As the era of ‘big data’ unfolds, researchers are increasingly engaging with large, complex data sets compiled from heterogeneous sources and distributed across networked technologies. The nature of these data sets makes it difficult to grasp and manipulate their materiality. We argue that moments of breakdown – points at which progress is stopped due to a material limitation – provide opportunities for researchers to develop new imaginations and configurations of their data sets' materiality, and serve as underappreciated resources for knowledge production. In our ethnographic study of data-intensive research in an academic setting, we emphasize the layers of repair work required to address breakdown, and highlight incremental innovations that stem from this work. We suggest that a focus on the breakdown–repair process can facilitate nuanced understandings of the relationships and labour involved in constituting data assemblages and constructing knowledge from them.

Ample scholarship demonstrates that data-intensive technologies have the capacity to cause serious harm and that their developers are obliged to address ethics in their work. This ethnographic paper tells the story of data scientists attempting to instantiate a carefully considered ethical vision into a data infrastructure while balancing competing priorities, negotiating divergent interests, and wrestling with contrasting values. I use their story to develop the concept of “ethical abduction,” which I characterize as an exemplary process by which actors can intentionally and systematically address ethical issues that arise during their day-to-day actions by making decisions with consideration for a foundational ethical worldview. It entails tacking back and forth between divergent but complementary ways of thinking: between establishing ideals and making decisions given practical constraints; between understanding historical context and anticipating future consequences; between acknowledging structural dependencies and accepting responsibility for moral agency.

As a nascent field within the academy, the contours, attributes, and bounties of data science are still indeterminate and contested. We studied how participants in an initiative to establish data science at a large American research university defined data science and articulated their relationships to the field. We discuss two contrasting visions for data science among our research participants. One vision is a transdisciplinary view portraying data science as a phenomenon with transcendent, appropriative, and impositional qualities that sits apart from academic domains. Another view of data science—one that was far more prevalent among our research subjects—casts data science as grounded, relational, and adaptive, emerging from crosspollination of numerous academic domains. We argue that this latter formulation represents a more quotidian reality of data science and positions the field as an extradiscipline, defined as a field that exists to facilitate the exchange of knowledge, skills, tools, and methods from an indeterminate and fluctuating set of disciplinary perspectives while conserving the boundaries of those disciplines. We argue that the dueling transdisciplinary and extradisciplinary visions for data science have important implications for how the field will mature, and that the extradiscipline concept opens novel directions for studying academic knowledge production in STS, contributing additional precision to the literature on disciplinarity and its permutations.

Introduction. Recent work on algorithmic systems has characterized such systems as relational entities, which are embedded in larger sociotechnical assemblages. We examine some implications of this framing for the practical accomplishment of transparency by organizations and communities. Specifically, we investigate a large astronomical survey project’s effort to render its observation scheduling algorithm transparent to a large international community of researchers. Method. We report on one part of a larger ethnographic study with the survey project, using interview data and documentary data as the primary empirical sources for this analysis. Analysis. The authors conducted theory-generative qualitative coding based on grounded theory methodology, developing three codes relevant to the organizational accomplishment of transparency in this case. Results. We argue that from the perspective of an organized group that is wrestling with the opacity of its algorithmic systems, accomplishing transparency is effortful in design and organizing work, it is ongoing, and it produces novel organizational relations. Conclusions. Relational conceptions of algorithmic and AI/ML systems highlight the work and challenges of accomplishing transparency beyond processes of publication, visibility, or access. Greater attention to this organizing work is warranted in the broadening rollout of such systems in scientific and social life.