Gender and Boyle's Law of Gases

19. Good Science

19 Good Science For many people, there is something deeply wrong with a book that says science is not always value-free. “Everyone knows” that good science is value-free and that if values in®uence scienti¤c work, the result is bad science. So this book must be wrong about Boyle. Since his work was good, he cannot have let his religious and political views in®uence his choice of theories. Or else the book is wrong because it implies without saying so that, since Boyle’s work leading to the Gas Law was in®uenced by his religious and political views, it was bad science after all. How could Boyle’s ideas about proper religious, social, and political arrangements, including gender arrangements, in®uence his choice of mechanism and so his laboratory work, his collection and interpretation of data, and still allow him to produce good scienti¤c work? In this chapter, we will see how good science can be in®uenced by gender and class politics and still be good science. But ¤rst we need to distinguish two kinds of values. Philosophers distinguish many kinds of values in connection with science, but two are relevant to our argument. One kind is said to be acceptable for science; these are called constitutive values (or epistemic values or cognitive virtues). Boyle referred to them as “Requisites for a Good [or “Excellent”] Hypothesis.” The list varies among philosophers but usually includes the theory’s being probably true, closely ¤tting experimental data, and having informational content, predictive power, simplicity, compatibility with established theories in related ¤elds, refutability, internal consistency, etc. Scientists use such criteria in evaluating their hypotheses and theories. The other kind of values is called contextual (or non-epistemic or non-cognitive ), and it includes political, religious, economic, and any other values not considered constitutive.1 Although we begin this discussion by making this distinction between two kinds of values, we will end by seeing that in the actual practice of science, there is no clear distinction . 172 One of the problems in dealing with the connection between science and contextual values is stating precisely what the nature of the connection is alleged to be. Longino distinguishes ¤ve ways in which contextual values can affect good scienti¤c work; they can 1. affect practices that bear on the integrity of science, as when the desire for pro¤t leads a scientist-entrepreneur to present his results ¤rst at a press conference rather than in a professional journal or at a professional conference; 2. determine which questions are asked and which ignored; 3. affect the description of data, that is, value-laden terms may be employed in the description of experimental or observational data and may in®uence the selection of data or phenomena to be investigated; 4. be expressed in or motivate the background assumptions facilitating inferences in speci¤c areas of inquiry; or 5. be expressed in or motivate acceptance of global, frameworklike assumptions that determine the character of research in an entire ¤eld. (Longino, Science, 86) Our study of Boyle’s work shows that contextual values in®uenced his choice to pursue a mechanistic research program (way 5) and so in®uenced his interpretation of data, e.g., the data con¤rming Boyle’s Law (way 4). We need a model of scienti¤c thinking that shows us how it is possible for scienti¤c work to be in®uenced by contextual values and yet still be good science. Such a model allows us to see that background assumptions about proper religious, class, and gender arrangements, along with assumptions about the natural world, constrained the choices Boyle made about how to pursue his experimental work. And, most importantly, such a model allows us to see that Boyle’s work was, nevertheless, excellent: predictive, fruitful, and with good explanatory power. There are several models or philosophies of science available that will serve these purposes. I have found Mary Hesse’s network model of scienti¤c theories useful for pursuing the gender politics in seventeenth-century science.2 Hesse understands a scienti ¤c theory as a system of laws which has a very complex relation to nature.3 When a scientist establishes a law, Hesse argues, he classi¤es phenomena on the basis of resemblances among them. Any scientist is, then, constantly faced with decisions as to whether two things Good Science 173 are similar enough to...