Go to Cyber Semiotic Institute home page
Go to Course Outline
Go to Lecture: One or Two or Four or Five or Six or Seven or Eight

Critical Semiotics

Instructor: Scott Simpkins

Lecture Three: The Implications of Codes

Assigned Readings:

Roland Barthes, S/Z: An Essay, Trans. Richard Miller (New York: Hill and Wang, 1974).


"...all codes are finally coercive..."
Robert Scholes, Semiotics and Interpretation

The Great Code

Despite its considerable shortcomings, S/Z has been extremely influential in many discussions in the human sciences, especially those inclined toward structuralist methodologies. It is particularly attractive because it not only outlines an extensive theoretical position, but it demonstrates an application of it as well. For semioticians, furthermore, it illustrates strategies for identifying (or perhaps, more accurately, constructing or delegating) codes relevant, and maybe even intrinsic, to a given social system. And, again, it reveals how a specific decoder might go about employing them when attempting to assess and comprehend the significant components of that system.

For the purposes of a critical semiotics, Barthes' text additionally illuminates the ways in which those interested in assessing the signs of an entity can naturalize their endeavor. To create, in other words, the illusion that semiotics exists as an established, homogeneous discipline with a universally accepted conceptual nomenclature.

The notion of the "code" itself, a concept borrowed from information theory, begs the question of such a phenomenon, particularly when it is camouflaged as the root of common terms like "encoder" and "decoder". After all, what is a code? It is sociologically revealing to survey the ways in which code theorists so frequently rely upon remarkably similar concepts in their discussions about the "code" yet give very little scrutiny to the attendant implications.

Perhaps the most revealing portrayal of potentially underlying implications of the "code," however, comes from Northrop Frye's commentary on the Bible as "the Great Code" (a notion he adopts from the British poet, William Blake). Frye argues that the Bible "has traditionally been read as a unity" and claims that it reveals "some traces of a total structure" (xiii). In addition to possessing "a beginning and an end," it has "a body of concrete images" that "clearly indicate[s] some kind of unifying principle." The Great Code that Frye identifies is, he says, "a unified structure of narrative and imagery." Frye's descriptions of the Bible in this sense all hinge on an intelligible system that operates as a whole. This intelligibility, presumably, can be decoded because it is encoded. Or rather, it operates according to a system of determined or organized codes.

Frye's use of "code" is largely consonant with its depiction usually offered by the code theorists discussed here. Roland Champagne, for instance, considers a code as "a network of ideas, images, and stylistic devices that have an internal cohesive principle" (35). Manfred Frank similarly employs a figuration based on regimentation: "One masters the [sign] system in question through being able to pass beyond its expressions to their significations while conforming to the rules, in other words, to recognize these expressions as signs" (158). It can even be taken so far as to embody an agency of governance.

Umberto Eco defines "code" as "a system of rules that would involve a fixed number of elements and that would exclude some combinations while allowing others" (Open Work, 56). For Robert Hodge and Gunther Kress, it is something governed by one or more "terms" (111) that enable it to "carry" (109) significance, presumably either away from the encoder or toward the decoder. Basil Bernstein, in Class, Codes and Control, views codes from the standpoint of power and the ways in which they function as "dominant principles of interpretation" (3.24) or regulative "social principles" (3.30). In this way, a code constitutes, or acts as, an "underlying regulative principle" of an entity (1.8). Keir Elam also views codes as "rules determining the encoding or decoding of texts" (52). Additionally, it is useful to consider several code-related genres and actions identified by Elam ("code rules" [62], "code-observing, code-making and code-breaking" [55]) and Roman Jakobson ("code signs" ["Aspects," 102] and "code units" ["Translation," 430]).

Robert Scholes says that "there are rules governing text production and interpretation" (Semiotics, 1) and "the semiotic [literary] critic" "looks for the generic or discursive structures that enable and constrain meaning." From this perspective, the code could be seen, metaphorically, as the ground on which communication is erected. In the case of Jakobson's model (also appropriated from information theory), codes are visually portrayed as functioning like the fulcrum of a lever, or perhaps more fittingly, a fulcrum for a seesaw as signs are passed back and forth between encoder and decoder ("Linguistics," 66):

This is reinforced by Jakobson's definition of "code" as information "fully, or at least partially, common to the addresser and addressee (or in other words, to the encoder and decoder of the message)" ("Linguistics," 66). Jurij Lotman echoes this view by asserting that "the receiver and sender use a common code. The common nature of the artistic language is unconditionally assumed, and only the message is new" (24). In both renderings, the code is considered beyond substantive control of either the encoder and decoder. Neither one can structurally alter the code itself; rather, they both have to use the code as it was codified prior to their engagement with it. Thomas Sebeok extends this common ground to a neat, expressed understanding, defining the code as "an agreed transformation, or set of unambiguous rules, whereby messages are converted from one representation to another" (465). Sebeok's definition reflects a widespread inclination in semiotic studies to presume that semiosis can be regulated to such an extent that it becomes little more than, to use Jean-François Lyotard's expression, the equivalent of a "business trip" (45).

The common-ground position can also, as Scholes contends, lead to the belief that the code can operate as "a principle structuring agent" of an entity (Semiotics, 100). Following Bernstein's extensive work on the code, Paul Thibault contends: "social semiotic codes function to classify and frame the relations between meanings, their realization, and the contexts in which these occur" (99). Similarly, he continues, semiosis is "regulated and controlled by higher-order classification and framing principles through which the social semiotic codes differentially distribute the material and semiotic resources of the social formation and the access of social agents to these" (165). While discussing the two "modes of arrangement" in a linguistic sign (combination and selection), Jakobson declares that "the addressee perceives that the given utterance (message) is a combination of constituent parts (sentences, words, phonemes) selected from the repository of all possible constituent parts" ("Aspects," 99). Jakobson goes so far as to claim that "the efficiency of a speech event demands the use of a common code by its participants" (97). In effect, he argues that the code possesses a type of parental, supervisory agency as it, in the case of viable phoneme combinations, "sets limitations on the possible combinations" of signifying units by establishing what is "permissible" or "circumscribed" within its practice (98). Jakobson maintains that this guidance is so firmly internalized by experienced, competent users of language, that "when faced with individual words, we expect them to be coded units" (98). Elam reiterates this position by asserting:

Formation and understanding (or encoding and decoding) of messages is made possible by the code...the ensemble of rules-known to both transmitter and destination-which assigns certain content (or meaning) to a certain signal. In linguistic communication the code allows speaker and addressee to form and recognize syntactically correct sequences of phonemes and to assign a semantic content to them. (35)

As these numerous examples attest, the sign user is seldom accorded an empowered status among at least these code theorists. (Barthes is frequently criticized for according too much power to the sign user in his assignment and idiosyncratic use of codes in S/Z.) Lotman is unusual in this respect in his portrayal of the actual use of codes as empowering, and potentially creative or constitutive, on the part of the decoder. In the case when "the listener tries to decipher the text using a code different from the one that the creator uses," three options exist, he says. "The receiver [can impose] his own artistic language on the text, whereupon the text is recoded (which occasionally involves even the destruction of the structure created by the sender)." Or, "the receiver [can attempt] to perceive the text according to familiar canons, but through trial and error is convinced of the necessity of creating a new code, one as yet unknown to him." A third phenomenon occurs when "an artistic text has a different meaning for sender and receiver," in which case the receiver has to "work out a code for deciphering that message" and "constructs a model" to do so (25).

The point to the above comparisons is that the code is always seen as a "key" to semiosis. The "mystery" of semiosis, in other words, is somehow solved, forced into stasis. The sign, then, is transformed from an amorphous, unreadable, atomistic entity into a classifiable, decodable "message". This is reflected in conceptualizations like that offered by Marshall Blonsky: the code is "the force that correlates an expression with a content" (442). In effect, the signifier doesn't acquire a meaningful context until it is frozen into a relation with another concept through coding. The obvious flaw with these renditions, though, is that they all necessitate the existence of an essential, or at least comprehendible, signified before an entity can participate in semiosis.

