Responsibilization through visions

ABSTRACT

Vision Assessments in technology assessment (TA) focus on how constructed futures (such as visions, scenarios and roadmaps) influence processes of innovation. To extend this perspective, we want to focus on the following question: What responsibilities do the practices of using such futures produce? Therefore, we offer a concept which analyzes visions as socio-epistemic practices in present re-arrangements of sociotechnical constellations which shape larger processes of innovation. Such visionary practices enable and produce particular responsibilizations between actors. Through three examples – visions as socio-epistemic practices in experiments with smart grids, in the global movement of FabLabs, and in discourses on Big Data – we distinguish three modes of responsibilization: re-forming responsibilities, inventing responsibilities and creating irresponsibility. Although not providing ethical criteria to assess whether an innovation process is ‘responsible’, our approach enables basic knowledge for research and practices of responsible innovation through contributing to an understanding of how responsibilities are produced and distributed in practice.

KEYWORDS:

• Vision assessment
• technology assessment
• responsibility
• smart grid
• fab lab
• Big Data

1. Introduction

Constructed futures (such as visions, scenarios and roadmaps) circulate in diverse arenas of sociotechnical processes of innovation, which are assumed to influence the processes – i.e. enabling and steering negotiations, decisions and practical coordination of the included and addressed actors in research policy, laboratory practices, the media and many more fields. From a variety of research on visions (e.g. Grin and Grunwald 2000; Ferrari and Lösch 2017), imaginaries of the future (Adam and Groves 2007; Jasanoff and Kim 2015) and expectations (e.g. Brown, Rappert, and Webster 2000) in technology assessment (TA) and Science and Technology Studies (STS), we know that such constructed futures have various impacts on processes in the present, which might result in innovations.¹ From our point of view, research and practices of responsible research and innovation (RRI) would benefit from such research on futures, if we can answer the question: Who is made responsible for particular actions by the practical use of such futures in the diverse arenas of innovation processes? The answer to this question is to date an unsolved challenge in RRI and related research in STS and TA. Our paper tries to address this challenge by following the analytical research question: What responsibilities do visions of the future produce in the practices of their use?

The European debate on ‘Responsible Research and Innovation’ (see, e.g. Owen, Bessant, and Heintz 2013; von Schomberg 2013; van den Hoven et al. 2014), for example, is itself driven by a future vision of new forms of distributed responsibilities in innovation processes which exceed the traditional distributions, typically among the actors of the market, towards a distributed responsibility of many societal actors. The visionary aim is to make innovations more ‘responsible’² through an orientation towards collectively held ethical guidelines and values. From the perspective of vision assessment in technology assessment (TA), we not only know that future visions influence political decisions and public deliberation processes, which have impacts on innovation processes and their societal embedding (see, e.g. Grin and Grunwald 2000; Lösch 2006b; Grunwald 2012); we also know that future visions of technologies themselves attribute responsibilities narratively or in the use practices of societal actors. That visions of future technologies do this is plausible, if we recognize that such visions are never limited to proclamations of technical changes. These changes always imply social transformations – that is, transformations of relevant actor-constellations that are situated in broader ‘sociotechnical arrangements’.³ Such changes imply new distributions and ascriptions of the responsibilities that the actors would or should have in the envisioned actor-constellations of the future. That is, visions in practice effect ‘responsibilizations’.⁴

Our paper argues that if we analyze what visions do in the corresponding practices of their use, we can also identify how visions contribute to the distribution of responsibilities in the practices of innovation processes, without which there would be no such processes in the first place. The analysis and assessment of such distribution of responsibilities in processes of change is important for both research into and the practice of responsible innovation. Therefore, we seek to know what visions enable and hinder in practical contexts and how their practices contribute to processes of responsibilization. Such processes are an effect of visionary discourses in which particular actors are addressed as responsible agents now or in the future. To enable, prevent or realize an envisioned future, particular actors are discursively addressed as capable and in need of doing something. Because responsibilization is an enabling condition of innovations, insight into such processes is needed for the realization of the above-mentioned RRI vision of distributed responsibilities in concrete fields of change. Such a perspective on visions focuses on the social processes of attributing and distributing responsibilities. It is not itself an ethical assessment of the visions, but it could provide orientation for ethical assessments of the normative implications of reponsibilization processes and how such processes could be shaped in better ways.⁵

Our paper offers an analytical concept for these theoretical and practical demands. We state that visions have to be analyzed as socio-epistemic practices in present innovation and transformation processes. With the concept ‘visions as socio-epistemic practices’, we focus simultaneously on the twofold effect of visionary practices that enable the production of knowledge and visionary practices that newly order sociotechnical arrangements. Taken together, both are a requirement for innovation and transformation processes. Accordingly, the hypothesis of this paper is that visions as such socio-epistemic practices enable and produce new responsibilizations of the included and/or addressed actors. That means the effects of visionary practices are the normative ascriptions of responsibilities between the actors (among both the participating and also the externally addressed actors). Insights in such processes are – from our point of view – not only important for technology assessments of present processes, they are (or should be) fundamental for research and also practices of responsible innovation. Research on responsible innovation has to know such practices of responsibilization because they are constitutive for all processes of innovation. In practice the knowledge derived from such an analytical view is important if one wants to create or use visions in a ‘responsible’ manner. And more: through our perspective it becomes clear that a ‘responsible’ shaping of innovations has to start where the visions are generated, used and distributed. Producers and users of visions are responsible – this should be evident on the basis of our elaborations – because visions lead to the ascriptions of responsibilities among the actors which collectively shape the processes. Only if we know the dynamics of productions and distributions of responsibilities can we know where or on which levels of a certain sociotechnical arrangement and process we could intervene.

