Goethe's celebrated maxim, "Seek nothing beyond the phenomena. They are themselves the theory" was deeply congenial to Wittgenstein. But Wittgenstein believed that a synoptic ordering of the phenomena ["übersichtliche Darstellung"] is grammar, and that essence is given by grammar, not by scientific theory. He claims that "Someone who agrees with Goethe believes that Goethe correctly recognized the nature of colour" (Remarks on Colour, 1977, I 71). He adds, "Nature here is not what results from experiments, but lies in the concept of colour." This is Wittgenstein's interpretation of Goethe's Farbenlehre: it is a grammatical inquiry mistakenly believed by Goethe to be an empirical scientific investigation. Yet Goethe did perform a mass of scientific experiments from which a picture of the nature of colours was meant to emerge. This paper is a defence of Goethe against Wittgenstein's linguistic idealism on the one hand and the crudities of the wavelength theory of colour on the other. In Remarks on Colour Wittgenstein raises a number of puzzle questions which, I claim, cannot be answered by a science with the foundational muddles inherited from Newton. I attempt to solve them by offering an übersichtliche Darstellung of the relevant facts, derived from Goethe's theory. My criticism of Wittgenstein is that he was the victim of one narrow image of what form science must take.