The code, in this sense, functions as a liaison between an encoder and a decoder. To decode something, according to these models, presumes that it was encoded. This would even apply to encoder-less signs (e.g., "natural" signs). After all, within these models, the encoder is not actually needed in order for semiosis to take place as long as something has a decipherable order or relation (e.g., certain clouds=likelihood of rain). And vice versa. For, who hears a soliloquy? Who is its decoder? The encoder! In the first example, the opposite takes place: the decoder becomes a retroactive encoder of a sign, producing a signified of one's own making.

Moreover, the models cited above evidently presume a monosemous "message". But, as Lotman argues, "one concrete text can submit on various levels to different codes" (25). If this is so, then no single "message" can necessarily correspond exclusively to the signified of a given sign. A common-ground viewpoint obviously meets a serious challenge with this contention.

An alternative view toward these code theories can be situated in the standpoint of governance-as-options decoding strategies. Jakobson maintains, for instance, that "in the combination of distinctive features" into signifying units, "the freedom of the individual speaker is zero: the code has already established all the possibilities which may be utilized in the given language" ("Aspects," 98). M. A. K. Halliday, also noting Bernstein's work, portrays codes as something akin to syntagmatic choices. Codes, he contends, are "principles of semiotic organization governing the choice of meanings by the speaker and their interpretation by the hearer" (67). The sign user, like a chess player, would never be free to create new move options beyond those prescribed in its rules (the queen can't move in an L-shaped manner like the knight can, etc.). Instead, a move decision would be limited by the allowable "moves" as dictated by the state of the game at the moment, the number of available pieces to select from, the piece-moving choices, and so on. Halliday argues that

codes act as determinants of register, operating on the selection of meanings within situation types: when the systemics of language-the ordered sets of options that constitute the linguistic system-are activated by the situational determinants of text (the field, tenor and mode, or whatever conceptual framework we are using), this process is regulated by the codes. (67)

Still, following Ruqaiya Hasan, Halliday suggests that the individual sign user is not prevented from creating unique codes (a view proposed in Saussure's commentary on language as the product of consensus, too). It's just that new codes cannot be recognized as such by other sign users until they assume the status of something like the common ground mentioned above. "Codes are not varieties of language, as dialects and registers are," Halliday asserts. "The codes are, so to speak, 'above' the linguistic system; they are types of social semiotic, or symbolic orders of meaning generated by the social system" (111). Accordingly, "'uncoded' means 'not (yet) fully incorporated into the system'" (180).

Another consideration regarding the mutability of the code is related to Eco's contention that codes, like signs in general, are vulnerable to changes imposed on them by sign users. "A semiotic theory must not deny that there are concrete acts of interpretation which produce senses that the code could not foresee," Eco suggests, "otherwise the principle of the flexibility and creativity of language would not hold" (Theory, 133). While drawing upon Peirce's commentary on abduction, Eco outlines two "different hypothetical movements" related to this phenomenon. (Not unlike Thibault's commentary on "particular coding orientations" [182].) Eco proposes "overcoding" as a "circumstantial selection" in which, "on the basis of a pre-established rule, a new rule was proposed which governed a rarer application of the previous rule" (Theory, 133). Or, a sign user presented with an initially incomprehensible sign system may sense "the feeling of organization that permits one to speak of a significant whole" (135). In this case, one could engage in "a sort of imprecise coding, a tentative hypothetical 'gesture' subsuming one or more large-scale portions of text under a given heading." This form of "rough coding" Eco identifies as "undercoding". The two operations contrast in that "overcoding proceeds from existing codes to more analytic subcodes while undercoding proceeds from non-existent codes to potential codes" (136). Furthermore, Eco suggests that possible confusion between the two operations can result because they are "frequently intertwined in most common cases of sign production and interpretation." When "it seems difficult to establish whether one is over or undercoding," one is in the position that Eco labels "extra-coding" (136). (For a much more detailed account see Eco's chapter on "Theory of Codes" in Theory, 48-150.)

S/Z: "Another Semiotics"

In an interview, Barthes remarked that he believed a significant shift had taken place in his semiotic approach between his influential essay, "Introduction to the Structural Analysis of Narratives" (1966) and S/Z (1970). In both cases, he felt had had engaged in different forms of semiotics, with the first instance oriented more toward a "scientific" methodology. A close examination reveals, however, that the seeds for the latter were in fact sown in the former. In his "Structural Analysis" essay, he had announced that he was pursuing the development of "another semiotics." "Linguistics knows this kind of frontier," Barthes remarks, "which it has already postulated-if not explored-under the name of situation." As an example of this province, he cites Halliday's concept of "situation" as, in the case of a sentence, referring to "the associated non-linguistic factors" (127). Barthes simply extends this viewpoint in S/Z, producing an increasingly polyvalent, indeterminate and fluid text in the process. What he does there is similar to his commentary on his analysis in the 1971 essay, "The Struggle with the Angel." "What is given here," he notes, "is not a 'result' nor even a 'method' (which would be too ambitious and would imply a 'scientific' view of the text that I do not hold), but merely a 'way of proceeding'" (127). This is what Barthes develops in S/Z, a vaguely theorized analysis substantially lacking in a unified, focused thesis. Instead, Barthes recounts his chronological reading of "Sarrasine," a novella by Honoré Balzac. "In a word, I radically abandoned so-called critical discourse to enter a discourse of reading, a writing-reading," he observed ("On S/Z," 73). Fredric Jameson elaborates on this argument: "S/Z is...his farewell to the attempt 'scientifically' to disengage from the infinite variety of human stories and tales some ultimate abstract narrative structure from which they are all generated" (45). S/Z is indeed a "farewell" along these lines. Between Barthes' haphazardly applied codes and the threadbare theoretical support for them, little more than a lively, intensely subjective reading of one person's reading of a novella is offered in the course of the text.

Even though S/Z is well known, it may be useful to quickly review its design. The text itself is derived from Barthes' discussions in a seminar at the École pratique des Hautes Études in 1968-1969. Barthes' method entails a section-by-section analysis of all of "Sarrasine" with the entire novella included in an appendix.

Balzac's text, published in 1830, is a frame narrative set in mid-eighteenth-century Italy. The interior, embedded frame of the narrative focuses on a French sculptor, Ernest-Jean Sarrasine, who becomes infatuated with a gorgeous singer, La Zambinella, while visiting Italy. As his desire increases, he attempts to "copy" her in his next statue, but that fails to satisfy his desire for her. At the same time, he is warned to withdraw his interest in La Zambinella and informed that she is under the protection of Cardinal Cicognara who, moreover, "doesn't trifle" with competitors.

At a party, Sarrasine is stunned to learn that La Zambinella is, in fact, a castrato and he kidnaps "her" to confirm whether this information is correct. He discovers the truth and is infuriated to learn that he has been the butt of a joke orchestrated by Zambinella's friends for their amusement. Before he can retailiate he is killed by the Cardinal's agents, Zambinella is rescued, and the Cardinal commissions a marble copy of Sarrasine's statue. The Lanty family subsequently commissioned a painting, based on the statute, of Zambinella portrayed as an Adonis.

This portrait functions as a link for the exterior frame narrative which takes place in a salon in the Hôtel Lanty. The unnamed narrator of this second narrative is amusing himself at a party held by the Comte and Comtesse de Lanty when his companion spots a strange elderly man who has attracted everyone's gaze. The narrator's companion, the Marquise de Rochefide, is horrified by the old man, and they retreat to a side room where they come upon the portrait of La Zambinella. Intrigued by the beauty of the Adonis, the Marquise asks the narrator about its background. At this point the interior narrative begins, and it turns out that the old man they had seen had, in fact, been La Zambinella, the source of the Lanty family fortune.

