From Infrastructure to Exostructure: Emergence and Transience in the Exploitation of Public Data

• In Event: Experimental Infrastructures: Material Insufficiency, Scale, and the Question of (Im)possibility - I

Abstract

Inspired by the open source and free culture movements, early efforts to exploit data in the public sector have led to the widespread adoption of open data platforms for municipal governments. These web-based reservoir-like infrastructures for government data are thought to be agnostic to use-case and user alike, such that any interested party can access open data for any given purpose. But recently, actors and observers in the open data arena have noted that the reservoir model of open data has largely failed to materialize as the boon to innovation that was promised. In this talk, I look at an alternative to this top-down approach for scaling the use of public data, one that is built instead on project-based collaborations intended to grow data infrastructures from the ground up through replication and customization of small-scale experiments. I argue that the extant treatments of infrastructure in STS scholarship do not entirely account for this mode of scaling data-intensive technologies, and introduce the concept of exostructures as a way of ontologizing the phenomenon. Exostructures differ from infrastructures in that they are designed to be transient, replaceable, mobile, and modular rather than durable, stable, embedded, and consolidated. In this talk, I draw on four years of ethnographic participant-observation in a Data Science for Social Good program to explore the consequentiality of exostructural arrangements for public data, attending in particular to the making and remaking of relationships and power dynamics between academia, government, business, and civil society.

Author

• Anissa Tanweer, University of Washington