We argue that the dream of a ‘perfect language’ – namely, a universal, unambiguous and semantically transparent medium of expression –, whose intriguing story has been told by Umberto Eco (1993), is deeply intertwined with the myth of instant rationality: the idea that a perfect language is one in which all logical relations becomeimmediatly visible, so that the language itself “does the thinkingfor us” (Frege 1884). In the first part of this paper we trace this versionof the dream in the works of Leibniz, Frege, Russell and Wittgenstein. In the second part we re-examine it in the light of more recent negative results in logic and theoretical computer science.