Given the rich re-framings that take place in "Sarrasine" (the embedded narratives, the gender confusion, the cultural code clashes, the subject of castration, etc.) it is easy to see why Barthes selected it as his text for a semiotic analysis of this nature. Barthes' method for writing about the text is similarly revealing. Shortly after S/Z appeared, Barthes reflected that "I had wanted for a long time to devote myself to a microanalysis, a patient and gradual analysis, in order to further structural analysis of the narrative" ("On S/Z," 69). Engaging in what he calls "a 'step-by-step' approach to the text" ("Interview,"135), Barthes divided the novella into 561 sections, or "lexias". "Each reading unit," he says, "corresponds approximately to a sentence, sometimes a little more, or a little less. The division into units can remain arbitrary, purely empirical, and without theoretical implications, if the signifier does not pose a problem in itself" (Grain, 71). The lexias, Barthes claims, consist of "blocks of signification," "units of reading" (S/Z, 13). The only controlling logic to his "artificial" division of the lexias is that "each lexia should have at most three or four meanings to be enumerated" (S/Z, 13-4). These quirky, impossibly ambiguous explanations do little to flesh out the theoretical support of S/Z, however.

But, it's this same outrageousness in Barthes' approach that may well be one of the most useful contributions he makes to semiotics, a field in which many of the truly insightful theoretical texts are exempla of lifelessness. Barthes' depiction of the criteria he uses for this selection usually entails almost unbelievable arbitrariness, often bordering on conceptual mayhem. He describes the selection as taking place "in the manner of a minor earthquake" (S/Z, 13). "This cutting up, admittedly, will be arbitrary in the extreme; it will imply no methodological responsibility, since it will bear on the signifier, whereas the proposed analysis bears solely on the signified." This abstract, lyrical gibberish highlights Barthes' rhetorical sleight of hand that underlies his entire approach toward crafting a narrative of his personal reading strategies for his own ends. Ultimately, he admits, the division "will be a manner of convenience." A related instance of Barthes' arbitrariness is connected with the arrangement of S/Z around 93 divagations. Louis-Jean Calvet reports: "'Why ninety-three?' a friend once asked [Barthes]. 'Because that's the year my mother was born,' he replied with a smile" (182).

While this dissection itself gained a lot of attention, perhaps the element of S/Z which has received the most commentary is Barthes' staggeringly chaotic application of five "codes" for his analysis of the novella. This, arguably, is what marks the development of "another semiotics" for Barthes. The "scientific approach" would have entailed "plac[ing] all texts in a demonstrative oscillation, equalizing them under the scrutiny of an in-different science, [and] forcing them to rejoin, inductively, the Copy from which we will then make them derive" (S/Z, 3). The alternative approach involves "restor[ing] 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, subjecting it from the outset to a basic typology, to an evaluation." This "coherence" derives from the five codes Barthes designates, but S/Z is, in the end, primarily just one reader's attempt to narrate an account of a slapdash reading, as Barthes himself acknowledges.

Barthes produces a "coherence" that represents the semiotic nightmare feared by those who denounce the anarchy of a ceaseless polysemy. For, Barthes attempts to produce what he calls a "writerly" text, one which exists only as the result of an active, even forceful, reading. He contrasts this with a reading operation he refers to as the "readerly," an acquiescent, complacent engagement with the text that only slavishly consumes, exhausts its monosemous sign field. The writerly reading, on the other hand, endeavors "not to give [the text] a (more or less justified, more or less free) meaning, but on the contrary to appreciate what plural constitutes it" (S/Z, 5). Of course, the aforementioned semioticians who decry this view are hardly to be expected to share Barthes' "appreciation" outlined here. (It is frequently asserted, too, that Barthes' further attempt to pair the readerly with the "classical" text [e.g., "Sarrasine"] and the writerly with the "modern" text [e.g., the nouveau roman] is seriously flawed. In the course of S/Z he demonstrates that the readerly and the writerly are actually different decoding operations and one can-as he does, in fact-produce either a readerly or a writerly reading of any text, depending on the approach employed.)

Scholes reflects the sentiments of many readers when he observes that "both the satisfaction and the exasperation one feels in reading S/Z are related to Barthes' use of the concept of code" (Structuralism, 149). It has become a familiar exercise in S/Z commentary to attempt one's own explanation of the five codes in order to strip the veil that Barthes cast over them (see Silverman, 250-283, Detweiler, 143, and Scholes, Semiotics, 99-104 for examples). (Barthes subsequently did this himself, as well-see "On S/Z," 74-75.) But the attempt to "clarify" the codes is really nothing more than a retreat into the readerly mode. In an effort to maintain the positive side of Barthes' thinly described codes, I will simply cite his initial explanations of them in my brief overview:

Hermeneutic code: "all those units whose function it is to articulate in various ways a question, its response, and the variety of chance events which can either formulate the question or delay its answer; or even, constitute an enigma and lead to its solution" (17).

Semic code: "the unit of the signifier" which creates or suggests "connotation" (17).

Symbolic code: "lays the groundwork" for a "symbolic structure" (17).

Proairetic code: "the code of actions and behavior" (18).

Reference code: "the knowledge or wisdom to which the text continually refers" (18); "references to a science or a body of knowledge" (20). (Barthes also calls this the "cultural code.")

It should be apparent why one of the most common responses to these five codes is to paraphrase them in a way that is more concrete and precise. A "better" grasp of the codes can be established by examining Barthes' applications and further discussions of them, however. One example of Barthes' designation of each code will suffice to illustrate this final point:

Hermeneutic code: "The title raises a question: What is Sarrasine? A noun? A name? A thing? A man? A woman?" (17).

Semic code: The title "has an additional connotation, that of femininity, which will be obvious to any French-speaking person, since that language automatically takes the final 'e' as a specifically feminine linguistic property, particularly in the case of a proper name whose masculine form (Sarrazin) exists in French onomastics" (17).

Symbolic code: Barthes quotes the lines recounting the engrossment of the narrator's companion in the painting of Adonis when she learns the model for it was a relative of Mme de Lanty. The narrator feels spurned: "I had the pain of seeing her rapt in the contemplation of this figure...Forgotten for a painting!" This evokes the symbolic code, Barthes concludes: "Marriage of the castrato (here, the union of the young woman and the castrato is euphorized: we know that the symbolic configuration is not subject to a diegetic development: what has exploded catastrophically can return peacefully united)" (78).

Proairetic code: Barthes quotes "Sarrasine"-"'To be loved by her [Zambinella], or to die!' Such was the decree Sarrasine passed upon himself"-and "decodes" this as the following action: "'To decide'"-"to propose an alternative" (117).

Reference code: Sarrasine discovers the truth about Zambinella after referring to him as a "she" while talking with the Roman Prince Chigi. "'Where are you from?'", the Prince asks him. "'Has there ever been a woman on the Roman stage? And don't you know about the creatures who sing female roles in the Papal States?'" This evokes the reference code, Barthes asserts: "History of music in the Papal States" (184).

The Discussion of S/Z

Like the threads that link the code-theory commentary examined here, the critical examinations of S/Z yield considerable illumination, especially in the repeated attempts to "clarify" Barthes' use of the five codes and their subsequent implications. In what is probably the most insightful analysis of S/Z written in English, Jameson argues that Barthes' differentiation between two classes of these codes "is not without its symptomatic value" (26). And, predictably, he describes this value ultimately in relation to his sense of the semiotician's political obligation. "Of the five codes, only three establish permutable, reversible connections, outside the constraint of time (the semic, cultural, and symbolic codes);" Barthes declares, "the other two impose their terms according to an irreversible order (the hermeneutic and proairetic codes)" (S/Z, 30). The first three, Jameson contends, "are essentially batches of what [Barthes] elsewhere calls indexes, that is, shorthand supplementary messages drawn from some more basic pool of shared cultural attitudes that permit us to decipher them" (26). The other two, he suggests, are "forms in time, and thus, passing now into musical figures, far more akin to melodic structures than to the harmonies of the previous 'reversible' types" (30).