After positioning our perspective in relation to other selected perspectives in STS, TA and RRI on the role of futures for responsible innovation and sketching the most important levels of our analytical concept (Chapter 2), we show the benefits of our concept by means of three exemplary cases. We analyze responsibilizations through visions as socio-epistemic practices in experiments with smart grids in Germany, in the global movement of FabLabs that aim to make digital fabrication accessible, and in the case of the discourses on Big Data (Chapter 3). In the conclusion, we compare the three cases and reflect on the consequences of our research insights for TA and especially responsible innovation in theory and practice. Here we also point out the responsibilities of TA-actors and RRI-actors in the unavoidable explicit or implicit visionary practices of every larger processes of innovation and transformation in the current society (Chapter 4).

2. Analytical perspective in the context of related research

Various research strands from Foresight, TA, STS and on RRI focus on relations between futures and responsibility in specific ways, but they do not see the processes of responsibilization which take place in the practical use of visions in the course of innovation processes. Through their construction of diverse potential futures (e.g. scenarios), foresight and scenario activities in TA try to produce future-oriented knowledge in order to enable ‘better’ decisions and actions in the present (see, e.g. Grunwald 2013; Dieckhoff 2015; Gransche 2015). They do not, however, reflect on the practices that underlie the use of their products, in which, as we state, responsibilizations between the participating actors are emerging. The hermeneutic extension of TA proposed by Armin Grunwald as a contribution of TA to responsible innovation aims to recognize the ascriptions of meaning through futures in current processes in order to enlighten society about the scientific and normative implications of these ascriptions (see, e.g. Grunwald 2014; Grunwald 2017). The aim here is to shape communications and decisions towards desired innovations in a ‘responsible’ manner. While this approach explicitly asks who is responsible for a given future, it leaves implicit whom a given future makes responsible in the practices of its use.

Against technoscientific orientations that try to shape innovations in orientation towards far-reaching visions (long-term futures of about 50 years), the philosopher Alfred Nordmann argues that a ‘responsible’ shaping of innovations is only possible through orientation towards problems currently experienced by society and through the ‘craft of anticipating’ innovative solutions (Nordmann 2014). But his elaborations do not address the question of who should be responsible for correctly crafting such anticipations. Answering this question would again require an analysis of how the processes of anticipation themselves produce and distribute responsibilities.

Finally, we look at René von Schomberg’s ‘vision of responsible research and innovation’ (2012, 2013) and the corresponding European debate on responsible innovation (see, e.g. Owen, Bessant, and Heintz 2013; van den Hoven et al. 2014). In von Schomberg’s vision, RRI should ‘lead to an inclusive innovation process whereby technical innovators become responsive to societal needs and societal actors become co-responsible for the innovation process by a constructive input in terms of defining societally-desirable products’ (von Schomberg 2013, 65). This vision of RRI, therefore, envisions the forming of responsible collectives of heterogeneous actors that shape responsible innovations. From our point of view, one needs to be aware that the responsible collective is produced rather than something that has a prior existence in society. The collective is the envisioned result of a distribution of responsibilities amongst actors, which is itself a product of innovation processes, in which visions are constitutive in the sense of socio-epistemic practices. That means, in innovation processes, responsibilities are always produced and distributed, that is also the case in processes of RRI. RRI, as sketched by the vision, is from this view a specific and, ideally, more reflexive form of responsibilization, but it is one form of responsibilization among others. To foster the RRI form of responsibilization demands first of all to know how responsibilizations are emerging when visions serve as socio-epistemic practices.

With this view, we follow Grunwald’s focus on the present meaning of futures, e.g. visions, and the responsibility of all actors shaping, distributing and using future visions (including actors of TA and RRI which analyze, assess and even co-shape the visions; see, e.g. Schneider and Lösch 2015; Sand and Schneider 2017). Yet, we focus on the examination of ascriptions of responsibilities to and among the diverse societal actors which are addressed in visionary discourses and shape and enable the ongoing processes of sociotechnical change. In such arrangements and processes, responsibilities are produced in the practical use of visions by the participating actors. From such a perspective, visions are means and practices which enable and create responsibility.

Generally, it can be stated that visions are imaginations of the future in which already known technological trends and societal constellations are rearranged around a vague point in the future. But they also emerge from practices which produce the normative premises about the desirability of the provoked change and which culminate in expectations and promises which are forged by general imaginaries of the society, as the STS-literature has shown (e.g. Brown, Rappert, and Webster 2000; Adam and Groves 2007; Jasanoff and Kim 2015). Here, we do not focus on the interrelations between visions and expectations of innovators, or between visions and overall sociopolitical imaginaries; nor do we develop a conceptual definition of vision in contrast to other concepts such as imaginaries or expectations and promises.⁶ Rather, we want to know what visions ‘do’ in certain practices of change – namely, in the case of this paper, practices of responsibilization.

