SOURCE: Metzidakis, Stamos. “Barthes's Image.”¹Neophilologus 71, no. 4 (October 1987): 489-95.

[In the following essay, Metzidakis traces a change in Barthes's use of the term “image” in his writings, asserting that it corresponds to a change of attitude in his critical thinking.]

Roland Barthes did more to change the way literature is read and taught than perhaps any other French critic of the last thirty years. Having begun his career with the simple desire to write, he passed through successive phases of aesthetic and intellectual evolution during which he was—to borrow his terms—a social mythologist, semiologist, textual critic, and finally, moralist.² Barthes has thus come to be regarded by many as something of a protean figure of modern critical thought. However, the images we have of him are so elusive that, often, we know neither what to call him, nor what to call the activities he performed at various moments of his life.

A similar problem of classification arises when we turn our attention to several other contemporary critics, and critical approaches to literature. What, for example, should we call a Michel Foucault, or a Jacques Derrida (to cite but two of the more problematic cases)? Are they philosophers or historians, archaeologists or social scientists, structuralists or post-structuralists? Since each of these questions deserves a more extensive treatment than we can give them in this essay, a more efficient way for us to proceed in dealing with them is to focus our attention on one figure who appears exemplary. Therefore, we shall limit our scope to the specific case of Roland Barthes, whose influence was greater, and more pervasive in the recent past than that of any other critic.

I would submit, though I am certainly not alone, that the single most important factor responsible for our terminological hesitation is the massive importation into literary studies of hermeneutic models from diverse areas of inquiry, areas like structural linguistics, psychoanalysis, and sociology. This appropriation by scholars of what one might call “foreign” models has given rise to a new field of intellectual investigation. Jonathan Culler, taking Richard Rorty's lead, has recently named this conceptual domain, “theory.”³ However, except for Culler's deceptively simple label, this new genre is, in fact, exceedingly difficult to characterize. Being neither literary theory per se, nor philosophy, nor even textual theory, “theory” would appear to be as hard to define as was that earlier catch-all term, “structuralism.” The latter, of course, has been used for years to denote many dissimilar interpretive stances which are considered untraditional, for one reason or another. (And no one in France during the heyday of structuralism was more untraditional, or more associated with the structuralist movement, than Roland Barthes.)

But, from the very start of his discussion of this new “genre,” Culler seems reluctant to define structuralism exactly. He insists instead that the time has come to stop distinguishing traditional scholars from their socalled structuralist counterparts, and to try to determine how structuralists differ from post-structuralists. He suggests that even though the earlier label has been applied with varying degrees of profit to thinkers who, in fact, do many different kinds of analysis, there seems to have emerged from within the “common” ranks of all these untraditional critics, two diametrically opposed types of readers. Culler emphasizes that the epistemological differences between these two readers are far more pertinent to modern scholars than any anachronistic debate over whether someone is a structuralist or not. Succinctly stated, the differences are that, “structuralists are convinced that systematic knowledge is possible; post-structuralists claim to know only the impossibility of this knowledge” (p. 22).

By examining the specific case of Roland Barthes' work, the present paper attempts to call into question the legitimacy of this distinction. My intention is not to deny the personal, ideological rifts which often occur between proponents of structuralism and those of post-structuralism. I should, however, like to suggest that classifications of certain individuals into one or the other camp may depend, in the final analysis, more on our own image of them, or on their own self-image, than on what they actually do. In this sense, my analysis should be seen as an attempt to rectify some recent taxinomic errors, and to recover a certain continuity in the fundamental procedures, and presuppositions of much modern thought.

In fairness to Professor Culler, it should be noted from the start that he, too, quickly rejects the dichotomy structuralist/post-structuralist in favor of the above distinction between readings and readers. The latter seems to him to be of much greater practical value to the modern theorist. Moreover, he is fully aware of how hard a task it is to classify Barthes, in particular. He admits that Barthes could be said to have belonged to both camps throughout his entire career (p. 30). Nevertheless, because post-structuralism has, for whatever historical and ideological reasons, become accepted as a major component of “theory,” it is essential for us to see how one of its most influential representatives⁴ actually undermines the very specificity of this conceptual domain. In the process, we will be forced to reconsider the common belief that certain structuralist tendencies have been transcended by great numbers of modern theorists.