To Jameson, this second employment of codes too often fails to make a substantial contribution to semiotic studies. Jameson complains that in the case of the latter two codes, "it should be observed, however, that to name a thing does not always suffice to explain it." In this way, Barthes may not have, as Jameson claims, done "any more than to designate the basic problem to be accounted for, namely, that of the diachrony or sequentiality of narrative discourse." But the first three codes, Jameson charges, are little more than attempts to systematically reinforce political quietism. "Concealed beneath a scheme of code classifications," he maintains, "we once again touch on that fundamental option of contemporary criticism (sociology versus psychoanalysis) which is itself a prime symptom of the fundamental split in modern life between the public and the private, the political and the sexual, between the untotalizably collective and the alienated experience of the individual" (29-30). "We may wonder," Jameson adds,

whether the procedure of assigning each of these dimensions to a different code really helps clarify this dilemma (in fact, it would seem to presuppose that each dimension of being had found adequate expression in a full code or sign-system of its own), or whether the concept of various codes here merely forestalls the problem and prevents it from being adequately explored. (30)

Ultimately, Jameson concludes, "in the case of the ideological materials" analyzed in S/Z, "it is clear that Barthes is concerned, in his later, semiotic period, to defuse this material and reduce it to data as inert and malleable as possible" (30). Jameson's contention, which comes across as a bit overwrought at times, is nonetheless more responsive than much of the critical discussion of Barthes' designation and employment of the concept of the code in S/Z. In this respect, he does successfully highlight the implications of code analysis that most commentators overlook in their own attempts to do, ironically, the same thing that Jameson claims Barthes was doing. For the discussion of S/Z reveals that most commentators have endeavored to strip it of its lively fluidity as a means of turning it into a useful heuristic, even when its extensive engagement with semiotic play is being acknowleged in the process.

Frequently, Barthes' use of the five codes is viewed as an only superficially liberating strategy for literary semiotics. Robert Detweiler is unusual in this respect when he argues that Barthes is engaging in an "anti-reductionist tactic" in S/Z, a position which would certainly seem warranted considering Barthes' repeated avoidance of a totalizing project (141). While he is setting up S/Z, Barthes concedes that "the reading of this text occurs within a necessary order, which the gradual analysis will make precisely its order of writing." "But the step-by-step commentary is of necessity a renewal of the entrances to the text," he adds. "It avoids structuring the text excessively, avoids giving it that additional structure which would come from a dissertation and would close it." Finally, he concludes, "it stars the text, instead of assembling it" (13). (By "starring," he evidently means drawing attention to or "separating" the text, as opposed to subdividing it in an arguably meaningful fashion, as in a taxonomy.) As Frank claims, Barthes tries "to acknowledge the text as a form of multiple meanings...by regarding the text as the intersection of codes often crossing and communicating with each other. Their basically open interaction is not determined by any rule that has been taken out of play" (156).

Of course, it cannot be denied that, as Frank observes, Barthes "operates, to be sure, with the structuralist category of the code, the system and the 'systematic mark.'" "But he multiplies the codes and works no longer with only one," Frank observes. "Every code signifies a systematic investigation with which every sequence of the story can be examined" (157). And, clearly, the code does serve as the base for Barthes' analysis. Although it is extremely generous to call Barthes' analysis of the codes "systematic", Scholes emphasizes the prominence of the code in S/Z when he suggests, while referring to Jakobson's schematization of the process of semiosis, "the basic tenet of Barthes's entire approach to literature may be stated in terms of that diagram" (Structuralism, 150).

Yet, like many other accounts of Barthes' endeavor cited here, Scholes accords the code a powerful role in presumably controlling signs. "For Barthes, there is no such thing as a pure context," Scholes argues. "All contexts come to man already coded, shaped, and organized by language" (Structuralism, 150). Unlike the discussions of the sign user's power vis-à-vis codes, this view elevates elements of the sign system to the position of exercising ultimate control over semiosis. Moreover, this contention often prevails in semiotic discussions, largely perhaps because of the concomitant establishment of fixed systems that one can then presumably study in a "scientific" fashion. Within this conception, the sign user "masters" the system, instead of contributing to-as well as being created by-its constitution. Kaja Silverman aligns herself with this rendition when she writes that "S/Z suggests that ideological imperatives express themselves through a multiplicity of codes which 'invade' the text in the form of key signifiers" (31). For Scholes, this invasion is commandered by the "code" and its army of sub-agents: "Each of these signifiers represents a digression outside of the text to an established body of knowledge which it connotes; each one functions as an abbreviated version of the entire system (code) of which it is a part" (31).

A "Natural" Alibi

A consensus becomes increasingly apparent within the commentary on S/Z that Barthes does engage in, at least to a certain (albeit limited) extent, the kind of explanatory analysis/synthesis that Jameson says it lacks. The code is arguably one of the best points of entry for this examination because Barthes' use of it here marks a progression beyond his earlier studies that had-in the case of Mythologies, for instance-focused more explicitly on ideological mechanisms of representation. While he explored several operations for locating different signifier positions in Mythologies, for instance, in S/Z he expands this approach into a larger, though non-totalizing, system. As in Mythologies, Barthes here attempts to emphasize the processes of connotation creation that are so ubiquitous that they become virtually invisible. Furthermore, the narrative trappings that accompany encoded ideologies serve only to reinforce this naturalization, contributing to what Barthes calls an "alibi" for its existence. "Driven to having either to unveil or to liquidate the concept" behind the myth, Barthes says, culture "will naturalize it" ("Myth,"129). Champagne, for instance, remarks that for Barthes, "the code is akin to the 'langue' in that it is representative of a given community...A whole cultural ensemble arrives with a code" (96). An ideologically sensitive reader such as Jameson might argue here that this "ensemble" is typically accepted as a "natural" characteristic of a culture, as opposed to the result of an ideological construction. While referring specifically to mimetic texts, Barthes highlights the explanatory agenda that often motivates code analysis:

The goal of all structuralist activity, whether reflexive or poetic, is to reconstruct an "object" in such a way as to manifest thereby the rules of functioning (the "functions") of this object. Structure is therefore actually a simulacrum of the object, but a directed, interested simulacrum, since the imitated object makes something appear which remained invisible or, if one prefers, unintelligible in the natural object.
("Structuralist Activity," 214-5)

As a result of this naturalizing agenda, elements within social systems appear to lack codified status. Like the natural sign, in other words, they appear to be invested with meaning only when it is imposed by an external, and obviously interested, party. "Our society evades as carefully as possible the coding of the narrative situation," Barthes suggests ("Structural Analysis," 128).

countless are the narrative devices which attempt to naturalize the subsequent narrative by feigning to assign it a natural occasion for its origin, and, so to speak, to "disinaugurate" it: novels in letters, manuscripts supposedly recovered, the author who has encountered the narrator, films which start their story before the titles. The reluctance to parade its codes marks bourgeois society and the mass culture which has issued from it: each demands signs which do not seem to be signs.

