ABSTRACT
It is tempting to dismiss worries about the underdetermination of theories as purely philosophical doubts. Of course we cannot prove beyond all doubt that electrons exist, but then we cannot prove beyond all doubt that the Sun will rise tomorrow, that all metals expand when heated, or that everyday objects such as tables are still there when we aren’t perceiving them. Hence, assuming that the mere possibility of error is not sufficient seriously to undermine ordinary claims to know, neither does it threaten scientific realism. Indeed, as we saw in the previous chapter, the very most that van Fraassen can claim to have shown is that it is not irrational for us to withhold belief in unobservables, but neither is it irrational to believe in them, and this is much weaker than traditional forms of scepticism about science. In this chapter, we will consider arguments for various kinds of antirealism, which are motivated by careful scrutiny of the practice and history of science, rather than by epistemological scruples. In different ways, facts about real science raise the question ‘what should we be realists about?’.
