Preface

Since the dawn of time and with a profusion that still inspires awe, creative men and women have inscribed configurations of lines, marks, and colors on the most diversified surfaces imaginable, providing for their transmission down to the present time. But these visual representations still remain enshrouded in the mysterious aura surrounding the semi-abstract, semi-figurative engravings on stones or the drawings wrought on the immense walls of obscure grottoes. They are as intriguing in their own way as any of the compelling transformations that mark the art of our own century. It could be said, after these millennia of production of visual representations, that the nature of this language and its modes of functioning still escape us.

Despite widely scattered efforts throughout the centuries, we can attribute to a scholar of our own century, Ferdinand de Saussure, a new awareness of the need to develop a general science of all the various systems of signs, that is, a ‘semiology’ in which verbal linguistics would constitute one element among many others.

In this perspective, where analyses effected upon other signs (verbal, gestural, mathematical, and so on) would be completed by a theory of visual signs, visual semiotics would contribute to the completion of this general theory of semiology, as uniquely capable of shedding light on the fundamental mechanisms of the symbolic function.

However, the extraordinary developments of verbal linguistics itself, from phonology to generative grammar and the first levels of grammatical formalization in the 1950s and 1960s, paradoxically, placed at risk this whole semiological project, specifically in relation to visual semiotics. The many difficulties encountered, as well as the contradictions surrounding this endeavor, have already been enumerated, notably by Hubert Damisch (1974) in an article summarizing arguments for and against the possibility of producing a semiology of painting.

At its most fundamental level, a semiology of visual language postulates that visual representations are signifying practices constituting a language. This concept retains the minimal definition of language set forth, for example, by Iouri Lotman: “By language, we mean any system of communication which uses signs arranged in a particular way” (1973; 34). Although any communication implies a speaker and a listener, Lotman stipulated that the communicative function can be exercised by any one individual alone, who is simultaneously the sender and receiver of the message which he generates and which “he speaks” in order to inform himself about an internal or external experience.

From the first tentative steps toward constructing visual semiotics, it is this specific linguistic function of visual language which was called into question. This problem resulted from comparisons between visual structures and those newly discovered structures of verbal language, construed as an absolute paradigm for any notion of language and a model for any semiotics.

The program was clearly defined by Roland Barthes: “The point is to draw analytical concepts from linguistics which we think a priori are sufficiently general to permit semiological research to be initiated” (1967). We can better surmise today, a quarter of a century later, the dangers inherent in borrowing various concepts of verbal grammar, which may appear valid at a given moment but not some time later, and in applying them volens nolens to visual language. Unfortunately, certain significant, if somewhat ‘unconsidered’ borrowings, as Luis J. Prieto (1976; 106) astutely pointed out, were made at that decisive moment when it was necessary to establish the basic concepts of a visual, that is a nonverbal, semiology.

It was readily apparent that analytical concepts so selected a priori could not ensure the construction and development of a semiology of nonverbal languages. As pointed out respectively by Damisch (1976), Eco (1976), Metz (1977), Greimas (1979), and Floch (1985), the research inspired by Barthean positions, while claiming the name ‘visual semiology’ or ‘visual semiotics’ (both terms interchangeable, for the purpose of this discussion) did not mark the beginnings of a new discipline, but simply perpetuated the iconology instituted by Erwin Panofsky in 1939.

These first attempts to formulate a theory of visual language established a dependence of its basic elements, as well as its syntax and semantics, on the constitutive categories of verbal language. The main preoccupations were the following: Did visual signs possess a double articulation? Were they constituted by ‘isolated and isolatable’ elements of a finite number, as were the characteristics of verbal phonetic elements? What about certain signs called iconic, that argue against the arbitrary character of the sign linked to verbal units by Saussure and appear to be motivated? If visual signs did not possess the characteristics of verbal language, they were thought as not constituting a bona fide language but rather a sort of secondary and restricted one, interpretable only through the mediation of verbal language.