In response to the firmly entrenched status of the myth, Barthes' contention here may well be, as Culler suggests, that "the task of the semiotician...is to penetrate the alibi and identify the signs" that constitute it ("Tourism," 155). "In S/Z the codes were hardly more than suggested avenues of association that, by their multiplicity and contention, might help to keep the signifier in play," Elizabeth Bruss adds. "The point was not to comprehend the codes-and perhaps make them seem inevitable-but to lay bare the semiotic work behind seemingly natural appearances" (420). This task is consistent with Jameson's plea for a committed semiotics that tries to explain the operations of this social mythification by acknowledging and examining its coded networks:

Textuality may rapidly be described as a methodological hypothesis whereby the objects of study of the human sciences (but not only of the human ones: witness the genetic "code" of DNA!) are considered to constitute so many texts that we decipher and interpret, as distinguished from the older views of those objects as realities or existants or substances that we in one way or another attempt to know. (18)

Strangely enough, Jameson collapses these two distinctions in his ultimate chastisement of the Barthes of S/Z. He applauds Barthes for investigating the presumably coded nature of (in this case) narratives, yet he upbraids him for failing to shape these investigations into a cohesive, coherent web of knowledge. However, in part, both sides of this desire are undermined by Jameson's own belief in the code, in the assumption that a key to a given text can be used to unlock its signified and fully illuminate its epistemological status.

Countenancing the Code

Code theory, by necessity, has to beg the question regarding the code's very existence. In many respects, "semiotics" exists because it establishes presumptions like "codes" as constitutive elements.

Silverman, for example, proposes a connective-model function for the code and accepts, without question, the concepts that supposedly ground it: "As Barthes explains in S/Z, a code represents a sort of bridge between texts" (239, emphasis added). The intertextuality that Silverman employs to defend this assertion never examines the initial premise of the code: what Barthes offers, from her perspective, is a revelation, not merely a contention, proposal, or designation. Like the function of a deity in certain theologies (the encoder of the Great Code?), the code has to be accepted a priori in order for it to be used effectively. Without the code, there can be no way of intelligibly framing a sign. It's as though the code is a prerequisite condition for comprehensibility, as suggested by Barthes' assertion: "Without the-always anterior-Book and Code, no desire..." (S/Z, 73). Any substantial questioning of the code effectively dismantles it as an analytical tool. Barthes illustrates the extent to which a culture, desperately in need of an ordering principle, will go to find one. He cites a hypothetical "extreme case":

I have here before me a collection of objects so lacking in order than I can find no meaning in it; it would seem that here, deprived of any previous meaning, the form could not root its analogy in anything, and that myth is impossible. But what the form can always give one to read is disorder itself: it can give a signification to the absurd, make the absurd itself a myth...Even the absence of motivation does not embarrass myth; for this absence will itself be sufficiently objectified to become legible: and finally, the absence of motivation will become a second-order motivation, and myth will be re-established. ("Myth," 126)

A parallel assessment of the examinations of S/Z currently in circulation can be constructed by simply substituting "code" for "myth" in the passage above. This also could pertain to extant commentary on code theory which is never "embarrassed" by unquestioningly employing the concept of the "code" to serve its purposes.

Some examples: Champagne illustrates this well by asserting that Barthes "discovers five codes" (96), or reveals "the codes" (75) in "Sarrasine." Scholes declares that Barthes "recognizes five master codes in the text" (Structuralism, 154). Culler: Barthes "identifies the codes on which [the lexias] rely" (Barthes, 84). Bruss: "By revealing the codes at work in Balzac's 'Sarrasine'..." (432). And Eve Tavor Bannet: Barthes "demonstrate[s]" that the "lexes are coded in terms of five codes" (59).

Even the terms that Barthes devises for his five codes appear rooted in the desire to validate their existence. Philip Thody is one of the few observers to comment on this as a strategy itself, as when he observes that Barthes "gives them [the five codes] impressive neo-classical names in abbreviated form" (115). While he employed playful neologisms in the earlier Mythologies ("basquity," "Sininess," "governmentality," "bouvard-and-pécuchet-ity," etc.), it appears that Barthes attempted to establish a semblance of "scientific" analysis by virtue of the apparent gravity behind his designation of, and to a far lesser extent, his theorizing and subsequent use of the codes. (Unlike, say, the linguist John [Háj] Ross, who identifies negating expressions such as-"He doesn't know squat about soccer!"-as "squatitives.") In this respect, Barthes himself may have collaborated with the discussion of semiotics by going along with its "alibi" of the code. As Frank argues, "The control to which Barthes submits consists in his observing each lexia by the standard of a set of 'codes'" (156). Nevertheless, in many respects, Barthes' wild employment of the codes effectively eliminates any of the "scientific", or at least sufficiently theorized, aspects of S/Z.

Admittedly, a skeptical strain also runs through much of the S/Z discussion in the form of citational restraint. Calinescu effectively draws attention to this by commenting that "Barthes proposes five codes" (211, emphasis added). The more emphatic manifestation of this uneasiness is demonstrated by Frank's quote in the previous paragraph: he places the "code" in quotes, as though he is unwilling to naturalize the use of a concept in his own discourse, to be a dupe to the "alibi." (Seymour Chatman, 115; Scholes, Structuralism, 150; Bruss, 420; Moriarty, 120; and Jameson, 25 do the same thing.) Jameson (35) and Barbara Johnson (6) take this another step by referring to Barthes' "so-called" codes, as does Thody, who comments that Barthes "makes great play with the five codes into which he claims that the statements in 'Sarrasine' can be classified" (115).

Barthes' designation of specifically five codes also has elicited negative feedback. "A number of criticisms can be made of this selection of codes," Scholes declares. "There is something too arbitrary, too personal, and too idiosyncratic about this method" (Structuralism, 155). Scholes additionally attacks the reification of the codes that grants them an almost essentialistic or totemic status. "Five is not a magic number," he argues (156). Along these lines, Thody charges him with being stingy: "to provide only five codes for an infinitely meaningful text is a shade miserly" (116). Surely Barthes picked this group of codes to keep S/Z from being even more diffuse than it turned out to be. But his one attempt at creating focus ultimately only succeeds in severely limiting, possibly even crippling, his code applications.

But Barthes was well aware of the potential for accusations of this nature. In an interview he said that he merely "distinguish[ed] five main semantic fields or codes." "Admittedly," he added, "I don't know if this selection has any theoretical stability; similar experiments would have to be done on other texts to find out" ("On S/Z," 74). Barthes noted as well that he had planned yet another code (or subcode) focusing on the author that he later decided to omit:

As for the author, if I have radically withdrawn Balzac from my commentary-for which reason, I add in passing, it is an error to see in my work a "reading of Balzac": it's a reading . . . of reading-it's because I thought it was important to show that one could "get to the bottom" of a text without laying it at someone's doorstep; besides, even if one did determine the provenance of a text, it would just be one critical code among many. I had even begun coding, in my work, all the possible references to Balzac's life and works as a unit of the scholarly and university code, a cultural code if there ever was one; would the literary historians and psychologists have been more satisfied with that than my silence? (80, Barthes' spoken ellipses)

Undeniably, Barthes' use of the "code" offers a vast number of potential footholds into-in this instance-literary semiotics, despite the various shortcomings that have been raised regarding his implementations of it. Possibly the greatest hindrance associated with the code, however, derives from the extent to which it can so readily be employed in reductive, and clearly "interested", ways. Even the polysemous applications of code theory that Barthes argues for may not avoid this problem. "Barthes in the final analysis does not break with the code model of understanding," Frank contends. While "he corrects the most glaring deficiencies[,]...a report that is codified in multiple ways is still capable of being systematically decoded. The 'plural text' is 'multiple'; open to interpretation it is not" (158). Indeed, it appears as though any one who draws upon notions of the "code" is destined to be implicated in the bad faith associated with its concomitant false rendition of ordered codification.

The Larceny of the Code

In his discussion of the distortion of cultural mythography in Mythologies, Barthes characterizes "myth as stolen language," "a language-robbery" ("Myth," 131). Codes could be said to function as cohorts in this larceny, depleting semiosis of its energetic and ceaseless flux through the imposition of a rigidifying, and decidedly unnatural, framework. They don't necessarily have to, though. It is possible that the commentary on, and the general critical reception of, S/Z itself has created a problem that didn't exist beforehand (or, perhaps, exists in other ways). Some observers, such as Eric Blondel, have offered much more fluid, and if possible, non-reductive renderings of Barthes' project that may provide substantial insight for a critical semiotics to profit from Barthes' S/Z project.

