On Re-reading Saussure for Literature

What follows is an engagement with the nature of literature from the point of view of a reader—as distinct from that of a speaker or a writer. The peculiar kind of grasping of a vivid image that yet eludes definition as clear and distinct, that is not quite a ‘idea’, seems to be of the essence in literature. In other words, literature has a “sign-character.” Recent interest in developing theories of reading¹ ranging from reception theory to hermeneutics and allegories of reading indicates the importance of uncovering the structure of the act of reading at this moment in our literary history. Our purpose here is not to review these theories but to examine instead the potential a semiotic analysis of reading might have for its structure and process. The contention is that any theory of reading eventually will have a theory of the sign-structure implicit in it. It is the structuralist sign that will occupy the central place in this chapter.

For some versed in semiotic theory, it is Peirce’s “pure rhetoric” that would provide an adequate model of the “reading” that is literary.² Peirce’s definitions have historical affinities with New Criticism, and the notion that literary work can be conceived as the kind of sign that gives birth to other signs (texts engender readings which, as interprétants, are also texts) has enormous appeal for literary formalists. For all his insistence on the concrete presence of the sign for the receiver, however, Peirce often elides the description of the act of reading the sign, of the ‘grasping’ of its meaning.

Reading is bound up with meaning, and it is meaning that concerns the other semiologist, Ferdinand de Saussure. In thinking over the question of reading, we have noticed several ironies: while Saussure’s work has been utilized by students of speech and students of writing (the semiolinguists, Lévi-Strauss, and Derrida) little has been done to explain, analytically, his importance for some of the greatest of modern readers, such as Barthes, Derrida, and Todorov. Moreover, it is Saussure’s, not Peirce’s, work that has itself given birth to what we can call “readings,” texts that are interprétants and that have by now become part of our literary history.³

It is our intent to open the discussion of Saussure’s essay “On the Nature of the Linguistic Sign” (1959) by drawing out its import for literary study. To be literate is to read; even though a reading may be wrong or unconsciously motivated or determined by the manner of presentation of the material. What is read are signs, and if the sign seems somewhat worn-out as a concept, it is nonetheless inextricably linked to the fate of reading.

In re-reading Saussure’s linguistic sign as a literary one, we do not hope to go over old debates. We hope instead to re-frame the Saussurian sign for the literary history of ideas. Moreover, this is not an attempt to return to a ‘sign’ in a mode that pretends to ignore Jacques Derrida’s brilliant critique (1967). On the contrary, our aim is to repeat the Derridean movement in which, it must be recalled, Derrida emphasizes his awe and respect for the sign as Saussure so radically conceived it, the sign prior to its metaphysical ‘perversion.’ Derrida tracked the sign’s corruption in the Course (its failure to maintain balance between signifié and signifiant) as a story of the inevitable repression of one aspect by the other. We take a different course: tracking the internal “deconstruction” of the serene sign within Saussure’s text, the movement from the apparent totality of the linguistic sign to the moment of its explosion in the associative relation. Derrida writes of the labor of uncovering the metaphysical dimension of the concept of one sign, up to the moment of disclosing its logocentric and ethnocentric limits:

We revert from Saussure’s final terminology of signifier/signified to his first approximation for naming the elements of the sign—image and concept. Many doors of the literary and philosophic tradition, in particular the epoch of pre-Romanticism, are open to Saussurian semiotics in this way. Use of image and concept also allows us to integrate criticism with the deconstruction of behavior (chapter 1) and to deal with the peculiar turn given literary criticism by the attention to the signifier. Finally, the current move toward rhetorical figures in literary criticism is seen as a bipolar movement.

SIGNIFIER/SIGNIFIED: IMAGE/CONCEPT

The Saussurian sign, everyone knows, has two components: the signifier and the signified. For Saussure, if for few of his successors, the two are absolutely mutually determinant: neither exists as signifier or signified outside of their mutual determination. The sign composed of signifier and signified is a concrete entity, and although it is not a ‘unit’ such as a rational science might desire, it is open to study: Saussure writes,

Like a chemical compound (Saussure’s analogy is water), the sign is radically conceived as the total determination by a relationship; and it is the sign that makes the signifier and the signified what they are in the situation of their relationship.

What is true of the sign holds, for Saussure, at the level of the linguistic system: there is no language where the two elementary planes of sound and idea do not relate to each other as sound-image and concept. Saussure illustrates,

[1966:12]

From both the standpoint of the sign and that of the system of signs the stress is on the idea of mutual determination of elements. Few statements, either scientific or poetic, have so underscored the idea of mutual determination—more than organic—in the production of the third. And not since Wordsworth, perhaps, has the third entity ‘produced’ proved to be a deceptively simple inversion of parental relations: the third element here (language-child), which seemed to have been engendered by the union of signifer and signified, has itself engendered the union of the father and mother (concept and image; not thought and sound).

