I’m co-convening a panel on “Data Worlds? Public Imagination and Public Experimentation with Data Infrastructures” at EASST2018 – the latest edition of the conference of the European Association for the Study of Science and Technology – which takes place on 25-28th July 2018 at Lancaster University, UK.
The panel is organised by several colleagues at the Public Data Lab, including Noortje Marres (University of Warwick), Carolin Gerlitz (University of Siegen), Tommaso Venturini (École Normale Supérieure Lyon) and myself.
We had a very interesting set of papers submitted in response to the CFP that went out in December, and further details about the panel (including talks and abstracts) are available here and copied below.
Data worlds? Public imagination and public experimentation with data infrastructures
- Location: Frankland Lecture Theatre (Faraday Complex)
- Date and Start Time: 27 Jul, 2018 at 09:00
- Sessions: 4
- Convenors: Jonathan Gray (King’s College London); Tommaso Venturini (École Normale Supérieure Lyon); Noortje Marres (University of Warwick); Carolin Gerlitz (University of Amsterdam)
How do data infrastructures distribute participation across society and culture? Do they participate in world-making, and if so how? Could they be utilised not just to close discussions, but also to open up public debate, imagination and experimentation?
This panel is about the world-making capacities of emerging data infrastructures – including their epistemic, social and political possibilities and limitations. It examines how emerging data infrastructures may distribute and redistribute participation in knowledge and world-making across society and culture – from the platform data of big technology companies to open data from public institutions, citizen data, sensor data and associated forms of journalism and activism. How are digital technologies entangled with social practices of classifying, counting, reasoning, narrating and making decisions? Rather than just extending the reach of certain pre-authorised ways of knowing, seeing and dealing with things, how might digital technologies support more substantive forms of interactivity and participation in order to open up public conversations, imagination and experimentation about how data is made and put to work?
One of the key strengths of research methods and design approaches developed across STS, participatory design and digital social research has always been their experimentality – the ways in which they seek to combine knowing and doing – representing and intervening in social life – in potentially new, creative ways. This session asks: What distinctive forms of engagement with data infrastructures do these methods and approaches enable, and what is their capacity to contribute to data world-making? This panel organised by the Public Data Lab (https://publicdatalab.org) will explore the capacity of STS, design and digital methods to take on the challenges outlined above, with the aim of identifying priorities for exploring and intervening around emerging data infrastructures today.
A citizen data app as an emergent para-site for official statistics: imagining citizens as more than data collectors and subjects
Authors: Francisca Grommé (Goldsmiths, University of London); Evelyn Ruppert (Goldsmiths, University of London)
We discuss how a para-site emerged around the experimental design of a citizen data app for official statistics, potentially enabling citizens to participate as co-producers of statistics. We furthermore highlight three aspects of this para-site relevant for producing ethnographic imaginations. ci Long abstract:
The idea to experiment with a ‘citizen data app’ emerged in an ethnographic project in which a team of six researchers followed the working practices of official statisticians across Europe. As national statistical institutes are taking up the challenge of adopting Big Data sources and techniques in their repertoires, we proposed that involving citizens as active co-producers of official statistics can help address issues of access, ownership, privacy, representativeness and quality that come with ‘smart statistics’ infrastructures based on Big Data. In an ongoing conversation between actors from official statistics, academia and technology design we developed the idea that a citizen data app could address existing gaps between citizens’ actions that create data, the interpretation of that data for statistics, and citizens’ identifications with the resulting statistics.
We show that the situated design practices (individual conversations, joint meetings and a collaborative workshop) that constitute this para-site (Marcus 2014) have enabled varying and sometimes contingent understandings and re-imaginations of statistical practice among its participants. Among others, notions of ‘citizen co-production’ emerged ranging from including citizens in existing infrastructures of statistical production to parallel infrastructures that challenge core categories in official statistics. We discuss three aspects relevant to this para-site as a method that enables ‘ethnographic imaginations’: 1) the provocative potential of workshop materials; 2) the value of engaging with the details and struggles of designing a concept; 3) how the dynamics of a workshop can over-rule a research team.
