and between what was generalised and what remained specific. For example, there had been widespread adoption worldwide of certain discrete ICT tools and component technologies. However this has not occurred across the board, and in particular it does not seems to apply in relation to more complex information and cultural offerings. This cultural conservatism, and emphasis on the local is, Preston (1999) suggests, particularly significant in relation to heavily cultural or symbolically-laden content types/ genres of multimedia information and content. It is in relation to this type of products (for example, those concerned with local community history and community development) in which questions of identity and self-presentation are likely to be salient, where local knowledge and authenticity are likely to be crucial. Despite the widespread expectations of the rapid take-off of new multimedia markets, the social learning perspective, and the question of appropriation in particular, suggest that the extension of use of new multimedia products may be rather slow overall. We note that the uptake of new technologies has often fallen substantially behind expectations, particularly in relation to large scale technological systems such as new ICT delivery systems (for example the failure of UK Prestel (Schneider 1991). However we also note the uneven-ness and fluidity of development - and consequent difficulties in predicting the future development and uptake of technologies. We must also take into account historical experiences which demonstrate that, in some circumstances, technology can be taken up far more quickly and completely than anticipated at the outset. In this process, the significance and use of artefacts is often transformed from that originally anticipated. For example fax took off very rapidly in the 1980s once the technology became cheap, robust and easy-to-use (Coopersmith 1993). A similar story can be told of mobile telephony in the 1990s. The even more explosive growth world-wide in the internet in the period of this study provides the most striking illustration of this point. The intuitively simple interfaces of world-wide web have become widely accepted. We can conclude that progress is likely to be rather uneven. We remain rather uncertain about the overall pace of change - though we can be fairly sure that the predictions by technology promoters of a rapid, across the board uptake of multimedia will not be fulfilled. the social learning of genres and metaphors and the normalisation of technology Perhaps with hindsight it will be evident that we are currently in a transition period, from an era when ICTs were largely the preserve of a technical elite built around its sub-culture, skills and perspectives, to an era that may qualify better for the description Information Society through the uptake of these technical capabilities by a wide range of groups across society. Norman (1988) has highlighted the possibility of ICTs becoming 'normalised' in social life. A number of 153 ICT artefacts have become 'transparent' in their use and utility – for example the telephone, the fax, the mobile phone – and widely adopted. He also criticises the failures of contemporary technology design to create computer-based technologies that are indeed readily usable, useful and convivial (Norman 1998). Multimedia technologies have not yet become stabilised. For example, even a group such as UK Bank's Research Department, possessing high-levels of technical skills, found themselves stretched in their attempts to configure together the notionally standard electronic mail systems, network and operating systems, telecommunications links, and personal computers that they needed for their homebased teleworking trial. There is certainly enormous scope for technologies to be better designed and more transparent. Consider the problem of what we might describe as the 'baroque' design of many contemporary artefacts such as the Video Cassette Recorder, with programming facilities that most users are unable to exploit. However the SLIM study highlights the fact that the development of these complex multimedia applications cannot be reduced to a matter of design, but depends equally upon wider social learning processes about the potential utility and use of supplier offerings, in which certain capabilities, presumptions and genres for using artefacts become widely communicated as a resource for design and appropriation efforts alike. In our discussion of technology design/ representation, we drew attention to the uneven-ness of progress in developing common understandings about tools and overall narrative structures and genres, that may help to inform the user's navigation around and use of the system as well as designer efforts. In particular we pointed to the rapid uptake of certain design/ use metaphors (buttons; windows, scroll bars) particularly those associated with the world-wide-web. It is interesting to consider how this global familiarisation has been achieved. Partly these have become disseminated as common components of computing - eg the graphic user interface. However, another, and perhaps more significant vehicle has been the transfer of elements between media. Television, in particular, has been quick to take up and popularise certain elements. For example TV programmers (especially those programmes geared towards children and the 'youth' market) were quick to include windows and buttons in the screen presentations (paradoxically in the case of buttons which the viewer cannot yet utilise with broadcast TV). Cross-media reinforcement - making TV seem more modern, and IT seem more dynamic - may play an important role in the borrowing and evolution of genres of multimedia use. They also reflect the fact that media producers have been one of the earliest users of advanced multimedia technologies - and quickly learned to incorporate these into their programmes. Buttons and windows provide rather discrete and simple metaphors for how to use a computer-based system- and ones that can be readily understood and applied in a whole set of areas. When it comes to developing richer concepts of how people might find a system useful, more complex metaphors may be needed. However one might expect that, under these circumstances, the processes whereby they become generalised may be more obdurate, complex and uneven. 154 An effective metaphor may provide a means of conveying more widely ideas about the meaning of the artefact - about its utility and its use. Thus the idea of the Digital City has proved a readily understandable device for communicating the possibilities of the Internet as a starting point for organising a virtual community. Indeed it proved to be an attractive device that could not only inform the activities of technical enthusiasts and wider publics, but also policy makers and the media. The Digital City metaphor was particularly important to the Amsterdam Digital City (DDS), helping the project more beyond its founding constituency, a small subculture of technology enthusiasts, and making it understandable to lay-people; something the media could write about. The success of DDS cannot be divorced from the way it became a cultural phenomenon (van Lieshout). Finally, returning once more to the question of the transparency of technology, we note that hitherto, multimedia technologies have remained largely in the hands of technical specialists, and an enthusiast culture of early adopters. However we should bear in mind that this will change profoundly when larger parts of the population 'buy-in' to understandings of the use and utility of multimedia products and services (with the establishment of new genres) and acquire the capabilities needed not only to use them but also to deploy and adapt them for their particular purposes. We should bear in mind that these capabilities are becoming much more widely available through various avenues. For example we note the efforts of educational establishments to utilise ICT capabilities, not only as a technical subject in education and post-school training but also as part of the generic skills offered at school and especially in Higher Education. Although, as our Education studies revealed, most teachers lack these skills (and in the short term had little prospects of acquiring them since the training costs would be prohibitive), as new generations of students come through the education system and come to constitute the new generation of primary and secondary teachers we can anticipate the much wider provision of such skills, to the extent that they become as much taken for granted as literacy and numeracy. Other routes of skill acquisition and circulation are no-less important - particularly since many of the skills required to use multimedia are contingent and experience-based (eg how to configure systems together, how to iron out bugs and find short-cuts in using packaged systems). Stewart (forthcoming) has pointed to an important informal knowledge economy, as individuals who have attained the experience and skills needed help neophytes overcome the 'barriers to entry' of getting involved with multimedia use and development. Similarly people may apply in their everyday life skills and familiarity acquired at work (Stewart forthcoming, Sørensen 1999). We suggest that the private and community use of computers often involves an informal intermediary (perhaps a friend, neighbour, workmate or relative - but someone informally involved rather than as part of their job). It is through these generalised exchanges that multimedia capabilities can in time be expected to become part of our general culture - in this sense to become transparent (in the same way perhaps that the way of using the telephone, and the skills involved have become taken-for-granted and indeed invisible to us today). 155 The SLIM project found some evidence of the dissemination of familiarity with multimedia and its usage and a growing level of technical skills and competence amongst different groups across society - as illustrated, for example, the emergence of sub-cultures of computer enthusiast young women (as evinced by the shift From Spice Girls to CyberGirls: see box). Such a development would have profound implications for the appropriation of multimedia. It also has implications not only for efforts to overcome gender inequalities in access to multimedia, but questions of social inclusion/ exclusion more generally. Though acknowledging the importance of existing barriers which may tend to exclude marginal groups, the SLIM studies note other dynamic processes that may serve to undermine or change as well as reinforce these barriers. Policies addressing social exclusion should be aware of and seek to exploit these potentialities - and should in particular encompass informal social and cultural processes as well as the established policy agendas which tend to be couched in terms of formal initiatives, e.g. for training. Indeed one implication of the 'CyberGirls' study and the 'turn to entertainment' noted in our cross-cutting study of Multimedia in the Home (Sørensen 1999), is that we should look at the potential use of these technologies for play and other pleasurable activities, rather than just their functional utilitarian applications, to encourage appropriation of these technologies and the acquisition and sharing of the formal skills and the important informal culture needed for setting up and using multimedia systems. From Spice Girls to Cyber-Girls This Norwegian study focuses on a group of computer uses that has not received much attention hitherto: computer fascinated girls who are enthusiastic in exploring uses of the computer and Internet. It points to a relatively successful reconfiguration of the gendering of computers. A key factor was the development of the Internet that has in a very profound manner changed the meaning of computers and made them far more attractive to girls. The computer is used for intimate social communicative roles, as writing a diary and writing poems. The girls perceive the computer as communicative device, by itself and as gateway to Internet. Surfing, for fun and chatting are amongst their favourite uses. The emergence of a subculture of computer enthusiast girls was underpinned by a change in the gender distribution of computer skills (this is not to suggest that the construction of girlhood has changed, since these young women emphasise their gender normality). These girls acquired their computer skills mainly through learning by doing. This may presents a paradox to education policy, since computers in education initiatives often seem to have had the effect of driving women away. To get girls interested in and fascinated by computers we may need exploit the attraction of playful activities - to explore the potential of the computer as a toy - and one which you can continue to play with as you get older (Nordli 1999). c) Appropriation of artefacts 156 Given that most of the multimedia applications being developed and used are based upon relatively standard technology platforms (typically a personal computer with CD-ROM drive or internet access), the level of technical reconfiguration during the appropriation process - and thus, for example the potential for innofusion in the technology platform - could be expected to be rather modest. This does not mean that their application was seamless and without effort. Indeed these technologies are still difficult to implement and use. They are by no means as 'easy to use' in the way that suppliers would have us believe - that they can simply be plugged in, switched on and used. Initiatives that promise to deliver hardware and software that are easier to use have been criticised for having little visible effect. 71 At the same time, the continued technical and commercial dynamism and turbulence surrounding the development of technical components can undermine attempts to stabilise technological development. As a result these offerings are often far from finished particularly in relation to their interoperability. We return to these points below in relation to standardisation. At present the competences needed to deal with these kinds of technical problems are not yet widely disseminated. Many of the skills required are highly contingent and relate to particular problems in operating and configuring together specific technology components. They involve rather informal kinds of knowledge, that can only be acquired through practical experience. 'Getting started' with a new system typically involves a rather substantial investment of time and effort in configuring a system and sorting out the many 'teething troubles' that arise. Contact with someone who had already been through this process could be extremely helpful. Novices often drew upon this kind of experience through informal channels.72 Reducing these entry costs was one of the motivations for people to go to cybercafes, where someone else had gone through this 'learning curve'. More generally overcoming such barriers in getting started and learning how to use a system constitutes one of the main reasons for the importance of 'appropriation intermediaries'. d) the mediation of consumption and usage Appropriation as an intermediated process 71 For example, Steve Ballmer, Microsoft's president, recently announced to the annual Windows Hardware Engineering Conference (WinHEC April 1999 Los Angeles) an initiative to deliver: 'Software that's easier to use ... (on) hardware that's easier to operate'. The Easy PC Initiative led by Intel and MicroSoft, also involves Dell, Hewlett-Packard, Compaq, Toshiba, NEC, IBM, Fujitsu, and Gateway. Ballmer promised "reduced hardware complexity, the right software support and a rebirth through Easy PC of the consumer computing experience". However a similar initiative, Simply Interactive PC, announced at WinHEC in 1996, has been criticised for having had little visible effect. Jack Schofield, 'Windows 98 reprieved', The Guardian On-line section Thursday 15th April 1999. 72Individuals have different levels of access to such local experts - though they turned out to be a very important - for example in the hypertext programme used in the IMMICS media education course, particularly for those students who were less committed to and experienced in the use of these technologies. 157 In Chapter 6 we flagged the importance of 'appropriation intermediaries' and suggested that they may play two key roles in the appropriation of multimedia: i) First is their potential enabling role in consumption and use - in reducing the barriers to access, entry costs and uncertainties for 'users, particularly in relation to non-specialist users. ii) The second possible role of appropriation intermediaries is their potential contribution to innofusion. Every implementation involves some local reconfiguration of systems. This may involve quite a substantial technical effort given the lack of stable standards and methods of operation. Knowledge of these activities can be an important resource for technology developers not only in identifying 'bugs' and problems but also in refining their products and developing new products. Their potential contribution can be evinced by the efforts of suppliers of ICT products to develop more sustained communication channels with their consumers through 'alpha' and 'beta' testing. Appropriation intermediaries could be used as a similar resource. We shall explore these in turn. consumption and usage Particularly in relation to new kinds of multimedia product, appropriation may need to be an active and mediated process. One of the most interesting examples of the role of appropriation intermediaries in consumption and usage is provided by the 'Virtual Museum of Colm Cille'. Before the development of this product had been completed, the developers, Derry Nerve Centre, embarked upon an active process of diffusion and supporting appropriation, in an attempt to build a market for example introducing local schools and community groups to the new and unfamiliar medium - recognising that otherwise uptake would be slow, limited and uneven. The Nerve Centre case highlights the fact that considerable resources may be needed to support this kind of broader dissemination/ appropriation effort.73 One obstacle it encountered was that public support for commercialisation (i.e. for the resources needed to foster widespread appropriation) proved more difficult to obtain than for the initial research and development of new technology products. Craigmillar Community Information Service faced a similar challenge of actively enrolling new user groups. Though its starting point was established community groups, these showed little interest in using the system as initially conceived. CCIS staff changed their strategy and sought to foster the development of particular groups of user such as the Senior Citizens (the CyberGrannies). What is of interest is the way that the potential relevance of CCIS is constructed around each group the senior citizens are assumed to want to use the web for ’remembering’ and to contact groups of their peers. CCIS also linked its information system up to external This case illustrates how intermediaries involved in cultural products sought to utilise existing popular cultural outlets such as film festivals in order to find cheap, low risk and trusted avenues for disseminating and popularising their products (Preston 1999). 73 158 information systems to make the system more relevant (a move which promoted controversy about the proper users and uses of the system [Slack 1999]). The final cases we shall discuss here involves the cybercafes (Stewart 1999). These were public sites for appropriation of new commercial multimedia offerings. The cybercafe managers were very active in creating an appropriately-designed and attractive setting - a safe and enticing place where particular groups could engage with new supplier offerings. They created an ordered context for multimedia appropriation.74 The managers were also active as market mediators – searching out what new games and other products were becoming available, evaluating them and selecting appropriate offerings. Finally they played a role in implementation and final appropriation - installing products and learning how to use them- thus reducing barriers to entry/ foreshortening the learning curve for new users to get involved with these various offerings (as well as, post-implementation, helping to build up and exchange local knowledge about how to play). Cybercafe managers thus provided a very important bridge in the appropriation process. They constitute particularly salient instances of the 'appropriation intermediaries' that we argue are critical in the current early stages in introducing new multimedia technologies. In this case their role was commercially motivated - users were willing to pay them (through usage charges) for their role in adding value to commercially available products and reducing search/ access costs, reducing risks and sharing costs of acquiring new commercial products. The cybercafes in our studies did not seem, however, to have been engaged as a vehicle for direct feedback to future technology supply. This might be an important resource for suppliers to tap. e) unanticipated use - and its potential as a resource for innofusion and domestication efforts Our study highlights the complexity and unpredictability of the appropriation process. In virtually all cases where we were able to examine implementation and use of multimedia we found more or less clear discrepancies between the concepts underpinning the design of the multimedia products and the richness and diversity of user appropriation experiences. Unanticipated uses and outcomes were frequent. We will review some of these findings here. They confirm our insistence on the importance, and active nature, of the appropriation process. From the point of view of social learning, a key question concerns whether and how we can integrate knowledge emerging from appropriation processes into future multimedia supply. Examples from Educational Multimedia 74In the cybercafes, as in the case of DDS, an important element of this appropriation process concerned the establishment of formal or informal rules for how participants (citizens or customers) should behave. Appropriation in collective settings involves certain types of decision about what kinds of behaviour are proper and appropriate. Such local rule making draws upon resources from the broader setting of norms and regulations. We point to the reciprocal interaction between wider rule-making activities/ regulation and local regulation in creating and sustaining technology regimes. 