What should give us pause in re-reading Saussure for literature is more than its pre-Romantic sophistication in understanding the structure of cultural “production.” Rather, it is the fact that this very text has been able to evoke a completely perverse reading of it by semioticians who convey the sense that the two elements of the sign are separable and free to persist as “singles” outside their relationship to each other. Thus we have “modernist”⁴ versions of the Saussurian sign among leading semioticians, as in Umberto Eco’s formulation:

One should note that Eco uses Hjelmslev’s expression/content division, not Saussure’s, and it is this assumption that eventually necessitates notions of subjectivity and form—again notions that Saussure’s formulations can defer. Moreover, despite the attractiveness of the mobility assigned to the sign-functive over the seemingly static identy of the Saussurian sign,⁵ Eco’s version only brings out one facet of the sign’s existence, or its actual functioning: its arbitrariness. For, although Saussure notes the necessary arbitrariness of the linguistic sign (as would anyone who had ever heard a language other than one’s own), he notes that the sign has an equally important existence that is traditional:

The transitoriness that is so striking and salient a feature of Eco’s ahistorical and scientistic attitude^* is in direct contrast to the Saussurian vision. The sign is arbitrarily instituted; to be sure, it is not motivated either by God or by nature, yet once instituted, any change in the sign is never less than catastrophic:

Theoretically language is sheer arbitrariness and the locus of a potentially free-wheeling “associating of any idea whatsoever with any sequence of sounds” (1966:76); it is not so in fact, in practice. We are not daily reminded that the absolute freedom to have been or to become anything at all is the chief constituent of language. Eco’s abstract portrayal is exciting but eventually more heavy-handed than Saussure’s: the sociological sophistication of Saussure’s supple handling of the traditional, social-institution aspect of language has been noticed by such critics as Jonathan Culler, who compares Saussure to Dürkheim (1976:71), and it is much less deterministic than Eco’s coding “rules,” whose préexistence ironically undermines the apparent freedom of the detachable functives. Clearly Saussure is heir to the post-Rousseauian, post-Kantian revision of the individual identity versus group identity split—its reformulation as a tension between person and community in which the definition of each depends on its relationship to the other.⁷(In this version language allows no subjective ‘freedom to mean’—its meaning is in the response to it; the identity of the self is in the relation to the other, etc.)

While it is possible to situate Saussure and his linguistic sign and sign-system within the movements of post-rational thought, as we have tried to do in the previous chapter, that is not our purpose here. The real question, the literary pertinence of the sign, can and should take as its point of departure the manner in which literary theorists have utilized it: if a pattern of resistance emerges, that in itself is important for uncovering the uses, misuses, and abuses of the sign for literary reading.

What then are the elements in the Saussurian text that have inspired theories of the sign (Eco is not alone) that are quite resistive to its spirit and direction? Why has there been no major and/or overt effort to analyze and utilize the sign as conceived by Saussure without first decomposing it into signifier versus signified—and then concentrating all one’s attention on one or the other in the absence of its mate? For although one might argue, as do Carontini and Peraya (1975), that one is either for or against the “ideologism of the sign,” it is much more correct to say that it is rather a choosing of either the signifier or the signified that divides recent theoreticians: Deleuze favors the signified, Barthes the signifier, etc. As with Eco, the “sign considered in its totality,” to borrow a chapter heading from Saussure, is something of an embarrassment to literary semioticians.

Derrida (1967) has dissected the phonological bias, the verbal/logical/mythic identity complex in Saussure; and he has noted the repression of the graphic signifier by the phonic one in semiology. But there is more than this repression of the signifier by another signifier, more even than the repression of the signifier by the signified—the sign is the locus of repression in general. What is repressed by the sign may not only be the material signifier—phonic or graphic—its becoming transparent to meaning; the signified, or rather signifieds, may also be repressed, not because the signified has been rendered transparent, but because it is apparent: the appearance as the signified represses alternative differential meanings that are latent in the sign.

What seems important in retaining the sign, as opposed to dividing it, is that it may be a potential model of the text and of literary relations—as bearing both the quality of the puzzling enigma and the sure sign. The association of a signifier with a signified is the exemplary literary act, in terms both of reading and writing, and the avoidance of the sign may well be one source of methodological difficulties in literary analysis.