Now, as if the problem of describing Barthes were not challenging enough a priori, it is merely exacerbated when one realizes how much he himself opposed the idea of qualifying people and things with labels. Since he always preferred “functions” to “meanings,” the thought of reifying his or anyone else's essence with a label was, doubtless, anathema to him. In this regard, we would do well to consider the following statement he made concerning the adjective, that part of speech whose very grammatical function it is to describe and affix essences: “a relationship which adjectivizes is on the side of the image (emphasis mine), on the side of domination, of death” (R.B. [Roland Barthes by Roland Barthes], p. 43).

Yet, if Barthes had been consistent in opposing any attribution of qualities to things or people (including himself) through use of images, then how does it happen, one must ask, that the term “image” itself recurs so frequently in his many texts? What is there about the concept and function image which merits such a harsh characterization by Barthes, and, at the same time, seems so irresistable and unavoidable to him? A close examination of his use of the term in several representative pieces from the phases mentioned above will provide a few answers to these questions. By means of this examination, I propose to show that Barthes never really changed his basic approach to the experience and interpretation of the world. Even though he thought differently about what he was doing, and about what he stood for at various points in his life, he never really altered his activity. An accurate image of Barthes should, therefore, include an extensive appraisal both of his image-repertory, and of the way he treats the concept/term “image” itself.

Writing Degree Zero (1953) is perhaps best known as Barthes's earliest attempt to define a rather complex notion of writing or écriture. This notion was to remain at the heart of much critical discussion in France, and elsewhere, throughout most of the 1950's and 60's. Among the numerous definitions of écriture advanced in the book, one of the first we find reads:

… writing is a hardened language which lives on itself and does not have the responsibility at all to entrust a mobile succession of approximations to its duration, but, on the contrary, to impose by the unity and shadow of its signs, the image (emphasis mine) of a speech constructed well before being invented.⁵

Even at this beginning stage of Barthes's theoretical and critical production, we see that he already depends on the word “image” to describe a concept which was inarticulated, and probably unrecognized, by critics before that time. In this first major work, “image” signifies a kind of reflection—as in a specular image—of what Barthes calls a “foreign circumstance” pre-dating any particular speech. This image or reflection constitutes the linguistic vehicle by which Barthes tries to express to his readers something abstract (écriture). He thereby tries to make more clear, more intelligible, what might have otherwise remained unspoken, hence, unknown. As he will say four years later in Mythologies:

Every object in the world can pass from a closed, silent existence to an oral state, open to appropriation by society, for there is no law, whether natural or not, which forbids talking about things.⁶

Thus, it is in this “talking about things” that we find Barthes utilizing the word “image” from the very start of his career. He utilizes the word in order to say that which is essentially unsayable. His reliance on it should be seen as an important prefiguration of his later inability to analyze other phenomena structurally, without recourse to analogous terms. What makes this instance of “image” important here is that it underscores Barthes's continual preoccupation as critic, theorist, and writer with the problem of how to render intelligible the specificity of writing, social myths, photographs, literary texts, pleasure, even himself. If the concept/term image reappears often in his writings, one must therefore suppose this to be the result of Barthes's fundamental incapacity to find a better vehicle with which to describe the various phenomena.

But, any explanation of writing, myths, and the like should be construed in Barthes's early work, at least, as meaning an explanation of their structures. To clarify what he meant by “structure,” let us look carefully at his 1963 essay entitled “The Structuralist Activity.”⁷ According to Barthes, structural man, or homo significans, aims at reconstructing an object under investigation, “in such a way as to manifest thereby the rules of functioning (the functions) of this object” (p. 214). The reconstructed object, in his view, takes precedence over, and even takes the place of, the object that was originally chosen for study. In this manner, the reconstruction, or better still, the “simulacrum,” or “imitation,” becomes the observer's real focus of attention. Structuralism in its most general sense can thus be defined as “an activity of imitation” (p. 215).

Now, as is pertinent for this discussion, Barthes points out that the word “image” should be linked to the Latin root imitari.⁸ Any act of imitation or repetition is consequently an instance of an image. Images repeat, imitate, re-present, or “resurrect” things which, by definition, they are not. In the process of imitation, the image comes to form the very “limit of meaning.” That is, meaning is contingent on the image itself. Meaning or, more precisely, the fabrication of meaning (signification) is the direct result of the repetitive process. It derives from the gap—Derrida would say the “difference”—between the object represented and its image. As Barthes says in the original preface to Mythologies: “I don't know whether, as the saying goes, “things which are repeated are pleasing,” but my belief is that they are significant.” (p. 12).

