Universal informatics : building cyberinfrastructure, interoperating the geosciences
- Author(s): Ribes, David
- et al.
The creation of cyberinfrastructure is an ambitious U.S. endeavour to build large-scale information infrastructure for the sciences. Dubbed 'revolutionary' by their advocates, cyberinfrastructure names the goal of building a unified information substrate to 'interoperate the sciences' and promote multidisciplinary research collaborations. This dissertation is based on a three-year ethnography of one such emergent infrastructure project: GEON, the geosciences network. I identify, as a principal research object, the logic of interoperability: an emerging set of techniques and technologies which seek to preserve the specificities of heterogeneous sciences while linking them. In principle standardization offers the benefit of making possible communication, data sharing and integrated computing systems; however, in practice such projects often fail or generate substantial opposition. I argue that the logic of interoperability seeks to blunt the politics of standardization while retaining its enabling properties. Rather than erasing disciplinary difference interoperability calls for the sciences to be known and mapped in order to make possible an automated crossing. In this vision, the specificity of the sciences are preserved while domains are linked through relations of mediation. Drawing from research in Science and Technology Studies and the methodologies of actor-network theory and ethnomethodology, I trace the enactment of the logic of interoperability in GEON at three scales of action: institutional, organizational, and technical. At each scale I sustain a focus on the material and organizing practices of members as they work to interoperate the earth sciences. At the institutional scale there is a growing impetus and increasingly sophisticated skill-set for the arrangement of multidisciplinary collaborations of domain and computer science. At the organizational scale new methods for constructing large-scale umbrella infrastructures are being invented. At the technical scale a set of technologies of interoperability are under development which seek to automate translations of the data, language, concepts, and knowledge of science itself. Together these point to a mounting confluence of efforts at interoperability seeking a 'revolution' of science at all scales of action and positing a new model of governance for science