Acts of digital parasitism: data, humanitarian apps and platform economies
Authors: Claudia Aradau (King’s College London); Tobias Blanke (Kings College London)
This paper explores the performative effects of digital humanitarian technologies developed for refugees in the wake of the so-called ‘refugee crisis’ in Europe since 2015. Drawing on the work of Michel Serres, we develop a method of ‘acts of parasitism’ as distinctive from ‘reverse engineering’.
We have seen a proliferation of digital technologies and data processing in the work of humanitarian actors. NGOs operate globally and rely on data infrastructures to connect, gather information and reach out to their target communities. ‘Digital humanitarians’ have tended to concentrate on the success or failures of their digital projects. Have digital infrastructures really helped the communication in organisations? Have migrants taken up the digital offerings of NGOs? In this paper, we propose to focus on the performative effects of digital humanitarian technologies developed for refugees in the wake of the so-called ‘refugee crisis’ in Europe since 2015. To this purpose, we develop a series of methodological experiments in interdisciplinary collaboration to research digital technologies. We show how ‘hacking the blackbox’ of digital technology has led to us to approach digital technologies as ‘parasitic’, drawing on the work of Michel Serres. We thus develop earlier suggestions to conceptualise the digital as parasitic in a methodological direction through ‘acts of parasitism’. We start with a discussion of the key methods used to open blackboxes and then show how we deploy ‘hacking’ as a collaborative method distinctive from reverse engineering. In a second stage, we discuss the insights that this method has yielded and their limitations. Thirdly, we argue that the understanding of refugee apps as ‘parasitic technologies’ allow us to develop a critical method as ‘acts of parasitism’.
Agencies in the database: the role of flagging
Author: Sarah Inman (University of Washington)
Database designers and archivists are central to improving natural resource management, deploying instruments and standards to achieve data interoperation. This paper explores practices and politics of making data open as a site for unpacking who data are for and what purposes they serve.
All data and metadata hold within them records of implicit debates, forgotten decisions, missing measurements, and casual compromises. Data are neither ‘raw’ nor ‘objective’ but always formatted in a way that makes them speak – in particular ways and to particular audiences. To explore the curation of data amongst a team of salmon and data scientists, we extracted data from their GitHub repository, a common archive used for programming work, and then analyzed 185 issues-a type of entry in GitHub-to identify the major topics around which negotiations took place. Through the work of cleaning, synthesizing, and archiving data, new possibilities are created as certain types of data are made visible while others slip into the background. In presenting our findings back to the scientists, we take their perceived mundane everyday work and reveal it back to them to highlight opportunities for broadening participation and building opportunities for wider uses of data downstream. In principle, one goal of these salmon scientists is to make their data open to the public -through our exercise we sought to surface competing conceptions of ‘the public’ and the scientists’ imagined uses of these data by that public. With an eye to what is absent from these discussions, this research sheds light on the daily environmental data practices and highlights the ways that scientific data are made to speak.
Death, taxes and gases: the issuefication of data and the datafication of issues
Author: Jonathan Gray (King’s College London)
How are digital technologies redistributing practices of making data public and making public data? This paper reflects on several cases where transnational data infrastructures become “matters of concern” around deaths, taxes and gases, leading to interventions and alternatives.
How are digital technologies redistributing practices of making data public and making public data? How are different actors challenging, contesting and creating alternatives to official data infrastructures and regimes of datafication? How do issue activists, civic hackers and others mobilise in order to change how issues are accounted for through data? This paper looks at what can be learned from several cases where transnational data infrastructures become “matters of concern”, leading to interventions and alternatives. Whilst the UN talks of “data gaps”, this framing over-emphasises the representational capacities of information systems, and does not do justice to the many different ways in which data can be created in relation to different issues. In order to account for the performative capacities of data infrastructures in rendering and shaping collective life, the paper examines what can be learned from ongoing mobilisations around public data – including around deaths (e.g. police killings and deaths in migration); taxes (e.g. the economic activities and tax contributions of multinationals); and gases (e.g. carbon emissions and air pollution). It looks at the methods, devices, technologies and practices through which alternative data worlds are created, maintained and seek public recognition and legitimation. It considers how controversies around data infrastructures may inform more ambitious forms of public involvement, intervention and imagination around processes of datafication, as well as suggesting possible unintended consequences of rendering life as data.