159 We will start by exploring our educational multimedia experiments, where we have fairly systematic information about appropriation. For example even though the Language Course case involved users in a tokenistic manner (teachers and students we brought in so that the pilot could be demonstrated at a forthcoming trade exhibition) it is striking that, despite this, the Language Course developers DID learn important lessons from these users about how their systems might be used in different ways to those anticipated, and the consequent need to reconfigure the system. For example the designers had been working with the presumption that the children would only use the system in the ways they anticipated – overlooking the rather well-known propensity of children to play and their creativity in inventing playful uses of the system. In the original design, a voice switch was used to control the source of the video-channel broadcast. The children invented a game of making noises to bring the attention of the videoconferencing systems to them and then hiding. To try to regulate the behaviour of the children, the teacher was given a device to turn off the child's microphone and the children were given a button to alert the teacher if they wanted to speak (Mourik, 1998).75 Often unanticipated uses related to working around the shortcomings as well as the affordances of the system as installed. Again with Language School the time delay between audio and video ruled out its intended use for pronunciation exercises. The teachers had to find new ways of teaching around the affordances of the system. The medium - for example of the electronic whiteboard and voice - was available to all. Teachers could not teach individually because then the others would not be able to learn, so the only way to help the slow pupils was by giving them more advice and less difficult assignments. Education changed from being teacher centred to student centred. The teacher became more of a facilitator and mediator between the information available in surplus through the system and the student. Each student in other words defined their own course and rate of progress (Mourik, 1998). The Danish Bornholm tele-education case identified similar types of change in the role of the teacher. Whereas some had expected the new medium to 'deliver' education by the teacher, the latter needed to adapt teaching methods and materials to this new communication medium (Hansen. 1998). Teleteaching on Bornholm The Bornholm Teleteaching project shows how users have to develop new means of interaction and presentation when moving to new media. The project was designed to deliver teaching to a relatively isolated island off the Baltic coast of Denmark through the use of video telephony. Social learning dimensions center around the need to translate the teaching content to the technology and to develop pedagogical styles that suit the medium. Practically, the introduction of It must be noted that this represents a rather mechanistic response to the 'problems' of unanticipated use, based on technical constraints rather than developing social conventions to regulate computer-mediated interaction. 75 160 a distributed teaching and learning context profoundly changes the relationship of teachers and students. Students and teachers are not co-located and there is a need for greater awareness on the part of teachers about classroom management issues: teachers have to ’host’ programmes on the video conferencing system as opposed to teaching as if the students were in the same room. The usual rules for conversational interaction do not apply in such environments and new formal rules need to be developed if the lesson is to proceed effectively. Such simple tasks as asking questions are made problematic in that the technology does not facilitate the use of gaze and intonation to gain students’ attention and to indicate that the question is for them. Teaching materials must also be changed to facilitate the use of distance teaching and learning in that the materials must explain fully what is required since the teacher will not always be available to elaborate on what the task demands. Teleteaching on Bornholm led to a change in the roles of teachers and students. The project shows that it is no simple matter to translate content across media since this involves a switch of genres and attendant representational conventions. To be effective, new skills must be learned by all involved. (Hansen. 1998). Broadly similar sets of issues came up in the Telepoly case using a broadband network for distance teaching in the Swiss Higher Education sector. Some teachers failed to adapt their teaching methods to the new medium, and consequently received negative feedback from the students and lost enthusiasm for the project. Two lecturers were successful - though in very different ways. One adapted to the particular style of communication imposed by the system, such as addressing the mobile camera, talking slowly, and thinking constantly about possible interactions with the virtual classroom, visible only on a TV screen. He had put a lot of effort into preparing this teaching including putting documents in electronic form for the whole course. The other lecturer installed his teaching on Internet with many hypertext links with demonstrations, which the students could consult whenever they wanted (Rossel, 1998). In the Cable School experiment in the UK, the eventual outcomes were far removed from the initial conception of teachers in different schools sharing worksheets and other resources electronically. Though this was defeated by the failure of the initiative to provide resources or institutional support and incentives for teachers to develop such sheets, 76 at the end of the project, some uses of the local servers and networks had become institutionalised within the participating schools, including Paradoxically, since Cable School rested upon collaboration between schools, there seems to have been little attempt to create a supportive culture and norms between the subject teachers involved across various schools. There were a few meetings of teaching management in the various schools involved but no direct contacts between the subject teachers who needed to be at the heart of the experiment. These teachers told us they felt inhibited from presenting their offerings on the system, for fear they might attract critical comment from peers in other schools. 76 161 email and the collection of world-wide-web pages as a teaching resource (Slack 1999). What is clear is that we are still at the early stages of the appropriation of multimedia within education (and elsewhere). The key challenges still surround the articulation of technical and pedagogical potentialities - and there have been some pointers from our cases in terms of the shift to student and resource centred learning approaches. Many of the educational initiatives we have examined have been primarily in the hands of technology specialists or enthusiasts - whose technically-driven visions have often proved 'out of kilter' with the circumstances and culture of most players in the appropriation domain. It remains the case, as noted earlier, that a key prerequisite for more effective utilisation of multimedia in education will be the involvement of teachers (and through them the students), in exploring and developing the potential of these media in different local educational settings. The potential contribution of teachers as actors in this process (rather than simple recipients of technological offerings) has not been properly recognised. Indeed Egyedi Lieshout and Bijker (1999) describe the support that teachers receive as 'shameful’ - noting that limited funding does not allow for training and support of the average teacher, and that enthusiasts from teaching staff often contribute in their in leisure time. Innofusion and other appropriation lessons could easily be lost because teachers were not properly institutionalised within project management structures that focused upon commercial suppliers and education managers. There were, however, differences in this respect between our cases. There was, for example, some indication that teachers in Higher Education tended to have more autonomy, and better access to skills and other resources needed to shape multimedia, than for example those in school settings. These kinds of creative efforts of innofusion and domestication were by no means restricted to these educational multimedia applications, but was found wherever multimedia was being appropriated. Multimedia artefacts remain unfinished – in the sense that they have not become stabilised or fixed in their meanings and usages. Users made different interpretations of the meaning and uses of artefacts depending upon their context and standpoint. In some cases they added functionalities to the artefact that were not foreseen (or even implied) by the designers, or changed the understanding of existing functionalities and their use. For example the UK Bank's videoconferencing experiment showed them which of the bundle of functionalities they had configured together from commercial packaged solutions would prove helpful for remote cooperative working (intriguingly, the video channel, though helpful in setting up a sense of telepresence, tended to get switched off, so that people could use the screen to simultaneously examine documents and spreadsheet data, and discuss it on the audio channel).The case also involved 'social innovation' through the elaboration of new sets of social conventions for such video-mediated communication - mainly to find ways of working around the restrictions imposed by the technology (eg difficulties of distinguishing people in group meetings; the need to take turns in 162 speaking77 and regulate who would go next; the need to prepare in advance for such meetings) (Cashin, Procter and Williams 1999). The innofusion processes we noted went beyond what has normally been flagged in discussions of 'supplier-user links'. The flexible and malleable character of artefacts that are based on ICT, means that the traditional borderline between innovation and diffusion has become blurred. This was particularly notable in cases such as Amsterdam Digital City (DDS), which exemplify the evolutionary model of system development, which went through a process of continual development and one which involved many players (Lieshout DDS). This kind of feedback from contexts of use to technology supply resumed by innofusion often - indeed mainly - takes place through informal channels. Many players are actors and contribute to this process. What is learnt is hard to predict, and often highly contingent to particular settings. Questions then arise about what avenues exist to carry and deploy this knowledge more broadly: as a resource both for the further appropriation of the artefact and for future technological supply. Multimedia trials and experiments constituted spaces for these different actors to meet and communicate. Intermediaries provided an important avenue for the further exploitation of these experiences in the form of embodied knowledge. There were, however, particular issues concerning feedback from final users, who may be far removed both spatially and in terms of their societal roles. Though the previous section we have focused mainly upon the substance of what was learnt in social experiments or commercial trials, such Multimedia experiments are about process as much as substantive outcome. Learning relates to both the design/ application of new multimedia products/ services, and the processes whereby they are developed and appropriated. Indeed learning about 'the technology' turned out often to be only a relatively minor component of the experiment; more important was finding out about the behaviour, concerns and commitments of other players (suppliers, intermediate and final consumers, promoters, regulators and policymakers). Thus trials and experiments are a way of learning how to develop new Multimedia products as much as they are about what these products and services will look like. This involves learning for example how to build constituencies of suppliers of component technologies (eg communication networks, terminals) and complementary products (particularly in relation to content); how to obtain the necessary resources and build commitments and alliances with players inside and outside the constituency, as well as how to conduct user trials etc. Many of the most important lessons from an experiment may be more general reflexive ones - about how to learn through experiments about constituency In its initial implementation, the UK Bank videoconferencing project also encountered problems with voice-switching of the video channel that bears some similarities with the Language Course case. When the systems was used in half-duplex mode - grunts of affirmation would transfer communication from the speaker to the listener. 77 163 building, how to address uses/ users etc.. This has important implications when we come to consider the implications of this study for public policy/ business strategy in Chapter 9. But first let us consider the implications for the organisation and management of multimedia experiments. 164 Chapter 8 The conduct and management of multimedia experiments In Chapter 6 we discussed the processes through which multimedia constituencies and constellations were developed in the course of multimedia experiments. Our discussion highlighted the active role of intermediaries in bringing together players with diverse interests and perspectives, to establish visions and incentives to motivate their involvement in order to secure the intellectual and material resources needed. We noted the diversity of settings for multimedia experiments - which constituted particular 'translation terrains' shaping multimedia development. We emphasised the contingent nature of many of the problems addressed and the creative way these were tackled by intermediaries. This poses particular problems in terms of how to draw lessons and policy implications from multimedia experiments - a point we return to in the conclusions in the next chapter. This chapter therefore seeks to analyse the choices in the way that multimedia experiments were conducted and managed. From the SLIM and related studies we can identify a set of recurrent issues and dilemmas that arose across a range of very different multimedia experiments. We pick out three: 1) 2) 3) 8.1) mode of experimentation and mode of control: in conducting an experiment there are important choices about how much control should be exercised (and when closure should be established) over participation and the goals of the project. We found that projects are located along a spectrum between experimentation and control. involving the end-user; despite the frequent espousal of the importance of 'involving the user', our cases highlighted the frequent lack of such involvement, and the difficulties in such an endeavour. the role of the state and managing the public - private boundary; despite rhetorics of market provision, the state continued to be a key player in promoting and providing resources for multimedia experiments; most multimedia projects are 'hybrids'. This raises issues about how to integrate public and private elements. Mode of Experimentation and Mode of Control in Multimedia Experiments The SLIM project has investigated a wide range of multimedia experiments. An important empirical variable between cases, across three broad application areas (Education; Public Administration; Cultural products), concerned their location on a spectrum between what we have described as the Mode of Experimentation and the Mode of Control (van Lieshout et al, Lobet et al. Preston et al.). This relates to the level of open-ness of the project both regarding participation in the project (and the scope for extending participation) and regarding its goals. In the mode of control, 165 typically, there may be an attempt to define the goals and expected outcomes of a project in advance and rather precisely, and also to prescribe who will participate in development decisions and the resources used. In the mode of experimentation, there is scope for a more open learning process, with greater open-ness to other participants and agendas. The benefit of the mode of experimentation is a greater willingness to change direction and approach in the light of experience; the 'downside' may be uncertainty and loss of predictability and control over the direction and outcomes of a project. Lieshout et al (1999) have distinguished a number of different dimensions of the modes of experimentation or control. Particular experiments may be differently located on these dimensions on a spectrum between experimentation and control. These elements, which may be more or less closely related, are summarised in Figure 8.1 Mode of Experimentation Relevant dimension Mode of Control open to change process oriented minimum of boundaries and constraints exploratory experimentation ---dominant attitude ---- goal/objective ----- environment ----- mode of practice --- open style of management --- style of management --- semi-open to change product oriented explicit boundaries and constraints purposeful experimentation hierarchical style of management Figure 8.1: Dimensions of mode of experimentation vis-a-vis mode of control 166 open to change IMMICS Bornholm teleteaching open-non hierarchical process Language course Cableschool exploration Telepoly no boundaries Figure 8.2: Diagram of Dimensions of Experimentation and Control Figure 8.2 seeks to display these different dimensions of the mode of experimentation and control on a single diagram (capturing for comparative purposes the various SLIM educational multimedia studies [van Lieshout 1999]). For each dimension, proximity towards the control end of the spectrum is at the centre of the diagram, while more experimental approaches are shown at the edge. 167 Designers Intermediate Users Proxy Users Final Users DESIGN /DEVT. Mode of control TRIAL USE control over goals and learning Mode of Experimentation open about goals, some learning USE Figure 8.3: Multimedia experiments – mode of experimentation and mode of control Control is of course very much a question about who is seeking to exercise control and over what issues. Figure 8.3 seeks to clarify the contrast between multimedia experiments conducted under the mode of experimentation and the mode of control by showing what these might look like in relation to the strategic options for multimedia experiments that we established in chapter 6. It shows how these differences in control and open-ness relate to two key dimensions: first the range of players involved and second the extent of their involvement in the various stages of project development. In other words, in the mode of control, control over the project is narrowly restricted (perhaps to a narrow technical or managerial group), and broader arrays of users are doubly excluded - both by not being party to the experiment and because the key design decisions have been taken prior to the rollout and use of the system. This exclusion and preclusion is less clear in the mode of experimentation. In some of our cases the adoption of the Mode of Control seems to reflect particular contingencies for the Multimedia experiment, notably in the case of projects with a stronger technical development focus, where the mode of control corresponded also to the imposition of a narrow technical division of labour One example was the Pericles public administration application. The adoption of the mode of control can be related to one of the goals of the project - to develop an on-line front-end to the local authority's internal information system which involved 168 a considerable development effort. However, this was slightly exceptional, amongst the public administration cases we examined. Most of the other SLIM public administration applications studied (and the SLIM cases more generally) are not so tightly geared to particular information systems and organisational practices, but are relatively independent stand-alone systems for information sharing. In addition we find that the development of some of the cultural content products would also seem to resemble the mode of control insofar as cultural production is carried out by professional designers. This links in to the points made in the preceding chapter about different modes of design. The more authorial model (mode of creativity) of design/ representation may also suggest a greater privileging of design than the mode focused upon accountability to the user. However there are many dimensions of control – and the relationship between mode of design and mode of control is unlikely to be simple. Indeed, the link that can be drawn with particular exigencies should not disguise the scope for choice that existed between experimentation and control. Thus, rather similar types of projects (such as our Digital City/ Public Administration cases) could differ sharply in this respect (Lobet 1999). Another immediate empirical observation from the cases is that there appears to be no simple relationship between the selection of a particular mode (of experimentation or design) and the success of the project. Our concern with social learning would, in principle, suggest that there might be important learning advantages from a more experimental approach. On the other hand we have already noted the potential costs and uncertainties that may arise from the latter’s unstable and dynamic development setting. The calculus of risks/ costs and benefits within a project of course represents only a part of the overall assessment. Indeed as Figure 8.3 reminds us, the formal conduct of a multimedia experiment, and the way its goals are initially set, only forms a part - and perhaps only a small part - of the broader domain for social learning in which multimedia is ultimately innovated and appropriated. The conduct of experiments thus needs to be addressed against this backdrop. The appropriateness of choices, for example between modes of experimentation and control, can be only partly assessed in the light of the experiences within a project, and