Abstract

Linguists are becoming more and more interested in the genealogy of linguistics. And in reconstituting the history or prehistory of their science, they are discovering numerous ancestors, sometimes with a certain astonished recognition. Interest in the origin of linguistics is awakened when the problems of the origin of language cease to be proscribed and when a certain geneticism—or a certain generativism—comes back into its own. One could show that this is not a chance encounter. This historical activity is no longer elaborated solely at the margins of scientific practice, and its results are already being felt. In particular, we are no longer at the stage of the prejudice according to which linguistics as a science was born of a single "epistemological break"—a concept, called Bachelardian, much used or abused today—and of a break occurring in our immediate vicinity. We no longer think, as does Maurice Grammont, that "everything prior to the nineteenth century, which is not yet linguistics, can be expedited in several lines."1 Noam Chomsky, in an article announcing his Cartesian Linguistics, which presents in its major lines the concept of "generative grammar," states: "My aim here is not to justify the interest of this investigation, nor to describe summarily its procedure, but instead to underline that by a curious detour it takes us back to a tradition of ancient thought, rather than constituting a new departure or a radical innovation in the domain of linguistics and psychology."2If we are to set ourselves down in the space of this "curious detour," we could not help encountering the "linguistics" of Jean-Jacques Rousseau. We would have to ask ourselves, then, in what ways Rousseau's reflections on the sign, on language, on the origin of languages, on the relations between speech and writing, and so on announce what we are so often tempted to consider as the very modernity of linguistic science, that is, modernity as linguistic science, since so many other "human sciences" refer to linguistics as their particular model. And we are all the more encouraged to practice this detour in that Chomsky's major references, in Cartesian Linguistics, are to the Logi cand General and Reasoned Grammar of Port-Royal, works that Rousseau knew well and held in high esteem.3 For example, on several occasions Rousseau cites Duclos' commentary on the General and Reasoned Grammar. The Essay on the Origin of Languages even closes with one of these citations. Thus Rousseau acknowledges his debt.1. Maurice Grammont, cited by Noam Chomsky, Cartesian Linguistics , p. 1.2. Chomsky, "De quelques constantes de la théory linguistique," Diogène, no. 51 ; my italics. See also Chomsky, Current Issues in Linguistic Theory , p.15 ff. There is an analogous gesture in Jakobson, who refers not only to Peirce and, as does Chomsky, to Humboldt but also to John of Salisbury, to the Stoics, and to Plato's Cratylus: see Jakobson, "A la recherche de l'essence du langage," Diogène, no. 51 .3. "I began with some book of philosophy, like the Port-Royal Logic, Locke's Essay, Malebranch, Leibniz, Descartes, etc." .Jacques Derrida, professor of the history of philosophy at the Ècole Normale Supérieure in Paris, is the author of, among other works, Writing and Difference, Of Grammatology, and Marges de la philosophie, from which the present essay is taken. His previous contribution to Critical Inquiry, "The Law of Genre," appeared in the Autumn 1980 issue. Alan Bass, a psychoanalyst, has published essays on deconstruction and psychoanalytic theory and practice