The Epistemology of Data-Intensive Science and Its Implications for Science Policy

The massive scale of scientific data production, combined with novel technologies and regulatory frameworks for the dissemination and analysis of big and open data, is having a profound effect on what counts as scientific knowledge and on how that knowledge is obtained and used. Scholarship in the philosophy, history and social studies of science is crucial to capture and articulate precisely what is new about data-intensive science, what the characteristics of this mode of research are, and what implications they bear for policy and society. Based on philosophical analysis and an extensive ethnographic and historical study of data practices across several research fields, this talk provides a systematic overview of the characteristics and implications of data-intensive science. I focus on how data are defined and processed, the significance of data curation strategies towards data-driven discovery, the extent to which data can and should be made open, and the Open Data divide that threatens to engulf data-intensive research. I conclude that the novelty of data-intensive science lies in the methods and principles used for data processing, rather than in the volume, variety, velocity and veracity of data themselves; and that far from being comprehensive, current data collections are highly selective and hard to evaluate and re-use, particularly for those who have no means to participate in the development of data infrastructures and related standards. These insights have far-reaching consequences for data governance and particularly for Open Science guidelines and implementation, as illustrated by my current involvement as advisor in the development of Open Science policies at the European and global levels (as a member of the Open Science Policy Platform of the European Commission and Open Science lead for the Global Young Academy).