Realism and Reason (Philosophical Papers, Volume 3)

The argument starts from a result in formal logic: any consistent theory with an infinite model has many models, so the theory's sentences can be reinterpreted systematically while every one of them stays true. Putnam applies this to an ideal science that satisfies all operational and theoretical constraints. If such a theory still admits unintended interpretations, then nothing in the theory, or in our use of it, picks out its intended reference. He concludes that the metaphysical realist's idea of a unique correspondence between words and mind-independent things cannot be sustained. Critics have disputed whether the argument proves this or merely assumes that use cannot fix reference.

Metaphysical realism holds that the world consists of a fixed totality of mind-independent objects and that truth is correspondence between a theory and that totality. Internal realism, as Putnam presents it, says that asking what objects exist makes sense only within a theory or description, and that truth is idealized rational acceptability, not correspondence with a world viewed from nowhere. He insists this is not relativism, since not every description is as good as another and experience constrains inquiry. Readers should compare it with his later natural realism, which he came to regard as a better way to keep objectivity.

Putnam is widely associated with externalism about meaning, the view that the reference of natural-kind terms depends partly on the environment and on a division of linguistic labor. Realism and Reason sits beside that work in the same Philosophical Papers series, and his stance here has created a puzzle for commentators: externalism seems to suppose that the world fixes reference, while the model-theoretic argument seems to deny that anything can. Putnam's reply is that the causal story is itself part of a theory and cannot supply an external guarantee. Scholars continue to debate whether this resolves the tension or sharpens it.

David Lewis answered that Putnam's argument assumes the only constraints on reference are those expressed in the theory, and that reality also contains natural properties that are privileged candidates for what our words refer to. On this response, an interpretation that assigns reference to gerrymandered properties is eliminated by nature itself, not by anything we stipulate. Putnam countered that appealing to natural properties adds more theory, which would be open to the same reinterpretation. The exchange is among the best-known disputes in analytic metaphysics and is a standard route into the literature surrounding the book.

Because Realism and Reason gathers essays composed for different occasions, its themes recur from several angles rather than building to one conclusion. Readers who want the most compact statement of internal realism will find it in Putnam's Reason, Truth and History from the same period, while the papers here offer technical detail and responses to critics. The collection also shows the evolution of his thinking over a few years, including hesitations that a finished treatise would have smoothed over. It is most useful to those already familiar with the main outlines of the debate.

A frequent misreading treats internal realism as idealism or relativism, as though truth were whatever a community or an individual decides. Putnam rejects both. He holds that our conceptual choices do not create the world arbitrarily, that experience and inquiry limit which descriptions are acceptable, and that there are better and worse ways of speaking. What he denies is the further claim that one description captures the world as it is independently of all conceptual contribution. Reading him as denying a real world overlooks the distinction he draws between rejecting a particular philosophical picture and rejecting ordinary realism.