From Work to Text

Roland Barthes

[Rptd. in Josue V. Harari, ed., Textual Strategies: Perspectives in Poststructuralist Criticism (Ithaca, NY: Cornell UP, 1979), 73-81.]

Over the past several years, a change has been taking place in our ideas about language and, as a consequence, about the (literary) work, which owes at least its phenomenal existence to language. This change is obviously linked to current developments in, among other fields, linguistics, anthropology, Marxism, and psychoanalysis (the word "link" is used here in a deliberately neutral fashion: it implies no decision about a determination, be it multiple and dialectical). The change affecting the notion of the work does not necessarily come from the internal renewal of each of these disciplines, but proceeds, rather, from their encounter at the level of an object that traditionally depends on none of them. Interdisciplinary activity, valued today as an important aspect of research, cannot be accomplished by simple confrontations between various specialized branches of knowledge. Interdisciplinary work is not a peaceful operation: it begins effectively when the solidarity of the old disciplines breaks down--a process made more violent, perhaps, by the jolts of fashion--to the benefit of a new object and a new language, neither of which is in the domain of those branches of knowledge that one calmly sought to confront.

It is precisely this uneasiness with classification that allows for the diagnosis of a certain mutation. The mutation that seems to be taking hold of the idea of the work must not, however, be overestimated: it is part of an epistemological shift [glissement] rather than of a real break [coupure], a break of the kind which, as has often been remarked, supposedly occurred during the last century, with the appearance of Marxism and Freudianism. No new break seems to have occurred since, and it can be said that, in a way, we have been involved in repetition for the past hundred years. Today history, our history, allows only displacement, variation, going-beyond, and rejection. Just as Einsteinian science requires the inclusion of the relativity of reference points in the object studied, so the combined activity of Marxism, Freudianism, and structuralism requires, in the case of literature, the relativization of the scriptor's, the reader's, and the observer's (the critic's) relationships. In opposition to the notion of the work--a traditional notion that has long been and still is thought of in what might be called Newtonian fashion--there now arises a need for a new object, one obtained by the displacement or overturning of previous categories. This object is the Text. I realize that this word is fashionable and therefore suspect in certain quarters, but that is precisely why I would like to review the principal propositions at the intersection of which the Text is situated today. These propositions are to be understood as enunciations rather than arguments, as mere indica- tions, as it were, approaches that "agree" to remain metaphoric. Here, then, are those propositions: they deal with method, genre, the sign, the plural, filiation, reading (in an active sense), and pleasure.

(1) The Text must not be thought of as a defined object. It would be useless to attempt a material separation of works and texts. One must take particular care not to say that works are classical while texts are avant-garde. Distinguishing them is not a matter of establishing a crude list in the name of modernity and declaring certain literary productions to be "in" and others "out" on the basis of their chronological situation. A very ancient work can contain "some text," while many products of contemporary literature are not texts at all. The difference is as follows: the work is concrete, occupying a portion of book-space (in a library, for example); the Text, on the other hand, is a methodological field.

This opposition recalls the distinction proposed by Lacan between "reality" and the "real": the one is displayed, the other demonstrated. In the same way, the work can be seen in bookstores, in card catalogues, and on course lists, while the text reveals itself, articulates itself according to or against certain rules. While the work is held in the hand, the text is held in language: it exists only as discourse. The Text is not the decomposition of the work; rather it is the work that is the text's imaginary tail. In other words, the Text is experienced only in an activity, a production. It follows that the Text cannot stop, at the end of a library shelf, for example; the constitutive movement of the Text is a traversal [traversee]: it can cut across a work, several works.

(2) Similarly, the Text does not come to a stop with (good) literature; it cannot be apprehended as part of a hierarchy or even a simple division of genres. What constitutes the Text is, on the contrary (or precisely), its subversive force with regard to old classifications. How can one classify Georges Bataille? Is this writer a novelist, a poet, an essayist, an economist, a philosopher, a mystic? The answer is so uncertain that manuals of literature generally chose to forget about Bataille; yet Bataille wrote texts--even, perhaps, always one and the same text.

If the Text raises problems of classification, that is because it always implies an experience of limits. Thibaudet used to speak (but in a very restricted sense) about limit-works (such as Chateaubriand's Life of Rance, a work that today indeed seems to be a "text"): the Text is that which goes to the limit of the rules of enunciation (rationality, readability, and so on). The Text tries to situate itself exactly behind the limit of doxa (is not public opinion--constitutive of our democratic societies and powerfully aided by mass communication--defined by its limits, its energy of exclusion, its censorship?). One could literally say that the Text is always paradoxical.