But how can we shed light on these practices? How can we analyze what the visions are doing? We claim that for analyzing and assessing such functions of visions in practice, we have to grasp them as socio-epistemic practices. To unfold our concept, we integrate and relate diverse analytical approaches already developed in, mainly sociological and empirical STS research into one concept. All the dimensions of such functions relate to each other or co-constitute each other in a functional manner, when visions serve as socio-epistemic practices. Through this, new knowledge and sociotechnical arrangements are produced. We distinguish four important dimensions of practical functions of visions (similar on this topic see, e.g. Lösch and Schneider 2016; Ferrari and Lösch 2017):

1. Visions can serve as interfaces which enable translations between present constellations and the future and through this open up imaginative and practical possibilities (see, e.g. Brown, Rappert, and Webster 2000; Adam and Groves 2007; Anderson 2010). In this dimension, they provide orientation and enable the identification of options for change.

2. Visions can serve as communication media between different actors and discourses to which all the involved or addressed actors can refer, even if they have very different interests and perspectives (Lösch 2006a; for similar descriptions of knowledge and boundary objects see, e.g. Knorr-Cetina 1997; Star and Griesemer 1999; Star 2010). In this dimension visions enable communication and interaction needed for change.

3. Visions can serve in different discursive or practical constellations as guiding visions, which enable coordination between different activities (see, e.g. Dierkes, Hoffman, and Marz 1996; Mambrey and Tepper 2000; Böhle and Bopp 2014). In this dimension, visions enable coherence between all the changing relations of actors and technologies in the corresponding sociotechnical arrangements.

4. Visions can activate behavior by serving as a normative force: the envisioned and proposed innovations are presented as desirable solutions for current and/or future problems or challenges (see, e.g. Nordmann 2010; Grunwald 2014; Jasanoff and Kim 2015). In this dimension, they also address the specific responsibilities of the participating actors or even external actors in society who should do this or that following the vision’s narrative.

For this paper, these practical functions of visions imply that we must make visible and explain how ascriptions and adoptions of responsibility are produced during the (1) translations, (2) communications, (3) coordinations and (4) normative activations that visions enable in practice. In this way, we can investigate what dynamic arrangement of distributed responsibilities are emerging through the work of visions as socio-epistemic practices. This distribution is enabled and shaped by the specific ways in which visions open up spaces of possibilities through the co-functioning of the four dimensions.

The following three cases analyze such processes of responsibilization enabled and produced by visions as socio-epistemic practices. With our selection of three different examples, we want to make plausible that our approach is valuable for empirical research across a variety of cases that are interesting for RRI. All the examples are typical cases for TA, which is mainly confronted with current processes of sociotechnical change, without knowing their future outcomes but where we can analyze how imagined futures produce particular arrangements of responsibilities in the present. The cases were chosen because they do not only differ in terms of the related technologies but also in the different innovation practices involved and the constellations of actors. The three cases also show how arrangements of responsibilities vary considerably and how this corresponds to the idea of visions as socio-epistemic practices. We identify and distinguish three modes of responsibilization: re-forming responsibilities (smart grid experiments), inventing responsibilities (organizing Fab Labs) and creating irresponsibility (Big Data discourses).

3. Cases of responsibilization through visions as socio-epistemic practices

3.1. Smart grid experiments: re-forming responsibilities

Engineers and IT-experts describe the future smart grid as

a vision of a future electricity grid, radically different to those currently deployed, where the bidirectional flow of both electricity and information allows demand to be actively managed in real time, such that electricity can be generated at scale from intermittent renewable sources. […] Unlike existing grids where electricity generally flows one-way from generators to consumers, [the smart grid] will result in flows of electricity that vary in magnitude and direction continuously. (Ramchurn et al. 2012, 86–89)⁷

Such a vision of a smart grid is very influential for the German Energy Transition and the transformations of energy systems worldwide because the vision serves as a solution for problems emerging in the course of the transition process and as a condition to enable desired changes (e.g. BMU and BMWi 2011, 19). In the case of the German Energy Transition, the vision of the smart grid is positioned as a future solution for the upcoming challenges to control and regulate an increasingly decentralized and dispersed energy system.

The vision addresses the increase in varieties of energy production and consumption, the volatility of renewable energies and the need for uninterrupted supply of energy. Different to the existing centrally controlled one-way system, which transports energy from particular sites of production to particular sites of consumption, the smart grid shall enable two-way flows of energy and information with sites of production and consumption changing in a decentralized system. A massive integration of digital technologies shall enable this. To test this vision, a variety of field experiments have been initiated and conducted in Germany and the EU with different experimental designs (Nyborg and Røpke 2013; Covrig et al. 2014; Goulden et al. 2014; Engels and Münch 2015). Beyond the most general idea of making the electricity grid ‘smart’ and increasing its automation, many different smart grid experiments have explored how this vision can be practically worked out. The overall smart grid vision links and integrates these different smart grid experiments.

One important and common insight gained through the German experiments is that the implementation of the principles depicted in the smart grid vision in the energy system would imply much more than new technical designs (e.g. BMWI 2014). The realization of the vision would imply far-reaching and multidimensional transformations not only of technologies but also of all of the actors, the knowledge flows, modes of governance and much more. The central insight is thus that the whole sociotechnical arrangement of the electricity sector has to change. Envisioned changes include consumers turning to prosumers and inventions of new forms of electricity markets, novel business models and new sustainable everyday practices. Each of these changes would involve massive changes in everyday routines. In short: the experiments point to massive demands for changing the existing arrangements of actors (users, suppliers, regulators, etc.) in such a way that they fit into the not-yet implemented but envisioned smart grid of the future.

