This chapter is probably irrelevant to scientific realism and so can well be omitted. It is about Putnam's important new ‘internal realism’, apparently a species of idealism. A switch from realism to idealism sounds central to our discussion, but it is not. Putnam is no longer engaged in the debate between the scientific realist and the anti-realist about science. That debate makes a keen distinction between theoretical and observable entities. Everything Putnam now says ignores that. So it should be. His is a philosophy founded upon reflections on language, and no such philosophy can teach anything positive about natural science.

To omit Putnam's developments would nevertheless be to bypass issues of current interest. Moreover, since he finds a predecessor in Kant, we can bring in Kant's own kind of realism and idealism. Kant is a useful foil to Putnam. If we simplify and pretend that Kant too is an ‘internal realist’ (or that Putnam is a ‘transcendental idealist’) we can imagine a Kant who, unlike Putnam, emphasizes the difference between observed and inferred entities. Putnam seems to be a scientific realist within his internal realism, while we can invent a Kant who is an anti-realist about theoretical entities within a similar setting.

Internal and external realism

Putnam distinguishes two philosophical points of view.