may ultimately depend upon the outcomes within this broader domain. On the other hand, as Sørensen (1999) points out in relation to mass-market products for the home, designers committed to the mode of control will tend not to be interested to examine social learning when their products are consumed; "users are seen as a challenge of persuasion, to make people buy the product, not as a resource to make improved designs." Much depends, therefore, on subsequent, broader processes of social learning outside the formal boundaries of the experiment. SLIM research has pointed to the success of the Evolutionary development model (exemplified by DDS), in which the experimental approach is maintained throughout the appropriation as well as developmental stage. What may be at issue here may revolve around rather 169 different views of the lifecycle of a multimedia artefact as on the one a finished product and on the other as a product that remains perennially unfinished. Whilst some products may become stabilised and ‘transparent’, the overall thrust of the SLIM studies is that multimedia applications and usages in general today remain profoundly experimental. Hypermedia Use in Education: Experimenting with Dis-closure This case is about the use of a hypermedia programme for a new visual culture course. The course has two aims. Students are to learn how war is represented in different visual media (e.g. photography and television), and they are to learn how to do research in a non-linear way. The use of hypermedia software is a means to make explicit the process of non-linear structuring of research data. The software is new to all concerned. Two teachers treat hypermedia use as an experiment. They do not define a best practice beforehand. They seek diversified hypermedia use and therefore steer as little as possible. Their attempt to exploit the interpretative flexibility of hypermedia use and postpone ‘closure’ is a main ingredient of what we term diversification experiment. It proves difficult for teachers not to give any pointers at all for the use of hypermedia for non-linear research. They partly succeed. At times, the lack of guidance frustrates students. The students’ capacities of home computer user and experienced learner affect their hypermedia use in the course. Some hypermedia uses are constrained by the software. In certain cases students find ways to work around these constraints. Unexpected hypermedia uses develop. Postponement of closure appears to encourage new uses. Furthermore, it requires frequent reflection on the use of hypermedia for non-linear research. The experiment has led to intense learning by students as well as teachers. Our emphasis on experimentation should not obscure the importance of some degree of control and coordination in multimedia projects. Indeed, the lack of a central player able to co-ordinate or control these experimental efforts could also prove problematic, as exemplified by the Cable School (Slack 1999) and other cases (Nicoll, forthcoming 2000). The overall lesson from these discussions is that what is required is an appropriate combination of experimentation and control. As our study of Educational Multimedia concluded: Control is needed to survey and evaluate what is going on, to offer clarity of responsibility, and to motivate those involved by showing innovative uses and by offering a platform to exchange ideas and experiences. Exploration is needed since educational multimedia are no ‘one size fits all’ products. They have to be explored on their possible uses, they have to find a place in the setting of use, they permit new and innovative uses that were not foreseen. (Bijker et al. 1999, chap. 13.) 170 Both experimentation and control are thus important. However there is a sense that technically focussed projects have often veered unduly towards the mode of control, and that the necessary experimentation has had to be ‘smuggled in’. Technological objectives have taken first place, and only after they have been realised has space emerged for experimentation about usages (for example in the above cases, about educational objectives [idem]). The presumption that the technology would provide a solution per se, meant that users have had to grapple with the constraints and affordances of new technologies under circumstances of use (for example in teleteaching with the Telepoly initiative). There was a failure to provide sufficient time and a safe context for users to ‘play around’ with the technology. In the case of Telepoly (and likewise CableSchool) the result was that teachers who had negative experiences lost their commitment to the project. Thus one of the important strategic choices in building and conducting a multimedia experiment concerns the dimensions and timing of control. 8.2) The involvement of final users In relation to broadening involvement, a key questions concerns the involvement of 'final users' in system development. The public rhetorics of ICT have been emphasising the centrality of the user for over a decade - and the idea of user-involvement has become almost a sine qua non in the contemporary teaching of systems development. In this context, one of the most striking and surprising findings of the SLIM project concerns the marked absence of multimedia constellations directly bringing together end-users as well as supply-side players. A number of reasons can be advanced for this.78 78This absence raises a raft of questions that we are not fully able to answer - in part because our sample (though fairly large in all) included virtually no instances of systematic and organised enduser involvement. It would not be adequate to explain this lacuna solely on the grounds of 'false consciousness' (i.e. that the misleading rhetorics about the speed and capacity of technological changes were sufficient to blind developers from the evident advantages of user involvement). Instead we need to consider a range of factors that may, in combination, help to explain this state of affairs: * the difficulties in integrating actual/ potential final users into multimedia experiments may make working with complementary suppliers and intermediate users a more feasible and costeffective option. * Suppliers do have methods of representing the user other than direct user involvement (eg market research, panels of potential users; representation by proxy by engineers etc.). Questions arise about how adequate these methods are? It is easy to criticise such methods of involving/ representing the user, particularly in relation to radical new multimedia applications. However, it is most likely that multimedia products and services will evolve incrementally - and therefore ultimately could be largely captured by these methods despite their limitations. * This kind of information may be collected by the private sector players involved, but is likely to be kept commercially confidential. Many of the above methods for tapping user experience are costly and require significant investment - and in turn yield potentially valuable market related information - for the commercial players involved. Firms are likely to keep secret their proprietary knowledges of and linkages with their own customer base. * We need to avoid the 'ethnographic fallacy' - the presumption that we need to build ever larger bodies of knowledge about specific user settings into applications - and the assumption that direct end-user involvement is the only means of doing this (for example, we know that artists 171 Commercial suppliers of course are increasingly to conduct market research and acceptability testing on their products as they come near to market. However it remains the case that such firms are likely to treat their own market research as proprietary - commercially in confidence - and are unlikely to share this information. Sharing access to the user has been experienced as a particular problem (see for example, Nicoll forthcoming, 2000). In many multimedia experiments the involvement of proxies for the final user remains the dominant approach - particularly where there is a significant development effort (i.e. where the 'mode of control prevailed over the mode of experimentation). We can relate this to the difficulties that may surround end-user involvement at the earliest stages of innovation. Experiences with user involvement in relation to industrial ICT applications show that such naive users may lack the technical skills and confidence to articulate their requirements effectively (Ehn 1988). Non-specialist users will often see technological determination where it does not exist (for example, seeing the particular features of designed artefacts as necessary and immutable) and tend to defer to technical specialists. They are not accustomed to 'technological fantasy' - developing visions of how technologies might be deployed in their activities. Following on from this we note that the various methods of acquiring knowledge of the user and / or involving the user provide greater or lesser opportunities for the user to become involved as an actor. In some cases the user is largely passive and primarily an object of study (Nicoll forthcoming, 2000). However, effective user participation may depend upon the users acquiring the skills and confidence needed to become actors - systems builders/ intermediaries - in their own right (Jaeger & Qvortrup 1991, Jaeger, Slack and Williams forthcoming, 2000). There is of course likely to be an asymmetry between final (or end) -users and other players, which bears upon their motivation and willingness to be involved in a development (see our discussion of the role of intermediaries). The core purposes of final users may not be linked to technology per se. Instead their interest revolves around the extent to which they may be able to appropriate multimedia for their particular purposes. In other words their interest is likely to be more contingent; and limited to a particular project (in contrast to players more closely tied to technology, who are more likely to find it in their individual career or organisation's interest to participate in various projects over time). One development that is particularly interesting here is the success of ICT suppliers to enlist the help of users in identifying problems in the technical performance and and journalists are very good at anticipating the responses of their audiences - as discussed in relation to the 'Creativity' Mode of Design / User Representation). Conversely we may need to consider the advantages of economising on social learning - where it is not absolutely necessary. Indeed multimedia providers may decide to step back from the complexity of learning about the user/ use and instead 'let the market decide', until at a societal level these questions gradually become resolved - by a more diffuse and less consciously directed process. 172 ease of use of their offerings at an early stage by setting up alpha and beta testing of their offerings. Many of those involved in such testing will, of course, be intermediate users of generic technologies who may anticipate benefits from having early access to new products (coupled perhaps with the status and credibility of being a member of this club). However alpha and beta testing extends remarkably far into the community of individual users of technology (including professionals and enthusiastic amateurs) who seem willing to give of their efforts in exchange for very modest material incentives. The phenomenon of 'early adopters' of technology is, of course, well known. We should, however, bear in mind that such a technology enthusiast culture may represent a poor basis for reliable extrapolation about the potential wider market future users. ICTs could of course provide an important medium for feedback from the widely-dispersed and anonymous customers of mass-market multimedia products. However outside 'technology enthusiast' cultures, such avenues are rudimentary, but can be expected to increase. Perhaps the best example from SLIM research was Girls ROM, where designer of this CD ROM product set up a comment facility on its associated web pages to receive input on the design of next version. Sørensen (1999) describes this as a 'weak' form of learning economy because, although there is interaction between design and use, this is 'electronically mediated, accidental in nature and separated in time and space'.79 8.3 the role of the state Particular issues arise regarding the role of the state. The various national information society policies produced in the last decade in the European Union, the USA and Japan are united in their insistence upon the leading role of the private sector in many areas of multimedia provision (Williams & Slack 1999). Alongside their laissez-faire emphasis, these policies discuss the need for state support in those cases where the self-interests of the players concerned will not guarantee development (i.e. where markets may fail). Candidates for public funding include support for groups which lack economic power to ensure their needs are met, support for the early stages of research and or development of a new technology where the costs and risks may exceed expected benefits and finally the many public services where the state is itself the customer. 79Thus Sørensen notes (Sørensen 1999, section 2.3): 'the home seems to be weakly linked to the major learning economies related to the multimedia industry. Even if the Internet allows new opportunities for users and producers to communicate in an informal way through web-pages and email, it is unclear whether these opportunities are made use of. In fact, one may suspect that most companies consider their home pages to be sites of advertising, information and downloading of updates, rather than as a real medium of communication with their customers/ users. Social learning of multimedia in households may be embedded in small user-to-user learning economies, but these economies do not appear to be linked to user-producer networks in any substantial manner. This means that producers miss out on opportunities to “ tap” core user groups for ideas, information and experience, while users do not get the information and advice from designers that they could have given them substantial benefits.' 173 However one striking finding from our review of multimedia experiments, which also characterised the vast majority of the SLIM cases concerns the key role of the state (at international, national and local levels) in supporting multimedia experiments (Jaeger, Slack and Williams forthcoming, 2000). In some areas, notably cultural content products with well-established commercial markets (games, television), private provision did prevail. However even here the state was often an important player whether as a key source of funding (as in the case of the Nerve Centre) or a provider (for example the public broadcaster RTE with its Irish 'heritage' projects). In the other areas we examined - education and public information - it is perhaps less surprising that the state was a central player in many if not all cases. However, the boundaries between public and private provision seemed to be shifting. Partly as a result, multimedia experiments were most often 'hybrid' initiatives, involving public and private players and resources (and we should not overlook here the contribution of voluntary actors, such as teachers and citizens, as well as the public administrators with responsibilities in the area). An important aspect regarding the continuation and growth of multimedia experiments concerns the establishment of a sufficient market of users to make a project self-funding. Dealing with these commercial issues and negotiating the public-private boundary may be a key issue in terms of the viability and success of a new multimedia product. This work has important policy implications, to which we now turn. 174 IV Conclusions and policy implications Chapter 9 Conclusions and policy implications This Chapter pulls together some overall conclusions and discusses the implications for public policy and for the strategies of practitioners involved in multimedia. It starts by reviewing the current challenge in terms of shifting away from supplydriven perspective and in reconciling the partially competing demands of technology dynamism and entrenchment. It then focuses upon the issues surrounding the exploitation of multimedia experiments. By focusing upon the key ‘social learning’ question of how the knowledge and experience acquired can be transferred more widely it suggests new approaches to considerations of transferability and of that take into account the many broader, indirect and longerterm outcomes of an experiment. The next section reviews our central policy recommendations in terms of how public policy can best support social learning around new technologies like multimedia. We then make a number of concluding observations about the achievements of the SLIM project. Finally, our integrated and cross-cutting studies of the application of multimedia across a range of social settings yielded a number of substantive policy recommendations. These are briefly summarised. 5 Challenges arising from the current context and dynamics of innovation of multimedia a) shifting away from a technology supply perspective The key policy challenge posed by the SLIM study surrounds the need, we identified, for a 'double shift' in the focus of technology promotion efforts, away from their rather exclusive concern with promoting the development and supply of core technologies, towards: • the appropriation activities of intermediate and final users; and, associated with this, • the need to support the development (and appropriation) of cultural and information content. The need for such a shift in focus has been taken on board to some extent within a number of recent public policy developments - including, notably, the Information Society Technologies programme of the European Commission Fifth Framework Programme - through its Key Actions in the areas of Systems and Services for the Citizen; Multimedia Content and Tools; and New Methods of work and e- 175 commerce. This is an important and welcome shift. However the IST programme still displays some features that seem to have been 'inherited', passed on from its roots in traditional EC technology-supply driven Research and Technological Development programmes, particularly in its emphasis on the creation of new technological artefacts as a means to solving a range of socio-economic problems.80 The SLIM project points, instead, to the need for public support to be geared towards multimedia innofusion and appropriation as well as, and perhaps as an adjunct to RTD projects. Moreover, support for multimedia uptake (which is currently largely cast in terms of 'awareness raising' - as if there existed a finished set of multimedia capabilities which potential industrial or private users merely need to be aware of) needs to recognise the experimental nature of appropriation. The implication would be that such initiatives should include provision for a creative effort and for dissemination of appropriation experiences to other appropriators and to future technology supply. The outcomes of artefact-centred approaches - multimedia experiments and RTD projects alike - have often seemed rather disappointing. They have certainly not fulfilled the widespread expectations that such artefacts, once developed, would move rapidly towards roll-out (bringing economic and social benefits). This draws attention to the limitations of the artefact design/development centred model. The low uptake in most cases (even those cases which pay more attention to the user than conventional technology-driven approaches, such as social experiments and user-centred design, which seek to build more user knowledge into the designed artefact) can be linked to the linear model which still underpins them with their linear model of the relationship between initial design and subsequent roll-out. This is not to deny that some products may and do emerge which are adequately aligned with current and emerging user/ market requirements, and which become commercially successful and socially significant. Given the difficulties in prefiguring evolving needs, such alignment is to some extent serendipitous, whether it was arrived at largely by accident or arising through careful and informed design. It is important to recall that we do not have sufficient information, and a sufficiently reliably understanding, at this stage, to determine which would be the most effective technology supply strategy between, on the one hand, user-centred design (building more knowledge of specific user contexts and purposes into design) and on the other, a laissez-faire strategy based upon maximising the launch of new products onto the market for selective uptake and adoption by users. The optimum may lay between these extremes – and will certainly be different for different kinds of technology application. Current orthodoxy in system design discussions favours the former. However we should not 80Similarly, though the Fifth Framework Programme overall gives welcome attention to the commercial and societal exploitation of RTD, the legacy of earlier supply-driven EC programmes live on in the very structure of the programme - for example, the separation between RTD and support measures. This separation, in IST and in 5FP as a whole, is extremely unhelpful. For example, the 'take-up' measures should be seen not just as a practical matter but as a resource for socio-economic research and even future socio-technical research and development. Thus, we would argue, studies of innofusion and the appropriation of earlier RTD projects could be no less important a source for innovation as further RTD. 176 overlook the potential of the laissez-faire strategy. Indeed if we examine the Japanese experience, success in consumer electronics appears to rest precisely upon very high levels of incremental innovation within their domestic market from which a number of innovations have emerged that have achieved global significance highlighting the importance of receptive and demanding markets as well as smart supply (Porter 1990). The mainstream response, exemplified by the EC Fifth Framework Programme (5FP) and Technology Foresight in the UK, has been to take on board the criticisms of supply-driven models by setting up a range of measures to support commercial exploitation of public funded research and development. This has been described as a 'linear plus' model, as it tries to redress the failings of linear models of innovation, which in effect tries to get innovation processes to work in an essentially linear manner, driven by the supply of technological knowledge (Tait and Williams 1999). Attempts to shift away from the linear, technology-driven model of innovation must engage with certain problems. In particular: a) roll-out and commercial launch typically takes an order of magnitude more resources than initial technical development. However resources for commercialisation have been less-readily available than for technology development. (Similarly, support for associated developments has often been overlooked – in the case of multimedia this relates particularly to the development of cultural content and information services which is labour intensive and expensive, but which has not been well-supported); b) the process of domestication/ appropriation typically occurs over a much longer time scale than technology innovation. The core technologies underpinning multimedia and the world-wide web have enormous technical dynamism, with annual new launches of new versions, and ever-shortening product cycles, to the extent that today the ICT-supply industry considers 2 5 year old offerings as obsolete. We have seen that the complete cycle of appropriation whereby earlier ICTs such as television became widely adopted and incorporated in everyday life took a generation or more; more recent ICTs such as the CD have been more rapidly adopted, but still took well over a decade to establish their market (Winston 1998). 