In recent years, there has been much discussion over how to assure scientific integrity. It has become clear that a few scientists have fraudulently collected or reported data, conducted harmful or unethical experiments, or practiced “unscientific” procedure. What are regulative bodies to do? The approach has been to define research misconduct and then use that definition to assess scientific practice.[1] But just how to define research misconduct and hence, regulate the conduct of scientists in research? The debate that resulted in (...) response to this question, and that led ultimately to the new federal definition (42CFR50), has both theoretical and political underpinnings. The political underpinnings have been greatly discussed. But the theoretical underpinnings (and their connection to the political) have not. To give a definition of “good” versus “bad” science requires some understanding of the scientific process itself. So theoretical ideas about what constitutes good or bad science—ideas that influence and help shape our ideas and applications of research misconduct definitions—have political implications for the regulation of science. However, as the debates within both the scientific and philosophical communities have made clear, there are significant limits to any appropriate legislation of what counts as “good” science. A definition of research misconduct (or “bad” science) that spells out the nature of science too stringently may stifle scientific innovation. Consider the case of sociobiology when it was first introduced in the 1970’s. The idea is to account for social behavior of various species (including Homo sapiens) in terms of evolutionary biology. To take one example from sociobiology, the greater tendency of males toward rape is to be explained in terms of the evolutionary advantage for the male’s genes. Not surprisingly, sociobiology met with great controversy; some considered it an inappropriate application of biological principles to the social-interpretive-human domain. Still, few did or would suggest that it is the province of any regulatory body to silence the advocates of the new field. We will endorse and reinforce the standard arguments that the scientific process is too complicated to provide mechanical rules of conduct.. (shrink)