It is important to mention that, with the changes of positions and relations of the diverse actors in the transforming sociotechnical arrangements (tested in the experiments), new distributions of responsibilities of the involved actors are emerging, tested and identified. For these processes of responsibilization, the practical functions of visions as socio-epistemic practices to produce new knowledge and to enable the testing of new sociotechnical arrangements are constitutive. This becomes clear if we look at expert evaluations of smart grid experiments. Experts from different sites of the energy system evaluate these experiments and do so in light of their interpretation of the vision of a future smart grid. Experts deploy the practical functions of the vision in order to develop their evaluations and their statements suggest that the visions are indispensable for the experimental practices and the resultant learning. Through the experts’ use of the vision, the experiments created new responsibilities for actors in the electricity sector. The effects learned in the experiments, e.g. the identification of further experimental demands, were and are always accompanied by recognitions of newly established or demanded responsibilities and ascriptions of new responsibilities by the experts to actors which in their view have not adopted the new responsibilities. Such ascriptions and adoptions of responsibilities are articulated by experts who helped conduct or evaluate the experiments.

The use of the vision as an interface between the future smart grid and the present situation, with which the experiments were confronted, enables for example, an R&D manager of one of Germany’s four ‘big’ energy suppliers to point at the new responsibilities of the energy suppliers to communicate with the customers. He mentions that his

first surprise was how difficult it was to make customers participate. We had to explain endlessly what we want to do to inform customers why we wanted them in the field test. For us the system and the related changes were relatively clear, it‘s obviously not clear to the customer out there. (Energy Supply Company 2013)

We can say he and the other experimenters used the vision as an enabling medium of communication between themselves and the other actors in the experiments, mainly the customers. The vision as an interface and medium of communications furthermore produces insights in demands for changing the behavior by the customers (especially the business customers) in order to get ‘actively engaged’ and to show a new kind of ‘flexibility’ (Energy Supply Company 2013). That means the customers are ascribed the responsibility to learn to change their behavior and routines in orientation to the vision of the smart grid and, in their turn, the old suppliers are responsible for enabling this learning process by initiating experiments. Through this, the suppliers themselves have to be flexible and willing to experiment. Here we see how the visions unfold their activating and normative force, the imperative of behavioral change, which does not only address the behavior of the customers, but at the same time the one of the suppliers.

This normative imperative is also demonstrated by another expert. For him, the main insight from the experiments was that regulatory bodies and politicians ‘have to change the legal framework in such a way that the grid operator can do research and development […], but also take risks’ since it became clear to him that ‘their perspective is not just the reduction of expenses and not just the avoidance of risk’ (Industry Association 2013). In this statement, not only is the new responsibility of the grid operators articulated; a new regulatory responsibility also becomes visible: the imperative to enable experiments with smart grids and not mainly to regulate in order to minimize risks corresponding to the problematized old behavior of actors such as grid operators or power suppliers. Here we see that the vision serves as a guiding vision in the sense that it leads to the identification of new modes of coordination (governance).

The experts’ emphasis on a new landscape of emerging and necessary responsibilities was enabled by the use of the smart grid vision in all of its functions as socio-epistemic practices together. As one consumer protection expert put it, ‘There is a completely different basic structure in the system and a regional responsibility and I think that is not yet properly communicated’ (Consumer Protection Association 2013). Similarly, an environmental association expert states, ‘Now we’re arriving at a certain stage where we want to turn the whole system upside down. There is an infinite number of actors that have to be included into the system’ (Environmental Association 2013). This new distribution of responsibilities is identifiable by observing the requirements for new social arrangements that the experimental practices of the smart grid vision probe and produce.

Statements such as these by experts from different positions in the existing energy system arrangements show that visions, by modulating the experiments through socio-epistemic practices, enabled and produced processes of responsibilization. While responsibilization was a central learning effect of the experiments, it also becomes a condition for experimentation itself. The visionary experiments point to specific responsibilizations as a basis for conducting further experiments and for translating the experimental results into the governance of innovation processes. The experiments produce the actors’ responsibilities for future smart grids in present practices of their probing and learning from options for sociotechnical rearrangements.

Responsibilization emerges out of the interaction of the practical functions of visions as socio-epistemic practices. But what is especially evident in this case is the re-formative mode of responsibilization, in which we see a reform of responsibilities in the corresponding sociotechnical arrangement. Following the experiments, none of the actors in the energy system will be excluded or will lose their responsibility. Some actors (suppliers, grid operators, regulators, etc.) remain responsible but have to extend and change their responsibility for new tasks and technologies. Other, more passive actors such as customers (both business customers and private households in their role as consumers) must take on new additional and active tasks (e.g. as prosumers). This re-arrangement in the landscape of responsibilities corresponds to the envisioned reform of the electricity sector under the conditions of the future smart grid.

3.2. Fab labs: inventing responsibilities

Next, we analyze FabLabs (‘fabrication laboratories’), which have become a global phenomenon of around 700 organizations that aim to provide simple and sometimes public access to digital fabrication, including 3D-printing⁸. They have been important in experiments with novel technologies in novel social contexts that are sometimes seen as a ‘democratization’ of technology and innovation (Walter-Herrmann and Büching 2013; Smith et al. 2016). In this analysis we show how the vision of FabLabs was transformed through socio-epistemic practices in material settings that linked FabLabs to other visions and in the process changed the responsibilizations. In particular, through such socio-epistemic practices of combining imaginations of the future and practices, inventive responsibilizations have taken place and over time created actors that unfolded FabLabs.

