It is often held that only by the time of the late Sophist did Plato discover a way of dealing with puzzles about the possibility of false judgement and false statement. Earlier dialogues such as Euthydemus and Theaetetus which introduce the puzzles are thought to labour under assumptions about how language relates to reality, born of inexperience in semantics, that stood in his way. Here it is argued that in both those dialogues Plato is in fact doing something subtler than captivity to a crude picture of the way language works would allow. A more attentive reading of these two texts makes it clear that he has already identified the structural relation between subject and predicate as the key: not only to understanding how false judgement is possible, but through that to bigger questions about the relation of thought and language to the world in general. The Euthydemus, in particular, shows us how many more ways there are for an argument to go wrong than are dreamed of in the logic books. It even suggests that a failure in logic may sometimes be simultaneously a failure in love.