The Notion of Literature Tzvetan Todorov (bio) Before launching into the awesome question of the nature of literature, let us begin, as a precautionary measure, by examining not literature itself but rather the kind of discourse which, like this very study, takes literature as its object. The difference will be one of approach rather than of objective; but who can say whether the course of the inquiry is not of greater interest than its final results? * * * We must first cast a doubt upon the legitimacy of the very notion of literature; neither the mere existence of the term, nor the fact that a whole university system is based upon it, can of itself justify its acceptance. The first grounds for doubt are of an empirical nature. No complete history of the word and its equivalents, in all languages and throughout time, has yet been undertaken; yet even a superficial inquiry into the question suffices to convince us that the term has not always been with us. In the European languages the word literature, in its present usage, is quite recent: it barely dates back to the nineteenth century. Could it be that we are dealing with an historical and not at all an "eternal" phenomenon? Moreover, many languages (those of Africa, for example) still have no generic term to designate literature as a whole; and while Lévy Bruhl would have sought to explain this absence by the so called primitive nature of these languages allegedly incapable of abstraction and hence devoid of any words designating the general rather than the specific, the time when we could accept such an explanation has long passed. Finally, we must also take into account the diversification of literature in our own countries; who would dare decide today what is literature and what is not, given the irreducible diversity of all the written works which, from infinitely different perspectives, tend to be regarded as literature? This argument is not decisive: a notion may be legitimate even though no corresponding word may as yet exist to designate it; it does, however, create a first element of doubt as to the "natural" character of literature. Nor is a theoretical examination of the problem more reassuring. Whence [End Page 1] do we derive the certainty that an entity such as literature really exists? From experience: we meet literary works in school, then in college; we find them in certain specialized stores; references to "literary authors" crop up constantly in our everyday conversations. That an entity called "literature" does function on an intersubjective and social level seems indeed unquestionable. Agreed. But what does this prove? That in a larger system—a society, a civilization—there exists an identifiable element referred to as literature. But does it prove that all the individual works grouped under this heading partake of a common nature which we can identify with equal justification? Let us call "functional" our first definition of this entity—the definition which identifies it in terms of what it "does" as an element in a larger system; and "structural" our second, whereby we seek to test whether all the individual works collectively regarded as literature in the functional meaning of the word partake of the same characteristics. The distinction between the functional and the structural points of view should be rigorously kept in mind, even though to pass from one to the other is perfectly permissible. In order to illustrate this distinction, let us take the example of advertising: its precise function within our society is undeniably clear; but what of its structural identity? It can express itself though the visual and auditory, as well as other media; it may or may not have a duration in time; it may be continuous or discontinuous; it may use techniques as varied as direct inducement, description, allusion, antiphrasis, and so on. The unquestionable functional entity—assuming for the moment that it is indeed unquestionable—does not necessarily have a corresponding structural entity. The one need not necessarily imply the other, although the affinities between them can be easily observed. The difference is more in the point of view than in the thing itself: if literature (or advertising) is discovered to be...