81 This points to a sharp disjuncture between the cycles of innofusion and the cycles of appropriation. Different parts of the innovation system may be operating with different dynamics. The slow pace of appropriation processes have often been unhelpfully miscast, from the viewpoint of the rhetorics of technology-supply, as a problem - an obstacle of user resistance or non-acceptance which needs to be overcome - rather than as an 81Even this only tells part of a much more complex story. CD technology, having failed in its original market as a video recording medium, became the industry standard for audio recording, and also a medium for computer storage - an observation which reinforces our argument about the scope for transferring multimedia components to new and different contexts (Winston 1998). 177 essential and central part of social learning around a technology. We cannot eliminate appropriation by fiat (though we may be able to organise social learning more effectively); the selective local uptake and experimentation around use are vital to technology domestication. Conversely it would be extremely unhelpful to conclude that the rate of technological development of core components of the information infrastructure should be should be chained to the slow pace, and rather conservative trajectories of appropriation processes for multimedia products and services Such interpretations of the 'linear plus' model of innovation would be very damaging. Instead, innovation policy needs to find ways of addressing the different dimensions and timeframes of social learning - between innofusion and appropriation. An integrated innovation strategy, of the sort suggested by linear plus views, does not necessarily imply a monolithic strategy. Policy may need to explore the different spaces in which social learning is taking place. Further we may need to find ways to uncouple particular elements of innofusion and appropriation, rather than try to link these mechanistically together. b) reconciling innofusion and appropriation; dynamism and stabilisation This brings us on to consider the potentially competing requirements that may beset a technological system, and how they may be reconciled. This encompasses the tensions between stabilisation and entrenchment versus the dynamism of development of a technology and between the creation and appropriation of technological capabilities, and also the tensions that may exist between different elements of a technological system (eg core component technologies; the technology platform/ infrastructure; applications). Let us start with the example of the development of a single artefact. Referring back to figure 6.3, when we consider a particular stage in the development of an artefact we can identify two potential vectors for its further evolution, between: * technical enhancement - efforts geared around technology supply (to deepen and refine the technological knowledges underpinning artefacts) and, * social appropriation - efforts geared around the domestication, consumption and use of artefacts. Figure 6.3 thus highlights something of the different directions in which the innovation of an artefact can be taken. These may be (to a greater or lesser extent) orthogonal, one to another. There may be a trade-off between expending effort on technical enhancement versus social appropriation. In chapter 6 we reviewed a range of strategies. At one extreme, development of an artefact may be taken primarily towards stabilising technological development, building markets and promoting social appropriation, and at the other, consolidation of uptake and use may be set aside in favour of the further development of the state of the art of technology. However, our analysis pointed to the variety of possible 'innofusion/ appropriation' paths, between these extremes, which involve different strategies for matching technical potentialities with user exigencies to get a 178 technology to work under specific user circumstances. This suggests that there may be a number of strategies to reconcile these different exigencies. Tensions may thus exist between innovation/ innofusion of technologies (for example the technology infrastructure/ delivery system) and the promotion of a stable and homogeneous market - particularly for information and cultural content. Innofusion processes will tend to lead to a patchwork of more or less unique configurations. This presents particular problems where an information infrastructure is emerging through innofusion. The internet is precisely such an example of a large-scale technological system arising through a process that is widely distributed across many players (including the huge numbers of 'final users' configuring their local ICT platform). These processes have significant consequences. High levels of dynamism and turbulence around technological trajectories and standards will tend to encourage consumers to defer purchasing decisions. This may not only slow the rate of adoption but, more critically, may undermine the very viability of new services where these depend upon attracting a critical mass of customers (Williams 1995). This has important implications for the information/ content market. It was precisely this kind of consideration that underpinned the decision by France Telecom to support and heavily subsidise the widespread adoption of its Teletel/ Minitel videotext system, thereby enabling the establishment of a healthy information market. The innofusion model, in contrast to such concerted models of technology adoption, though allowing for the emergence of de facto, industry standards, remains extremely heterogeneous. The outcome of innofusion processes in the technology platform will be an extremely heterogeneous technology infrastructure through the crystallisation of many specific local configurations as different groups implement technology in the course of a dynamic process of further technical innovation together with highly dispersed and differentiated processes of the social appropriation of artefacts across a range of settings. Such a socio-technical setting would constitute a very uneven and fragmented market for complementary products. We see this already in the problem presented to information service/ application providers of supporting a plethora of standards and platform capabilities.82 This also affects the usability of the system as well as posing particular problems for design. 83 One commercial and public policy implication that follows from these considerations is that different components of the multimedia system may be subject to different exigencies and requirements. The development of a mass 82For example digital cities and community information systems migrated up to web-based standards, but still faced problems as to which functions and which generations of browsers to support. As we saw in Chapter 7 UK Bank needed to make difficult decisions about the basic technology platform it would expect for customers using its internet banking system. 83For example not all browsers currently support scroll bars. Should designers forego top-end functionality or conversely create products that are not geared to or usable by 'bottom-end users'? Can designers simultaneously design for different levels of user functionality. Conversely users may find it frustrating if the functionality of their system does not match that of the content provider, and confusing if the look and feel of applications differs depending upon which technology platform/ point of access is being used. 179 information market may call for stabilisation and concertation of the technology infrastructure. Whilst the relatively small numbers of public and commercial infrastructure providers may be able to cater for the heterogeneity that will consequently exist across the network (through gateways and standards) this will be particularly an issue for the local technology platform, where the mass of end-users lack the technical and other resources needed to ensure they are aligned to standards. On the other hand, optimising technological innovation in the multimedia platform may be best served by innofusion. Public policy needs to grapple with this rather contradictory terrain. This will not be easy – as we see below there may be competing demands between different policy goals and between different technical fields. In relation to technology supply, we have already noted the emergence of commercial and technological strategies which try to reconcile the dynamism of supply of new technologies with the stability of the market defined by the existing installed base of complementary products. This includes the development of 'architectural technologies' ([Morris and Ferguson 1993] offering longitudinal compatibility between different generations of technological artefacts) to allow a supplier to innovate some elements whilst keeping the product compatible with other elements (eg the existing installed based of applications and software) (Williams 1997). Though the classic examples of such standardisation and architectural technology have been in relation to core 'technical' components (eg microprocessors and hardware more generally) we have suggested that such factors will increasingly come to play in relation to the 'user surface' (the human-computer interface and 'the look and feel' of an application, the organisation of information and navigation metaphors and signposts) (Williams 1993). The segmentation of the technology infrastructure into different levels, the stabilisation and standardisation of the interface between them, and the emergence of interoperability standards more generally can all be seen as ways of allowing technology dynamism to take place within particular 'black-boxes', without imposing unwanted change and new learning requirements upon other players (Fincham et al. 1994, Brady et al). This kind of 'black-boxing' effort can be seen as a strategy for economising upon social learning. [And we should not forget that innovation and social learning is expensive of time and effort; it may be frustrating - and needs to be economised on wherever feasible.] In this connection a repeated finding from SLIM studies of the early Community Information Systems (based for example on text processing and bulletin board technologies) is that the changes in the human-computer interface, to take on board graphic capabilities and the design metaphors of the world-wide web, caused a lot of disruption and uncertainty for users (notably DDS and Craigmillar Community Information System). This was particularly an issue for groups that were less technologically confident and committed. Change imposes new uncertainties and learning costs upon existing users who have invested in learning and becoming familiar with languages and procedures. Change in the look and feel of the user surface (for example about how to operate the user interface and how to navigate 180 uncertain information terrains) could present a considerable obstacle to these groups – particularly those lacking confidence and with a narrow skills base. Of course, other users may become bored with a system that seems dated – forcing providers to constantly redevelop their offerings. One idea that might follow on from this concerns whether it would be possible to provide continuity in change – with (at least the option of) some stability in the look and feel of earlier versions in new launches of information systems and content product. 9.2) Exploitation and Transferability of Results of Multimedia Experiments a) Introduction This section explores the issues surrounding the exploitation of multimedia experiments, and Research and Technology Development initiatives in the field of multimedia products and services more generally. It then examines the implications of our empirical findings, and the social learning perspective, for policies for promoting the development and uptake of multimedia. Transferability is a central concern in discussions of social learning. The concept of social learning seeks to analyse not only local learning experiences but the ways in which they may be communicated more widely and 'translated' to make them applicable in other contexts. Social learning to be relevant has to involve some 'change of state' or practice - to mobilise (or stabilise) a development. Particular experiences need to be carried forward and communicated more widely. This raises questions - about how learning takes place, how the lessons that have been learnt and transferred and, finally, about what structures and incentives exist that may enable this learning to go on? These considerations have very immediate implications for public policy and commercial strategy - both in relation to the problems in exploiting multimedia experiments, and in the kinds of policy setting that might be needed for effective experimentation and exploitation of experimental outcomes. Public and private funders of multimedia experiments have the commendable desire to ensure that the projects they are supporting do lead to useful outcomes, and can be more generally applied. The pursuit such demonstrable research outcomes seems, paradoxically, to have often had the consequence of promoting a search for concrete 'deliverables', and in particular channeling expectations towards the idea that the transferability of project outcomes will be through the wide dissemination and use of artefacts.84 The 84Arguably, one reason for the continued attractiveness of the linear model of innovation, and the associated idea of technology exploitation through the dissemination of artefacts, despite the many critiques that have been advanced, is that these seem to hold out a simple and apparently compelling metaphor both for the ways in which technology development effort can be exploited (and likewise of how this exploitation process, and the cost-effectiveness of research investments, can be monitored). The search for evidence of such linear exploitation is however often fruitless (and can lead to false judgments about the lack of impact of research). (Rayman-Bacchus, Williams & Bechhofer, 1998, Tait & Williams 1999). 181 focus on concrete deliverables has been unhelpful, we argue for a number of reasons. First it may discourage the kind of experimentation about usage that we have identified as key to the domestication of technologies. Second, the complexity and uncertainties surrounding commercialisation and appropriation of technology mean that this form of exploitation is often unsuccessful, at least in the short term. Third, and following on from this, production of a material artefact is but one of a number of potential outcomes from a research project (or even a commercial trial), and is not necessarily the most significant transferable element. Equally important, potentially, are non-material outcomes. These include the development of specific social relationships amongst diverse players involved in technology development and use, and knowledges of the processes of developing these relationships and technologies. Though the non-material outcomes of a multimedia experiment may be far more significant than the material artefacts generated, there may, however, be particular difficulties in capturing and conveying them. In particular, as we explore below, many of the lessons are extremely contingent - and therefore may be hard to communicate and generalise. Some areas of national and European technology policy have begun to address these issues about the wider exploitation of RTD - notably through measures to support collaboration between ICT suppliers and commercial user organisations in the development of ICT applications, which reflects the growing recognition of the importance of such linkages in the 'learning economy'. However the broader types of outcome have not always been fully recognised by technology policies and strategies. They also present particular difficulties for wider transferability and exploitation, for several reasons: many of the lessons appear highly contingent to local circumstances; many outcomes are tacit; the full relevance outcomes may only become clear in the longer term. Given this element of unpredictability, the attribution of success of failure to projects may be a perilous task - as we see in the next section, many of the most important lessons may arise from a project that ostensibly was a failure. As a result, public policy and RTD and exploitation strategies need to be designed around the social learning and innovation exigencies of particular application fields. b) Artefacts – a promise of transferability Most pilots and trials are geared towards the development of new multimedia artefacts. This is partly because they may be funded under RTD programmes (particularly in relation to the European Community - given its long-standing responsibility there), which have historically received substantial state funding. In contrast, greater difficulties have been experienced in obtaining the resources needed for the development of cultural products and multimedia content, which is also expensive (let alone the resources needed for disseminating and building markets for such products (Preston 1999), where public support has been less wellestablished. 182 However, the vision of developing a new multimedia artefact holds out the prospect of the wider dissemination of the product or service - offering an apparently clear and self-evident metaphor for the exploitation process, and one that contains within itself the promise of social benefits and economic gain. However, a brief examination of multimedia experiments - including those studied by SLIM and more generally - shows that such a linear exploitation trajectory is more the exception than the rule. For start, the costs of commercial development are enormous, even in those cases where technical innovation has been successful in creating a new artefact. The evidence from the mainly commercial players is that the costs of roll-out and commercial launch of new, mass market platforms are probably an order of magnitude higher than the technical development costs. Such investments are only likely to be supportable by the very largest global players. One consequence of the prevalence of the linear model, and the focusing of public support for the development of artefacts, has been the relatively limited availability of funding for exploitation.85 RTD support tends to be finite - and there is a danger that, once a project comes to the end of the funding 'conveyor belt', there may not be resources to carry forwards the development and commercialisation of the artefacts (and, no less importantly, to maintain the expertise and relationships developed amongst the collaborators) . The danger, of course, is that developers may be drawn by the availability of RTD support into the further technical development of an artefact, or a wholly-new development project, simply by the need to find follow-on funding - and away from the commercialisation and societal appropriation of existing offerings. Other modes of support (with larger levels and longer timescales of support) may be needed for the commercial launch and wider uptake of multimedia artefacts. This has further implications for the way in which public support is provided - for example, as we discuss below, RTD projects need to have more elaborate 'exit strategies' for when public support ceases. The focus on artefacts has had a number of other unhelpful consequences from the point of view of social learning around multimedia products and services in terms of the way in which multimedia experiments are operate. There seems to be a general tendency for funders to seek to fund novel elements - and a tendency to see such novelty primarily in terms of technical novelty. Partly in consequence many multimedia projects seek to utilise or carry forwards 'state of the art' technology. This may impede the appropriation and market uptake of multimedia products and services to the extent that what is developed that may be out of alignment with current market opportunities which are rooted in the skills, the expectations and the equipment possessed by potential users. Equally, we have found some evidence that projects are subject to limited degrees of freedom - and that an emphasis on technical development may divert energies and resources away from technology domestication. The conduct of multimedia experiments with a strong technology development focus was more likely to be resemble the mode of control than the mode of experimentation. Our research suggests that if the goal is to promote the 85Though there have been important exceptions including the TIDE programme funded under the EC Fourth Framework Programme. 