Scientific Realism, Scientific Practice, and the Natural

Both sides in the debate about scientific realism have argued that their view provides a better account of actual scientific practice. For example, it has been claimed that the practice of theory conjunction presupposes realism, and that scientists' use of multiple and incompatible models presupposes some form of instrumentalism. Assuming that the practices of science are rational, these conclusions cannot both be right. I argue that neither of them is right, and that, in fact, all scientific practices are compatible with both realism and instrumentalism. I also repudiate van Fraassen's argument to the effect that the instrumentalist account of scientific practice is logically weaker, hence better, than the realist account. In the end, there are no scientific practice arguments on the table that support either side of the debate. It is also noted that the deficiencies of van Fraassen's argument are recapitulated in Putnam's miracle argument for realism. My pessimistic assessment of the state of the debate is reminiscent of Arthur Fine's. However, Fine's argument for the 'natural ontological attitude' once again repeats the problems of van Fraassen's and Putnam's arguments.