"Sometimes there are differences of opinion as to what the data mean": bureaucracy and facticity in the face of inconclusive science