183 economic and social benefits of wider uptake and use of multimedia, a rather more effective starting point for public policy might be with strengthening the local multimedia appropriation capabilities, and promoting innofusion rather than with technology development, per se. Though in this account we have stressed the factors that tend to undermine the prospects of linear, research driven processes, we should not downplay the possibility of the exploitation of such artefacts, but instead should show how these emerge from complex interactive social learning processes - and notably of innofusion. For example an important aspect of the astonishing technological dynamism in relation to multimedia development has been the way in which artefacts can be unbundled from particular applications and contexts of adoption and relocated into a different socio-technical context. In the recent history of ICTs, we can point to the rapid uptake of certain generic models regarding communication services (email, www) and concepts of the human computer interface (windows, buttons, hypertext navigation models). What seems to be important in this process is the 'unbundling' of components with a generic applicability from their particular context of development and use (as exemplified by the way in which the world-wide-web shifts from being a vehicle for scientific interaction, to a tool for more general information search and exchange). Elements are 'decontextualised'. Specific features of applications and tools, relating to particular contexts and contingencies of adoption, are removed or made removable or reconfigurable. These components can them more readily be incorporated as elements of new technological configurations. Often what is taken up is a concept of use, rather than a technical component - for example a metaphor, such as that of the digital city. Again a job of reinvention is needed to make it more generically applicable. As we see in the next section, this process of unbundling different learnt elements is critical to understanding how social learning takes place (especially regarding the transferability of lessons from multimedia experiments). c) Non-Material Outcomes Some of the most valuable outcomes of a multimedia experiment or trial are nonmaterial, including, notably, its contribution to the development of specific social relationships and the acquisition of new areas and combinations of knowledge. The circumstances for developing and exploiting these may be very different than those for material artefacts. Relationships The development and uptake of new multimedia products and services typically relies on the activities of a range of different players (suppliers of complementary products - eg technical components; content). This was rather evident in the studies of cultural content development in which different organisational and industrial cultures of ICT and of media/ cultural activities were brought together (Preston 184 1999, Nicoll 2000). An important element is the development of knowledge of other relevant players - and the building up of mutual understanding, and possibly commitments and alliances. Further exploitation of a multimedia development project may depend upon the continuation of these relationships - perhaps through a follow-on project - directed towards further 'technical' development or as part of an effort directed towards the dissemination and exploitation of a developed product. The challenges of effective exploitation may be very different from those surrounding development. There may be a need to transform existing relationships. In particular, studies of Multimedia and other 'network technologies' involving complementary products is that the benefits of co-operation in the development phase (reduced uncertainties; shared costs) may be outweighed by the growing pressures for competition in the exploitation phase. The shifting balance between co-operative and competitive incentives represents just one way in which relationships may need to be transformed. A critical feature in relation to social learning is that direct relationships with (end) users are unlikely to be transferable to other settings (either in commercial exploitation or further development). One reason for this is that user-relationships - particularly for 'end-users' - are likely to be localised (users may be drawn to take part in projects in their immediate geographical or institutional vicinity), and partly because they may be ephemeral (consumers drawn to consume particular products and services may do this as a 'one-off'; they are not necessarily interested in technology per se, and may not wish to move on to series of other multimedia projects). These problems of end-user relationships do not have the same force in relation to business users of multimedia products and services - particularly larger firms whose interest in exploiting emerging new technological opportunities may motivate them to become involved. Smaller firms have experienced difficulties in this respect, as they may lack the financial and technical resources to engage the supplier community. The failure of the 'learning economy' in this area - evidenced by the weaknesses of supplier-user links and engagement - has been addressed by public policy initiatives specifically directed to promoting such linkages and innofusion (eg in the UK and in Sweden), for example in relation to industrial ICT applications or multimedia product development. Such initiatives might usefully be extended in relation to multimedia content and service provision. Processual knowledge The acquisition, sharing and deployment of various different kinds of knowledge is crucial to these processes of building and maintaining relationships and collaboration around the development, innofusion and appropriation of multimedia. A key motive for collaboration is the need for groups with different specialisms and expertise to share knowledge in specific areas needed (for example about the responses of users to particular offerings; about the capabilities, culture and offerings of other providers). This is well illustrated by the study of Compuflex which highlights the importance to this global players of having a ‘local’ partner 185 who understood local media consumption patterns and the differences between markets. In addition such collaborations may offer more generic knowledge about the processes involved in developing and exploiting multimedia products and services, building socio-technical constituencies etc. There are significant problems about the transferability of both these kinds of knowledge: i) Some of the knowledge may be highly contingent to particular circumstances and stages of innovation; questions arise about whether and how it may be possible to from extrapolate this knowledge and apply it in other similar circumstances and more generally; ii) Some of the knowledge may be tacit and difficult to formalise and communicate more broadly. The need for a managed approach to knowledge transfer Given the tacit nature and local 'stickiness' (von Hippel 1988) of such knowledge and experience, it may be difficult to capture and broadly disseminate such knowledge in a way that is useful and applicable. Conventional approaches to the dissemination of formal (e.g. 'technical') knowledge may not be effective. To give an example, lists or databases of multimedia experiments - though potentially a useful resource in identifying other players in a field and this facilitating contacts and networking - are unlikely to provide the basis for conveying the complex understandings generated in experiments. Even simple case descriptions, which will tend to be 'heroic' accounts of success stories, are unlikely to convey what was learnt from an experiment. In relation to such complex conceptual deliverables, a more active strategy is needed both for building the knowledge-base and managing knowledge transfer. 'Observatories' and interest groups may provide a useful vehicle for this kind of managed dissemination process (Lobet and van Bastelaar 1999) .86 These could provide the basis for the more critical and analytic approach which is needed for digesting and communicating multimedia project experiences and their outcomes. For example, given the importance of the local context, there is a need to address the contingencies surrounding the conduct of projects, linking these, where possible to particular features of the case (eg regarding the particular actors involved or their local context) and the difficulties and weaknesses encountered as well as strengths of the case. Equally important would be meta-level analyses of One example of how this might proceed arises from the SLIM 'Integrated Study' of Public Administration/ Digital Cities, which culminated in the development of a proposal for a 'vademecum of digital cities', to the 4th European Digital Cities conference in Salzburg in October 1998. SLIM research findings have attracted interest from builders and future inhabitants of digital cities. Drawing on these findings a number of propositions have been advanced which aim to provide useful and practical advice to system developers. The next stage, currently under way, will be to validate these proposals more broadly for other digital cities. Rencontre réelle de villes virtuelles (FUNDP-Namur-June 16, 1998), PAI Seminar on State, Citizens and Market (FUNDP-Namur-June 25, 1998), 4th European Digital Cities Conference (Salzburg, October 29-30, 1998). 86 186 generic, and potentially generalisable aspects of the conduct and outcomes of experiments. However these analyses should consider the influence of specific historical contingencies, and the consequent limitations and uncertainties surrounding the application of these findings more broadly. Similarly, as we see below in discussing success and failure, particular care must be exercised in using concepts such as 'success' or 'best practice', since these are always judged against particular frames (eg from earlier technological developments) which may be contested. The possibility of such change requires a level of humility in policy pronouncement (for example the criteria chosen for addressing such features as success and failure, and their presuppositions, must always be defined as precisely as possible). It further suggests a shift in the way we gather information to support decision making. Instead of collecting information to reinforce particular expert prescriptions we may need to allow for a more reflexive process of gathering rich descriptions of experience and context that can be a resource for actors in creatively addressing their particular problem settings. Given the difficulties in formalising and communicating social learning experiences, an important avenue for exploitation may take the form of 'embodied knowledge'. An important pre-requisite for the exploitation of experiences may be the existence of enduring actors with the interest and capacity needed to exploit these experiences and apply these more generally. This reinforces our argument about the key role of intermediaries in this social learning, in conveying and applying experiences and lessons more broadly, and brings us back to some of the questions raised in Chapter 6 about which actors can take on board the role of bringing out these lessons (transforming this knowledge; extracting from complex experiences, particular simplifications which can have wider applicability). d) Success and failure of a Multimedia Experiment One of the implications of the social learning perspective is that attribution of 'success and failure' may be rather perilous affairs - not least because different parties are likely to have different commitments to and perspectives upon a project, and may draw different kinds of benefits and lessons. Attribution of success or failure may thus be a relative question - and is often contested (as we saw in the case of the Craigmillar Community Information Service). Indeed, whether a project is formally defined as a success or a failure may tell you very little about the outcomes of the project, but may primarily reflect local micro-political contingencies, regarding which groups have control over the retrospective attribution of success to a project. Our study of multimedia education projects highlighted the way that dominant actors were able to impose their criteria and notions of success and failure. However, they noted “ success was much more paramount than failure. Failure was denied and neglected while success was announced and exploited.” (Bijker et al. 1999). We conclude from this and other studies that: it is only under very particular circumstances that sponsoring groups find it in their interests to define a project as a failure. 187 The search for 'success factors' and to identify 'best practice' can therefore be misleading, as well as potentially uninformative. Success is determined against the particular sets of criteria and frameworks, often based upon our knowledge, which may be more or less stereotyped, of established technology products and services. It will be clear from the preceding sections that questions inevitably arise about how far we can reliably extrapolate from the templates and exemplars on which these criteria are based. Are these metaphors reliable in different settings?87 Have the rules of the game changed over time? The conclusion of our review of 'Digital City' initiatives was that was not possible to identify a simple account of 'best practice' either regarding the model of development, the reorganisation of the administration nor the management of the technical changes. The factors that seemed to underpin particular successes were often highly contingent to the particular circumstances of the case in a way which impeded extrapolation to other settings. The exploitation of experiences arising from experiments was thus rather complex - and knowledge about change processes seemed to provide a more reliable basis for transferability than correlations between specific factors and outcomes (Lobet and van Bastelaar 1999). The outcomes of multimedia are often rather mixed. For example Microsoft Network has become the third largest online service in the world – but it is still losing a lot of money. The Videotron interactive TV system has closed – but the experience gained is seen as valuable (Curry 1999). The assessment of success may thus be rather complex. However, from the point of view of social learning, what is more important is not whether a project is assessed as successful, but whether it provides for flows of information and alignments of players. From this perspective, it is not necessary for a Multimedia experiment to proceed to roll-out/ widespread use for it to make a helpful contribution to social learning. Some of the most important learning may take place in relation to projects that failed to meet the objectives set from the outset or that did not proceed to roll out. However public support for multimedia initiatives (as with many other areas) is structured around more straightforward (and indeed simplistic) ideas about project outcomes, reflected in the criteria of success and failure that are adopted. Public funding for multimedia (and other) initiatives increasingly seems to take the form of competitive tendering for awards for discrete projects and initiatives (in contrast to earlier models of public service based upon direct, long-term funding of in-house Thus, as we saw in Chapter 7, one of the problems that beset the Frihus 2000 feasibility study has been linked to the inappropriate resort to templates from other projects ( including 1980s Scandinavian telecottage experiments as well as projects in New Brunswick, Canada and Salford, UK [Brosveet 1998]) that were very different in size and ambitions and whose success had been debatable. This highlights a more general problem that these kinds of metaphors and exemplars are, inevitably stereotypical and simplified accounts of developments 'elsewhere'. A crucial pitfall in the search for best practice exemplars is that assessments of the success of a project may differ sharply – for example between those directly involved, those in the immediate proximity who have some dealing with it, and those at a greater remove, who may be forced to rely on public pronouncements and third party assessments. In this process there is much scope for the circulation, and acceptance as factual of interpretations that others, more directly informed, might consider erroneous. 87 188 developments/ services). This mode of project-based funding has proved rather flexible, cost-effective and effective in orienting players towards multimedia and in promoting new kinds of collaboration.88 However there may be pitfalls. For example, proposers face the temptation to oversell project achievements (and, linked to this, may seek to oversell technical capabilities). Both funders and proposers may unwittingly contribute to a consensus about what can be achieved which is unrealistic and does not pay cognizance to the difficulties and uncertainties that may beset developments in this field. Competitive funding of this sort seems to be strongly geared towards achieving public success, and this will tend to be defined in terms of fulfilling the goals set at the outset. Those involved in experiments may feel under pressure to conceal certain problems which beset an experiment (and thus also conceal potentially valuable lessons concerning how these were addressed) in order to keep external funders confident and happy with the project. External sponsors are likely to look for evident indicators of success in the projects they have funded - which as we have seen tends to be seen in terms of concrete deliverables. Public (and thus politically directed) funding may thus have some unintended consequences that may impair social learning. 89 A general consequence of social learning - and an important feature of multimedia experiments - is that it is not possible to fully prefigure the outcomes of a project at the outset. The conditions for promoting social learning may indeed be in conflict with the desire of project funders to seek to establish project outcomes in advance. We see this, for example in the tendency of many multimedia projects to be 'verification experiments' - geared towards confirming the value of multimedia products in meeting preconceived success criteria - rather than more open 'diversification experiments'. This was particularly notably in the Educational multimedia projects we studied which were unduly oriented towards the development of multimedia artefacts rather than what was judged to be the critical task of locally integrating multimedia within particular educational settings and activities (Egyedi Lieshout & Bijker, 1999). e) unpredicted and longer term outcomes Finally, the rules of the game in multimedia innovation - and thus the criteria for judging an outcome successful - may be changing. The history of ICT shows that time and time again, we have based our criteria for judging new technologies upon experiences of the last generations of technology implemented. Somehow the historical lesson - that the eventual significance and uses of many ICTs were far removed from initial concepts - has been repeatedly suppressed by the taken-for granted 'self-evident' nature of current perceptions of a technology's utility (Norman 88 For example European Commission and other public support has had considerable influence in promoting multimedia initiatives in the education sector (Egyedi Lieshout & Bijker 1999). 89We are by no means arguing that private funding is immune from such problems. However, the external goals of the private sector are rather narrower and more pragmatic. It could further be expected that commercial players have an interest in seeking commercial advantage from particular knowledges and links obtained in a project, whatever its public outcome, and may be happy to accept and draw private advantage by learning from failure. 189 1998, Winston 1998, Williams 2000, forthcoming). As a result, what is today seen as a paragon of success may tomorrow be called into question. The case of Teletel/ Minitel in France provides a perfect illustration. As already noted, the substantial support by the French state and France Telecom for the widespread uptake of the technology platform, particularly through subsidised public acquisition of Minitel terminals, and the creation of a broader infrastructure have been widely acknowledged as a recipe for developing a large and vibrant market for information services. In contrast videotext remained restricted in other countries to smaller niches of business users (Schneider 1991). However Schneider (2000 forthcoming) has subsequently pointed out the difficulties that subsequently arose in France in adapting to the emergence and global dominance of internet standards. In contrast in Germany, where, by accident of history a standard personal computer has been used as the domestic platform rather than a dedicated Minitel terminal, it has been much easier to open up videotext to the burgeoning information services of the world-wide-web. When we come to assess the implications for assessing the outcomes and significance of a multimedia experiment this has several implications. First, the outcomes of a multimedia experiment have to be examined in the light of their context. Second, the outcomes, and thus the ultimate significance, of a Multimedia project are likely to be unpredictable and only fully-revealed in the long-term. Thus we saw that early projects in the field of community information systems such as the ideas of telecottages (allowing telecommuting and distribution of technical skills to the rural periphery, that had been developed in social experiments in Scandinavia in the 1980s) and the Manchester Host and Amsterdam Digital State were widely discussed and taken up as best practice exemplars and templates for contemporary multimedia projects. Our examination of how these projects were used as best practice exemplars in various SLIM cases also highlighted difficulties and pitfalls in determining the appropriateness of such exemplars (as was illustrated by the FRIHUS 2000 project in Fredrikstad, Norway, in which the influence of ther above was seen as unhelpful). This brings us finally to the question of the verification and assessment of experimental outcomes. It is clear from the above that this is a complex process. We must bear in mind indirect and unanticipated outcomes which may result over long timeframes and in other settings. Benefit can be drawn not only from experiments that succeeded in meeting their goals but also from those that apparently failed. This brings us back to our finding that the utility of a multimedia experiment cannot be determined solely in relation to how the experiment is conducted, but has to be assessed against its broader context and history of socio-technical change, and the specific challenges addressed. Let us examine a case where an experiment yields new artefacts and practices. Other broader and more diffuse forms of social learning may take place - for example as the system is rolled out. The utility and value of these direct outcomes will ultimately be verified by whether they are carried forwards and built upon or alternatively rejected and refuted. In other words, the value of an experiment is related to the extent to which the intense, focused and organised forms of social learning it affords adequately emulate the more diffuse 190 'organic' forms of social learning that may take place subsequently and across the full array of groups that may be involved. 