The concept of FabLabs emerged at the Massachusetts Institute of Technology (MIT) at the Center for Bits and Atoms where researchers around Neil Gershenfeld investigated 3D-printing and other computationally controlled processes to manipulate matter. This research was particularly influenced by visions of Nanotechnology that framed matter as formable based on single pieces, ‘atoms’, which corresponded to the single pieces of digital information, ‘bits’, that would provide the form. Furthermore, drawing parallels to the personal computer, the research was partly seen as contributing to the vision of ‘personal fabrication’ that would allow ‘anyone to make anything, anywhere’ (Gershenfeld 2012, 57) through machines that digitally transform matter. Through a funding program in 2001 that encouraged the ‘outreach’ of science to society, the institute set up a handful of initial FabLabs with the aim to ‘deploy proto-personal fabricators in order to learn now about how they’ll be used instead of waiting for all of the research to be completed’ (Gershenfeld 2005, 11). These FabLabs were small workshops hosted by MIT or partner institutions that made relatively expensive, yet small-scale digital fabrication machines accessible to the public. They manifested a vision of an increasing capability of these machines to produce things and to become usable by anyone in the future.

In this first phase of FabLabs a vision as interface between present and future was created and used that entails a particular configuration of responsibilities. The interface is established between existing digital technologies, personal computers and contemporary versions of digital fabrication machines that would both become more powerful in the future as their trajectories intersected. In the future, ‘anyone’ is therefore seen as responsible for personal fabrication in which people make ‘anything’, whereas the responsibility to reach this future is ascribed to research. This is, therefore, also enacted in the sociotechnical arrangements whereby MIT deployed FabLabs in public settings as prototypical arrangements of ‘personal fabrication’. This was then still a rather typical constellation of responsibilities with a strong separation between research and society, between developers and users of technologies.

This rather constrained form of responsibility was to change a few years later. More FabLabs were set up with formal links to MIT and often funded through states and large organizations. In 2007, however, in the Netherlands an initiative of media and design organizations agreed with MIT to independently open their FabLabs without paying a fee or becoming a formal partner of MIT. And a few years later, also in the Netherlands, in 2010, the first ‘grassroots’ FabLab was set up by a group of citizens, including artists and activists with an initial budget of around €5000 (Troxler 2014). By now hundreds of grassroots FabLabs, often run by volunteers, are providing access to shared digital fabrication machines. What has happened to enable this diversification of the organizational foundations of FabLabs, which is also a diversification of responsibilizations? In the decade that passed, two changes in the environment of FabLabs have taken place that entangled their history and vision.

First, in 2004, similar to the guiding vision of personal fabrication, the open source project RepRap started to develop open source 3D-printers and has since grown into an important hub for small-scale and inexpensive 3D-printing (Söderberg 2014). Initially, RepRap envisioned a radical transformation of the economy through its goal to make machines that make machines that would fundamentally decentralize the production of material goods. By now, most RepRap designs that circulate publicly are machines that produce small plastic objects. However, through the open source approach of the project, thousands of enthusiasts have positioned themselves as responsible for developing, using and experimenting with this form of 3D-printing and have thus opened a path of low-cost 3D-printing beyond industry (Tech, Ferdinand, and Dopfer 2016).

Typically, these open source 3D-printing enthusiasts, hobbyists and professionals alike also position themselves in relation to a vision of the emergence of a decentralized economy that they anticipate and experiment with in their practices (Alvial Palavicino 2016) and that they consider themselves responsible for. Second, and relatedly, the first decade of the twenty-first century has also seen more diffuse discourses on the ‘web 2.0’ and related visions of ‘prosumers’ and online collaborators that would lead to the ‘democratization of innovation’ (von Hippel 2005). These discourses do not only draw upon older forms of Internet utopianism (see, e.g. Dickel and Schrape 2017a; Turner 2006) but also address ‘everyone’ as a possible responsible agent in such processes. In relation to the production of material objects this diffuse discourse brought the figure of ‘the maker’ into being that is seen as a prosumer who produces their objects empowered through Internet collaboration and typically uses 3D-printing (Dickel and Schrape 2017b). Here, we find suggestions that responsibilization made ‘makers’ responsible for the further expansion of FabLabs.

The grassroots FabLab in 2010 was also based upon a normative vision of a new society through the Internet. In the years following, many more grassroots FabLabs emerged and conversely influenced the other visions, such as those just mentioned, as well. At present, the FabLab landscape is based upon a huge diversity of organizations – including universities, businesses and nonprofits – that operate FabLabs. Many more open source projects have emerged and governments and businesses are increasingly launching strategies to ‘open’ innovation processes (Tkacz 2015). This diversity is also reflected culturally in many different goals that can be found within the FabLab landscape, entailing ideas of individual empowerment, novel forms of education or even a rejuvenation of industry (Schneider 2017). The vision that is part of the socio-epistemic practices of FabLabs thus has a normative force based upon desires to decentralize, democratize and to ‘open’ innovation processes. A new, more inclusive and collaborative way to invent and produce material objects is envisioned and explored – still with many constraints – in FabLabs.

