Science off the books: institutional friction and the invisible costs of research governance

Abstract Conventional research evaluation systems measure scientific outputs but remain largely blind to the cognitive conditions under which these outputs are produced. This blindness generates no corrective pressure on the institutional processes shaping research work, progressively favouring metrically visible activities over organisationally complex or operationally intensive ones. This perspective introduces the concept of institutional friction to describe how legitimate governance mechanisms—such as peer review, reporting, and coordination—generate cumulative and unobserved cognitive costs through their interaction. These costs do not arise from individual processes, but from their simultaneous operation on the same actors, creating substantial but systemically invisible cognitive load. A three-criterion framework—epistemic reliability, coordination necessity, and cognitive proportionality—is proposed and applied across research contexts. The analysis highlights three implications: making cognitive costs observable, redesigning coordination mechanisms, and treating attention as a limited infrastructural resource.