Few claims in the recent history of philosophy in its relation to literary studies are as notorious or have been as much misunderstood or misrepresented as Derrida's assertion that `there is nothing outside the text' (Derrida, 1976, Of Grammatology, Johns Hopkins University Press, Baltimore, p. 158). Like other, less remarked-upon but similar assertions, `the thing itself is a sign' (Derrida, 1976, Of Grammatology, Johns Hopkins University Press, Baltimore, p. 49) and `there never was any "perception"' (Derrida, 1973, Speech and Phenomena, Northwestern University Press, Evanston, IL, p. 103), it appears to commit Derrida, and deconstruction as whole, to a radical form of linguistic idealism in terms of which everything is dissolved into language. I want to argue that, contrary to common perceptions, this is not Derrida's position. Derrida insists on neither the collapse of the world into language nor on language as untrammeled free play. He has said this unequivocally on numerous occasions. However, the truth of his denials needs to be demonstrated, not simply stated. In the following pages I therefore offer a brief analysis of some of Derrida's early texts, in order to show that, from the very beginning, Derrida argues that language is essentially imbricated in the world, and that it is produced out of a set of very specific, even `historical', determinations, even if his conception of such imbrication and determination is philosophically unorthodox, and therefore likely to be mistaken for the positions that he in fact attacks.