Within this environment, the visionary concept of FabLabs is also a communication medium between different actors that offers a concrete example of how such novel organizations that make new and emerging technologies increasingly public and inclusive could look. Precisely through such mediated communications, the vision of FabLabs addresses and integrates different actors and creates an arrangement based on responsibilizations. FabLabs potentially address ‘everyone’ as responsible in their further unfolding and FabLabs as organizations are responsible for reaching out to everyone. This ‘everyone’ includes individual enthusiasts who pursue their own projects, education institutions and businesses that try out novel business models. Since their beginning, however, the vision of FabLabs has enabled concrete FabLabs as experimental spaces to explore such visions and in turn transform themselves and the visions they were associated with (Troxler 2015; Schneider 2017). Through visionary discourse, organizational development and technical practices, this process of self-transformation also creatively unfolded novel responsibilities in relation to digital fabrication technologies. FabLabs thus appear to be a case of the invention of a constellation of responsibilities that emerged through a creative process engendered through visions as socio-epistemic practices.

3.3. Big data: creating irresponsibility

Like every technological vision, the Big Data vision (Boyd and Crawford 2012; Schrape 2016) addresses not only technological changes, but also the necessity of social changes that in turn imply new responsibilities or a redistribution of responsibilities for the actors (Morozov 2013 Crawford, Miltner, and Gray 2014;).⁹ Insights into responsibilization processes related to Big Data are gained from the analysis of the use of the Big Data vision in discursive practices and its effects on potential strategies of action. Big Data is a vision of the future on the one hand, but on the other hand, it has already become a social reality. As a socio-epistemic practice, the vision performs as a fundamental means of orientation in discourses, thus constituting its own subject as it contains technological developments and challenges under a common term. In doing so, Big Data at least to a certain degree takes on the legacy of digitalization. The technical part of the Big Data vision is comparatively trivial: digitalization has progressed to a point where the resulting data are far too extensive, too complex and too heterogeneous for conventional methods of storage and analysis. The matter is in many cases further complicated by the requirement that the data should be processed in real time. This is the ‘non-visionary’ part of the Big Data vision rooted in the present, as it merely describes existing technical challenges resulting from current trends. But the Big Data vision also assigns those current trends to the future as a larger (social, economic and political) horizon of meaning.

As it classifies and interprets current trends in light of this horizon of meaning, the Big Data vision, seen as a socio-epistemic practice in the respective discourses, also develops the normative force of an imperative: according to the vision, data is not simply there and somehow needs to be processed; rather, its existence is considered to be something good and desirable. From this, an ethical obligation to produce more and more data is inferred for a variety of social actors. Through this normative postulate, the vision develops a normative activation function in discourses that does not merely encourage certain patterns of action, it almost ‘commands’ them. In the process, the Big Data vision also works as an interface that enables translations between present trends and challenges and the future. At the same time, it serves as a communication medium between the participating actors and develops specific principles to coordinate actions as a guiding vision, the implementation of which it declares as a normative imperative of responsible action.

It is important to stress the point that the vision is not limited to improvements in data processing, but rather provides a legitimation for generating, collecting and processing additional data on one hand as well as the integration of existing data collections (Prodhan and Nienaber 2015) and their transfer to different or entirely new contexts of use on the other hand.¹⁰ From the unchallenged and unconfirmed epistemological postulate, ‘More data and additional connections produce more knowledge’, the vision infers a straightforward ethical obligation and moral responsibility: we therefore have to produce more data, improve the integration of existing data and use it in the most versatile way possible. Those who refuse to do so act unethically as they do not assume the responsibility that has been attributed to them. But how does this attribution of responsibility work?

Not only does the vision structure the field of argumentation by providing concepts, it also defines what counts as legitimate action. It positions itself as essentially positive on the one hand and as inevitable on the other. The vision does not deny negative consequences, but it presents them as mere side effects (that should be avoided as best as possible) of an essentially good and inevitable development. In short, the Big Data vision fuses (1) factual authority (those trends are existing in the present) with (2) a technodeterminist core assumption (those trends will continue whether we like it or not) and (3) a normative postulate (those trends are essentially good) into a powerful narrative.

Larry Page, one of the founders of Google, provides an example of this when he says: ‘Right now we don’t data-mine health care data. If we did we’d probably save 100,000 lives next year’ (Manjoo 2014). Page does not simply frame the evaluation of personal data by a private firm as a good thing; he implies that any behavior to the contrary is unethical, as it is inconsistent with the responsibility that the vision attributes to a potentially all-encompassing ‘we’. Those who object to the vision are, by implication, potentially responsibility for the deaths of 100,000 people per year. While the structure of argument is not novel, its embedment in the normative frame of the Big Data vision helps it gain enormous momentum. Responsibility for Big Data is effectively delegated to society in the form of an imaginary ‘we’. Therefore, only those who make their data available for the sake of a presumed enhancement of the common good and are willing to generate additional data act responsibly. The use of fitness trackers constitutes a current example for such possibilities of action in everyday life. This way the vision is confirmed by everyday practices that are already established in the present and would have very little to do with Big Data were it not for the vision. Technodeterminism and diffuse attributions of responsibility for a postulated common good are thus functional complements of each other.

Attributions of responsibility as enabled by the Big Data vision lead to irresponsibility of the actors who actually participate in the collection, evaluation and commercialization of data. For instance, Big Data changes the perspective on the responsibility of businesses. Some companies already describe or justify the collection and evaluation of data they engage in as not merely a business activity, but as a service to the common good of society (Morozov 2013). The entrepreneur is presented or presents himself as making responsible use of his abilities and the revenue of his business in order to enhance society in accordance with the inevitable requirements and possibilities of Big Data. In short, the entrepreneur as part of the collective ‘we’ fulfills his collective responsibility as generated by the vision.

