FOUR

Reference

4.0. A discourse that wishes to confront the problem of literature’s rapport with the world necessarily implies taking a position with regard to the various linguistic theories on the sign-referent rapport. Still, the position I have already taken with regard to culture as a complex of “systems” that constitute the subject’s fundamental structures, either as world view and behavioral norms or as confuted and negated institutions—which for this very reason are always present—thereby excludes of itself certain traditional theses.

a) It obviously excludes the Platonic position, according to which linguistic and extralinguistic facts are connected by a fixed, natural rapport—a transcendent rapport in which man has no part. It is excluded because it conceives “systems” as cultural products, destined to vary and to be substituted in time.

b) It also partially excludes the Aristotelian position, according to which knowledge of the world preexists and is independent of language, which does no more than fix such knowledge by means of an arbitrary, even if conventional, rapport with things. I have spoken, in fact, of culture preexistent to and culturally inherited by the subject, thereby constraining the subject to structure itself within culture, to see things in conformity with its paradigms. But I have also recognized in the subject the power to enrich, to modify, and also to destroy systems.

c) It excludes as well the thesis held by certain linguists according to whom one must speak of a fixed bipolar rapport between “sign” and “signified.” My pragmatic discussions on the production and reception of the sign are closer, if anything, even if not integrally, to the theses of the behaviorists: a large part of the signified is caught precisely in the “behavior” of the subject who produces and uses the sign and of the subject who responds to the sign’s stimulus.

d) My position also refuses the thesis of those who postulate a direct rapport between “sign” and “referent.” It accepts, rather, the proposal made by Ferdinand de Saussure, according to whom the sign does not have a direct rapport with things but establishes it through the concept (“signified”). It therefore includes the triadic rapport among “signifier” and “signified” (concept)—constitutives of the “sign”—and among “sign” and “referent.” And it includes, therefore, what is established by J. Stenzel, according to whom “the significative signs of the verbal expression acquire in their turn an objectivity that stays within the knowledge of the speaker and the signified object . . . separating them . . . and nevertheless making the tie between them possible” (1934, 35). This intermediary “objectivity”—that J. L. Weisgerber then defined as “Zwischenwelt” (1962, 58)—is the “vision of the world” constituted by “cultural systems.” For this reason we accept the principle—sustained today by many linguists—that language is not a passive mirror of reality, but, on the contrary, a structure projected upon reality and conferring upon it the forms of its entities and parts, of its relationships and its segmentations (Humboldt, Cassirer, Trier, Peirce, Morris, Wittgenstein, Hjelmslev, etc.). (The principle is beyond every suspicion of Platonism, since this intermediary sphere is not something ontogenetically archetypical, but nothing other than an historical product, a cultural given.)

I have added that within the “intermediary” sphere, along with the “social,” “cultural,” fixed contents we must recognize the “subjective,” variable contributions, either in the form of personal empirical data that fill word outlines with experience, or subjective concepts, or intimate psychological elements, or connotative emotional facts.

It remains, now, to specify that my perspective puts the accent on the fact that intermediary “conceptual systems” in no way exhaust the real, even if culturally they tend to give that impression. “Systems” and “models” simply construct relationships out of the world’s infinite availability. Speaking of scientific “models,” Enrico Bellone recently said (1973, 16):

From the beginning of the nineteenth century, the production of scientific knowledge developed by criticizing every representation of the so-called “external world” which sought to reduce the real to an immutable block of things, given once and for all and characterized by an external absoluteness; the criticisms that science has addressed to this image have always brought forth more evidence that the so-called “external world” may rather be a dynamic organization which systematically escapes every attempt at explanation founded upon criteria of absolute truth, whatever the degree of internal coherence of such attempts may be.

The “model,” then—including the “model” produced by art that does not differ, in procedure, from other types of modelization—does not “explain” the object, which is never reducible to the subject, and is never to be conceived as the elusive shell of an ultimate and eternal “hidden structure” (for which categories of knowledge cannot be absolute, but flexible and relative), but simply organizes the object and constitutes a vision of it. It could simply be defined as a systematic opening of the intellect and the senses to the world (without any implication of ontological truth). Its “cognitive” power consists in the fact that it shows how to see an object, and therefore allows things to be seen that were not seen before.

