There is a relatively simple argument which shows that science is not, in any important sense, value free. Further, if science is not value free, it may be that the values involved in science rest upon moral considerations. In this paper the author examines one argument for the claim that science essentially involves value, and then looks at the types of values which may be an essential facet of scientific inquiry.
Two apparently unproblematic premises lead to a quite controversial conclusion about the relationship between science and value. It seems readily apparent that science, in some essential way, involves making decisions; and it is generally agreed that assessing the rationality of any decision involves appraising how well that decision promotes or fosters the achievement of the values sought. If these two premises are granted, then it follows that any scientific decision which claims to be rational must be so because of the relationship between that decision and the values involved in science.
Obviously, it is possible to deny that science essentially involves decision, as Richard Jeffrey 1 did in response to the claim by Richard Rudner 2 that science involves decision making, hence value. But then the question which arises is what such a 'decisionless' science would look like? It is, I think, clear that such a view of science is not adequate. To think of science as a wholly decisionless enterprise involves not only Jeffrey's denial that a scientist properly 'accepts' hypotheses, but also requires that one deny that scientists properly choose measurement procedures designed to yield an estimate of a parameter within a (chosen) degree of error, make choices among various experimental designs, decide upon fruitful questions to be pursued, and so on. The list could be extended indefinitely, but the point has been made before and need not be elaborated here. 3 The necessity of admitting that science essentially involves making decisions can be most clearly seen by noting that a science without decisions would be simply an indefinitely long series of