Emerging data infrastructures and ethnographic world-making
Authors: Maja Hojer Bruun (Aalborg University); Astrid Oberborbeck Andersen (Aalborg University)
What kinds of social worlds are imagined by different parties developing data technologies such as blockchain and secure multiparty computation? The paper presents a collaboration between mathematicians, computer scientists, control engineers and anthropologists engaging interventionist ethnography.
Data-controlled automated systems that integrate cyber-, physical and social worlds offer great potentials for the improvement of communication and resource use, e.g. the transition to renewable energy. So-called ‘smart’ projects, such as smart cities, smart homes and the smart grid, establish communication and coordination between large numbers of entities. Such interaction between entities requires exchange of privacy-sensitive information and raises critical questions about users’ trust and participation in data systems. New data sharing technologies, such as blockchain and secure multiparty computation, have the potential to disrupt existing systems of interaction and exchange. What kinds social worlds are imagined by the different parties involved in the development of these emerging technologies that may become the cornerstones of future connected societies? What ideas about confidentiality, trust and privacy are produced and circulate among developers and potential users of emerging data systems?
This paper presents an ongoing interdisciplinary research collaboration between mathematicians, computer scientists, automation and control engineers and anthropologists who engage ethnographic studies of developers and potential users of emerging cyber-physical systems. As part of the interventionist ethnographic research, public events are staged and methods developed to study and engage participants in the creation of technologies that are not yet part of our everyday lives. Crucially, the paper asks what roles are relegated to humans. Will we see models of co-creation, transparency and self-organization or centralised, hierarchical paradigms? And how can we as social scientists contribute to fostering public debate and participation in decision-making concerning future technologies and data infrastructures?
Fintech apps and data-driven irrationalities: speculation in the face of precarity
Author: Rolien Hoyng (Chinese University of Hong Kong)
This paper analyzes financial-technological apps and the (ir)rationalities of data-centric speculation. It draws on interviews with stakeholders and app analysis. It further probes tactics of data activism to conjure publics and tell stories that enable relations of solidarity.
Fintech apps (financial technology applications) contribute to the rise of the speculator-citizen, for whom taking risk on the basis of speculative calculations, or being governed in terms of risk calculations, becomes part of everyday life and, ironically, the search for stability under conditions of precarity. Even though fintech apps advance datafication and data enclosure, they also promise financial participation by redistributing data-centric perception and cognition. Accordingly, in Hong Kong—a colony turned neoliberal testbed—fintech apps do not target the rich but rather millennials (i.e. students), elderly housewives, and the unbanked (i.e. migrant domestic workers). This paper analyzes the (ir)rationalities of data-centric speculation, taking into account the ungovernability of financial-technological systems (Lotti forthcoming; MacKenzie 2014; Pasquale 2015; Speculate This! 2013). Methodologically, it draws on interviews with stakeholders and app analysis, along with probing tactical data activism. First, I explore how the expediency of data is implicated in social stratification and self-responsibilization, for instance to secure one’s (personal) future. I distinguish three “classes,” namely user positions informed by distributions of data-centric perception and cognition: the predicting class equipped with robo-advisors; the predicted class self-identifying through credit scoring; and the unpredictive class, negotiating exclusion and anonymity. Second, I juxtapose articulations of uncertainty as quantified expression of risk and as cultural signification and affective experience of precarity. Turning to data activism, I ask whether and how personal and financial data could be made to conjure publics and tell stories that underscore the radical openness of the future or enable relations of solidarity.