9.3 How can public policy support social learning? The social learning perspective poses a number of challenges regarding the policy process. On the one hand it criticises work which attempts to draw mechanistic policy conclusions, for example, around the linear model of technological change, and highlights the uncertainties that surround the development of policy. At the same time it opens up new policy opportunities. However the critical scrutiny that the social learning perspective encourages also problematises the process of policymaking. In particular it encourages a reflexive turn, and one that shifts away from making particular substantive policy recommendations towards an approach which seeks to foster certain policymaking processes. A key issue here concerns the extent to which public policy can support social learning (and in particular the knowledge flows needed both to transfer social learning achievements and to apply them more widely). However raising the importance of social learning does not resolve the policy problem. Social learning processes are far from straightforward - they raise new sets of considerations. Although in theory all experiences may yield lessons that are in some sense useful, this does not imply that all social learning outcomes are equivalent. We have seen how the concern by project funders to be seen to have supported 'successful' projects might have negative consequences in terms of favouring less experimental approaches, as well, potentially, as the suppression or dispersal of potentially valuable experiences from projects which appeared to be 'failures'. This reminds us that social learning is a contradictory and 'lumpy' process, and points to the possibility of 'unhelpful learning outcomes'. As we saw in the preceding section there is a danger of inappropriate mimicry of external examples rather than adaptation to local circumstances. This might arise for example where developers and promoters become too closely committed to particular concepts of technology which may prove neither feasible or appropriate (Clark and Staunton 1988). a) Modes of Public Support We have argued that the key to social learning is whether players find it in their self-interest to act as carriers of this knowledge - applying or disseminating it more broadly. But how is such self-interest recognised and sustained? What is the role of the state? Some interesting policy questions follow on from this. For example: * how does the constitution of an experiment contribute to the prospects for the subsequent wider application of experience obtained in the project. 191 * does the institutional setting provide support and incentives for such longer knowledge transfer and application? In other words, * is it in the continued interests of constellations to stay together? * will key intermediaries continue to exist (or will bodies be dissolved and actors transferred to different settings when project funding ceases, preventing the further utilisation of the knowledge they had acquired?) Of course, where experiments are privately-funded, the question of whether a development is in the mutual self-interest of participants is, at least in principle, resolved from the outset. With state support, the allocation of resources becomes in principle an issue which must be addressed. We are interested in the consequences of state resourcing both in relation to the conduct of an experiment and its further exploitation. As we saw in Chapter 8, despite the laissez-faire rhetorics of public information society/ information superhighway policies, the state seems to play a key role across a wide range of areas of multimedia development. both in relation to public information and services and to areas of 'market failure'. The latter reminds us of high costs and deep uncertainties surrounding multimedia developments and markets. Private players may not find it in their short term interest to embark upon developments, particularly where there are uncertainties about whether a critical mass will be attained to make a project viable. It is in this context that the resources provided by the state seem to be crucial.90 State resources - knowledge and coordination efforts as well as financial support - underpin the bringing into being of many of the multimedia experiments. Furthermore such public support often provided the main resources deployed by the intermediaries that were at the heart of multimedia constellations, in building and maintaining experiments. b) Unhelpful Learning Outcomes The first question to be addressed concerns whether modes of public support help (or may inhibit) social learning. We have already seen that public support may have 'unhelpful' learning outcomes (for example the 'suppression of public failure' which may impede 'learning from failure'). Similarly, the survival contingencies of particular projects that are dependent upon external income may have unintended and undesired consequences on the behaviour of particular actors or groups. learning to be fundable: For example, one challenge for many projects, especially those depending upon state-subsidies, concerns the need to learn to be fundable (Slack, 1999). To enlist the This was, of course, recognised at an early stage in discussions by the Bangemann report, which noted (p. 12) that "Competition alone will not provide such a mass, or it will provide it too slowly". This was why it was necessary for the European Union and the member states to stimulate and fund social experiments.. since "Initiatives taking the form of experimental applications are the most effective means of addressing the slow take-off of demand and supply" (ibid.: 18). 90 192 support of external funders it may be necessary to cast a project proposal in terms which match the presuppositions of potential funders. This may be reflected in the adoption of unrealistic or inappropriate goals in order to bring funders on board. Sometimes the result could be that sponsors' goals (driven perhaps by their particular visions of the benefits of multimedia) took precedence over user needs. Once a project has been established, how can it continue to secure the resources it needed? It proved relatively easy for many of the public-funded projects we studies to get money for new project development. They found it much harder to obtain long-term funding, e.g. to establish an ongoing service or continued support for an existing project (particularly since an increasing proportion of central state funding is allocated, often very-competitively, on a project by project basis, as part of an attempt to improve the effectiveness of state support). Selectors look for novelty and additionality from new services - which may force those already involved in particular experiments and services to continually recast their offerings according to the exigencies of funds being currently disbursed to obtain further funding. Where project actors have to continually reapply to do new and novel things to gain renewed funding, there are dangers of a drift in goals - where instead of projects developing cumulative achievements, new goals were adopted that undercut fulfillment of goals from previous rounds of funding - and 'project fatigue' where those involved felt that they had lost their sense of direction and mission. For example the Craigmillar Community Information System (UK) experienced difficulties in maintaining continuous funding from external sponsors - and was forced to continually recast itself throughout its history (for example it's latest existence is as a 'Teleport' - a base for high technology training and point of access for business facilities). learning to be viable Conversely, as a project moves towards completion, there may be a need to learn to be viable . This has historically been a problem for many state-funded demonstrator projects (Jaeger, Slack and Williams forthcoming, 2000). One criticism that can be made of guaranteed state funding for an experiment is that it may divert participants' attention from the need to assess and build the future market needed to ensure the wider uptake and perhaps commercially viable of a new product and service. This had certainly been an important failing of certain historical social experiments (Cronberg et al. 1991). Some of our cases achieved viability by building an effective market of 'users'. Amsterdam Digital City provides a shining exemplar, as it succeeded in taking on commercial and other income generating services, weaning itself away from state funding as this waned, and becoming self-funding. However this remained the exception. Other cases are still dependent upon state support - and there would be questions about their future viability were this to be withdrawn. c) Modes of public support to promote social learning? 193 Social learning may be said to reach fruition when a technology becomes disseminated and embedded within particular social settings. However, current modes of public support for multimedia may not cater well for such evolution. In particular, there is a lack of fit between the relatively short-term discrete funding models that are typical of much public support (under the project-based mode of support) and the longer-term requirements of multimedia experiments driven by the dynamics and timescales of social learning. As we saw, after the Nerve Centre had developed its Colm Cille multimedia product it experienced problems in building a wider market - and identified the need to embark upon an active process of encouraging its appropriation. However public support for this 'post-project' stage proved difficult to obtain. This illustrates a problem experienced by many multimedia experiments in obtaining the kinds of longer-term funding needed to support the exploitation of technologies they had developed and in particular support for the appropriation, entrenchment and commercialisation of their offerings. The prevalent model of finite, project based funding may give little scope for a project to become established, bed down and provide the basis for more farreaching societal learning process. One implication might be that current regimes of project-based funding might need to be reformed specifically to cater for these long-term requirements - for example by encouraging longer-term multi-phase funding - either from the outset or through extension awards as a development project comes to an end. In many countries there are more or less covert modes of public support for commercialisation. However these tend to be separate from research support. The social learning perspective calls into question this traditional separation between (imputedly precompetitive) research and development and a competitive commercialisation phase (which by implication does not involve research). In shifting away from technology-supply-driven policy frameworks may suggest the need to support the production of complementary products. The case of cultural products may be a special case here. Their high up-front development costs may create disincentives to producing content geared towards minority cultures and small markets, raising the risk of market failure. There could be a synergy here between policies geared towards technology appropriation and those concerned with the preservation of cultural diversity in the face of the globalisation of the media economy how best to integrate public and commercial funding On the other hand provision of long-term public funding is not without problems as it may divert the attention of participants away from the need to become viable by building a user base. There are dangers of a culture of dependency and more significantly through a failure to address the market for products seriously, a tendency towards an approach dominated by technology supply. The question arises of whether different funding strategies can be devised that balance between public and private support, and between long-term security and the necessity for 194 short term survival - to offset such potential problems and encourage more effective social learning? Here we need to look at the integration of public and private support. As already noted, many multimedia projects were hybrids, involving various private and public sector players. We have already highlighted the need for systems of resourcing to provide incentives for social learning, in terms of the exploitation of experiences gained within a project. We should consider here both 'direct' exploitation - for example in terms of the entrenchment/ commercialisation of a new multimedia product and service - and indirect exploitation in further trials and experiments. These two may be in tension, with the former perhaps requiring that existing project constituencies to remain broadly intact and grow. For the latter this is not necessary. Closure of a project could of course be a vehicle for dispersing staff and disseminating their embodied knowledge more broadly. However, for this to happen, it is necessary that the players continue to be involved in the broad field, in roles where this experience can be applied.91 Either way the key question is whether the intermediaries involved continue to exist albeit in a different setting, with the necessary incentives and context to apply their experiences more broadly. Overall we conclude that further consideration is needed of modes of public support for multimedia. A more balanced approach is needed. Three important shifts are needed: i) away from the prevailing focus on technology development, towards its appropriation, and; ii) away from a preoccupation with hardware towards content and service development. iii) away from the provision of technical infrastructure towards the development and exchange of skills and experience. To illustrate the latter, participants in Cable School complained that it was easy to get hold of computers (which were kindly donated by a neighbouring IT manufacturer), but almost impossible to obtain the training they believed would be needed to allow teachers to explore the educational potential of this technology (Slack, 1999).92 9.4 Post Script Having explored a mass of empirical findings and considered their implications, it is useful to conclude by looking back over the intellectual journey that the SLIM 91For example when Cable Co decided that it no longer needed an local capability in promoting interactive services, but would concentrate instead on its core TV service, the key player left the field altogether and the huge body of contacts and expertise he had developed was lost – to the company and to the wide array of collaborative projects in which he had become involved. 92Important criticisms have been made of multimedia in education programmes precisely because of this emphasis on hardware - and neglect of the huge costs of retraining existing teachers (Soete et al.). Pleasingly, recent initiatives, such as the UK's National Grid for Learning (Department for Education and Employment 1997) seem to have gone some way in taking these criticisms on board. 195 project has undergone and reflect upon what has been achieved, and what still needs to be addressed. At the outset of SLIM research, when the research questions and perspectives of this project were being elaborated, they were based upon prevailing knowledge and opinion then available about the problems in developing multimedia products and services and how these might be addressed. These highlighted the problems in developing new kinds of multimedia application and use. A major current of thinking saw this in terms of acquiring a more adequate knowledge of the setting and purposes of different kinds of 'users', and thereby building a more adequate model of the user into the design of multimedia offerings. With hindsight we can see that this was an idealised and simplified schema. We found, for example, that the presumption that designers possessed relatively clear representations of the user and the user's setting was in almost all cases incorrect; instead designers were working with rather implicit and poorly specified models of the user based on incomplete and often unreliable sources of information. Our research further called into question the 'ethnographic fallacy' with its presumption that the best way to improve the design and usability of multimedia artefacts was by enhancing inputs into prior design. We found instead that there were a variety of social learning strategies including laissez-faire and evolutionary models in which users contributed directly to system configuration in the innofusion/ appropriation of artefacts rather than prior design. And supplier offerings are not ‘solutions’ to social need. Instead they are elements which ‘users’ selectively appropriate and configure with other artefacts, meanings and practices, around their evolving purposes. In this sense multimedia offerings are always ‘unfinished’, to a greater or lesser degree. In short, though the SLIM project showed that social learning is important - it revealed a rather different empirical reality from that presumed at the outset. For example, another initial presumption was that multimedia experiments would provide a context in which suppliers, users and other players could meet and negotiate their requirements. We found that it was necessary to take this point further, and see all multimedia developments as being in some way experimental, involving various forms of collaboration within and between organisations. The diverse forms for multimedia initiatives confirmed both our sense that social learning was important, also that there was considerable uncertainty about how such learning should be organised. In using the term social learning, we do not wish to convey an air of smooth and progressive accumulation and sharing of knowledge. To the contrary we found that the institutional arrangements were in most cases fragmented and subject to deep fissures and rifts - suggesting a terminology of socio-technical constellations rather than the more orderly and focused image conveyed by traditionally adopted concepts such as socio-technical constituencies or systems. We further found that these constellations tended to be clustered around either technology development or more frequently the configuration and appropriation of technology - and there was a marked absence of constellations bringing together users and developers. 196 In this context we identified the importance of intermediaries as key players in social learning, linking together players from different sites and acting as conduits for the exchange of knowledge and other resources. We also found that the importance of intermediaries and their potential contribution was often overlooked. Indeed certain informal intermediaries were not properly recognised in the multimedia experiments we examined (for example the contribution of teachers to the application of multimedia in education). Public policy needs to pay more attention to intermediaries as a resource for innofusion and for the appropriation/ domestication of multimedia. Many contemporary multimedia developments are configured from combinations of more or less standard technology components. Development of novel technical artefacts may be less important than this local appropriation effort (and one particular finding of the study concerns the importance of Appropriation constellations which have received little attention hitherto). There is a consequent need to shift effort away from technology development, narrowly conceived, towards an appropriation effort, focused upon innofusion and domestication. The development of multimedia content is also resource intensive and requires support. We should remember that the SLIM study was conceived in a context in which linear, supply-driven concepts of the innovation system were coming under criticism, and more integrated approaches were being sought which gave more attention to commercialisation efforts and building markets, as well as the development of technology artefacts. However the implication of that emerging view was that policies could be developed for better supplier-user coupling that might secure a harmonious integration of technology development and diffusion. In contrast this study has revealed the tensions and contradictions surrounding multimedia development, for example between the different dynamics of development of component technologies and of the appropriation of applications. Whilst an integrated approach to innovation policy (and business strategy) is indeed needed, this does not imply the adoption of a monolithic model that couples these tightly together. Instead loose coupling may be needed between some elements within an overall framework. For example, strategies may need to acknowledge the differing dynamics between the development of core technologies, delivery systems/ platforms, applications and the information and cultural content they hold, and thus allow greater autonomy between policy interventions across these fields. Finally we have addressed the relationship between local and global. Technology innovation has traditionally been seen in global terms - and European technology policy has often been couched in terms of combating and competing with global competitors from the USA and Japan. In contrast the SLIM study has demonstrated the enduring importance of the local - particularly in relation to certain kinds of cultural content concerned with local identities.93 Rather than the dominance of the 93Perhaps the most striking example in the SLIM study is the proliferation of cultural products geared towards local Irish history and culture. However this case immediately illustrates the ways that local and global relationships are being constantly transformed and re-worked, since these products have attracted enormous interest from the Irish diaspora in the USA, Canada and Australia in particular, who have access to these products through the internet. 