Blondel proposes a intriguing entrance into a discussion of this nature by utilizing a potentially illuminating paradigm drawn from the discourse of psychoanalysis that proposes the "latency-plurality-indeterminacy of the code(s)" (77). "Interpretation implies the plurality, indeed, the indetermination of the code," Blondel contends, "perhaps even its latency, in the the sense that the code is implicit, hidden by the plurality or by the unconsciousness, a failure to recognize: indeed, where the code is hidden by the text" (76). This view of the code proposes a trajectory that cannot be traced, cannot be structured retroactively. It may exist like dormant desire, but it can never be seized in a manner that, with certainty, reveals an imminent system of codification. While using an obviously questionable inside/outside distinction, Blondel asserts that "interpretation violates the text in a certain sense, corrupts it, imposes itself on the text from the outside, without any guarantees purely intrinsic to and acknowledged by the text." As a result, to continue this argument, Blondel adds that it is "necessary that [interpretation] have a virtual multiplicity of latent, indeterminate codes, and that it says something about the context, an outside of the text, which can charge and then alter the meaning of the text."

This view can easily be turned into a detriment, as it often is by detractors of contemporary literay theory who bemoan analyses based on indeterminacy, play, and polysemy. Bruss, for instance, appears to offer a substantial criticism of Barthes' text by asserting that "it is an easy matter to find passages, from S/Z on, that seem to celebrate the end of certainty, indeed, of meaning itself" (419). "Particularly in S/Z," she comments, "Barthes manipulates the language of the text until it exposes its own hollowness and contradicts its own desire for solid and stable signs." This is a familiar lamentation by now, but it is based on a grossly imprecise account of semiotic plurality in texts like S/Z. In fact, it could be suggested that the real larceny underlying a position of this nature is found in what it celebrates: an impoverished, though complacently "certain", monosemy. Polysemy is viewed, correspondingly, as a lack of determinate/determinable meaning, a plurality of signifiers without evidently corresponding signifieds. This is quite similar to what Jacques Derrida refers to as the "structuralist thematic of broken immediacy" which is "the saddened, negative, nostalgic, guilty, Rousseauistic side of the thinking of play" (292). It is a form of "sure play," he declares, "that which is limited to the substitution of given and existing, present, pieces." This is the form of play characterized by loss, especially the loss of a logical underpinning that could only be called codification, the codable, or codability. The other side of this play is the "Nietzschean affirmation, that is the joyous affirmation of the play of the world and of the innocence of becoming, the affirmation of a world of signs without fault, without truth, and without origin which is offered to an active interpretation" (292).

To return to the epigraph of this lecture, it could be said that it's not codes that are coercive as much as it is the ways they are applied, the status they are accorded for controlling either the encoder or the decoder, or semiosis in general. The ways in which, finally, they are used to enforce a sure play, needlessly.

"One Last Freedom"

When he establishes the decidedly loose parameters of S/Z in the beginning, Barthes outlines a series of "freedoms" to which he will avail himself to produce what he described elsewhere as "another semiotics." His declaration of drawing upon "one last freedom" ("that of reading the text as if it had already been read" [15]) could easily apply to his overall project in S/Z. Despite the frustration that he recorded regarding the reception of this text, Barthes nonetheless expressed satisfaction with its heuristic function as a hermeneutical catalyst. In an interview, he says the letters he received from readers of S/Z, offering further readings and so on, "showed that I had succeeded, even timidly, in creating an infinite commentary, or rather a perpetual commentary" ("Interview," 140). The critical reception of S/Z reflects this also, in that most critics attempt to fine-tune or re-apply his method and theory even in the course of expressing dissatisfaction with them.

Unlike his more overtly structuralist analyses, Barthes clearly endeavored to avoid suggesting that elements such as codes functioned in universal, essential ways. In S/Z, he told an interviewer, he was "rejecting the idea of a model that would transcend several texts, not to mention all texts, in order to suggest...that each text was in a way its own model, and ought to be dealt with through its own difference" ("Interview," 134). In effect, those who complain about the flamboyantly "local" nature of Barthes' reading of his reading of Balzac's text might miss the point he was trying to stress the most. If anything, he was attempting to, it appears, escape the replicability of structuralist (or, perhaps, the "scientific") method because of the larceny it enacts upon the uniqueness of each text. "The present problem," Barthes remarked to Stephen Heath in 1971, "consists in disengaging semiology from the repetition to which it has already fallen prey. We must produce something new in semiology, not merely to be original, but because it is necessary to consider the theoretical problem of repetition" ("Interview," 129). Again, the element that seems to garner the supposedly superior, "objective" (or at least empirical) status for "science" was the one least likely to responsibly account for the phenomenon of decoding, specifically in this instance, the literary text. Indeed, it would be more of a case of "murdering to dissect," than replicating the living sense of active, and unfettered semiosis, that offers the greatest potential for a future semiotics (or "another semiotics").

This alternative semiotics would evidently abandon the considerable zeal that fuels desires such as "decoding," an operation that implies the code is roaming free and has to be captured, domesticated if not killed, in order to harness its otherwise unintelligible semiosis. Regarding S/Z, Barthes remarked that "the text is endlessly and entirely crisscrossed by codes, but it is not the fulfillment of a code (for example, the narrative code), it is not the 'speech' of a (narrative) 'language'" ("Interview," 134). This depiction of his apparent grasp of the code focuses intently on ongoing semiosis, hardly the longing for rigidity and certainty that so often characterizes critical commentary on S/Z, in particular, and code theory in general. Still, one of the problems frequently raised about Barthes' project is its patent irreconciliability. However, another way to frame this frustration (and perhaps more convincingly) is to attribute it his readers' inability to accept a pluralistic semiotics. They throw their hands up in frustration over his seemingly incompatible designs when he says things like: "my concern is...to pursue a general and systematic enterprise, polyvalent, multidimensional, the figuration of the symbolic and its discourse in the West" ("Interview," 129).

Ultimately, the easiest way to disengage Barthes' incongruous heterogeneity is to employ the strategy (mentioned earlier) that myth uses for its own purposes to create order where none exists. Witness Jameson's characterization of the sociological importance of Barthes' text: we can "situate" S/Z as an instance of "the postmodernist 'theoretical' text" (66). Through this gesture, he contends, "the description of Barthes' own discursive structure ceases to be a matter of weighing various critical alternatives against their object of study (Balzac), but has a specific cultural and historical object of study in its own right." The restorative, clarifying, homogenizing, contextualizing terms that Jameson uses to describe this action all point in the same direction: once delimited (or "sutured") by a generic code, S/Z can then be decoded because its controlling order-"postmodernism"-has purportedly been discovered. Jameson is ostensively prompted to this action by an ethical necessity to forestall the chaotic flux of semiosis that impedes the fulfillment of explanatory desire. He charges Barthes with a "moralizing valorization of critical pluralism" which he views as "at best a refusal to go about the principal critical business of our time, which is to forge a kind of methological synthesis from the multiplicity of critical codes" (59). This unforged order can be forged, though, simply by the convenient gesture of typologically coding S/Z, as Jameson demonstrates.

But this is only one of numerous other contentions that have competed without resolution over the "genre" of S/Z: It's a structuralist analysis. It's an example of deconstruction. It's a reader-response analysis. It's a psychoanalytical study. And so on. (Barthes himself identified its "genre" not as "semiology," but as "textuality," and grouped it with two other studies: Sade, Fourier, Loyola and The Empire of Signs. His "semiological" works, he declared, were Elements of Semiology and The Fashion System [Roland Barthes, 145].)