Grounding public engagement in data: some experiments to ‘issue-fy’ vertical farming
Author: Laurie Waller (Technical University of Munich)
The concept of “vertical farming” has recently provided a focus for debates about the technology and infrastructure of sustainable food. This paper discusses some engagement experiments involving social media data and attempts to ‘issue-fy’ vertical farming as a concern for sustainable food publics.
The concept of “vertical farming” has provided a focus for recent debates about the technology and infrastructure of sustainable food. In digital publicity – from news coverage to search engines to social media – vertical farming has been widely associated with images of plant-filled skyscrapers, stacked trays of leafy greens grown under intense purple LED lighting, and reports of high-profile venture capital investments. The sensational character of much vertical farming discourse, and the highly instrumental treatment of ecological issues in vertical farming media coverage, presents a challenging situation for researchers and practitioners experimenting with indoor growing environments. In controversies about vertical farming, it is not uncommon to find technology set in opposition to concerns of ecology and community and staged as a source of public antagonism over what makes food sustainable. This paper will discuss an on-going public engagement process in which, as part of a consortium of vertical farming researchers/practitioners, I’m attempting to use data gathered from social media research to map relations between vertical farming and sustainable food issues. I aim to explore the affordances of vertical farming data from different social media platforms for engaging publics concerned by sustainable food. The paper will reflect on the extent to which ‘disruptive’ technoscience publicity can be reassembled as material that can both ‘issue-fy’ innovation processes and ground a speculative concept like vertical farming in more earth-bound public problems.
Instrumental, collaborative, agonistic: a spectrum of civic data projects in Los Angeles
Author: Morgan Currie (University of Edinburgh)
This paper analyzes civic data projects. The world-making capacities of these projects fall on a spectrum, with one end exemplifying participation that colludes with the administrative state, and the other end a means to political action by social movements advocating for systemic change.
This paper proposes a theoretical analysis of civic data projects at the municipal level. The world-making capacities of these projects’ data infrastructures range on a spectrum, with one end representing forms of participation that are collusive and by which individuals take part in reproducing the administrative state, and, at the other end, more monitory and antagonistic activity by social movements calling for broader political change.
Civic hacking, for instance, contributes to the design of public service provisions that complement or even supplement government policies. Civic hackers often focus on designing service infrastructure and are necessarily collaborative with city administrations; they design services per policies already in place. On the other end of the spectrum, grassroots data projects engage in monitory and antagonistic forms of data activism, either by calling for greater government transparency or by criticizing state-based statistical representations of an issue or group. Such projects may be part of broader social movement campaigns advocating for systemic change.
For evidence of these distinctions I draw on participant observation at civic hacking events and grassroots projects in Los Angeles’ over a five-year period, from 2013 to 2017. This research entailed field visits and interviews at civic hacking events and meet-ups, and participant observation with civic data activism by three community groups: the Los Angeles Bike Coalition, which counts cyclists, and two social justice organizations, the Youth Justice Coalition, who count the dead at the hands of police, and Stop LAPD Spying, who track government data collection on poor and immigrant communities.
Intervening in data infrastructure using ontological switches
Author: Christian Nold (University College London)
The paper presents an ethnographic method for intervening in technology practices using ‘ontological switches’. This sensitises researchers towards seeing data infrastructures as fluid and allows them to see where they can intervene to allow more actors to participate in world making.
This paper presents an ethnographic method for observing and intervening in technology practices using ‘ontological switches’. This extends Winthereik & Verran’s (2012) notion of an ethnographic switch that can be used to performatively and disruptively shift between different generalisations. My paper reports on a multi-year study of four participatory sensing devices from design, usage and outputs. In the study, a single technology device would go through radical changes where it came to enact different practices of public health, environmental activism, behavioural experimentation and transhumanism. Using Mol’s (1999) notion of ontologies as multiple and enacted in practice, the paper identifies material and organisational choices such as GUI sliders that functioned as a ‘switch’ between different ontologies, making some possible and foreclosing others. The ‘ontological switch’ is a way to sensitise researchers and designers towards seeing data infrastructures as fluid and allowing them to see where they can best intervene. Using this approach, the paper shows how targeted interventions at these switch points can be used to multiply realities and allow a broader range of actors to participate in world making.