The consequences that followed were that visual figures could only be determined and stated through the efficiency of words, and that the regions of the visual field which remained unnameable were excluded from any analytical visual discourse. Ignoring the syntactic interrelations which construct these visual figures, this icono-semiology conferred on them merely denotative meanings through their connections with verbal lexicology, or connotations, as Panofsky intended, at the level of intertextuality, that is, the ensemble of textual sources: literary, philosophical, religious, political, and so on.

The strict dependence of this icono-semiology on verbal linguistics placed it in the same impasse as the one confronting verbal semantics which has failed to attain the same maturity or scientific validity as did phonology or syntactic theory. Even before the advent of contextualism and the influence of the cognitive sciences on linguistics, visual semioticians experienced the same fearsome dread that Hjelmslev observed among linguists, when the time came to analyze the contents of verbal language: “. . . confronted by an unrestricted number of signs, the analysis of the content appeared to be an insoluble problem, a labor of Sisyphus, an impassable mountain” (1963; 67).

The research was sidetracked momentarily by more or less fortuitous games of association in the face of visual images, following the hypothesis that Umberto Eco accredited in L’Oeuvre ouverte (1965), but drawn now from a modernized iconologic discourse resulting from the development of human sciences (anthropology, psychoanalysis, Marxism, rhetoric, and so on). But soon one became aware of the epistemological fragility of these most varied interpretations of visual images, be they photography, advertising illustrations, or artworks. Toward the end of the 1970s, many investigators no longer believed in the possibility of constituting a semiology of visual language on relatively scientific grounds.

The main reason for this epistemological failure is tied to the refusal to recognize the inability of verbal language—through its links with Aristotelian metaphysics and logic as well as by its mode of production of two-dimensional, discrete signs—to conceive of spatial experiences and to account for the functioning of a spatial, three-dimensional language such as visual language.

This discrepancy between the “surface” levels of verbal and visual languages, which was duly recognized by Saussure, had never been considered fully significant. In a civilization dominated by the Logos this oversight has largely contributed to excluding from human consciousness the proper correlative of visual language, namely the experience of space. Succinctly, spatiality can be defined as the apprehension of a simultaneous coexistence of multiple elements in an autonomous form of organization, which is considerably different from that of the temporal order of these elements.

We know that Freud (1979; 285) linked the experience of space to the phenomenon of the unconscious. Just as temporality is unacknowledged by the unconscious, so space would be ‘nonexistent’ to the conscious mind.

Moreover, whereas the human being has two implements, undoubtedly necessary, to achieve access to consciousness, these being the representation of words and the representation of things, the contribution of the latter is misapprehended, being reduced to a temporal and linear succession of iconic objects, that is, of figures resembling some objects of external reality and identifiable by words. In this orientation, the spatial dimension of things, that is, their mode of integration and function, is entirely lost in an ensemble where their meaning would depend on interactions between spatial coordinates.

Basically, the notion of space, which is preeminently pluralistic, must be thought of as anterior to any notion of the object or the construction of the object. M. Merleau-Ponty (1968; 234) suggested this view when he wrote that space is “a communication with the world older than thought.” R. Thom (1972; 327) suggested also that one should consider spatial thought as the basis of man’s conceptual thinking. He advocated a much wider use, in natural and human sciences, of diverse geometrical intuitions as basic to the understanding of a fundamentally spatial reality.

This proposition is not altogether new in the field of art theory. The essential message of P. Francastel (1965), an art historian and the initiator in France of the sociological approach to art, has been that artistic representations of each epoch are indissolubly wed to the geometrical hypotheses of different societies. In following the path thus opened, we have laid down some foundations for a semiology or science of visual signs which finds its basic spatial intuitions in the geometrical discourse of topology (Saint-Martin, 1980). This allowed for the present development of a general syntax describing the modes of functioning of visual language.

Topological relations not only constitute the first mediation by which a human being constructs his notions of reality and space, as described by Piaget (1956, 1960), but they also offer, he showed, the means for modeling fundamental organic spaces. These sensorial, concrete perceptual spaces constitute the fluctuating and pluralistic field to which every spatial representation specifically refers. As has already been well demonstrated by the works of Kurt Lewin in the 1930s, topological models seem to be the most appropriate for the representation of human psychological behavior relating to perceptual activities, affective organizations, or linguistic processes of representation.