(3) Whereas the Text is approached and experienced in relation to the sign, the work closes itself on a signified. Two modes of signification can be attributed to this signified: on the one hand, one can assume that it is obvious, in which case the work becomes the object of a "science of the letter" (philology); on the other hand, one can assume that the signified is secret and ultimate, in which case one must search for it, and the work then depends upon a hermeneutic, an interpretation (Marxist, psychoanalytic, thematic, for example). In brief, the work itself functions as a general sign and thus represents an institutional category of the civilization of the Sign. The Text, on the contrary, practices the infinite deferral of the signified [le recul infini du signifie]: the Text is dilatory; its field is that of the signifier. The signifier must not be conceived as "the first stage of meaning," its material vestibule, but rather, on the contrary, as its aftermath [apres-coup]. In the same way, the signifier's infinitude does not refer back to some idea of the ineffable (of an unnamable signified) but to the idea of play. The engendering of the perpetual signifier within the field of the text should not be identified with an organic process of maturation or a hermeneutic process of deepening, but rather with a serial movement of dislocations, overlappings, and variations. The logic that governs the Text is not comprehensive (seeking to define "what the work means") but metonymic; and the activity of associations, contiguities, and cross-references coincides with a liberation of symbolic energy. The work (in the best of cases) is moderately symbolic (its symbolism runs out, comes to a halt), but the Text is radically symbolic. A work whose integrally symbolic nature one conceives, perceives, and receives is a text.

In this way the Text is restored to language: like language, it is structured but decentered, without closure (here one might note, in reply to the scornful insinuation of "faddishness" which is often directed against structuralism, that the epistemological privilege presently granted to language proceeds precisely from our discovery in language of a paradoxical idea of structure, a system without end or center).

(4) The Text is plural. This does not mean just that is has several meanings, but rather that it achieves plurality of meaning, an irreducible plurality. The Text is not coexistence of meanings but passage, traversal; thus it answers not to an interpretation, liberal though it may be, but to an explosion, a dissemination. The Text's plurality does not depend on the ambiguity of its contents, but rather on what could be called the stereographic plurality of the signifiers that weave it (etymologically the text is a cloth; textus, from which text derives, means "woven").

The reader of the Text could be compared to an idle subject (a subject having relaxed his "imaginary" [1]): this fairly empty subject strolls along the side of a valley at the bottom of which runs a wadi (I use wadi here to stress a certain feeling of unfamiliarity). What he sees is multiple and irreducible; it emerges from substances and levels that are heterogeneous and disconnected: lights, colors, vegetation, heat, air, bursts of noise, high-pitched bird calls, children's cries from the other side of the valley, paths, gestures, clothing of close and distant inhabitants. All these occurrences are partially identifiable: they proceed from known codes, but their combination is unique, founding the stroll in difference that can be repeated only as difference. This is what happens in the case of the Text: it can be itself only in its difference (which does not mean its "individuality"); its reading is semelfactive (which renders all inductive-deductive sciences of texts illusory--there is no "grammar" of the text) and yet completely woven with quotations, references, and echoes. These are cultural languages (and what language is not?), past or present, that traverse the text from one end to the other in a vast stereophony.

Every text, being itself the intertext of another text, belongs to the intertextual, which must not be confused with a text's origins: to search for the "sources of" and "influence upon" a work is to satisfy the myth of filiation. The quotations from which a text is constructed are anonymous, irrecoverable, and yet already read: they are quotations without quotation marks. The work does not upset monistic philosophies, for which plurality is evil. Thus, when it is compared with the work, the text might well take as its motto the words of the man possessed by devils: "My name is legion, for we are many" (Mark 5:9).

The plural or demonic texture that divides text from work can carry with it profound modifications in the activity of reading and precisely in the areas where monologism seems to be the law. Some of the "texts" of the Scriptures that have traditionally been recuperated by theological (historical or anagogical) monism may perhaps lend themselves to a diffraction of meaning, while the Marxist interpretation of the work, until now resolutely monistic, may be able to materialize itself even further by pluralizing itself (if, of course, Marxist "institutions" allow this).

(5) The work is caught up in a process of filiation. Three things are postulated here: a determination of the work by the outside world (by race, then by history), a consecution of works among themselves, and an allocation of the work to its author. The author is regarded as the father and the owner of his work; literary research therefore learns to respect the manuscript and the author's declared intentions, while society posits the legal nature of the author's relationship with his work (these are the "author's rights," which are actually quite recent; they were not legalized in France until the Revolution).

The Text, on the other hand, is read without the father's signature. The metaphor that describes the Text is also distinct from that describing the work. The latter refers to the image of an organism that grows by vital expansion, by "development" (a significantly ambiguous word, both biological and rhetorical). The Text's metaphor is that of the network: [2] if the Text expands, it is under the effect of a combinatorial, a systematics [3] (an image which comes close to modern biology's views on the living being).