The act of imitation, or what was just called the “fabrication of meaning,” can be divided into two major operations or phases: dissection and articulation (“Activity,” p. 216). These operations are Barthes's versions of the now classic distinction drawn by Saussure between the paradigmatic and syntagmatic axes of language. In the first phase, the observer cuts up into mobile fragments an object to be discussed, analyzed, or otherwise given to the “simulacrum-activity” (p. 215). Barthes emphasizes that the individual units in question are not anarchic, but are governed by a conceptual paradigm in which they reside. They are then placed, during the second phase or operation, into a new kind of association with each other. The rules governing this procedure are, of course, difficult to determine because, to quote once again from Barthes, “The syntax of the arts and of discourse is, as we know, extremely varied …” (p. 217). What counts, however, is that the rules or constraints chosen allow the new object, the simulacrum, the copy, the image, in a word, to remain stable and to function.

If, for instance, the essay “The World of Wrestling” in Mythologies succeeds in locating any “meaning” in wrestling, this is because the dissected parts of that societal ritual have been fitted back together on the basis of the formal rule or paradigm, Good vs. Evil. The rule permits the “functional” category of the phenomenon (i.e., the wrestling plus its “myth” or speech) to remain stable for the observer. Without such stability, it would simply not work, which is to say that it would have no “social usage” (Mythologies, p. 109). Neither would it function, nor literally signify anything to anyone.

Given this view of structuralist activity, we must now ask ourselves whether Barthes ever really gave it up. Since our interest here lies not in what he said or believed he was doing, but rather in how he was accomplishing “it,” the question of his desire to abandon a certain brand of structuralism is obviously irrelevant. As Jonathan Culler indicates elsewhere,⁹ Barthes was certainly aware of a change in structuralism that corresponded to his passage from “Introduction to the Structural Analysis of Narratives” (1966) to S/Z (1970). This change consisted of a rejection of any transcendental narrative models à la Propp or Lévi-Strauss, and his subsequent postulation that every text institutes in some sense “its own model.”

But, insofar as Barthes retains this notion of model, one must still assume that its concomitant copy exists somewhere else in his work. The model/copy pairing, of course, reinvokes the entire problematic of repetition, imitation, and images already established in his “pure” structuralist phase. To be sure, Barthes seems intent on abolishing this problematic when he says at the start of S/Z that a choice has to be made with regard to texts. It is necessary either to force them

… to rejoin, inductively, the Copy from which we will then make them derive; or else to restore each text, not to its individuality, but to its function, making it cohere, even before we talk about it, by the infinite paradigm of difference …¹⁰

Yet, what all this imagery points to is that, for some reason—ideological pressure?—Barthes wanted to have his cake, and to eat it too. For let us not forget that to restore to a text its “function”, and to make it cohere, were among the basic tasks of the structuralist activity. Moreover, it is paradoxical to speak of a “paradigm of difference.” By definition repetitive, a linguistic paradigm accepts differences only insofar as it negates them, and puts into relief their common conceptual denominator.

One gets the impression, therefore, that Barthes tried very hard to assimilate a “post-structuralist” attitude¹¹ against representation, even though he apparently could not rid his style of its constituent parts, model/copy. Clearly, this fact bothered him on some level. In Roland Barthes by Roland Barthes, for example, he asks his imaginary self (whose words are meant to be “those of a character in a novel,” let us remember):

Who is still a structuralist? Yet he is one in this, at least: a uniformly noisy place seems to him unstructured because in this place there is no freedom left to choose silence or speech … how on such a day can I give a meaning to my silence, since, in any case, I cannot speak?

(p. 117)

Eighteen years after Mythologies, this passage recalls the idea that one has to choose between two ways of dealing with the world: one either gives meaning to otherwise silent phenomena, i.e. fabricates images of them through imitation, representation, and the like; or else, one leaves them silent. I am, therefore, tempted here to ask Barthes an obvious counter-question to his initial inquiry, one that might read as follows: “Have you as yet become a post-structuralist?”