This stultifying certainty is, however, precisely the imprisonment Barthes evidently tried avoid through his various "freedoms" in S/Z. Along these lines, Michael Moriarty says: "These codes, which structure the text, are not, however, themselves structures, that is, they are not closed sets of oppositions" (120). In S/Z, "the code is not a system or langue put into operation in the parole of the text; it is rather a perspective opened up by the text." In keeping with this contention, one could posit that Barthes uses codes as "post-structures," in that they are delegated so crazily that they seem immune to the limitations of structure.

By employing a vague generic designation, refusing to "synthesize" his decoding, and using his codes in perversely idiosyncratic ways, Barthes may have engaged strategic indeterminacy as a means of keeping his text as open as possible. This would be consistent with Culler's assertion that Barthes is demonstrating the "citational play of codes" in an intertextual arena of signification in S/Z (Barthes, 85). As Barthes says in S/Z: "if we make no effort to structure each code, or the five codes among themselves, we do so deliberately, in order to assume the multivalence of the text, its partial reversibility" (20).

This potential employment of sprezzatura, an order designed to appear disorderly (or Barthes might say, "natural"), is aligned with Bruss' observation on Barthes' use of "deliberately arbitrary order" in his later works (436). This would sort well with Barthes' description of his approach to semiotics as the employment of a "loosening method" ("Inaugural," 476). As he suggests in S/Z, he is "concerned not to manifest a structure, but to produce a structuration" (20). To whatever extent an author's testimony can be brought to bear on an issue in terms of evidence, it would seem that this is exactly what Barthes had in mind. For, despite appearances to the contrary, he claimed that he had spent considerable time organizing S/Z into its final form after working through it in his seminar. "I labored over what used to be called the composition, i.e., the arrangement, the organization of the lexias and their commentaries, the digressions," he said in an interview. "I wrote and rewrote, I took a great deal of trouble over it, with passionate interest...I have a vivid memory of the time I spent struggling to piece the book together" ("Interview," 139).

Yet, at the same time, Barthes comes across as developing S/Z in accordance with what he later outlined as "my semiology" ("Inaugural," 471). He described "his" view as a "negative semiology" (475), "apophatic" in that it "denies that it is possible to attribute to the sign traits that are positive, fixed, ahistoric, acorporeal, in short: scientific" (473). "Semiology is not a grid;" he argued, "it does not permit a direct apprehension of the real through the imposition of a general transparency which would render it intelligible" (474). His semiology "seeks instead to elicit the real, in places and by moments, and it says that these efforts to elicit the real are possible without a grid," he suggests. "It is in fact precisely when semiology comes to be a grid that it elicits nothing at all." Clearly, this "grid" can be said to consist of many of the common elements in the discussion of semiotics, and as this lecture has suggested, perhaps it applies in particular to the "code".

As with his unsynthetic arrangement, his use of codes may be seen as similarly liberating, an instance of Barthes taking advantage of a seemingly inevitable paradigm-as opposed to being taken advantage of by it. This is reflected in Barthes' comments on his university work in semiotics: "what can be oppressive in our teaching is not, finally, the knowledge or the culture it conveys, but the discursive forms through which we propose them" ("Inaugural," 476). Need it be said that the "code" is precisely such a form? Scholes asserts an evidently acquiescent acceptance of code power over the sign user when he asserts that authors (or encoders) and readers (or decoders) are actually "constructed" by, and even "at the mercy of" "the interpretive codes of their culture" (Semiotics, 14). These agents "are traversed by codes that enable their communicative adventures at the cost of setting limits to the messages they can exchange" (110). Sign users are empowered, however, when they "see them as codes," as opposed to essential entities (14). Similarly, Silverman argues that "S/Z suggests that we can escape from the symbolic field which we presently inhabit by first mastering its codes, and then recombining them to form a new one-by moving from a passive to an active discursive position, from repetition to innovation" (249).

Barthes contends that the multivalent nature of the text cannot be represented responsively by an analysis founded on a reductive codification. "For those of us who are trying to establish a plural," he observes, "we cannot stop this plural at the gates of reading: the reading must also be plural, that is, without order of entrance" (S/Z, 15). For Barthes, codes are too often, too carelessly used to reify an artificial status. This results in an accumulation that diminishes, if not exhausts, the energetic flux of semiosis. He remarks that he is using "code"

not in the sense of a list, a paradigm that must be reconstituted. The code is a perspective of quotations, a mirage of structures; we know only its departures and returns; the units which have resulted from it (those we inventory) are themselves, always, ventures out of the text, the mark, the sign of a virtual digression toward the remainder of a catalog. (20)

One of these structuring agents-possibly the most important one, at that-is the contribution made by the decoder in the course of activating, inventing, and testing (through "undercoding" and "overcoding"?) various code responses in order to enter satisfactorily into a given sign's semiosis. In his commentary on S/Z, Culler argues that codes "dwell" not in the text, nor through some origin associated with the encoder, but within the decoder (Pursuit, 102). "The reader becomes the name of the place where the various codes can be located: a virtual site," he maintains. "Semiotics attempts to make explicit the implicit knowledge which enables signs to have meaning, so it needs the reader not as a person but as a function: the repository of the codes which account for the intelligibility of the text" (38). Finally, "It is not that each convention or moment of a code had a determinate origin which the accidents of history have obscured," Culler argues. "Rather, it is part of the structure of discursive conventions to be cut off from origins" (102).

It may be in this respect that Barthes' employment of codes in S/Z is especially useful in terms of contributing to the construction of "another semiotics." The greatest source of complaints about his use of codes would, paradoxcially, serve as the most important contribution he makes for the study of signs in S/Z. He uses codes irresponsibly, unclearly, aberrantly; assigning them in impossible, inpenetrable ways that lead readers to abandon it in despair or "rewrite" S/Z so that it makes sense-at least to them. But, again, this is where semiotics becomes flesh (or what he calls "the real"), where it becomes embodied (but not corporeal) through a simulacrum of the translucency, if not opacity, of semiosis. As Barthes says about grid applications, the moment when the grid fits transparently is the same moment when it no longer reveals anything other than the neatness of the imposed organizing frame. Only when the pedestrian use of the "code" is applied to S/Z does it become undesirably insufficient which, once more, accounts for the frequently dissonant responses to it. By using a conventional paradigm in an extremely unconventional manner, Barthes beats the larceny of the code at its own game in S/Z.

Despite Jameson's claim about Barthes departure from an analysis that could offer a structural paradigm for understanding all literary texts (and, by extension, any "text"), Barthes may have done just that in S/Z. And, because it's so anarchic, he also may have demonstrated the only way to do so. Early on in S/Z he remarks that "if we want to remain attentive to the plural of a text (however limited it may be), we must renounce structuring this text in large masses, as was done by classical rhetoric and by secondary-school explication" (11-12). To the contrary, he argues: "no construction of the text" (12). "Everything signifies ceaselessly and several times," he declares, "but without being delegated to a great final ensemble, to an ultimate structure." This, and related passages already cited, is where those who criticize Barthes for abandoning structuralism would point to as a means of demonstrating his segue into post-structuralism. But right after this he comments on analyzing a single text that significantly troubles this viewpoint. He remarks that there are "several implications and several advantages" connected with "the idea, and so to speak, the necessity, of a gradual analysis of a single text."

The commentary on a single text is not a contingent activity, assigned the reassuring alibi of the "concrete": the single text is valid for all the texts of literature, not in that it represents them (abstracts and equalizes them), but in that literature itself is never anything but a single text: the one text is not an (inductive) access to a Model, but entrance into a network with a thousand entrances...