197 global over the local, the SLIM study points to the interweaving of local and global elements - and the continual reworking of the relationship between them. In this context there are enormous opportunities for players at local, regional, national and European levels to contribute to the social learning processes that will enable us to genuinely domesticate the global ICT capabilities captured by the term multimedia and put them to the services of the people of Europe, and to bring about the benefits promised by the Information Society rhetorics of wealth creation, knowledge sharing, enhancing services, strengthening community and enriching culture. 198 V Dissemination Dissemination Strategy This project started in May 1996, and ran to the end of January 1999. Dissemination was given a high priority throughout the project. The dissemination strategy for SLIM was elaborated in the research proposal, and reviewed in the course of the research project. Dissemination was standing item for discussion at network management meetings. The strategy emphasised: 10 the early dissemination and discussion of research findings to policy users and wider audiences through our own products such as working papers and web pages, and 11 dissemination through academic fora, notably conferences and the publication of journal articles and books. Each research centre in the SLIM network conducted its own dissemination through its existing, nationally focused, mechanisms. In addition, the lead centre at Edinburgh University coordinated a network-wide dissemination effort. It is difficult to do justice to this extensive dissemination effort. We can briefly review Dissemination to date Co-ordinating Site and International Events A web-site has been developed, with a home page at Edinburgh University, and links to web-sites from the other SLIM centres. In this way working papers were made widely available in electronic form. This allows very rapid dissemination of findings to a wide variety of audiences globally. Dissemination of each of the three phases of the SLIM project was initiatied in turn, as soon as findings from the phase was available. Thus early deliverables from phase 1 of the study were published as: Williams, Robin (1997) 'The Social Shaping of Multimedia' Chap. 18, pp. 299 - 338, in H. Kubicek, W. Dutton and R. Williams (eds.) The Social Shaping of Information Superhighways, Campus Verlag, 1997. The national studies were published by the Norwegian National University for Science and Technology (Slack and Williams [Eds.] 1998), in addition to being published in native language as working papers Three comparative analytical papers were prepared, addressing the influence of the national context on multimedia development; the policy context; and reviewing social experiments. It is expected that these will form the basis of journal articles or book chapters. For example, the Norwegian study will appear in The Information Society (2000 forthcoming). 199 A major international conference was organized on 'The Social Shaping of Multimedia' (The University of Edinburgh, 27 – 9 June 1997), jointly supported by SLIM and the COST A4 initiative on 'The Social Shaping of Technology'. This brought together researchers and others from a variety of international networks. Papers from the workshop were edited (Slack, Stewart and Williams [Eds.] 1998) and have been published in the COST A4 series. Robin Williams edited a special edition of The Information Society, (the US journal edited by Rob Kling, University of Indiana) drawing on the conference papers and SLIM papers, on the theme of public policy and societal learning around multimedia. Birgit Jaeger organized a workshop on Technology Studies: Social Learning and Multimedia, in conjunction with Prof. Steve Woolgar, director of the UK ESRC Virtual Society initiative at the meeting of the Society for Social Studies of Science (4S), Tucson, Arizona, October 23-26, 1997. This is the premier international professional meeting for scholars in the field of Science and Technology Studies. A number of papers were presented drawing on SLIM findings. Birgit Jaeger: SCOT in Action Knut H. Sørensen: Fishing for Fun and Profit: Norway domesticates Multimedia Wiebe Bijker: The Sleeping Beauty. Multimedia in The Netherlands A further workshop was organised at the meeting of the Society for Social Studies of Science, San Diego, October 28 - 30, 1999, bringing together findings from the SLIM stream concerned with organisational applications of multimedia and findings from the TSER PRECEPT project on Business Process Redesign. A major focus of dissemination activities following completion of the SLIM research has been with publishing the mass of findings from the three integrated studies. All the integrated studies have yielded substantial monographs, which are being revised by the coordinators for each of the areas for publication, either as a special journal edition or as edited collections. In addition the overall findings of the SLIM project will be reported in a single book, being prepared by Edinburgh. This represents a substantial dissemination effort that is likely to have a major impact on European discussion in the field. There will doubtless also be further spin-off in terms of book and journal articles, conference presentations etc. Early findings from the study were discussed with interested officials from DGXII and DGXIII at a workshop (Brussels, 22-3 January 1999). Presentations of SLIM research findings were made to the European Socio-Economic Research Conference (Brussels, 28-30 April 1999). A presentation on Lessons from the SLIM Project formed the keynote for a workshop on «Socio-Economic Research on Multimedia Content and Tools» (Luxembourg, 24.3.99) organised under the EC 5FP Programme on Information Society Technologies to explore the scope for Socio-Economic Research within their Key Action on Multimedia Content and Tools. 200 National Centers Each SLIM Centre was encouraged to publish its research findings in its own working paper series and through its web pages. Each centre has links with national industry and policy constituencies, and other relevant audiences. Each centre exploited their own networks to ensure dissemination and discussion of these findings outside academia. COMTEC, Dublin City University, Ireland 'The Social Shaping of Multimedia in Ireland; Actors, Networks and Relationships' which formed part of the COMTEC Research and Policy report series. COMTEC, 1997. This report was publicized on the Irish Interactive Multimedia Association's web-site and circulated to policy makers, academics and interested parties in Ireland. The Director of the COMTEC team, Paschal Preston, has been involved in advising on the design and development of a major 'Information Age Town' project in Ireland. Largely funded by Telecom Eireann, this project will provide all homes in the selected town with a personal computer as well as access to the Internet/ WWW and other innovative ICT-based services. Paschal Preston will also lead research and evaluation work related to the 'Information Age Town' project. Centre for Technology and Society, Norwegian National University of Science and Technology, Norway The SLIM work done in Trondheim have to a large extent been distributed through ordinary academic channels and been met with interest. Some of it, mainly the stuff on gender and multimedia, has been popularized orally as well as in written form for several different audiences, in particular teachers, school authorities and other communities with professional interests in young people. SLIM has also provided very important intellectual input to a series of new R&D initiatives focussed on ICT, social sciences and the humanities, locally as well as nationally. CITA, University of Namur, Belgium. The National report on multimedia in Belgium has been distributed to the policymaking community in the country and also to other interested parties in government and industry. It is notable that the report was the first systematic inquiry into the state of multimedia in Belgium and as such has been well received. In addition, the overall framework of SLIM reports enabled a comparison to be made between the situation in Belgium and other participating nations, adding value to the extant analysis for policy makers. One notable example was the role 201 that comparative analysis played in the assessment of the work carried out in multimedia experiments projects A number of international workshops were organized by CITA in 1998 and in 1999. Three of these centered on the presentation of SLIM results: June 16, 1998 - Namur: «Rencontre réelle de villes virtuelles» (real meeting of digital cities) June 25, 1998 - PAI Seminar (Federal Office for Scientific, Technical and Cultural Affairs Inter-University Attraction Pole Framework), Namur: Information Society, Citizens and Administration. Special SLIM workshop on digital cities with the participation of other SLIM researchers Marc van Lieshout (NL), Birgit Jaeger (DK), Roger Slack and James Stewart (both UK), and Olivier Glassey (CH). June 28, 1999 - Namur : « Deuxième Rencontre réelle de villes virtuelles » (2nd real meeting of digital cities) With these workshops, publications and other activities on digital cities in which members of CITA participated, CITA has gained an important expertise in the field of digital cities, and is recognised as such in Belgium as well as abroad. CITA has been asked to collaborate to the implementation of the digital cities policy of the Walloon regional authorities at least for 2000 and 2001. RCSS, University of Edinburgh Edinburgh SLIM Centre organised a joint workshop (Edinburgh March 1998) on ‘Multimedia Futures’ in conjunction with the Interactive Multimedia Alliance for Scotland (IMAS - the specialized industry association in the field), and also involving EDINAS - Edinburgh University's knowledge transfer agency in the field of Information Technology). Further collaboration possibilities are being explored. Roskilde University, Denmark In June 1998 Birgit Jaeger presented early findings from the Danish Digital City case study at the conference on New Public Management in the Local Authorities. The conference was arranged by Roskilde University with civil service participants from local authorities in Denmark. Dissemination achievements This review of dissemination activities has pointed to something of the diversity and success of the SLIM dissemination effort. The formal deliverables to which the project was committed have been produced and are being publicised more widely mainly through academic/ commercial 202 publication, and pending this through the world-wide web. These achievements are summarised in the Executive Summary (Figure ES.1). These activities form only the tip of the iceberg of a wider dissemination effort. The full list of out outputs is included in the Annexes to this report. This shows a remarkable level of outputs including no less than 6 books and special journal editions, together with 18 book chapters, 18 journal articles. This points to the creativity and quality of the project findings and the success of dissemination efforts. The breakdown of these outputs by type of deliverable and year is shown below. Further outputs can be expected. Breakdown of SLIM outputs by type of output and by year Book Chapters Journal Articles Working Papers Conference Papers 1996 1 2 1 2 1997 5 3 7 3 9 7 Books and Journal Issues 1998 1 5 3 1999 2 4 3 2000 3 3 7 1 Totals =78 6 18 18 18 6 18 203 VI References Akrich, Madeleine (1992) `Beyond social construction of technology: The shaping of people and things in the innovation process', Chapter 9, pp. 173 -190 in Dierkes and Hoffmann (Eds.) New Technology and the Outset: Social Forces in the Shaping of Technological Innovations Frankfurt/ NY: Campus/ Westview. Akrich, M (1992a): "The description of technological objects", in W Bijker & J Law, Eds.: Shaping technology/building society, Cambridge, MA: MIT Press Andersen, E S & B-Å Lundwall (1988): "Small national systems of innovation facing technological revolutions: an analytical framework", in C Freeman & B-Å Lundwall, Eds.: Small countries facing the technological revolution, London: Pinter, p. 9-36 Arrow, K (1962): "The economic implications of learning by doing", Review of economic studies, 29, 155-173 Arthur, W. Brian (1989) `Competing Technologies, Increasing Returns and Lock-in by Historical Events' The Economics Journal Vol. 99, (March), pp. 116-131. Aune, M (1996): "The computer in everyday life: Patterns of domestication of a new technology", in M Lie & K H Sørensen, Eds.: Making technology our own? Domesticating technology into everyday life, Oslo: Scandinavian University Press, p. 94-124 van Bastelaer, Béatrice, with Claire Lobet-Maris (1999) 'Development of Multimedia in France', in R. Williams and R. S. Slack (Eds.) Europe Appropriates Multimedia, Trondheim: Norwegian University of Science and Technology, 1999. van Bastelaer, Béatrice, Claire Lobet-Maris and Jo Pierson with Jean-Claude Burgelman, Yves Punie and Frank Neuckens, (1999) 'Development of Multimedia in Belgium ', in R. Williams and R. S. Slack (Eds.) Europe Appropriates Multimedia, Trondheim: Norwegian University of Science and Technology, 1999. Berg, Anne-Jorunn and Margrethe Aune (Eds.) (1994) Domestic Technology and Everyday Life - Mutual Shaping Processes, COST A4 Vol. 1, Social Sciences, European Commission Directorate-General Sciences Research and Development, Luxembourg: Office for Official Publications of the European Communities. Berg, Anne-Jorunn (1994a) 'Technological flexibility: Bringing gender into technology (or is it the other way around)? Chapter 5, pp. 94 - 110 in Cockburn, Cynthia and Furst-Dilic, Ruza (Eds.) (1994) Bringing Technology Home: Gender and Technology in a Changing Europe Milton Keynes: Open University Press. Berg, Anne-Jorunn (1994b) 'A gendered socio-technical construction: the smart house' Chapter 9 pp. 165 - 180, in Cockburn, Cynthia and Furst-Dilic, Ruza (Eds.) (1994) 204 Bringing Technology Home: Gender and Technology in a Changing Europe Milton Keynes: Open University Press. Berg, A-J (1996): Digital feminism, Trondheim: Centre for technology and society, Norwegian University for Science and Technology Bijker, Wiebe (1993) `Do Not Despair: There Is Life after Constructivism', Science, Technology and Human Values, Vol. 18, No. 4, (Winter), pp. 113-138. Brosveet, Jarle (1998), 'Frihus 2000: Public Sector Case Study of a Norwegian IT Highway Project', Center for Technology and Society, Norwegian University of Science and Technology, Trondheim, STS Working Paper 1/98 (SLIM working paper No. 3) Brosveet, J. and Sørensen, K. H. (1999) 'Fishing for fun and profit: Norway domesticates multimedia', in R. Williams and R. S. Slack (Eds.) Europe Appropriates Multimedia, Trondheim: Norwegian University of Science and Technology, 1999. Bijker, Wiebe (1994) 'Sociohistorical Technology Studies'. Chapter 11, pp. 229 - 256 in Jasanoff, Markle, Petersen and Pinch (Eds.) Handbook of Science and Technology Studies Thousand Oaks, London & New Delhi: Sage Publications. Bijker, Wiebe & Law, John (Eds.) (1992) Shaping Technology/Building Society: studies in socio-technical change Cambridge/ MA, London: MIT Press. Brady, Tim, Margaret Tierney and Robin Williams (1992) 'The Commodification of Industry Applications Software' Industrial and Corporate Change, Vol. 1 No. 3 pp. 489514. Bruce, Margaret (1988) 'Home Interactive Telematics - Technology with a History', in van Rijn, Felix and Williams, Robin (Eds.) Concerning Home Telematics Amsterdam: North-Holland, pp. 83-93. Callon, M (1987): "Society in the making: The study of technology as a tool for sociological analysis", in W Bijker, T Hughes, T Pinch, Eds.: The social construction of technological systems, Cambridge, MA: MIT Press, p. 83-106 Callon, M (1991): "Techno-economic networks and irreversibility", in J Law, Eds.: A sociology of monsters. Essays on power, technology and domination, London: Routledge, p. 57-102 Callon, Michel (1993) `Variety and irreversibility in networks of technique conception and adoption' Chap. 11, pp. 232 - 268, in Foray, D and Freeman, C. (Eds.) Technology and the Wealth of Nations: The Dynamics of Constructed Advantage London: Pinter. Campbell-Kelly, Martin (1989) ICL: a Business and Technical History, Oxford: Oxford University Press. 205 Caracostas and Muldur (1998) Society, The endless frontier: A European vision of research and innovation policies for the 21st century European Commission DGXII Science Research Development Studies EUR 17665, Luxembourg: Office for Official Publications of the European Communities. Castells, Manuel (1996),The rise of the network society Oxford : Blackwell. Cawson, Alan, Leslie Haddon and Ian Miles (1995) The Shape of Things to Consume: Delivering IT into the Home, Aldershot, Avebury. Christoffersen, Mads and Bouwman, Harry (Eds.), (1992) Relaunching Videotex, Dordrecht: Kluwer. Clark, Peter & Staunton, Neil (1989) Innovation in Technology and Organization London, Routledge. Clausen, Christian and Robin Williams (Eds.), The Social Shaping of Computer-Aided Production management and Computer Integrated Manufacture, Vol. 5, Proceedings of International Conference, COST A4, ISBN 92 828 1569 2, European Commission DGXIII, Luxembourg 1997. Cockburn, Cynthia (1993) 'Feminism/ Constructivism in Technology Studies: Notes on Genealogy and Recent Developments' paper to workshop on European Theoretical Perspectives on New Technology: Feminism Constructivism and Utility Brunel University, September 1993. Cockburn, Cynthia and Furst-Dilic, Ruza (Eds.) (1994) Bringing Technology Home: Gender and Technology in a Changing Europe Milton Keynes: Open University Press. Cohen, Eli (1992), Le Colbertisme "high tech". Économie des télécom et du grand projet. Paris Collingridge, David (1980) The Social Control of Technology. London: Pinter. Collingridge, David (1992) The Management of Scale: Big Organizations, Big Decisions, Big Mistakes London/ NY: Routledge. Collinson, Simon (1993) 'Managing product innovation at Sony: the development of the Data Discman'. Technology Analysis and Strategic Management, Vol.. 5, No. 3, pp. 285-306 Collinson, Simon et al. (1996) Forecasting and Assessment of Multimedia in Europe 2010+, Final Report to European Commission, Edinburgh: University of Edinburgh. Collinson, S. and Molina, A. (1998) Reorganising for knowledge integration and constituency building: product development at Sony and Philips. pp. 76-107. In 206 Cooms, R., Green, K., Richards, A. and Walsh, V. (Eds.) Technological Change and Organization Cheltenham: Edward Elgar. Collinson, Simon (1999) 'The Development of Multimedia in Japan' in R. Williams and R. S. Slack (Eds.) Europe Appropriates Multimedia, Trondheim: Norwegian University of Science and Technology, 1999. Coombs, Rod, Saviotti, Paulo & Walsh, Vivien (1987) Economics and Technological Change Basingstoke: Macmillan. Coopersmith, Jonathon (1993) 'Facsimile's false starts', IEEE Spectrum, February, pp. 46-49. Cowan, Robin (1992) 'High technology and the economics of standardization' Chap. 14, pp. 279-300 in Dierkes and Hoffmann (Eds.) New Technology and the Outset: Social Forces in the Shaping of Technological Innovations Frankfurt/ NY: Campus/ Westview. Cronberg, Tarja; Duelund, Peter; Jensen, Ole Michael & Qvortrup, Lars (1991) (Eds.): Danish Experiments - Social Constructions of Technology. Copenhagen: New Social Science Monographs. Cronberg, Tarja (1992) Technology in Social Sciences: the Seamless Theory, mimeo, Technical University of Denmark, Lyngby, 20 pp. Cronberg, T. and Sørensen, K. (Eds.) (1995) Similar Concerns, Different Styles Technology Studies in Western Europe Proceedings of the COST A4 workshop, Ruvaslahti, Finland 13-14 January 1994 Luxembourg: Office for Official Publications of the European Communities. Curry, Andrew (1999) ‘Learning the Lessons of Videoway: The corporate economy of new media trials’ pp. 73 – 86 in Slack, Stewart and Williams (Eds.), The Social Shaping of Multimedia, COST A4, European Commission DGXIII, Office for Official Publications of the European Community: Luxembourg 1999. David, Paul (1975) Technical Choice, Innovation and Economic Growth: Essays on American and British Experience in the Nineteenth Century, Cambridge: Cambridge University Press. Department for Education and Employment, (1997) Connecting the Learning Society: The National Grid for Learning. London: UK Department for Education and Employment. Dosi, G. C. Freeman, R. Nelson, G. Silverberg and L. Soete (Eds.) Technical change and economic theory (pp. 38-66). London: Pinter Publishers. Dunlop, Charles and Rob Kling (Eds.) (1991) Computerization and controversy : value conflicts and social choices, Boston : Academic Press, 207 Dutton, W.H. (1995) 'Driving into the future of Communications? Check the Rear View Mirror' pp, 79 - 102 in S. J. Emmott (Ed.) Information Superhighways: Mutlimedia Users and Futures, London: Academic Press. Dutton, W.H. (Ed.) (1996) Information and Communication Technologies: Vision and Realities, Oxford: Oxford University Press. Edge, David (1988) `The Social Shaping of Technology' Edinburgh PICT Working Paper No. 1, Edinburgh University. Egyedi, T. M., Lieshout, M. van & Bijker, W. E. (1999) Education Final report: SLIM Education Stream. Ehn, P. (1988) Work-Oriented Design of Computer Artifacts Arbetslivcentrum, Stockholm . European Commission COM (96) 607 (1996) Europe at the Forefront of the Global Information Society: Rolling Action Plan, Communication from the European Commission, Brussels, Commission of the European Communities 27 Nov. 1996. European Commission (1996) Preliminary guidelines for the Fifth Framework Programme of research and technological development activities. Inventing Tomorrow: Europe's research at the service of its people. COM(96) 332 European Science Foundation/ Economic and Social Research Council (1991) Social Sciences in the Context of the European Communities Strasbourg/ Swindon: ESF/ ESRC. Faulkner, Wendy and Jacqueline Senker, with Lea Velho (1995) Knowledge Frontiers: Industrial Innovation and Public Sector Research in Biotechnology, Engineering Ceramics, and Parallel Computing Oxford: Clarendon Press. Fincham, Robin, Fleck, James, Procter, Robert, Scarbrough, Harry, Tierney, Margaret & Williams, Robin (1995) Expertise and Innovation: Information Strategies in the Financial Services Sector. Oxford: Oxford University Press/ Clarendon. Fischer, Claude (1992) America Calling. A social history of the telephone to 1940, Berkeley: University of California Press. Fleck, James (1988) `Innofusion or Diffusation? The nature of technological development in robotics' Edinburgh PICT Working Paper No. 7, Edinburgh University. Fleck, James (1988a) 'The Development of Information Integration: Beyond CIM?' Edinburgh PICT Working Paper No. 9, Edinburgh University. A digest of this paper, prepared for the Department of Trade and Industry, is available as 'InformationIntegration and Industry', PICT Policy Research Paper No. 16, Economic and Social Research Council, Oxford, 1991. 