Why We Should Not Reject the Value-Free Ideal of Science

1. Introduction

In recent years, the value-freeness of science has come under extensive critique. Early objectors to the notion of value-free science can be found in Rudner (1953) and Churchman (1956), later objections occur in Leach (1968) and Gaa (1977), and more recent critics are Kitcher (2001), Douglas (2009), and Elliott (2011). The goal of this paper is to examine and critique two arguments opposed to the notion of a value-free science. The first argument, the uncertainty argument, cites the endemic uncertainty of science and concludes that values are needed to give direction to scientific investigation. The second, or moral argument, cites the fact that scientists have moral obligations just like everyone else, and concludes that the decision to accept a scientific conclusion incorporates values by taking these moral obligations into consideration. My goal is to undermine these arguments, thus removing significant opposition to accepting the value-freeness of science.

2. The Uncertainty Argument

The uncertainty argument for the (moral) value-ladenness of science draws from Richard Rudner’s famous argument in his (1953), which Issac Levi (1960) succinctly formulates as follows:

1. 1. The scientist qua scientist accepts or rejects hypotheses.

2. 2. No amount of evidence ever completely confirms or disconfirms any (empirical) hypothesis but only renders it more or less probable.

3. 3. As a consequence of (1) and (2), the scientist must decide how high the probability of a hypothesis relative to the evidence must be before he is warranted in accepting it.

4. 4. The decision required in (3) is a function of how important it will be if a mistake is made in accepting or rejecting a hypothesis. (1960, p. 347).

Thus, scientists in their capacity as scientists must make value judgments. In Rudner’s original argument, the phrase “in the typically ethical sense” is inserted after “important” in (1953, p. 4). Levi drops this phrase in his reconstruction of Rudner’s argument and so his version of Rudner’s argument is more general, citing the need to invoke values of some kind, not necessarily moral (i.e., social, cognitive, and so on), to address the uncertainty inherent in confirmation. As I discuss later, the more specific, moral argument for the value-ladenness of science proceeds independently of the issue of whether confirmation is uncertain.

There are various ways in which one can respond to the uncertainty argument. To begin with, one can question premise (1), as Richard Jeffrey (1956) does. For Jeffrey, instead of saying that scientists accept or reject hypotheses, “the activity proper to the scientist is the assignment of probabilities [to these hypotheses] (with respect to currently available evidence)” (1956, p. 237). This position James Gaa calls “minimalism” (1977, p. 519): scientists simply determine the degree to which scientific hypotheses are confirmed on the basis of the evidence, and do not go further in either accepting or rejecting these hypotheses. They simply pass along these degrees of confirmation to decision-makers who, using these probabilities, arrive at decisions regarding appropriate modes of action. This approach has been criticized for not recognizing that scientists in fact regularly accept or reject hypotheses (a point made by Gaa 1977, p. 519, and by many others). Additionally, Rudner notes that the probability assignments which stand for the degrees to which hypotheses are confirmed are themselves the subjects of (meta-) hypotheses, a worry that Jeffrey asserts cannot easily be answered (1956, p. 246; see also Heather Douglas 2009, p. 54). In response to these criticisms, it is to be admitted that scientists are committed to many hypotheses in their work, and that often these commitments are based on valuational considerations, such as advocating a hypothesis because the scientific community supports it, or because one wants to get along with others or because it is a hypothesis that attracts a great deal of research funding. However, nothing in the thesis of scientific value-freedom conflicts with these obvious truths. Value-freedom, rather, is a normative ideal that scientists strive to attain: whereas a hypothesis may contingently be held for valuational reasons, it is epistemically preferable if it is held for evidential reasons...