The "model" (the amodel?) that Barthes does offer could arguably be one that perfectly captures the ineluctable modality of the decoder's always local, always irresponsible, always interested decoding practices. It is a model of difference, in other words. By "manhandling the text, interrupting it" (15) ("malmener le texte, à lui couper la parole" [22]), Barthes may well produce the closest simulation of actual semiosis that is possible. And, in so doing, he provides code theory with a new paradigm to work with, a process that significantly challenges the smug surety that so often results from presumptions of the "code" and codification.

As Elam asserts, in general, "the very concept of the code remains problematic and ambiguous in most of its applications" (52). This helps to explain the "shortcomings" that are cited in S/Z. However, Barthes' conception of "a triumphant plural, unimpoverished by any constraint of representation (or imitation)" (S/Z, 5) could just as easily apply to S/Z and foster a radical reframing of the "code" and its potential agencies. "In this ideal text," he adds,

the networks are many and interact, without any one of them being able to surpass the rest; this text is a galaxy of signifiers, not a structure of signifieds; it has no beginning; it is reversible; we can gain access to it by several entrances, none of which can be authoritatively declared to be the main one; the codes it mobilizes extend as far as the eye can reach, they are indeterminable...the systems of meaning can take over this absolutely plural text, but their number is never closed, based as it is on the infinity of language. (5-6)

"For the plural text," he concludes, "there cannot be a narrative structure, a grammar, or a logic" (6). Barthes would have saved himself a lot of grief if he had simply added one more component to that list: "...or a code." But semiotics would certainly have been the poorer for it.


Bannet, Eve Tavor. Structuralism and the Logic of Dissent: Barthes, Derrida, Foucault, Lacan (Urbana: University of Illinois Press, 1989).

Barthes, Roland. "Inaugural Lecture, Collège de France," A Barthes Reader, Trans. Richard Howard, Ed. Susan Sontag (New York: Noonday Press, 1982): 457-478.

--- "Interview: A Conversation with Roland Barthes," The Grain of the Voice: Interviews 1962-1980, Trans. Linda Coverdale (New York: Hill and Wang, 1985): 128-149.

--- "Introduction to the Structural Analysis of Narratives," The Semiotic Challenge, Trans. Richard Howard (New York: Hill and Wang, 1988): 95-135.

--- "Myth Today," Mythologies, Ed. and Trans. Annette Lavers (New York: Noonday Press, 1972): 109-159.

--- "On S/Z and Empire of Signs," The Grain of the Voice, 68-87.

--- Roland Barthes by Roland Barthes, Trans. Richard Howard (New York: Hill and Wang, 1977).

--- "The Structuralist Activity," Critical Essays, Trans. Richard Howard (Evanston: Northwestern University Press, 1972): 213-220.

--- "The Struggle with the Angel: Textual analysis of Genesis 32:22-32," Image Music Text, Ed and Trans. Stephen Heath (New York: Hill and Wang, 1977): 125-141.

--- S/Z (Paris: éditions du Seuil, 1970).

Bernstein, Basil. Class, Codes and Control, 3 vols. (London: Routledge & Kegan Paul Ltd, 1973-1975)

Blondel, Eric. "Interpreting Texts With and Without Nietzsche," Transforming the Hermeneutic Context: From Nietzsche to Nancy, Ed. Gayle L. Ormiston and Alan D. Schrift (Albany: State University of New York Press, 1990): 69-87.

Blonsky, Marshall. "When Chains of Difference Intersect: A Lesson," On Signs, Ed. Marshall Blonsky (Baltimore: The Johns Hopkins University Press, 1985): 441-443.

Bruss, Elizabeth. Beautiful Theories: The Spectacle of Discourse in Contemporary Criticism (Baltimore: The Johns Hopkins University Press, 1982).

Calvet, Louis-Jean. Roland Barthes: A Biography, Trans. Sarah Wykes (Bloomington: Indiana University Press, 1995).

Champagne, Roland. Literary History in the Wake of Roland Barthes: Re-Defining the Myths of Reading (Birmingham: Summa Publications, Inc., 1984).

Chatman, Seymour. Story and Discourse: Narrative Structure in Fiction and Film (Ithaca: Cornell University Press, 1978).

Culler, Jonathan. The Pursuit of Signs: Semiotics, Literature, Deconstruction (Ithaca: Cornell University Press, 1981).

--- Roland Barthes (New York: Oxford University Press, 1983).

--- "The Semiotics of Tourism," Framing the Sign: Criticism and Its Institutions (Norman: Univeristy of Oklahoma Press, 1988): 153-167.

Derrida, Jacques. "Structure, Sign, and Play in the Discourse of the Human Sciences," Writing and Difference, Trans. Alan Bass (Chicago: University of Chicago Press, 1978): 278-294.

Detweiler, Robert. Story, Sign, and Self: Phenomenology and Structuralism as Literary Critical Methods (Philadelphia: Fortress Press, 1978).

Eco, Umberto. The Open Work, Trans. Anna Cancogni (Cambridge: Harvard University Presss, 1989).

--- A Theory of Semiotics (Bloomington: Indiana University Press, 1976).

Elam, Keir. The Semiotics of Theatre and Drama (London: Metheun, 1980).

Frank, Manfred. "The Interpretation of a Text," Transforming the Hermeneutic Context , 145-176.

Frye, Northrop. The Great Code: The Bible and Literature (New York: Harcourt Brave Jovanovich, 1982).

Halliday, M. A. K. Language as Social Semiotic: The Social Interpretation of Language and Meaning (Baltimore: University Park Press, 1978).

Hodge, Robert and Gunther Kress, Social Semiotics (Ithaca: Cornell University Press, 1988).

Jakobson, Roman. "Linguistics and Poetics," Language in Literature, Ed. Krystyna Pomorska and Stephen Rudy (Cambridge: Belknap Press, 1987): 62-94.

--- "On Linguistic Aspects of Translation," Language in Literature , 428-435.

--- "Two Aspects of Language and Two Types of Aphasic Disturbances," Language in Literature, 95-120.

Jameson, Fredric. "The Ideology of the Text," The Ideologies of Theory: Essays 1971-1986, Vol. 1: Situations of Theory (Minneapolis: University of Minnesota Press, 1988):17-76.

Johnson, Barbara. "The Critical Difference: Barthes/Balzac," The Critical Difference: Essays in the Contemporary Rhetoric of Reading (Baltimore: The Johns Hopkins University Press, 1980): 3-12.

Lotman, Jurij. The Structure of the Artistic Text, Trans. Gail Lenhoff and Ronald Vroon (Ann Arbor: Michigan Slavic Contributions, 1977).

Lyotard, Jean-François. Libidinal Economy, Trans. Iain Hamilton Grant (Bloomington: Indiana University Press, 1993).

Moriarty, Michael. Roland Barthes (Stanford: Stanford University Press, 1991).

Scholes, Robert. Semiotics and Interpretation (New Haven: Yale University Press, 1982).

--- Structuralism in Literature (New Haven: Yale University Press, 1974).

Sebeok, Thomas. "Pandora's Box: How and Why to Communicate 10,000 Years into the Future," On Signs, 448-466.

Silverman, Kaja. The Subject of Semiotics (New York: Oxford University Press, 1983).

Thibault, Paul. Social Semiotics as Praxis: Text, Social Meaning Making, and Nabokov's "Ada" (Minneapolis: University of Minnesota Press, 1991).

Thody, Philip. Roland Barthes: A Conservative Estimate (Atlantic Highlands, NJ: Humanities Press, 1977).

Lecture Four:
The "problem" of controlling the decoder.


Eco, Umberto. The Role of the Reader: Explorations in the Semiotics of Texts (Bloomington: Indiana University Press, 1979).
copyright Scott Simpkins 1997
Send comment or question to Scott Simpkins: scotts@unt.edu Go to Cyber Semiotic Institute home page
Go to Course Outline
Go to Lecture: One or Two or Four or Five or Six or Seven or Eight