The conceptuality of language would have to be seen, therefore—to recur to a metaphor dear to the General Semantics school¹—as a “map” with respect to the territory to which it refers. The map is indispensable for finding one’s way around the territory, but cannot be identified with it. Continuing the metaphor, I’ll say that the history of culture consists in a continuous variation of the map, with reorderings, accumulations of detail, destruction, and reconstruction, all intended to satisfy the contingent heuristic demands of the culture itself.

4.1. The situation therefore appears to be as follows: the conceptual screen of language determines our experience of the world, but it is also altered, modified, enriched, specified, by non-linguistic experience. Although one cannot negate the hegemonic position language assumes in cultural systems—above all, because all other systems can be translated into language, and further because it is principally to language that a culture entrusts the transmission of its own contents—it is nevertheless undeniable that language is not the only system of modelization. There are forms of knowledge that are not linguistic.

In a very interesting article—limited, however, to the visual world alone—Greimas mentions the point of view of neopositivist logicians, according to whom signification is constituted by a reference of “proper names” to the world’s objects, and to the point of view of the linguists, who in general point to the incidence of language in their constructions and significations. He then goes on to insist upon the malaise of the major part of those who have seen the material universe exhausted within the closed universe of linguistic concepts. (It might be recalled that Peirce and Morris attributed to language a function that not only represents the world, but creates it directly.) He suggests that in addition to a semiotics of natural languages, it is necessary “to postulate the existence of a semiotics of the natural world and to conceive the relationship between signs and linguistic systems, on the one hand, and signs and systems of signification of the natural world on the other, not so much as a reference of the symbolic to the natural. . . but more as a network of correlations between two levels of signifying reality” (1970, 52).

Undoubtedly the position of linguistic radicalism that idealistically exhausts the world in language, with its two fundamental theses—the one of “linguistic determinism” (language determines thought) and the other of “linguistic relativity” (every language realizes a definite vision of the world)—invokes some specification. It is true that language determines thought, but it is also true that thought, as conceptual result of experience, determines language. Every day, modern semiotics produces examples of structurations and non-linguistic systems that, in much the same way as natural language, establish units—that is, they segment the continuum of the world—they bring elements into relation with each other, structure them into systems, and fix them as values.

4.2. These illuminations, then, reflect upon the material world to show it as a complex of semiotic systems. Greimas states that a “common characteristic of natural signs,” that of “directing to the other from the self” (1970, 53), together with the idea (stemming from cultural typology) that these constructive and cognitive systems allow delimitation of cultural areas insofar as they represent “types” of culture, necessarily lead us to accept the thesis that in any case one speaks of systems of signs, even though there is an exception that we shall see shortly. Thus, when we refer to the rapport between the world of the literary work (of art, in general) and the extraliterary world, we must speak above all of a relationship among systems.

In the first place, given that the literary work is a linguistic phenomenon—and given further that the non-linguistic semiosis of the world is commonly translated into the semiotic system of language—the relationship work/world will be fundamentally constituted by the rapport between literary language and natural language (besides, of course, the rapport between literary language and other literary languages), a relationship, this, that has long been recognized, even by those who have thought along different lines. Literary language, for example, was spoken of as écart in comparison with standard language (Spitzer, 1948; Devoto, 1950, etc.).

But in the second place, and in the light of what I have sustained above, it must be observed that this rapport is not sufficient and exhaustive. The system of the literary work is not comparable only with that of natural language, but also with all other cognitive systems which are not necessarily translated in the common language system but rather are often in contrast with it, and are potentialities of its transformation from stale, closed ideology to up-to-date, open ideology. Further, the literary work can express semiotic systems, not yet present in the culture, which the work reveals by means of its own modelizations. Literature is at the same time “systematic” and “modelizer,” in the sense that “it actualizes systems” (collective) but also proposes “models” (singular). An attentive observation of the phenomenon “art,” in general, will indicate the importance that unclearly systematized “margins” assume. Lotman recognizes these, as we have said, around distinct centers of culture, and, just because they are peripheral, amorphous, or unknown, these limits show the subsistence of a portion of the world beyond cultural codification, a residue of great semiotic systemizations. Among its diverse activities art includes the venture into these indistinct confines of the world. To say that language structures the real should not—as we have said—necessarily imply that the structuration is complete and exhaustive. The real allows—even invokes—further structurations beyond the culturally given.