In reviewing the Saussurian sign for literature we return first to the chapter in the Course (l,i) on “The Nature of the Linguistic Sign.” This essay is renowned for two things, its differentiating language from nomenclature and its model of the sign. “The linguistic sign unites, not a thing and a name, but a concept and a sound-image” (1966:66). Few readers have remarked (see D. MacCannell 1976 and chapter 3 of this book) on Saussure’s intitial formulation of the signifier as an image. Saussure is actually quite adamant on this point: “The latter [the soundimage] is not the material sound, a purely physical thing, but the psychological imprint of the sound, the impression that it makes on our senses,” and he concedes that he will call the signifier “sounds and syllables ... provided that we remember that these names refer to the sound-image” (1966:66). This mental impression, or writing, is usually seen as a part of Saussure’s ‘formalism’—yet it is so bound to its ‘content’ that it would hardly warrant the term ‘form’. And the definition of the sign Saussure makes on its basis is important enough that he presents the first of a series of diagrams of the sign thusly: fig. 2 (66)

And then the second: fig. 3 (67)

The supreme irony—or a joke? The initial formulation of the linguistic sign is as a picture, the so-called phonic substance is a graphic image, and the entire sign, often assumed identical to the word or verbal unit, is an icon or hypotyposis, an existential graph. It is only on the third try that this graphic entity incorporates the notion of representation—the natural object, the tree, is there—yet it is hardly clear what is representing what. The tree, one notes, is the concept and not an image in the relationship; yet the concept itself is markedly only an image:(It is possible that this third figure is apocryphal; nonetheless its use would not have been inconsistent with Saussure’s terminology.)

We find from the outset two striking tfrngs about the ‘nature of the linguistic sign’: (1) as “image” the signifier can substitute sight for sound or sound for sight, depending on the temporal or spatial mode (1966:70), and (2) it is well nigh impossible to distinguish “concept” from “image.” The concept, as in the version of the classical figura, partakes of the same mode of being as the image. In breaking new linguistic ground Saussure has in effect undone both classical rationalist versions of how language works (language is the picture of thought) and Enlightenment versions (thought is the product of the linking or placement side by side of images.)⁸ He has placed two images together, but with the result that each acquires a new existence—as signifier and as signified—from the relationship. The loss of identity as a separate element in each case results from the relationship.

What follows is an effort to relate the Saussurian ‘linguistic’ sign to the ‘literary’ one; debates concerning the verbal and/or iconic character of literature will, we hope, benefit from casting the discussion in this way; so also will the less semiotechnical and philosophic movements in the contemporary rebirth of literature-as-rhetoric.

Applying Saussure’s semiology to literature has led in two distinct directions and may require yet a third. Attention to the signifier has dominated much of the methodologically self-conscious Continental literary semiotics; both the verbal narrative, the ‘linear’ signifier unfolding in time and the graphic, spatial image have been exhaustively examined by semiologically oriented critics. Even those concerned with figures can be distinguished on the basis of their implicit conception of the figure as aligned more with the image or with the concept side of the kind of sign Saussure was attempting to uncover. Since it is the figure that appears throughout our literary history as intimately bound up with the ‘nature of the literary sign,’ it seems essential at this point to look at what has happened to the sign-as-figure as a result of the division of signifier from signified in recent theoretical developments.

Figures of Speech

If we split the image-concept bond we lose whatever it was Saussure was trying to convey by the sign, something like a “meaning to be read.” This loss of the sign, with its vague innuendo of a meaning that is always somehow ungraspable, or dissolves into being a mere image, has seemed to many to be a gain. By using the terms concept and signified, Saussure suggested the sense of seizing or grasping, of ‘getting’ the meaning—only to withdraw this suggestion with the revelation that the concept is fundamentally imagistic in nature.^* The Sartrean bad faith of such a gesture does not apply if we leave off the problem of the concept and restrict our semioanalysis to signifiers, not to signs. And for this Saussure provided the exemplary methodology: analysis of the signifier for him proceeds along clear-cut lines.

The signifier is distinguished by its syntagmatic (sequential opposition) character. Both temporally constituted speech and the graphic spatial instance of writing are for Saussure in the realm of the signifier: even though it is Saussure’s subordination of the latter to the former that has occasioned some controversy, it is important to recall that Saussure at no point ignores either form of the signifier (or image):

Saussure’s first imperative in discussing the signifier is to deny it the capacity to be perceived in “simultaneous groupings in several dimensions,” as can “visual signifiers”—auditory signifiers can only be sequential. The bias in favor of the auditory signifier at this point appears to restrict analysis to the narrative dimension, to the flow of discourse.

It was the genius of structuralism to have brought out the paradigm as a way of accommodating “simultaneity in several dimensions” even for the analysis of signifiers—signifiers that owe their existence to the syntagmatic or chained (sequential) oppositions. For Lévi-Strauss, for Barthes, paradigms are groupings of opposites in sets which can be superimposed to show that they are so many variations on a single theme (Lévi-Strauss 1970:339-41); Barthes discusses paradigms similarly in what may be his most structural study, Système de la mode (1967).

The paradigmatic outgrowth of the signifier (exclusive of the signified; destructive of the sign) has been used for literary criticism. On the one hand, it appeals directly to the sense that literature exists at many different levels, much like Lévi-Strauss’s description of the structure of a myth:

The idea of superposition, of overlapping levels of the literary figure seems naturally to fit the paradigmatic structure; Jakobson, for example, describes “poetic language” in this way: “Poetic language draws attention to itself by forcing us to consider the paradigmatic aspect of language. This is precisely what we do when we are aware of the presence of a figure” (cited in Scholes 1974:161). Literary critics may take differing positions as to whether what is layered are, as in Lévi-Strauss, “meanings” in the guise of a palimpsest (semi-hidden meanings à la Genette). Or they may, objecting to the hiddenness of meaning, purify the paradigm of the relation to meaning and retain a structure of superposition and figuration (Todorov 1971:246-47 and Shapiro and Shapiro 1976:1-6).