Likewise, the Big Data vision treats non-human entities in its sociotechnical arrangements, especially algorithms, as if they were responsible actors. This shifting of responsibility from human actors to artifacts is justified by the alleged objectivity of data and algorithms, on one hand, and the complexity of the digital world, on the other—a complexity that has become impossible for humans to control and that thus requires a delegation of decision-making to algorithms.

The diffusion of responsibility within society as well as the attribution of responsibility to things (like algorithms) developed by the specific narrative of the Big Data vision as a socio-epistemic practice, rules out traditional (individualizing) attributions of responsibility. Meanwhile, the fact that neither data nor algorithms are ‘objective’ (natural facts) is neglected. The decision as to which data are or are not collected and in what way they are processed is already based on the preferences and interests of specific actors and thus inevitably flawed by values. At this point, many questions concerning possible attributions of responsibility remain open. The attribution of responsibility to a diffuse universality and the assertion of technodeterminist inevitability generated by the vision turn out to be an irresponsibilization rather than a responsibilization of groups of actors who are practically involved in the development and exploitation of technologies and processes accounted to Big Data by means of its vision.

In line with both other cases (Smart Grid and Fab Lab), the Big Data vision can be conceptualized as a socio-epistemic discursive practice. Furthermore, while the other cases re-arrange existing responsibilities and invent new distributions of responsibilities, respectively, the Big Data case introduces irresponsibilization through its normatively charged diffusion of responsibility.

4. Conclusion and outlook

Visions as socio-epistemic practices help to trace and analyze ongoing processes of responsibilization that create constellations of responsibilities in innovation and transformation processes. These are necessary starting points for analyses and interventions that aim to foster ‘responsible innovation’ or ‘responsible research and innovation’. We have shown how there are always dynamic processes engendered through visions that address particular actors as responsible or not responsible for particular positions or actions. That is not to say that these processes always produce ethically ‘responsible’ actors such as those that are envisioned in programs of RRI. Yet, the world of innovation and transformation processes is always already filled with particular ways and arrangements of making and contesting responsibilities within which any form of research or intervention takes place. Our concept and focus provides a promising perspective to better understand the complexities and performativities of responsibilizations through which particular actors are made responsible or not.

The three exemplary cases showed how such responsibilizations can have different effects. First, in relation to smart grid visions and experiments, we observed and analyzed socio-epistemic practices to reform arrangements of responsibilities with the aim to adapt the existing sociotechnical arrangement, including its dominant actors, to an emerging and still envisioned smart grid. Second, the analysis of FabLabs showed how a vision of personal fabrication over time combined with other visionary developments and increasingly turned the responsibility for the future of personal fabrication away from science and research. Instead, FabLabs have started to address everyone as responsible and provide the experimental settings to practically test such decentralized constellations of responsibilities. In this sense, we see visions of Fab Labs as socio-epistemic practices which invent arrangements of responsibilities. Third, in the case of the vision of Big Data, we showed how this vision, through constructing a technodeterministic narrative linked to a presumed common good, renders responsibilities invisible and brings about irresponsibilization through which Big Data is presented as irreversible and irresistible.

Our analysis of these three cases provides fine-grained if limited insights into how visions as socio-epistemic practices produce constellations of responsibilities. Due to practical, methodological and empirical limitations, we provide only situated rather than generalized views of responsibilizations in these empirical cases. That said, the extensiveness of the constellations of responsibilities in innovation and transformation processes that we reveal here transcend individual research projects and furthermore, as we have shown, are constantly being worked on by heterogeneous actors and discourses. For RRI purposes and practices, understanding the particular workings of responsibilizations is itself an important result, not least because it may help inform where and how to intervene in the attempt to foster ethically ‘good’ and ‘responsible’ innovation processes among diverse actors.

Furthermore, these ongoing changes engendered by visions as socio-epistemic practices become assessable for TA and RRI by looking at the new sociotechnical arrangements in the present – even if they are geared towards an imagined and not-yet materialized future full of uncertainties and unknowns, as is typical in projects of TA and RRI. Although we refrained from a normative assessment of the three cases, we want to suggest that the processes of responsibilization which we have analyzed also significantly differ in their openness to RRI practices. Since RRI aims at including many different stakeholders and values in innovation processes, such inclusions might be easier where there is already an ‘opening’ of responsibilities – such as in the FabLab case. Where powerful actors are still responsibilized as central in a changing landscape as in the smart grid case, or where responsibilities are made invisible and inevitable as in Big Data case; however, stronger interventions might be necessary to foster responsibilizations as envisioned by RRI.

Moreover, analyzing responsibilizations helps produce knowledge to orientate decisions and actions in the present. This kind of a real-time vision assessment is essential for research on and practices of responsible innovation. For instance, if one wants to assess whether an emerging re-arrangement of responsibilities fits with what is defined ethically as a good innovation process, one has to know how responsibilities are distributed and how they will or should get distributed by the envisioned changes. Or if one wants to initiate or create new settings of arrangements and processes, which should result in better innovations, one has to know how one’s own ideals and visions of responsible innovation may change in constellations of responsibilities as well as which new or modified constellations they push.