4.3. To my mind, then, the thesis closest to the truth will be constituted by a third position that, along with that of the logicians and the radical linguists, will moderate oppositions. Such moderation is made possible, however, by admitting the possibility of non-linguistic experience, according to which reality is intuited as objects, classes of objects, and relationships among objects not yet nominated and established, but that invoke nomination and structurization. In knowing the world, then, dialectical exchange between direct experience and linguistic mediation, between linguistic and non-linguistic structurations is possible. This takes nothing away from our recognition of the formidable function of language, which becomes not only the point of departure, but also the point of arrival, of knowledge, thus constituting an indispensable apparatus of psychic, sensorial, and intellective organs, and the fundamental means of segmenting the continuum of the world and of establishing differences and relationships among things.

This position is also supported by various criticisms of the “linguistic relativity” hypothesis, which is identified in the perhaps arbitrary association—all things considered—of Whorf with Sapir.² It has been objected that in his conception of language Whorf was too tied to verbal atomism, obscuring the fact that the signified is not so much constituted by the value of the single word but by complexes and relationships of words (the word itself—as Wittgenstein specified—assumes a precise value only in context), and that, in theory, word complexes can express, in all languages, the shadings of which he speaks.³ A language may not have words that designate “snow that falls,” “snow on the ground,” “muddy snow,” “packed snow,” or “snow that flies pulled by the wind,” distinctions that the Eskimoes possess (Whorf, 1971, 216)—but any language may, on occasion, express such concepts and can do so by means of its own locutions, if not otherwise than in the particular uses of a poetic language that creates appropriate segmentations, distinctions, and relationships, even if such designations, within the boundaries of a specific culture, do not constitute an ordinary social need. This observation is particularly important with regard to poetry, which seeks to make subtle and delicate distinctions of this kind. The value of poetic language lies precisely in the fact that it has expressive possibilities that do not belong to standard language. It is in this sense that Mallarmé opposed the language of poetry to the language of the “tribe.”

Thus, even with all its recognizable dilettantism, Korzybski’s thesis assumes a certain validity, although we cannot share his principle of a completely prelinguistically structured reality: language is always insufficient with respect to the real, the “map” always incomplete with respect to the territory to which it refers, because in the real, as in the “territory,” there are infinite possibilities of observing reliefs, levels, and gradations that had not been considered in conventional linguistic conceptuality.

4.4. In such a perspective, the scientific, philosophic and artistic activities that have been separated and opposed with such care naturally begin to come together again; all formulate constructive hypotheses of the real. It is commonly said that scientific hypotheses always function to verify practice, while artistic hypotheses subtract themselves in principle from every possibility of verification. But is it really true that the latter cannot be verified? This is also a point that must be attentively discussed in an effort to overcome certain prejudices that aesthetics has long entertained.

It is well known that linguistics has never been comfortable with the idea of reference, so much so that the major part of its conceptions leave out the sign-referent rapport as a question that belongs to peripheral fields of investigation. Semiotics has also pushed away the problems of referentiality, as Umberto Eco has categorically done in his delimitation of its specific field. And there is no doubt—as has already been pointed out—that Eco is right, given that his discourse attempts to found a science, and a science cannot be founded without establishing epistemological limits. But can the semiologist of literature found a science of literature—or even a theory of literature—on the basis of an analogous delimitation? It would be to isolate literature in a sterile vacuum that would reduce its function to that of a “semiotic technique” (while the “user” of literature does not receive a “technique” but its “effects”) or to restrict its function to one of mere diversion, without recognizing the fundamental values that tie it to experience, to feeling, to ideas of the world, and even in some cases to their transformations. The decadent principle of art for art’s sake, a phenomenon torn away from life, is quite overdone. With all probability it reflects—as has been observed—nothing more than a typically bourgeois attitude that, by means of such a relegation of art into pure ideal, conjures away every effusiveness on the practical level.