As the structural figure (metaphor) erected from the material of signifies of syntagmatic origin, the paradigm performs an extremely important function for structuralist literary criticism. Paradigms are relations between signifiers that can replace each other, alternative pairs that seem distinct; by retaining the clarity and distinctness of these oppositions on a métonymie base—in which value is created by the relations to preceding and succeeding terms—structuralism is in a sense, enabled in relation to literature. The figurative aspect of literary language can be shown to be based in the syntagmatic axis of language—as the paradigm’s “distinctness” is a kind of provisional resting place in the flow, the syntagmatic unfolding of signifiers. The paradigm is a kind of illusory or transitory revelation of central ‘theme’ or set of contrasts—the Greek etymology of the term is ‘to show side by side’—and it is the syntagmatic axis of language that can reintegrate the paradigm into the discursive chain. Literature can be shown to be, finally, discourse, and literary methodology can be enriched with the work of the linguists who study speech acts.

Figures of Thought

The question of meaning—of the signified—is deftly sidestepped by the structuralist critic’s affection for the signifier, an attachment perhaps akin to the Romantic respect for the image skillfully analyzed by Paul de Man (1960). The place of meaning in cultural productions is difficult to ascertain: Saussure designated it in the sign as the “signified” while undermining any suggestion that it have any existence by itself. “Meaning,” the meaning perpetrated by signs, often appears as anything but a neutral term, politically and morally; and in a critic like Barthes, for example, one senses resistance to imposed meanings throughout his entire oeuvre, resorting to the signifier as a primary defense. For others, like Todorov and Lévi-Strauss in the quote on the “levels” in myth cited above, meaning is a purely relational term. For still others meaning is inherent in form. Does restriction to the signifier effectively dispel the illusion of the immanence of meaning in the sign, at least for literary critics? Here it seems to us to be the capital question for literature. We wish to trace now several alternative critical patterns in those who limit analysis to the signifier, and we will then turn to those who deal more directly with the problem of the signified, particularly hermeneuticists and post-structural theorists, in the analysis of the other mode of rhetoric, the trope, or figure of thought.

The Formalist Tendency in the Attention to the Signifier

Even the most neutral-seeming signifiers acquire meanings for us: one need not assign a transcendental signified to them, nor go to the sophistication of a Freud to support this statement. One need only look close to home and to the subject at hand, literary theory, to find “meaning” intertwined with signifiers. William Wimsatt and Monroe Beardsley are perhaps justly famous for their attack on “intentionality” (for discussion, see de Man 1971), yet theirs is an attack that in no way precludes “meaning”:⁹

The assumption that poetic language “contains” meaning despite the intentions of the author is clearly in contrast to the structural sense of literature as parole, the subjective speech act. Yet there where the structuralist would see only signifiers from which implications of “meaning” are healthily absent—in rhythms, meter, sound-patterns—Wimsatt locates meaning: “Sentence patterns recur, like declensions and conjugations; but they are still expressive forms,” writes Wimsatt (Wellek and Warren 1942:179), and his approving fellow formalists remark, “Rhyme has meaning.... Several aspects of this semantic function of rhyme can be distinguished” (Wellek and Warren 1942:160).

One can imagine Todorovian horror at some of these formalist propositions, but nevertheless they indicate the profundity of Saussure’s insight: there where a signifier is so will meaning (the signified) be.

The structuralist alternative to the formalist feeling that meaning inheres in isolated units is clear and simple: “meaning” is a feature of literature, to be sure, but must be viewed, according to structuralist poetics and semiotic theory, as a relation between levels. As in Lévi-Strauss’s description of mythic meaning cited above and Emile Benveniste’s similar notion of linguistic levels, Tzvetan Todorov attacks formalist versions of meaning for literature.

The syntagmatic/paradigmatic solution is appealing: yet it works only on several conditions. Chief among these is the structuralist capacity to sidestep certain obscurities in Saussure’s text which develop in the analysis of the sign as a totality, as distinct from the analysis of signifies. For example, Saussure repeatedly indicates that the “linguistic unit” is not a clearly identifiable isolate (1966:105-7), particularly from the point of view of synchronic linguistics, or the analysis of la langue. (On this point Benveniste is in disagreement, since he claims that the smallest units of language are ‘meaningful signs’ [1969:11-12].) Somewhat ironically, the structuralist response to the formalist assumption that meaning inheres in isolated units eventually will appear as parallel. Recall that Wimsatt writes that “Poetry is a feat of style by which a complex of meaning is handled all at once” (1960:199); now hear Todorov:

Meaning is a relation among signifiers, signifiers lifted from the syntagmatic combinations, and arranged paradigmatically to produce the figure: the overlaying of different, but transparent levels, traversible by an integrating vision. The meaning will not have a locus’ either in itself nor in the perceiving mind: it will be a set of structured relationships based on paradigmatic contrasts and syntagmatic combinations. Yet the ‘product’—’the work as a whole’—which finally appears in the process bears at least a temporal affinity with Wimsatt’s own version of textual presence: ‘all at once’.

