Objectivity versus relevance in studies of scientific advance

A conceptual framework is suggested within which various techniques for studying scientific advance may be viewed. The two axes arerelevance of the technique to a “true” measure of the rate of scientific advance, versusobjectivity of the technique. It is suggested that a situation exists somewhat analogous to the Heisenberg uncertainty, principle; the most objective technique, a simple publication count, is the least relevant to a true measure of scientific advance, while the most relevant technique, interviews with an eminent and knowledgeable scientist in the field, is the least objective. Between these two extremes lie a group of scientometric techniques which should be capable of producing analyses which are both satisfactorilly relevant and satisfactorily objective.