Data management plans as anticipatory discursive infrastructure

The Data Management Plan (DMP) and similar plans are required by research and scientific agencies all over the world to ideally achieve the goals of open science. This paper focuses on the case of DMPs for the United States' National Science Foundation (NSF). The concept anticipatory discursive infrastructure shows how DMPs illuminate the policies, rules, procedures, and laws that shape data cultures. The DMP is anticipatory because it imagines but doesn't determine the future of science, and it is discursive in the sense that the guidelines that it points to are interpreted and enacted in data management processes. It is also infrastructural in that the DMP underpins the production of knowledge infrastructures. We contrast how the DMP policy requirement was adapted by two program areas of research: Science and Technology Studies, called STS, and Ocean Sciences. Through analysis of guidelines discussed in 423 Ocean Science and STS DMPs from funded NSF projects, our analysis demonstrates how DMPs both constitute and map different anticipatory discursive infrastructures: Ocean Science researchers referred to field best practices, policies from well‐funded data repositories, and specific NSF guidelines; STS referred to human subjects rules and general university resource guidelines for data management and data access.