It is in the poetics of structural literary theory that the ‘modernism’, or the sense of radical arbitrariness of the sign, in Saussure is lost for literature. Just as Eco’s overemphasis on the mobile ever-changing quality of the sign devalues tradition, so too the over-identification of literature with Discourse finally has a traditionalizing and conservatory effect that belies the modern. Foucault writes that discourse is what it is because it is rule-governed (1971:38), restricted by convention, or in Saussure’s terms, ‘tradition’. By deriving its models entirely from the analysis of the signifier, and in particular the auditory signifier, structural literary theory seems to go on to become Neo-Classical, arguing for ‘literariness’ by opposing literary discourse to ordinary discourse. Before Todorov wrote to distinguish literature from “everyday speech” (1977:24), Wordsworth was championing the reverse—literature’s counter-adoption of the “very language of men” (1961 [1802]:8). For Wordsworth it is poetic diction that is abusive and capricious: “The reader is utterly at the mercy of the poet respecting what imagery or diction he may choose to connect with the passion” (21); and the abuse results from the arbitrary bond between an image (“imagery and diction”) and a concept (“passion”: for Wordsworth, our thoughts are “the representatives of all our past feelings” [6]). The poet abuses everyday speech when in fact he should use it. The drive for “rule-governed creativity,” latent in Todorov, manifest in critics like Jonathan Culler, arises at least in part from using the model of speech, the analysis of the signifier (paradigm and syntagm), and the deferral of the question of meaning to the question of levels. The model of discourse generally entails the distinction between ‘ordinary’ or ‘everyday’ speech and ‘literary discourse’, to be sure, but the former bears more than its share of the burden of the definition of literature in structuralism.

On the one hand, that is, “everyday speech” is positioned by the unconsciously assimilated set of rules that language is; on the other, it is the source of change and creativity. Culler writes, “Change originates in linguistic performance, in parole, not in la langue, and what is modified are individual elements of the system of usage” (1976:41). Neo-Saussurian semiologists see language, la langue, as did Hjelmslev, as an abstract set of norms and forms fulfilled or realized in the parole or speech act. The tendency is then to identify poetic or literary language with the parole, inasmuch as for Saussure, as well as for the later Chomsky, it is performance that is the realm of creativity and change: a “rule-governed creativity” to be sure (Culler 1976:41, 84). Poetics, or literary structure, then would be logically akin to Saussure’s langue— the abstract norm scheme.

Logic fails in cultural institutions—they are bizarre and not rational—and it seems inevitable that the clear-cut parallels drawn above could not hold for so complex a cultural production as literature. From the point of view of literature, it is often in the realm of “everyday speech” that norms, forms, and rules are not latent, but patent. Decorum, etiquette, and politesse show in everyday talk; the ‘creative’ side of parole is, even for Saussure, limited to pure change. It is entirely possible that the identification of literature with speech and poetics with langue needs to be inverted. This seems true at least for some literary epochs, as the citation from Wordsworth, among others, would indicate. At any rate, for at least some literary epochs, the plane on which image and concept are bonded is associated with the “arbitrariness” and freedom of will that is linked to literature. Change, here, is not in the signifier, but in the signified, or to be more exact, in the relationship of signifier to signified. In the concept, turned, twisted, or deviated, the arbitrary shows: such is the version of literature as trope, or figure of thought.

STRUCTURALISM AND HERMENEUTICS

The important point to retain from our analysis of structural reading here is that, with the device of the figure-as-paradigm the structuralist reader’s aim is total clarity, or a kind of totum simul. Radicalizing Saussure’s presentation of the concept as merely another image makes of meaning, as it can be clearly mapped at different levels, a transitory illusion produced by the arrangement among (syntagm) and between (paradigm) images (signifiers). One might call the structural reading explicative rather than understanding, to revert to Diltheyian terms. Reading is the revelation of a purely relational “meaning”—it is not interpretation. Born at the moment when relationships are laid out, the figure-paradigm is quite other than the palimpsest model for literary texts: in the former nothing is hidden; in the latter meanings are hidden. It might be said that the joy of structuralism is the fact that meaning is so completely sustained by the perceptible relations between signifiers and can expand, like the concrete universal, from the simple signifier to the totality.