It is at this point necessary to admit that one thing does indeed change in Barthes' use of the term “image” in the later phase of his work. As suggested above, the change is in his attitude towards it. The longer he writes, the more he takes pleasure in the images he employs to render the world intelligible. Let us consider, for example, the opening line from Roland Barthes by Roland Barthes: “To begin with, some images (emphasis mine): they are the author's treat to himself, for finishing his book.” Significantly, the pleasure which the image now procures for him is simultaneous to the uneasiness he felt in the knowledge we alluded to earlier: to wit, that the image was “on the side of death.”

From the late 1960's to the end of his career, however, Barthes began to insist that he no longer meant “representation” or “copy” when he spoke about an image. As he explains in one of his last books, “the image is that from which I am excluded.”¹² Of course, we must not forget that within this context Barthes was attempting to voice the “untreatable” in the language as if it could be staged (mise en scène). His intention was not to explain a discourse, but to speak it, as it were. To say that this or any other discourse is excluded from any image is nevertheless tantamount to producing another image! In essence, Barthes presents the un-imaginable in what can be understood solely as an image. This is why he can not help but fail to escape performing an essentially structuralist activity. For even in the introduction to A Lover's Discourse, where he tells about the way the book is written, Barthes says that he intends to substitute for the description or analysis of this discourse, its simulation. With that, we are back to simulacrum activity.

I shall add a final note on one other important displacement (not rejection) that we can trace in the evolution of Barthes's conception and usage of the image. Because the main emphasis in his work gradually shifted from the text to the reader, as with the whole of literary theory and criticism, it was perhaps inevitable that the notion of his own image became the ultimate focus of his attention. After all, it was always Barthes who projected himself onto the earlier objects which he analyzed in a “truly” structuralist manner. He never once pretended to attribute anything other than his own meanings to the things he studied. To quote a last time from Mythologies: “Is there not a mythology of the mythologist?” (p. 12). As we are now in a position from which to answer this question affirmatively, let us conclude by saying the following. For the outside observer, there will always be an image of the image-producer. And this image, in its turn, is independent of the latter's self-image.¹³

Notes

1. An earlier draft of this article was written for the Faculty Seminar on Roland Barthes that took place at Washington University, St. Louis during Spring 1984. It was also delivered as a lecture in a special session on the literary image held at the Modern Language Association Convention in Washington, D.C. on December 28, 1984.

2. Roland Barthes by Roland Barthes, trans. Richard Howard (New York: Hill and Wang, 1977), p. 145. Further references to this work are noted in the text by the letters R.B. and the page number.

3. On Deconstruction, Theory and Criticism after Structuralism (Ithaca: Cornell University Press, 1982), p. 8. Further references to this work are included in the body of the text with page numbers.

4. Since Barthes appears first in Josué Harari's important and influential collection of post-structuralist essays entitled Textual Strategies: Perspectives in Post-Structuralist Criticism (Ithaca: Cornell University Press, 1979), it is clear that he is considered to be a post-structuralist by certain people.

5. Le degré zéro de l'écriture (Paris: Seuil, 1953), p. 19. All translations are mine.

6. Mythologies, trans. Annette Lavers (New York: Hill and Wang, 1972), p. 109. Further references noted in text.

7. In his Critical Essays, trans. Richard Howard (Evanston: Northwestern University Press, 1972), p. 213-220. Hereafter cited as “Activity,” followed by page number in text.

8. In his Image, Music, Text, trans. Stephen Heath (New York: Hill and Wang, 1977), p. 32.

9. See his Structuralist Poetics (Ithaca: Cornell University Press, 1975), p. 242.

10. S/Z, trans. Richard Miller (New York: Hill and Wang, 1974), p. 3.

11. I adapt the idea of such an attitude from Harari's Textual Strategies at the beginning of which is reproduced Barthes's article “From Work to Text.” Harari underlines a crucial fact in relation to this structuralist/post-structuralist dichotomy:

    The denunciation of the concept of representation is necessarily based on the structuralist institution of the sign; it relies on structuralist premises in order, paradoxically, to show that structuralism has not fully pursued the implications of those premises. The post-structuralist attitude (emphasis mine) is therefore literally unthinkable without structuralism (p. 30).

12. Fragments d'un discours amoureux (Paris: Seuil, 1977), p. 157. Translations are mine.

13. The essential distinction I have tried to make here between critical attitudes and critical activities recalls a similar distinction made by Philip Lewis. In his article, “The Post-Structuralist Condition,” Lewis distinguishes post-structuralism from a so-called “post-structuralist condition.” See Diacritics, vol. 12 (1982), p. 2-24, esp. 22-23.