Abstract

Philosophers concerned with the character of scientific disputes tend to divide into two camps. On the one side there are those who hold that scientists can always settle their differences by appealing to shared assumptions; on the other side there are those who maintain that in many cases scientists must resort to (nonrational ) persuasion to establish their views. The trouble is that for all their strong points both approaches labour under enormous difficulties. Scientific disagreement is often much deeper than the first conception allows while persuasion, propaganda and the like are far less prevalent than the second one requires. On examining scientific practice is becomes clear that philosophers on both sides err in supposing that scientists can resolve their differences in a purely scientific fashion just to the extent that they can single out one of their views on the basis of assumptions that they hold in common. For one thing scientists make progress even though they rarely employ the strategy of reversion to neutral territory. And for another the strategy of reversion to neutral territory is at odds with the important methodological requirement that all relevant information be taken into account. Instead of thinking of scientists as resolving their differences by reviewing their respective positions in the light of their shared assumptions, we should think of them as fashioning positions on which they can agree, as arguing to shared conclusions rather than from shared premises. The relevant picture is one of scientists gradually narrowing the distance that separates them, of their converging on a mutually acceptable viewpoint. Far from being prerequisites for scientists' resolving their disputes, shared assumptions are what result from their managing to resolve them. How scientists who disagree deeply make progress is thus much less of a problem than is usually supposed. What determines whether scientists' conclusions constitute advances is not wether they were obtained by the expedient of reversion to neutral territory but whether each of the available options was fully ventilated. When scientists take into account all the relevant considerations (including those introduced into the discussion by their opponents and those that they can reasonably be expected to know), the question of whether they can be said to have made progress no longer arises. More generally, we misconstrue the manner in which scientists resolve their differences because we allow ourselves to be guided by our philosophical preconceptions about scientific knowledge and we devote insufficient attention to what happens in practice. Instead of exploring how scientists reason to mutually acceptable conclusions, we wrongly assume that they cannot but proceed by reverting to neutral territory. What is needed is not a general account of consensus formation but rather a fresh look at the multitude of ways in which scientists make progress by dint of argument and debate. Considered as a phenomenon in its own right scientific disagreement poses no special problem; it is puzzling only when it is considered within the framework of traditional epistemology.