ABSTRACTThe determination of an appropriate level of support for research has led to hard questions about the value of research and to realization that assessment of research value must be based on careful evaluation of a wider range of social benefits and costs than has hitherto been customary. Defining and measuring social aims poses a number of difficult but not insoluble problems. We also need to think more scientifically about the values we place on different types of knowledge‐how these enter into decision making and how they affect communication with non‐scientists and the transition between research and practice. A dose of humility and compromise is in or