In particular, topological models allow us to think of the visual-spatial field as a field of forces or dynamic tensions that are in a state of continual change. In addition, they permit us to account for the material energies transported by plastic elements in the context of the dynamic processes that animate perceptual activity itself. In this respect, it would appear necessary to reappraise the contribution of Gestaltian theory, which has been too often reduced to the phenomenon of recognition of the ‘figure-on-ground’ or the drive for ‘good form’ (Saint-Martin, 1990). Gestaltian research has in fact described visual perception as being realized through a series of ‘visual movements’ of expansion and contraction, of interactions between percepts, which continuously modify already established configurations in the visual field (Kohler, 1940). It provides an essential framework for the elaboration of a visual grammar based on the specificity of visual perceptual processes.

Since the aim of any grammar is to study how a given ensemble of signs is organized in order to convey meaning, the syntactic theory of visual language that we propose here, if far from constituting a genuine semantics, could not be constructed without basic assumptions about the very meaning of visual language. These assumptions also necessarily question the phonocentric dogmatism which has dominated semantics as well as visual syntax in the recent past.

We allude, in particular, to the developments of “visual semiotics” by Thürlemann and Floch, which remain anchored in the structural semantics of Greimas (1966) and his “semiotics of the natural world” (1979). Following the productive research undertaken by Umberto Eco (1970) and R. Lindekens (1971), this semiotics acknowledges the existence of a “plane of expression” in visual representation, in the sense that Hjelmslev defined it as different from the “plane of content.” This visual plane of expression is markedly dissimilar to the verbal plane of expression and thus necessitates an autonomous description of a relatively syntactic order. This description is still in a rough-draft stage of development. Such a syntactic orientation, which must take into account all existing regions in visual representations and their specific interrelations, has in this context no precise function, insofar as it does not contribute to the interpretation of meaning. Indeed the interpretation of the contents of a work is there defined as being essentially tied to the “figurative,” that is, to a verbalized iconic form. This recourse to a natural world, conceived as only furnished by identifiable, nameable forms, overlooks, on the one hand, the decisive analysis of Eco against the use of the iconic sign as a basic element of visual language (1976). It confirms, on the other hand, the hegemony of verbal linguistics with respect to nonverbal semiotics.

Certainly, visual semiotics uses, it its very constitution, the instrumentality of verbal language as well as that of graphic models, but this natural language is no longer a language which refers to natural objects. It is rather a metalanguage, a language which talks about another language. The notions that it develops, the hypotheses and vocabularies that it uses, do not concern the same objects as those referred to by visual language. Those objects of visual representation which are, in some way, marked out by words, are in fact placed between parentheses and observed from points of view and levels of abstraction different from those which constitute them in visual language.

The essential characteristic of a visual language, a language of space, renders it suitable to modelize a large variety of distinct perceptual spaces by which human beings construct their relations with reality (tactile, kinesthetic, thermic, auditory, and so on). Though related, these spaces regroup stimuli quite different from those constructing strictly visual space. These spaces are not denoted or connoted by signs which would resemble, in a mimetic fashion, objects isolatable in the natural world. They are, on the contrary, manifested by structural interrelations of elements, revealed by syntactic analysis of visual texts as representations of nonvisual experiences and constructs.

This syntactic theory allows us to understand how visual language represents, in presenting them in the strict sense, and in “reproducing” them, those dynamic processes linked to our sensory, emotive, and conceptual experience, preverbal or paraverbal as the case may be, in order to render them accessible to a level of linguistic representation which makes consciousness itself possible.

It is, in effect, through the spatial constructs of visual speakers, that human beings can directly recover contact with, deepen, or master their nonverbal experiences. Through them, they can experiment with new modes of integrating the internal/external objects and drives. As new equilibria are attained, the visual spatial constructs can operate upon energies and tensions which, having never been “spoken” at the substratum of being, remain dormant, fixed, and largely unconscious.

We hope that, in opening the way to a more adequate comprehension of one of the most important nonverbal languages—namely, visual language—visual semiotics will bring to it a credibility that can provide a strong balance to the one-dimensional tendency of the Occidental Logos and, in this way, promote the development of a more humanistic civilization.