What is absent from the discursive view of literature is nevertheless the other version of the figure, pointed to above: the trope. Tropes are twists or turns, deviations in meaning, figures of thought; and for some it is in the realm of the signified that one should look for the model of literature. The trope is, in a sense, the figure whose meaning is incapable of being grasped only at the level of the signifier, as in the structuralist vision. But it is also that figure whose trajectory through levels cannot be traced in a clear, open, and direct manner. The palimpsest to which Todorov objects as a description of reading—the layered set of significations or hidden meanings—is nevertheless appropriate as a description of reading in relation to understanding. The trope js not contemporary to, nor coextensive with, nor even coeval with the signifier. Like Rousseau’s “figure” in the Essay on the Origin of Language, in which the “proper meaning” is found later, the trope exists by virtue of the bar—temporal essentially—dividing signifier from signified. Hermeneuticists may deal with the trope as a deviation from an original meaning; post-structuralists like Derrida may deal with it as a mixture of deferred and repressed meanings (plural). In the former, the hermeneutic, reading the trope is an act of faith in the potential for understanding (retrieving or conceiving) meaning that is not immediately present; in the latter, reading is the threat of misunderstanding the meaning. What was once a tentative pact linking structuralism and hermeneutics (Barthes interfolds a hermeneutic code as one among several required for understanding) threatens to break down with the purely discursive model of literature in the younger structuralist critics. And although it is the differences between “hermeneutics” and “deconstruction” that are today being highlighted, our semioanalysis here has perhaps allowed us to divide both from structuralism on the basis of the attention to the signifier and to the process of reading.

Non-structural reading is not clear and distinct, nor even ‘innocent’; signs, unlike paradigms, do not only have sides that show. Signs, that is, are not ad hoc, pristine, and happenstantial: they carry, by Saussure’s still unexcelled description the germ (in the pernicious as well as the reproductive sense) of meaning in their hearts. They have past histories—the elegiac past and/or the accretion and repression of past (often pathological) associations; and they have futures—the hermeneutic expectation of understanding and/or the deferred never-tobe-clear-and-present later Derridean signified.

In point of fact Derrida remains more respectful of the Saussurian signifié than have some hermeneuticists. We refer not to the very important debate between Ricoeur and Derrida over the relationship of metaphor and metaphysical truth, which we mention in the notes. We are thinking of this line from Gadamer confronting meaning: “Understanding means primarily to understand the content of what is said, and only secondarily to isolate another’s meaning as such” (1975:262; our emphasis). Here the hermeneut has divorced content from meaning; how different is this from Wimsatt’s locating meaning in form? Few writers indeed have sustained as Derrida has Saussurian insights into the absolute bonding through difference of signifier and signified and the catastrophic import of separating them.

In many areas of existence the syntactical difference between signifiers is precisely the same as the difference between the signifier and the signified: a mere difference (see Merleau-Ponty 1963:121-22 and Dean MacCannell 1976). In some situations it is the relation or relay of meaning from signifier to signifier (as in structuralism) that captures our attention as analysts: we can “map” social intercourse, “suspend” it for inspection. In the case of literature, however, it seems that the focus is, or ought to be, on the differential relation of signifier to signified. Saussure’s icon of the sign, itself an ironic conflation of verbal and pictorial, shows the impossibility of defining either image or concept, signifier or signified, external to their relation in the sign. That is, it is not possible outside of their relation to each other in the sign, to define what a signifier or a signified is. By itself, the Saussurian signified or concept is just an image, an image like , which can be said neither to contain nor to be a meaning. In the same way the image-signifier “arbor” could always be taken as a meaning and not just a graphic image. Saussure’s differential bar between the images is what must be emphasized, for this sign is, unlike the paradigm, the locus of an almost pure difference without systematic opposition based on substitution or replaceability of one part for the other. For while it is true that arbor can, in any of the graphs presented here, be taken as either signifier or signified, in any one of the specifically designated relationships this is not the case. Whenever an image relates to another through differentiation, it loses its ‘nature’ as an image and becomes a signifier or a signified in regard to the other. There where a signifier is so shall a ‘signified’ be. A mere image, that is, can appear to convey and also to be a meaning.

One can operate on the sign and demystify it as a sheer conjunction—happenstantial as it were—of two distinct images, or signifiers, in which ‘meaning’ is only always an irrelevant illusion. Yet what seems to us to be important about the sign and for literature is that it is precisely the meaning which is unavoidable: not the meaning, or the only meaning, but a ‘meaning’ that arises out of the differential relation and that always remains an enigma. The elementary analysis of the linguistic sign brings out the mechanism of meaning-production, yet without thereby doing away with the meaning.