Mol, A. (1999) ‘Ontological politics. A word and some questions’, in Law, J. and Hassard, J. (eds) Actor Network Theory and After. Oxford, MA: Blackwell Publishers, pp. 74-89.
Winthereik, B. R. and Verran, H. (2012) ‘Ethnographic Stories as Generalizations that Intervene’, Science Studies, 25(1), pp. 37-51.
Open data infrastructures: the tension between attachment, detachment and reattachment
Author: Antoine Courmont (Sciences Po)
This paper introduces an analytical framework for open data policies based on the tensions produced by the attachment, detachment and reattachment of data. It highlights that technical choices made in the implementation of a data infrastructure have strong impacts on the making of (open) data worlds
Open data policies rest on the ideology of information liberalism which claims that information must circulate freely in order to solve various problems in the context of cybernetic theories. The information liberalism is based on the assumption that data exist, are autonomous and can easily circulate.
Based on a four-year ethnographic investigation inside the Metropolis of Lyon (France), I question this assumption following an approach of sociology of quantification. Indeed, the analysis of the open data policy in the making reveals a tension between attachment and detachment that needs to be addressed to allow a smooth circulation of data. Attached to vast socio-technical networks, data must be detached from their initial environment to circulate, before being re-attached to new users. Following data from their production to their reuse, my communication will consecutively highlight the attachment between data and local policies, the trials of detachment to make data circulate, and the data reattachment to secondary uses.
The description of each of these steps highlights the technical choices made in the implementation of an open data infrastructure that have political consequences in the making of (open) data worlds. Indeed, moving from one social world to another one, the open data obtain new characteristics that enable (or restrain) their attachments to alternative users. The case of Lyon underlines a focus on economic re-users. As a consequence, the open data infrastructure reflects these prefigured users and limits the emergence of new forms of political participation through open data.
Research portfolios as tools for opening up deliberations on priority setting
Author: Ismael Rafols (Universitat Politècnica de València)
Different types of research data are becoming increasingly available. They are used to conduct analyses of the research portfolios of organisations and funding agencies. These portfolio analyses can be used in for opening up deliberations on priority setting – although current use remains unclear.
Funding agencies use the term research portfolio analysis to describe analyses that map the ensemble their activities. Data providers offer data infrastructure on publications, funded grants, patents, twitters of publications and clinical trials. In parallel, some funding agencies have set up internal mechanisms for portfolio analysis. By classifying this research data into suitable categories (disciplinary, disease type, institutional type), it is now possible to make estimates of the amount of resources spent on given topics, or performed by given units or organisations.
Most agencies are not open about the extent and goals of portfolio analyses conducted. Anecdotal evidence from interviews suggest that the main purposes of portfolio analysis are: i) to provide agencies of baseline information about the areas they support and ii) to internally justify resources spent in terms of outputs obtained.
In this study, however, we aim to explore how portfolio analysis can also be used as part of wider deliberative processes of priority setting. We will present some examples WHO and the Dept. of Health of Catalonia. These wider practices of priority setting require the development of technical expertise in order to manage research data as well as estimates of societal needs. Equally, they also require institutional learning for managing processes of deliberation and integration of knowledge from diverse stakeholders.
We propose that portfolio analysis may constitute an example of how large data can be garnered to visualise opportunities for making research choices – highlighting ambiguities and uncertainties- and thus facilitate deliberative processes of priority setting.
Authors: Noortje Marres (University of Warwick); Brit Winthereik (IT University of Copenhagen)
Social data science has not come to terms with socio-technical transformations of life in a digital age. To address this, we must make the situation the unit of analysis. But most situational approaches assume a “fieldsite.” How do we move from situational analysis to situational analytics?
