In my view, the problem with Rorty is that he fails to pay sufficient attention to practices outside of language that serve to strengthen or weaken a person’s — or a whole linguistic community’s — conception of truth. Clearly language is very important, but even the most persuasive and cohesive description that fails to describe what eventually happens or already happened when it uses its internal logic to speculate on past, present or future will lose persuasive force. Without this kind of activity, science would never have progressed beyond scholasticism.
Rorty may not be metaphysically logocentric, but he is pragmatically logocentric. Only some of our truth-instauration practices are linguistic.