208 Forester, Tom (Ed.) (1985) The Information Technology Revolution, Oxford: Blackwell. Forester, Tom (Ed.) (1989) Computers in the Human Context: Information Technology, Productivity and People, Oxford: Blackwell. Fransman, M. (1998) "Convergence, the Internet and Multimedia: Implications for the >Evolution of Industries and Technologies". In Proceedings of the >International Telecommunications Society 12th Biennial Conference, Freeman, Christopher, (1974) The economics of industrial innovation Harmondsworth/ Baltimore: Penguin, Freeman, Chris, Clarke, J. & Soete, L. (1982) Unemployment and Technical Innovation: A Study of Long Waves in Economic Development London, Pinter. Galison, Peter.(1997), Image and logic : a material culture of microphysics, Chicago ; London : University of Chicago Press. du Gay, Paul, Stuart Hall, Linda Janes, Hugh Mackay and Keith Negus Doing Cultural Studies: The story of the Sony Walkman, London/ New Delhi: Sage 1997. Gibbons, Michael, et al. (1994), The new production of knowledge : the dynamics of science and research in contemporary, London : Sage. Gilfillan, S (1970): The sociology of invention, Cambridge, MA: MIT Press Graham, Ian, Graham Spinardi, and Robin Williams, (1996) 'Diversity in the emergence of electronic commerce' Journal of Information Technology, Vol. 11 No. 1 pp. 161 - 172. Green, Eileen, Owen, Jenny, Pain, Den (Eds.) (1993) Gendered by Design? Information Technology and Office Systems, London/ Washington DC, Taylor & Francis. Hård, Mikael (1994) `Technology as Practice: Local and Global Closure Processes in Diesel Engine Design', Social Studies of Science, Vol. 24, No. 3 (August), pp. 549 - 585. Haddon, Leslie (1992) 'Explaining ICT Consumption: The case of the Home Computer' chap 5 pp. 82 - 96 in Silverstone and Hirsch (Eds.) (1992). Hansen, F (1998) Teleteaching on Bornholm. SLIM Education Stream Report. Herbold, Ralf (1995) ‘Technologies as Social Experiments. The Construction and Implementation of a High-Tech Waste Disposal Site’ Chap. 9, pp 185 – 197 in Rip, Arie, Thomas J. Misa and Johan Schot (Eds.) (1995) Managing Technology in Society: The approach of Constructive Technology Assessment London/ NY: Pinter. 209 High-Level Group on the Information Society (1994) [Bangemann Report] Europe and the global information society: Recommendations to the European Council, Brussels May 26 1994 Hirschman, A (1970) Exit, voice, loyalty, Cambridge, MA: Harvard University Press Von Hippel, E. (1988) The sources of innovation, New York ; Oxford : Oxford University Press. Von Hippel , E. (1994) “ Sticky Information” and the Locus of Problem Solving: Implications for Innovation. Management Science. 40(4); 429-439. Howells, John & Hine, Jim (Eds.) (1993) Innovative Banking: Competition and the management of a new networks technology London/ NY: Routledge. Hughes, Thomas (1983) Networks of Power Baltimore, MD & London: Johns Hopkins University Press. Jaeger, Birgit & Qvortrup, Lars, 'Community Teleservice Centres', In: Cronberg, Tarja; Duelund, Peter; Jensen, Ole Michael & Qvortrup, Lars (Eds.): Danish Experiments Social Constructions of Technology, p. 27-41. Copenhagen: New Social Science Monographs, 1991. Jaeger, Birgit and Finn J. S. Hansen, (1999) 'Multimedia in Denmark' in R. Williams and R. Slack, (Eds.) Europe Appropriates Multimedia, Trondheim: Norwegian University of Science and Technology, 1999. Jaeger, Birgit, Roger Slack and Robin Williams, (2000 forthcoming) 'Europe Experiments with Multimedia: An Overview of Social Experiments and Trials' to appear in The Information Society Jakobs, Kai and Robin Williams (1999) Proceedings of First IEEE Conference on Standardisation and Innovation in Information Technology (SIIT '99) Aachen, Germany, September 15-17, 1999, Aachen University: Aachen. Jenkins, (1975), 'Technology and the market: George Eastman and the origins of mass amateur photography', Technology and Culture, 15, p. 1-19. Kahin, Brian, and James Keller (Eds.) (1995) Public Access to the Internet Cambridge MA/ London: MIT Press. Kahin, B., & Wilson, E. (1997), (Eds.) National Information Infrastructure Initiatives, Vision and Policy Design. Cambridge, Mass: MIT Press. 210 Kemp, R, J. Schot, and R. Hoogma (1998): “ Regime shifts through processes of niche formation: the approach of strategic niche management” , Technology analysis and strategic management 10. Kerr, Aphra and Paschal Preston (1999) 'The Development of Multimedia in Ireland', in R. Williams and R. S. Slack (Eds.) Europe Appropriates Multimedia, Trondheim: Norwegian University of Science and Technology, 1999. Kranakis, Eda (1988) `Technology Assessment and the Study of History' Science, Technology & Human Values Vol. 13, Nos. 3 & 4 (Summer & Autumn), pp. 290 - 307. Kubicek, Herbert, William Dutton and Robin Williams (Eds.) (1997), The Social Shaping of Information Superhighways: European and American Roads to the Information Superhighway, Frankfurt: Campus Verlag. Kubicek, Herbert, Ulrich Schmid and Bernd Beckert (1999) 'The Development of Multimedia in Germany', in R. Williams and R. S. Slack (Eds.) Europe Appropriates Multimedia, Trondheim: Norwegian University of Science and Technology, 1999. Latour, Bruno (1986) Science in Action Milton Keynes: Open University Press. Latour, Bruno (1988) `How to Write "The Prince" for Machines as well as Machinations',, pp. 20-43, in Elliott, Brian (Ed.) (1988) Technology and Social Process Edinburgh: Edinburgh University Press. Law, John & Callon, Michel (1992) `The Life and Death of an Aircraft: a Network Analysis of Technological Change', pp. 29-52, in Bijker and Law (Eds.) Shaping Technology/Building Society: studies in socio-technical change Cambridge/ MA, London: MIT Press. van Lente, Harro (1999) 'Development of Multimedia in The Netherlands', in R. Williams and R. S. Slack (Eds.) Europe Appropriates Multimedia, Trondheim: Norwegian University of Science and Technology, 1999. Lie, M & K H Sørensen, Eds. (1996), Making technology our own? Domesticating technology into everyday life, Oslo: Scandinavian University Press Lobet-Maris, Claire (1997), 'European Policy in Multimedia Development', First report for the cross-national comparative analysis. Mimeo, FUNDP Namur. Lobet-Maris Claire, with Beatrice van Bastelaer and Bart Cammaerts, 'On the Role of Government in the Information Society', in Bart Cammaerts and Jean-Claude Burgelman (Eds.), Paving the way for a new public (tele)communication policy, VUB press, 1999. Lobet-Maris Claire and Beatrice van Bastelaer 1999 Digital Cities Final Report 211 McCracken, G (1988) Culture and consumption: New approaches to the symbolic character of consumer goods, Bloomington, IN: Indiana University Press Mackay, Hughie & Gillespie, Gareth (1992) `Extending the Social Shaping of Technology Approach: Ideology and Appropriation', Social Studies of Science, Vol.22, pp.685-716. MacKenzie, Donald & Wajcman, Judy (Eds.) (1985) The Social Shaping of Technology: How the Refrigerator Got Its Hum Milton Keynes, Open University Press. Mansell, Robin E. The new telecommunications : a political economy of network evolution, London : Sage, 1993 Mick, David G., and Susan Fournier (1998) ‘Paradoxes of technology: Consumer Cognizance, Emotions and Coping Strategies’ Marketing Science Institute Working Paper, Cambridge MA: MSI. Miles, Ian (1990) Home Telematics: Information Technology and the Transformation of Everyday Life London: Frances Pinter. Molina, Alfonso (1989) The Social Basis of the Microelectronics Revolution Edinburgh: Edinburgh University Press. Molina, A. (1992) 'Competitive strategies in the microprocessor industry : the case of an emerging versus an established technology'. International Journal of Technology Management, 7, 6/ 7/ 8 [Special issue on the Strategic Management of Information and Telecommunication Technology], p.589-614 Molina, A. (1994) 'Insights into the successful generation of a large-scale European initiative : from mis-alignment to programmatic alignment in the build up of sociotechnical constituencies.' pp. 90-120, in Mansell, R. [Ed.] The Management of Information and Communication Technologies: Emerging Patterns of Control, London, ASLIB. Morris, Charles R. and Charles H Ferguson (1992) 'How Architecture Wins Technology Wars' Harvard Business Review Vol. 71, No. 2, pp. 86 - 96. Mourik, Ruth (1998) Language Course. SLIM Education Stream Report. Myervohld, N. (1998) Transcript of Interview given to BBC Radio 4 In Business. Nelson, R R & S Winter (1982) An evolutionary theory of economic change, Cambridge, MA: Belknap Press Nicoll, D. W. (1999). Users as currency: Technology and marketing trials. pp. 55-72 In Roger Slack, James Stewart and Robin Williams (Eds.), The Social Shaping of Multimedia, 212 COST A4, European Commission DGXIII, Office for Official Publications of the European Community: Luxembourg. Norman, D (1988) The psychology of everyday things, New York: Basic Books Norman, D (1992): Turn signals are the facial expressions of automobiles, Reading, MA: Addison-Wesley Norman, D A, The invisible computer Cambridge, MA: MIT Press, 1998. Oakley, Brian (1990) Look back in Alvey : why support for R&D is not enough. Edinburgh PICT Working Paper, no.18, Edinburgh University; Edinburgh, Pacey, Arnold (1983) The culture of technology, Oxford : Blackwell, . Perez, Carlotta (1983) `Structural Change and the Assimilation of New Technologies in the Economic and Social System', Futures, Vol. 15, pp. 357-75. Pinch, Trevor & Bijker, Weiber (1984) `The Social Construction of Facts and Artefacts: Or How the Sociology of Science and the Sociology of Technology might Benefit Each Other', Social Studies of Science, Vol. 14, No. 3 (August), pp. 399-441. Porter, Michael (1990) The competitive advantage of nations, London : Macmillan. Preston, P. (1999). Culture final report. SLIM Culture Stream. Procter R. N. and R. Williams (1996) 'Beyond Design: Social Learning and ComputerSupported Cooperative Work: some lessons from Innovation Studies', Chap. 26, pp. 445 - 464, in Shapiro, D., M. Tauber and R. Traunmueller (Eds.), (1996) The Design of Computer-Supported Cooperative Work and Groupware Systems, Amsterdam: North Holland, 1996. Procter Robert N., Robin Williams and Les Cashin (1999) ‘Social Learning and New Strategies for the Management of Innovation in Multimedia’ pp. 155 – 169 in : Roger Slack, James Stewart and Robin Williams (Eds.), The Social Shaping of Multimedia, COST A4, European Commission DGXIII, Office for Official Publications of the European Community: Luxembourg. Quintas, Paul (Ed.) (1993) Social Dimensions of Systems Engineering: People, Processes, Policies and Software Development, NY/ London: Ellis Horwood. van Rijn, Felix and Williams, Robin (Eds.) (1988) Concerning Home Telematics Amsterdam: North-Holland. Rip, Arie, Thomas J. Misa and Johan Schot (Eds.) (1995) Managing Technology in Society: The approach of Constructive Technology Assessment London/ NY: Pinter. 213 Rip, Arie (1995) Introduction of New Technology: Making Use of Recent Insights from Sociology and Economics of Technology' Technology Analysis and Strategic Management, Vol. 7, No. 4, pp. 417 - 431. Rip, A., and J. Schot (1999): “ Anticipation on Contextualization: Loci for Influencing the Dynamics of Technological Development” , forthcoming in Sauer, D., and C. Lang (Hrsg.): Paradoxien der Innovation. Perspektiven sozialwissenschaftlichter Innovationsforschung, Frankfurt/ New York, Campus Verlag, 1999. Proceedings, workshop Verbund sozialwissenschftliche Technikforschung, München, December 1998. Rogers, E M (1983) Diffusion of innovations, New York: The Free Press Rosen, Paul (1993) 'The Social Construction of Mountain Bikes: Technology and Postmodernity in the Cycle Industry', Social Studies of Science, Vol. 28, No. 3 (August), pp. 479 - 513. Rosenberg, N (1979) Perspectives on Technology, Cambridge: Cambridge University Press Rosenberg, N (1982) Inside the black box: Technology and economics, Cambridge: Cambridge University Press Rosenberg, Nathan (1994) Exploring the Black Box: Technology, Economics and History, Cambridge: Cambridge University Press. Rossel, Pierre (1999), 'Multimedia in Switzerland', in R. Williams and R. S. Slack (Eds.) Europe Appropriates Multimedia, Trondheim: Norwegian University of Science and Technology, 1999. Rossell, Pierre (1999a) Telepoly case. Report for the Organisations Stream. Schmutzer, Rupert (1999) ‘The “ Social Construction” of Mew media Characteristics and Effects’ pp. 209 – 218. in Slack, Stewart and Williams (Eds.), The Social Shaping of Multimedia, COST A4, European Commission DGXIII, Office for Official Publications of the European Community: Luxembourg 1999. Schneider, Volker et al., (1991) 'The Dynamics of Videotex Development in Britain, France and Germany: A cross-national comparison' European Journal of Communication Vol. 6 No. 2, June 1991 Schneider, Volker (1997), 'Different Roads to the Information Society? Comparing U.S. and European Approaches from a Public Policy Perspective' Chap. 19 in Herbert Kubicek, William Dutton and Robin Williams (Eds.) The Social Shaping of Information Superhighways: European and American Roads to the Information Superhighway, Frankfurt: Campus Verlag, pp. 339 - 358. 214 Schon, Donald A. (1983), The reflective practitioner : how professionals think in action, London : Temple Smith. Schneider, Volker (2000, forthcoming), ‘Evolution in Cyberspace: The Adaptation of National Videotext Systems to the Internet’, The Information Society. Schot, Johan (1992) `Constructive Technology Assessment and Technology Dynamics: the Case of Clean Technologies', Science, Technology, & Human Values, Vol. 17, No. 1 (Winter), pp. 36-56. Silverstone, Roger (1991) `Beneath the Bottom Line: Households and Information and Communication Technologies in an Age of the Consumer' PICT Policy Research Papers No. 17, Swindon: Economic and Social Research Council. Silverstone, R (1994) Television and everyday life, London: Routledge Silverstone, Roger and Eric Hirsch (Eds.) (1992) Consuming technologies : media and information in domestic spaces, London : Routledge. Silverstone, R, E Hirsch & D Morley (1992), "Information and communication technologies and the moral economy of the household", in R Silverstone & E Hirsch, Eds.: Consuming Technologies. Media and information in domestic spaces, London: Routledge Slack, Roger (Forthcoming) Learning in Cable School. To Appear in van Lieshout, M., Egyedi, T. and Bijker, W. E. (Eds.) Social Learning in Multimedia: Education. Aldershot: Ashgate. Slack, Roger and Stewart, James (1999), 'Multimedia in the UK', in R. Williams and R. S. Slack (Eds.) Europe Appropriates Multimedia, Trondheim: Norwegian University of Science and Technology, 1999. Slack, Roger, James Stewart and Robin Williams (Eds.), The Social Shaping of Multimedia, COST A4, European Commission DGXIII, Office for Official Publications of the European Community: Luxembourg 1999. Sørensen, K H (1994): "Technology in use. Two essays on the domestication of artefacts", STS Working paper 2/ 94, Trondheim: Centre for technology and society Sørensen, Knut H. (1994a) The Car and Its Environments: The Past, Present and Future of the Motorcar in Europe, Proceedings of COST A4 Workshop, Trondheim, Norway, 6-8 May 1993, COST Social sciences, Vol. 2. Luxemburg: European Commission DGXIII Science Research and Development. Sørensen, Knut H and Berg, Anne-Jorun (Eds.) (1992) Technologies and Everyday Life: Trajectories and Transformations, Proceedings from a Workshop in Trondheim, May 28- 215 29 1990, Report No. 5, Oslo: Norwegian Research Council for Science and the Humanities. Sørensen, K H, M Aune & M Hatling (1996): "Against linearity. On the cultural appropriation of science and technology", STS Working Paper, 8/ 96, Norwegian University of Science and Technology : Trondheim. Sørensen, Knut H. (1994) "Adieu Adorno: The moral emancipation of consumers", in A-J Berg and M Aune, (Eds.) Domestic technology and everyday life - mutual shaping processes, COST Social sciences vol 1, Luxemburg: European Commission DGXIII Science Research and Development, pp- 157-169. Sørensen, Knut H. (1996) ‘Learning technology, constructing culture. Socio-technical change as social learning’ STS working paper no 18/ 96, University of Trondheim: Centre for technology and society. Spacek, Thomas R., (1997), 'How Much Interoperability Makes an NII?' Chap. 4, pp. 69 - 77 in Herbert Kubicek, William Dutton and Robin Williams (Eds.) The Social Shaping of Information Superhighways: European and American Roads to the Information Superhighway, Frankfurt: Campus Verlag. Spilker, H. S. (1998) ‘Girlsrom- A ROM of One’s Own’. SLIM report for Cultural Content Stream. Star, S. L. and Griesemer, J. R. (1989) "Institutional ecology, 'translations,' and boundary objects: amateurs and professionals in Berkeley's museum of vertebrate zoology 1907-1939". Social Studies of Science Vol. 13, pp. 205-228. Stewart, J. (1999 In press) Cafematics: communities in the computer. In Gurstein, M. (Ed.) Community Informatics Toronto: Idea Group. Stewart, J. (forthcoming) Networks and the Use of New ICTs. Draft PhD. University of Edinburgh. Steenhoudt, Frans (1997) De nieuwe cyberpalen: gebruiksonvriendelijk (Techno Antwerpen Deel 2), in De Morgen - Metro, 3/ 3/ 97, 8. Suchman, Lucy A (1987) Plans and situated actions : the problem of human-machine communication Cambridge : Cambridge University Press. Swann, P. & Lamaison, H. (1989) 'The Growth of an IT Network: a Case Study of Personal Computer Applications Software', Discussion Papers in Economics, No. 8807, Brunel University. Swann, P. (1990) 'Standards and the Growth of a Software Network', pp. 383-93, in J. L. Berg & H. Schumny (Eds.) An Analysis of the Information Technology Standardization Process Amsterdam: Elsevier Science/ North-Holland. 216 Tait, Joyce and Robin Williams (1999), 'Policy Approaches to Research and Development: Foresight, Framework and Competitiveness' Science and Public Policy, 1999, Vol. 26, No. 2, pp. 101 - 112. Thierry Vedel (1997), 'Information Superhighway Policy in France: The End of High Tech Colbertism', in Kahin, B., & Wilson, E. (Eds.). National Information Infrastructure Initiatives, Vision and Policy Design. Cambridge, Mass: MIT Press. Thomas, Graham and Ian Miles (1990) Telematics in Transition London: Longman. Vincenti, Walter G. (1994) `The Retractable Airplane Landing Gear and the Northrop "Anomaly": Variation-Selection and the Shaping of Technology', Technology and Culture, Vol. 35, pp. 1 - 33. Webster, Juliet (1990) Office Automation: the labour process and women's work in Britain Hemel Hempstead: Harvester Wheatsheaf. Webster Juliet and Robin Williams 'Mismatch and Tension: Standard Packages and Non-standard Users' in P. Quintas (ed.), Social Dimensions of Systems Engineering: People, Processes, Policies and Software Development 1993, Ellis Horwood, pp. 179-196 West, Joel (1996), 'Utopianism and National Competitiveness in Technology Rhetoric: The Case of Japan's Information Infrastructure' The Information Society, 12, pp. 251 272. Williams, Raymond (1990) Television, Technology and Cultural Form, London: Routledge. Williams, R. (1993) Information Technology in Organisations, Report to European Commission DGIII Industrial Policy Directorate, IT Strategy Unit, as part of a study on the Socio-Economic Aspects of a European Information Space; ,pp. 33 Available as Edinburgh PICT Working Paper No 54, (1994) Edinburgh University/ RCSS, Edinburgh. Williams, R. (1995) (Ed.) The Social Shaping of Inter-Organisational Network Systems and Electronic Data Interchange, Proceedings of International Conference, COST A4, European Commission DGXIII, Luxembourg 1995, pp. vi, 216 Williams, R. and Edge, D. (1996) 'The social shaping of technology', Research Policy 1996, Vol. 25, pp. 865 - 899. Williams, Robin (1997) 'The Social Shaping of Information and Communications Technologies' Chap. 18 in H. Kubicek et al. (Eds.) The Social Shaping of Information Superhighways: European and American Roads to the Information Society, NY/ Frankfurt: Westview/ St Martins Press. Williams, Robin (1997a) ‘Globalisation and Contingency: Tensions and Contradictions in the Mutual Shaping of Technology and Work Organisation' Chap. 8, pp. 170 - 185, 217 in McLoughlin, I and D. Mason (Eds.), Innovation Organizational Change and Technology, London: International Thompson Business Press. Williams, Robin, Wendy Faulkner and James Fleck (Eds.) (1998) Exploring Expertise, Basingstoke: Macmillan. Williams, Robin, (1999) 'The National Appropriation of Multimedia' Chap. 1, pp. 1 - 19 in R. Williams and R. Slack (Eds.) Europe Appropriates Multimedia: A study of the National Uptake of Multimedia in Eight European Countries and Japan, Senter for Teknologi og Samfunn (Centre for Technology and Society), Report No. 42 Norwegian University of Science and Technology, Trondheim: Norway. ISSN 0802 3581 42 Williams, Robin (2000 forthcoming), 'Public choices and social learning: the new multimedia technologies in Europe' accepted for The Information Society Special Edition on Multimedia. Winston, Brian (1998) Media Technology and Society, A History: From the Telegraph to the Internet, London/ NY: Routledge Woolgar, S (1991): "Configuring the user: the case of usability trials", in J Law, (Ed.), A sociology of monsters. Essays on power, technology and domination, London: Routledge, pp. 57-102 218