It is argued that digital data and tools enable the reconciliation of foundational oppositions in social research, such as that between content and context. Analytic operations that were previously assumed to require field work are brought within the frame of data analysis, with the aid of computational methods like natural language processing (Nelson 2017). But it is becoming clear that data analytics throws up methodological problems of its own: what operations enable the specification of events, issues and actors? how do capacities for intervention emerge from this? This paper proposes that if the digital has transformed the status of ‘the field’, this is due to socio-technical transformations of the settings of everyday life as much as to the analytic prowess of data science. Digital architectures make latent aspects of social life re-portable and share-able, and this renders social life more artificial, less spontaneous, and thinly structured, giving rise to we call “semi-fields” (Kelly, 2012). How to attune our analysis to these transformations of the very composition of social life in the wake of the proliferation of digital devices? To conduct social inquiry by digital means, we must counteract social ontologies designed into digital architectures, by making the situation our unit of analysis. In this we follow pragmatist and performative approaches and in particular Adele Clarke (2005). However, most situational approaches still assume the “field”: what elaborations are needed to apprehend situations as they unfold in semi-fields? How do we move from situational analysis to situational analytics?
Teaching to problematise digital infrastructures through data activism
Author: Tommaso Venturini (École Normale Supérieure Lyon)
This paper reflects on use of data activism as a pedagogical tool to train students in problematising data infrastructures and digital technologies
Far from being neutral, data generate political effects at every stage of their production, cleaning, analysis and presentation. Such effects are sometimes manifest, for example in the use of demographic statistics for the justification of public governance, and sometimes, subtler for example in the way we classify knowledge in our libraries.
While the bias of the data is easy to proclaim, it is more difficult to observe and even more to teach. Information systems hide their political attachments, not necessarily maliciously, but simply because such opacity makes them more efficient in their tasks of knowledge management. It would be impossible to search a piece of information on the Web if, every time, we had to discuss Google’s business model or to wonder why its algorithms privilege some results. Yet, data infrastructures have fundamentals political consequences that must be part of any curriculum related to digital technologies.
Data activism arises precisely from the desire to expose the power asymmetries inherent to information systems. It seeks to promote access to data; to investigate the conditions of their production; to explain the constraints they generate; to propose alternative ways of redistributing their social consequences. Precisely because it raises awareness about the political dimension of digital technologies, the practice of data activism can be beneficial to make students sensitive to the effects of the sociotechnical infrastructures.
In this contribution we discuss the benefits and the difficulties of teaching data activism drawing on the experience of a course at ENS Lyon, France.
Where’s the database in digital ethnography? Exploring database ethnography for open data and smart cities research
Authors: Ryan Burns (University of Calgary); Grace Wark (University of Calgary)
Open data infrastructures are increasingly used to make cities “smarter”, but can deepen and (re)produce existing structural urban inequalities. Here we propose the database ethnography as a methodological intervention into the contestations, politics, and promises of open data platforms.
Recent research has shown that the social, political, and epistemological relations underpinning open data platforms frame the platforms’ implications in the world. As “smart cities” increasingly turn to these data infrastructures, researchers are looking to the ways open data might (re)produce existing structural urban inequalities. Such inequalities range from the struggles around how data capture knowledge, places, and people; to the uneven material distribution of resources in a city toward areas adept at data analysis. While many long-established methodologies exist for illuminating these processes, the digital spaces of open data infrastructures pose critical new challenges for researchers.
In this paper, we develop the database ethnography as a rich methodological resource for open data research. This approach centers the database as a key site for the production and materialization of social meaning. The database ethnography draws attention to the ways digital choices and practices – around database design, schema, data models, and so on – leave traces through time. From these traces we may infer lessons about how phenomena come to be encoded as data and acted upon in urban contexts. More specifically, we argue that open databases limit data types, categorize and classify data to align with technical specifications, reflect the database designer’s episteme, and (re)produce conceptions of the world. We substantiate these claims through a database ethnography of the open data portal for the city of Calgary, in Western Canada.