The meditation on literature’s “autonomy” has also contributed to the conception of a literature separate from the world. There was talk, in fact, of “autosufficiency,” “autoreflection,” and so on, concepts that—as we shall see—are not to be abandoned, but neither are they to be taken as absolute. Too often—to refer to times nearer at hand—we are reminded of the Jakobsonian formula of the literary message’s “autoreflexivity,” forgetting that Jakobson himself proposed it as a principal, but not exclusive, characteristic of literature.

Recently, Paul Ricoeur (1975), reconsidering the problems of the poetic sign and remembering the teachings of Nelson Goodman (1968), energetically sustained the necessity for a denotative theory of metaphor. Not only “saying,” but also “representing” is referential. A painting denotes the real as much as a scientific discourse. The metaphor (understood, naturally, in a wide sense: the work is itself a “global metaphor”) sinks the primary sense (impossible) to let another perspective rise to the surface. This too is referential and, as such, is liable to “verification.” By destroying literal sense, the poem renews sense that does not remain closed in the sphere of poetic discourse but provokes a new referential horizon. Such references do have a hypothetical nature (“suspended referentiality,” 278) but one that, fallen into an extraliterary context, is recognized as “true” or “false” (“metaphoric referentiality”).

To go back to the distinction between “signified” and “sense,” again proposed by Antonino Buttitta and Mario Giacomarra (1972), the “signified” of signs becomes “sense” when the sign has fallen into the socio-cultural context: “The sense of a sign is given by its relationship with a determined extralinguistic situation; it depends, that is, upon the function that a signifying set performs in the heart of a situation” (32). Transferring this concept to our consideration of literary work, recognized and studied in its original historical context, or in later, diverse historical contexts, we can simply say that the verification of a work’s referentiality lies in whatever “sense” it demonstrates at the moment of reception. To have a “sense,” and not only a “signified” (which, with a maximum of historico-philological approximation can always be reconstructed) means that some connections have been made with the referents that belong to the receiver’s socio-culture. The above-cited authors conclude as follows:

It is thus from an investigation into the situation, inasmuch as it is a socio-cultural context, that the sense of a word, a gesture, a work of art emerges. Just as society promotes the creation and use of an historically determined language, so only knowledge of the sociocultural context will allow us to further semiotic investigations, beyond the moment of the simple signified (32).

This, to my mind, is the way to found a semiotics not only of the signified, but of the referent as well.

4.5. These concepts of “metaphoric referentiality” and “sense” open, as we see, a clear perspective on the question of the work of art’s referentiality but, according to the directives of our discourse, the phenomenon can be further specified.

I spoke above of an intermediary sphere (Weisgerber’s “Zwischenwelt”)—hereinafter referred to as (Z₁—constituted by the conceptual plane of language, a conceptual plane that is placed between the subject’s knowledge and the world, and gives to this latter order and sense, insofar as it is a determinant of experience. But we have also added that the autonomous experience outside language (E) is a determinant of language itself, just as is the other non-linguistic conceptuality constituted by the semiosis of the natural world (Z₂). Now, in literary language one verifies naturally the same conceptual function with regard to the real that one recognizes in natural language and in the semiosis of the natural world, but the conceptual plane of literature is another conceptual plane (Z₃) equipped with its own laws, that establishes a triple rapport with (Z₁), (Z₂), with (E):

From one point, the conceptual plane of literature refers to the world organized by common language and by the modelizations performed by philosophy and science (Z₁) from another, to the world of natural semiotic systems (Z₂), from still another, directly to empiria (E), with operations that explore zones which are not culturally systemized, denoting the non-denoted. By means of such neologistic operations, literature inserts itself into cultural “vacuums,” offering itself as an extremely refined means of picking up shades of meaning, of creating “words” for the most subtle and linguistically unexplored experiences. And in that sense it returns to the origins of the world, upsets the ordinary denomination of “natural language” and “artificial language,” inverting their values. “Artificial” language becomes, if anything, standard language.