Therefore, no vital "respect" is owed to the Text: it can be broken (this is exactly what the Middle Ages did with two authoritative texts, the Scriptures and Aristotle). The Text can be read without its father's guarantee: the restitution of the intertext paradoxically abolishes the concept of filiation. It is not that the author cannot "come back" into the Text, into his text; however, he can only do so as a "guest," so to speak. If the author is a novelist, he inscribes himself in his text as one of his characters, as another figure sewn into the rug; his signature is no longer privileged and paternal, the locus of genuine truth, but rather, ludic. He becomes a "paper author": his life is no longer the origin of his fables, but a fable that runs concurrently with his work. There is a reversal, and it is the work which affects the life, not the life which affects the work: the work of Proust and Genet allows us to read their lives as a text. The word "bio-graphy" reassumes its strong meaning, in accordance with its etymology. At the same time, the enunciation's sincerity, which has been a veritable "cross" of literary morality, becomes a false problem: the I that writes the text is never, itself, anything more than a paper I.

(6) The work is ordinarily an object of consumption. I intend no demagoguery in referring here to so-called consumer culture, but one must realize that today it is the work's "quality" (this implies, ultimately, an appreciation in terms of "taste") and not the actual process of reading that can establish differences between books. There is no structural difference between "cultured" reading and casual subway reading. The Text (if only because of its frequent "unreadability") decants the work from its consumption and gathers it up as play, task, production, and activity. This means that the Text requires an attempt to abolish (or at least to lessen) the distance between writing and reading, not by intensifying the reader's projection into the work, but by linking the two together in a single signifying process [pratique signifiante].

The distance separating writing from reading is historical: during the era of greatest social division (before the institution of democratic cultures), both reading and writing were class privileges. Rhetoric, the great literary code of those times, taught writing (even though speeches and not texts were generally produced). It is significant that the advent of democracy reversed the order: (secondary) school now prides itself on teaching how to read (well), and not how to write.

In fact, reading in the sense of consuming is not playing with the text. Here "playing" must be understood in all its polysemy. The text itself plays (like a door on its hinges, like a device in which there is some "play"); and the reader himself plays twice over: playing the Text as one plays a game, he searches for a practice that will re-produce the Text; but, to keep that practice from being reduced to a passive, inner mimesis (the Text being precisely what resists such a reduction), he also plays the Text in the musical sense of the term. The history of music (as practice, not as "art") happens to run quite parallel to the history of the Text. There was a time when "practicing" music lovers were numerous (at least within the confines ofa certain class), when "playing" and "listening" constituted an almost undifferentiated activity. Then two roles appeared in succession: first, that of the interpreter, to whom the bourgeois public delegated its playing; second, that of the music lover who listened to music without knowing how to play it. Today, post-serial music has disrupted the role of the "interpreter" by requiring him to be, in a certain sense, the co-author of a score which he completes rather than "interprets."

The Text is largely a score of this new type: it asks the reader for an active collaboration. This is a great innovation, because it compels us to ask "who executes the work?" (a question raised by Mallarme, who wanted the audience to produce the book). Today only the critic executes the work (in both senses). The reduction of reading to consumption is obviously responsible for the "boredom" that many people feel when confronting the modern ("un-readable") text, or the avant-garde movie or painting: to suffer from boredom means that one cannot produce the text, play it, open it out, make it go.

(7) This suggests one final approach to the Text, that of pleasure. I do not know if a hedonistic aesthetic ever existed, but there certainly exists a pleasure associated with the work (at least with certain works). I can enjoy reading and rereading Proust, Flaubert, Balzac, and even--why not?--Alexandre Dumas; but this pleasure, as keen as it may be and even if disengaged from all prejudice, remains partly (unless there has been an exceptional critical effort) a pleasure of consumption. If I can read those authors, I also know that I cannot rewrite them (that today, one can no longer write "like that"); that rather depressing knowledge is enough to separate one from the production of those works at the very moment when their remoteness founds one's modernity (for what is "being modern" but the full realization that one cannot begin to write the same works once again?). The Text, on the other hand, is linked to enjoyment [jouissance], to pleasure without separation. Order of the signifier, the Text participates in a social utopia of its own: prior to history, the Text achieves, if not the transparency of social relations, at least the transparency of language relations. It is the space in which no one language has a hold over any other, in which all languages circulate freely.

These few propositions, inevitably, do not constitute the articulation of a theory of the Text. This is not just a consequence of the presenter's insufficiencies (besides, I have in many respects only recapitulated what is being developed around me); rather, it proceeds from the fact that a theory of the Text cannot be fully satisfied by a metalinguistic exposition. The destruction of metalanguage, or at least (since it may become necessary to return to it provisionally) the questioning of it, is part of the theory itself. Discourse on the Text should itself be only "text," search, and textual toil, since the Text is that social space that leaves no language safe or untouched, that allows no enunciative subject to hold the position ofjudge, teacher, analyst, confessor, or decoder. The theory of the Text can coincide only with the activity of writing.


1. "Qui aurait detendu en lui tout imaginaire·" Imaginary is not simply the opposite of real. Used in the Lacanian sense, it is the register, the dimension of all images, conscious or unconscious, perceived or imagined.--Ed. (Josue V. Harari)

2. Barthes uses here the word reseau. I have chosen to translate it by "network" (rather than "web," for instance) at the risk of overemphasizing the mechanical implications of the metaphor.--Ed.

3. Systematics is the science (or method) of classification of living forms.--Ed.