Literary study based in the linguistic source of meaning-production with the alogical possiblity of meaning prior to referentiality would thus be tropic: the study of vicissitudes, turns, of a necessary écart, or deviation, from a literal or specific referent in the sign relation. It is no accident, perhaps, that although they conflict with each other dramatically, post-structuralism and hermeneutics often agree to disagree about the same texts,¹⁰ while structural readings tend to be more serene. For whether they deal with the trope as the deviation from an original referent that has somehow lost its rapport with the present text, or whether the trope is seen as a mixture of deferred and repressed referents in the never-to-be understood character of the Derridean sign, it is at least the sign-character of the text that is at stake both for the hermeneuticist and the post-structuralist. The former may still have faith in the possibility of understanding meaning, the latter may assume that reading is possible only on condition of misunderstanding the meaning—but it is still meaning that is the problem, not reference or opposition.

Such is not the case for structural reading, which is clear and distinct because meaning is based in systematic opposition. Meaning is that which is ad hoc and relational, and by definition it cannot exceed what is given in the discursive version of texts. The alternative or poststructural reading is that meaning spills over into a textuality in which oppositions are derivative and not conditional: meaning, a product of the sign, of the difference that makes a sign, is never yet systematizable as being opposed to and substitutable for the signifier. This is because it is not only an arbitrary opposition to the signifier that makes it a signified: it is because meaning also has to bear the burden of tradition—to revert to Saussure’s terms once again. Sign meanings have “histories,” although these histories are not necessarily empirical nor yet the diachronic story of transformation and change.

This brings us to the final and necessarily speculative section of our chapter. Saussure raised the possibility of introducing a temporal dimension into the analysis of the synchronic language state, the state semiologists have up to now seen clearly as the realm allied linguistically with grammar and literarily with poetics. It is our feeling that even within the synchronic analysis of langue Saussure was able to describe one of the two constitutive relations of language—the association and the syntagm—in such a way as to make the former relevant to the type of ‘history’ that pertains to literature.

Rewriting Saussure: Associative and Syntagmatic Relationships in Language

Saussure did not tarry overlong with the unitary, isolated sign. He moved quickly into the analysis of the structured relationship of signifiers to each other, and eventually to concepts of the overall creation of linguistic values out of the system of differences. As distinct from signification or meaning, value (“significance”?), the relative salience of one sign over another (“hierarchy”? or “prestige”?), is a possibility afforded by a mechanism in language. It is here that we may look for a model of the figure or the trope that seems essential to literary work, yet differs from the figure derived from the syntagmatic and paradigmatic opposition. Saussure discovers, at the level of la langue, a type of relation that supplements the oppositional-structural one that holds for signifiers. This supplementary relationship is called “associative.” The association (1966:125-27) is a kind of knot of differential values and meanings that make it pivotal in linguistic (and potentially literary) creation or production.

As in all his other discussions, Saussure has recourse to an icon or diagram. He distinguishes the syntagmatic relation by means of several features: it is present, formed by a series of oppositions, and it occurs in discourse. His image for this is that of the relationship, oppositional but supportive, between a column and an architrave. On the other hand, associative relations are “in absentia,” and their terms are united only in a “potential mnemonic series” (123). They do not require a speaking presence, and they are not solidly structural as are syntagmatic relations. They have only the imaginary being of memories: as when a “Doric column ... suggests a mental comparison of this style with others (Ionic, Corinthian, etc.)” (124). Based on a kind of material, but not simply historical, “memory,” the associative relationship can be schematized spatially. Contrasting it with the paradigm figure based on syntagmatic relations (the paradigm being clear groupings of ‘theme parts’—e.g., noun stem, suffix, etc.—in a defined and finite series), Saussure diagrams the associative relationship: [126]

A constellation: a “point of convergence of an indefinite number of co-ordinated terms ... [in] indeterminate order and indefinite number” (126). A bundle of differences prior to or at least coequal with those based in opposition. There is no simple pattern here, no structural way in which members of any one of the strands or series could be transformed into and/or substituted for members of another series. Yet this is, for Saussure, one of the two constitutive relationships in language. And Saussure underscores the distinction he is making between this type of relation and the paradigm—a point no structuralist to our knowledge has developed (Culler 1976:48-49; Barthes 1967, prior to his Pleasure of the Text [1975], sees “associations” as impossible in the Système de la mode) . 1. As distinct from the paradigm, which is the arbitrary arrangement of signifiers in oppositional and coequal pairs in a defined structural series (there is no reason, Saussure writes, [1966:127] why the “nominative case” should be “the first one in the declension series”), the associative relation is based on both signifiers and signifieds. We have enseignement linked to éducation (conceptual); to armement (auditory and graphic signifierer-ment) to enseigner, enseignons (a structural verb paradigm); and to clément, justement (both on the basis of the signifier-ment but also vaguely in terms of values, the values of education.)