With this conception of relationships, (Z₃) becomes an experience field organized in itself that reflects and clarifies world situations and can also constitute itself as a new and opposing systemization, different both from that of ordinary language and that of natural semiosis and science.

If what we have said is true, we see that at the level of the rapport with the extraliterary world, literature also activates an “introjective doubling” symmetrical to that of the other two vertices of the communication model expounded in the first and second chapters. In its complexity, the phenomenon can be schematized as follows:

R . . . n = external receivers

4.6. Perhaps it is still necessary to add something about this and about a possible misunderstanding that could arise from what we have stated up to this point. The possible misunderstanding regards the concept of the “autonomy” of literature which could be subverted if one were to put the accent decisively upon the literary text’s “referentiality.” However, the concept of autonomy remains valid even if the referential rapport we have been discussing goes back to reconsider the function of art as “mirroring.” Undoubtedly, art counts among its activities that of using, in the terms and modes proper to it, the states of the objective world. Art can also be—and is—imitation. But at the moment when thoughts, feelings, characters, sensorial data, happenings, and cultural systems enter into structure, these assume a reality that is (a) only hypothetical, and (b) obeys laws that are its own. Pierre Macherey adduces an elementary example:

The “Napoleon” of whom Tolstoy speaks in War and Peace escapes historical criticism. If we read the book attentively, we are aware that the name hardly designates a real person, but has a sense only in relation to the textual set in which it appears. The writer’s act . . . raises an object and at the same time institutes the unique evaluation norms to which the given object may be subordinate” (1966, 57)

It will be well to add however that Tolstoy’s “Napoleon” is a fantastic hypothesis of the Napoleon of history and is not conceivable without reference to reality. In fact, Macherey himself affirms that art is “autonomous” but not “independent”; “it institutes the difference that permits it to be what it is only by establishing some relationships with what is different from itself” (65,67).

The autonomy of literature is above all tied to the laws of language. Even when the work proposes to “mirror” or “imitate” a real object—let us say “describe” it—it must not only make “choices,” given that not everything can be described, but must also obey lexical and syntactical laws, and must then arrange these choices in syntagmatic order. While the real object is shown in its totality and simultaneity, the literary object is “read” as a complex of relationed signs. It happens thus that the word is a way of organizing the real, in the sense that by constituting itself as structure it produces an order that is at once the order of both subject and object. We repeat that the signs of literature do not differ in this from those of natural language. It is only that literature does this on a non-conventional, a different level from that of ordinary language, and separates itself from the latter inasmuch as it institutes another conceptual level that is set before the ordinary plane of reference. The referential plane of literature represents a proposal, or project, to organize (structure and give meaning to) the world different from any plan one may obtain by means of ordinary linguistic structures. The plane (Z₃) consequently becomes a fresh experience of things, free from habit, the obtuse and distracted torpor of ordinary experience. It seems to me that this substantially coincides with what Cesare Segre has recently written on the theme of mimesis and lie:

Departing from reality, fiction renders our perceptions of the real more refined and sensitive, corroborates our critical faculties, reveals, through paradox, forces and motivations. So much the more can the reality from which we exit be truly interpreted by the irreality into which we enter (1978, 185).

In brief, the autonomy of literature is one obtained with the materials of language; but these undergo fresh actualizations and transformations within language itself. And the conceptual plane which they produce does not establish a direct rapport with the contextual situation (natural language is always “in situation”) but is separated from direct contextuality, to actuate an indirect rapport—metaphorically referential, Ricoeur would say—that, so to speak, reproduces the real on an imaginary and hypothetical plane.

Lazar Osipovic Rèznikov (1964; 1967, 69) has stated that the conceptual plane of language (signified) is constituted by a certain quantity of information surrounding the object (referent), with which it never identifies itself. “The sign designates the object—he says—and expresses the signified” (70). Well, now, still in brief, and in this perspective that substantially coincides with what I have myself indicated, I shall say that the function of art rests in its capacity to furnish a non-conventional quantity of information about its referents.