Such practical applications of our proposed view of visions require self-critical reflection on such tasks both in the vision assessments of TA and the research and practices of responsible innovation. When TA assesses what opportunities and risks are emerging in a certain constellation of change and this knowledge is used for policy advice and other forms of intervention, TA is in the position of a responsible actor in the changing arrangements. Through its work, TA attributes responsibilities to other actors, as inevitably our empirical analyses have done in the three cases above. Furthermore, in innovation processes, TA is increasingly addressed as being responsible by the other included actors. And if such a TA expertise is used in responsible innovation practices, these practices also attribute and distribute responsibilities. Both TA and RRI have their visions of ‘good’ and ‘responsible’ designs and processes of innovations. And these visions also serve as socio-epistemic practices in the constellations where they are used for assessments or practices of innovation.

To integrate corresponding tasks of a self-reflexive vision assessment in concepts and practices of responsible innovation or responsible research and innovation is still an ongoing challenge, which calls for new designs in research and practice. TA and RRI in a way have to responsibilize themselves through their visions of better innovation processes in the future and find corresponding practices.

Acknowledgements

We would like to thank the guest editors for their work and the two anonymous reviewers for their feedback on this text.

Notes on contributors

Andreas Lösch, PD Dr phil., is a sociologist with focus on STS, and is senior researcher at the Institute for Technology Assessment and Systems Analysis (ITAS/KIT). Currenlty, he is leading the project ‘Visions as socio-epistemic practices – Theoretical foundation and practical application of Vision Assessment in Technology Assessment’. His recent publication with C. Schneider is: Transforming power/knowledge apparatuses: the smart grid in the German energy transition. Innovation: The European Journal of Social Science Research (2016), pp. 262–284.

Reinhard Heil, M.A., philosopher and researcher at the Institute for Technology Assessment and Systems Analysis (ITAS) at the Karlsruhe Institute of Technology (KIT). His working areas are Big Data, Life Sciences (Synthetic Biology, Epigenetics) and Transhumanism. His recent publication is: Frank, D.; Heil, R.; Coenen, Chr.; König, H. Synthetic biology’s self-fulfilling prophecy – dangers of confinement from within and outside. Biotechnology Journal 10(2015)2, pp. 231–235.

Christoph Schneider, M.A., is sociologist at the Friedrich Schiedel Chair of Sociology of Science, Technical University of Munich. His work revolves around the political economy of technoscience and related contested forms of future-making that culminated in his research on open digital fabrication for his PhD. Recent publication: Sand, M., Schneider, C. Visioneering Socio-Technical Innovations – a Missing Piece of the Puzzle. Nanoethics 11(2017), 19–29.

ORCID

Reinhard Heil http://orcid.org/0000-0001-7265-7376

Christoph Schneider http://orcid.org/0000-0003-1516-158X

Notes

1. For the current research in the field of vision assessment see the ITAS-project ‘Visions as socio-epistemic practices’: https://www.itas.kit.edu/english/projects_loes14_luv.php

2. To prevent conflations we highlight this ethical denomination of ‘responsible’ in quotation marks throughout the text. Whenever we speak of responsible as in the sense of made responsible through responsibilization we do not use quotation marks.

3. With the term ‘arrangement’ we denominate relational configurations of heterogeneous elements in a certain field, which are changing in the course of innovation processes. They are sociotechnical arrangements because they include actors from diverse areas of society, such as politics, science, industry and civil society and also technical artifacts, processes, building structures and more.

4. Especially Michel Foucault and the governmentality studies following his approach have emphasised how individuals are made responsible for particular aspects of their lives through discourses (Biebricher 2011; Lemke 2002). We are inspired by such Foucaultian perspectives and add further concepts to make an analysis of processes of responsibilization in innovation processes possible.

5. From a similar perspective on risk-discourse it is analyzed how the communication of non-knowledge as risks (through risk as a media of communication) produces new distributions of responsibilities in the case of nanotechnology see Lösch (2014). For the case of nanotechnology and the governance of such future technologies Mario Kaiser analyzed assessments (such as TA and Ethics) on the future impacts of future technologies as ‘assessment regimes’, which also invent responsibilities for the present society through their assessments of potential futures (e.g., Kaiser et al. 2010).

6. In our view, it does not make sense for this kind of approach to offer an a priori definition of vision since their essence as socio-epistemic practices emerges from specific practical impacts of visions, which we can observe only empirically.

7. The following insights in this chapter draw upon Lösch and Schneider (2016) where a more detailed analysis of an empirical study of smart grid visions in practice based on a document analysis and qualitative expert interviews can be found. The empirical work was conducted within the research project ‘Systemic risks in energy infrastructures’, one of the projects of the Helmholtz alliance ‘Energy-Trans’ (http://www.energy-trans.de/english/68.php; accessed 19 July 2016).

8. This analysis is based upon an in-depth study of the emergence of FabLabs and the foundation of a grassroots FabLab that draws upon mixed qualitative methods such as participant observation, document analysis, action research and interviews (see Schneider 2017; Schneider forthcoming).

9. The basis for this chapter is a literature research in 2016 and the results of an expert workshop with participants from science, industry, politics, non-governmental organizations and data protection within the framework of the BMBF project ‘Assessing Big Data’: https://www.itas.kit.edu/english/projects_grun15_abida.php

10. An example is the use of social network data for credit scoring (see, e.g., Wei et al. 2016).