None of the above associations is awarded priority in terms of truth or correctness, none dominates—at least not in the tableau given. The sign enseignement is here liberated’ in a sense, it is no closer to being restricted or limited in its possible relations either by its signifier or its signified. There is as well a sense that in the strands or links of association any one of them may, because it is not structurally and simply opposed to any of the other strands, equally become a ligation, or a liaison dangereuse. It can acquire dominance in the definitional chain and in effect repress the claims to relationship of the other strands or filiations. 2. The central sign here, enseignement; appears clear and distinct, but there is a sense of diminishment as we move to the periphery. We are taken with the double “etc., etc.” at the end of each associational string: implying and denying the infiniteness of the line.^* This diminishing in power or light is indeed what first reminded us of Lacan’s “points de capiton”—the privileged signifiers that achieve (transitory) prestige over other signifiers. Only in this instance it is a sign that appears central, and then it is only through the impossibility of assigning structural dominance or prestige to any one over the other of its associative links. More relevant perhaps than the Lacanian is the Freudian notion of the “condensation point,” which suggests both a tying up and an unravelling of strands. And more than a simple indeterminacy being constitutive or a source of figurative power here, we have, in a sense, overdeterminacy, or a kind of instability in which all is still possible, nothing is decided. Saussure’s diagram, in both the simultaneous clarity of the conceptual image and the indistinctness (“etc., etc.”) which reminds us of its sign-character, reminds us of its capacity to allude to something besides itself.

The suggestion here, then, is that there is a surplus of both signifiers and signifieds; it seems to us that this kind of supplementarity is parallel to the relationship between image and concept in Saussure’s first and deceptively simple diagram of the sign. Concepts are surpluses over images, and they do not differ in nature from images; but they differ in the sign relation. In the associative relation in language, we have the double and reverse: Concepts and images are surpluses over signs, and they differ in nature from the sign, but in the associative relation they are equivalent or the same. The multiple associative relation is also a “sign” of the type that one reads in “literature.” Not yet oppositional, not yet simply conflictual, its scene is that of a bounded multiplicity. And it is a scene in which ‘meaning’ has the suspended character of a wish, or a ‘meaning to’.

The singular sign, the one::one relation of image to concept in the Saussurian sign, has proved to be an enormously transparent obstacle to literary semioticians; in overattending to the simple linguistic sign and the paradigm/syntagm needed to analyze it, we have failed to explore the system of langue as a potentially more fruitful ground for uncovering literary sign systems.¹¹ Langue, we believe, is not only a set of abstract norms that structure parole; it is in langue (which also means tongue) that the axes turn that twist language structures into new shapes.

It should by now be clear that Saussure’s text has in effect deconstructed itself. Having produced language through the operation of the syntagmatic relation—so tellingly visualized as a Greek column (of a sacred, social institution, of a temple, perhaps)—Saussure proceeds to give us the catastrophe by which the solidity of structure undoes itself. The mannerist conjunction of differing styles (Doric, Ionic, etc.) is not simply an innocent or neutral association of ideas, a mental comparison; it is a profanation, a violence done to the classical, structured moment. The icon of the “association” is not only that of a flowering but of an explosion as well, an original repetition of the movement from classic to romantic....

In the (twisted) movement between the two axes of Saussure’s fictive ‘language-state’ we can trace not a firm structure, but a deconstructing process, that of making the linguistic into the literary, the fixed paradigmatic figure (metaphor) into the open trope. The story of this movement always generates ‘history’ (a fictive history), condemned ever to restage the same scene: the passage not from culture to nature (as in the sign) , but from “culture” (the stable temple column) back to “nature” (the neonature of pure difference prior to systematic oppositions). This passage—variously from social structural institution to free association, from the traditional to the arbitrary, from poetics to literature, from rhetoric to tropology—is the mode of being of modern culture.

Conclusion

Guided by our interest in reading we have here attempted a review of Saussure’s sign(s). Only when one has signs (signifiers with signifieds) does one have reading; only when one has the potential for relationships based on absence or lack (as in association) does one have literature. As Saussure once wrote, literature’s fundamental identity problem lies in the fact that it “continually draws from [prior signs] new meanings” (in Culler 1976:105). And whenever one has the potential to read one has the potential to be uncertain. In the absence of ‘total’ signs one can have criticism, one can have poetics, one can have forms and structures, but one cannot read in the specific sense of relating an image to a concept. And reading always seems to be both below structure and beyond subjective certainty. As an icon of ‘reading’ the ‘associative relation’ seems to us to begin to capture the literary relation, the suspense of not quite knowing, of not yet deciding the hierarchical value and meaning of a sign and its subsystem—its text. In a manner typical of literary obliquity we have most likely worked our way back to Peirce: with Saussure’s sign now seen as a reading, we can see the parallel with Peirce’s “interprétant.”

The great structuralist labor of working out the relevance of the parole for literature—the analysis of the signifier, the image, and the syntagm/paradigm—must never be underestimated. In bringing up the association we may dissipate somewhat the elegant and classic clarity of the purely structural approach, but we may regain language from grammar. Literature surely proffers both figures of speech and figures of thought; and if we can aspire to even half the analytic potential the work on parole has given us by our now having recourse to Saussure’s description of langue for literature, we will be fortunate indeed.