Interpreting images is at the core of daily experience. We just have to look around to recognize the action of images. They circulate plebeian, but nimble, throughout the social environment. Images simulate situations that help us reach decisions, whether political or scientific. Personal identities spring from the pictures on driver's licenses that certify and authorizes us. Images help to build us up. The intensive action of representations in daily life is reason enough to discuss the prevailing assumptions in reading images. The purpose of this paper is to criticize the two dominant presumptions about images -- one, mimetic; the other, conventionalist -- as well as to unveil an alternative.

Toward Mimesis

Recently it has been presumed, and with little dispute, that to understand what images do we must go beyond notions of mimetic correspondence. Not surprisingly, mimetism, the cardinal conception in the Western tradition of evaluating images, could not hold its ground. The mimetic criterion for the production of images is tied to oudated historicist expectations: It indicates that representations have the exacting goal of securing a perfectly realist rendering of nature.

Nowhere this is as clear as in a tale from Pliny's (AD 23-79) encyclopedic Naturalis Historia. The tale may be be apocryphal, but it can also be deemed the expression of an extreme fascination with mimesis as criterion of skill and principle of judgment in the production of images. To win a competition of dexterity in painting, Zeuxis finishes a bunch of grapes so perfect that birds come to peck them. Responding to the challenge, Parrhasius paints a curtain so perfect that Zeuxis himself asks to have the curtain drawn so that the picture could be displayed. But there is no painting underneath; the curtain is the painting. The painter of curtains won; he had duped more than just birds; he deceived a fellow-artist. In any case, the story reveals that mimetic image making is essentially a technical affair, whose objective is to replicate reality. Norman Bryson (1983, p. 1) considers this story the central anecdote summing up "the essence of working assumptions in Western Painting."

It is important to realize that, in the anecdote, the major reference was not to an error of observation in the case of birds or even a skilled painter. Pliny simply chose to express wonder for the technical competence of painters; their efforts were to satisfy our desire of illusion to the point of confusing image and reality. What led us to believe that images should aspire to replicate their objects? Surely, it is not difficult to recognize that mimesis is necessarily an after-effect of the disposition of non-mimetic shapes and forms in a given visual field.

In its immediacy, an image is a visual configuration resulting from the placement of its material elements. A brush touches a white surface; the point left on the paper disturbs and changes the extending surface. The Modernist and non-figurative painter Wassily Kandinsky (1866-1944) would say that the point is perfect concision, the frontier hedging silence and words. Thus, in the visual record of language, at the end of a written line, and after the point signifying a stop, we plunge into silence.

The dead light of a star, bells ringing, the drip-drop coming from a roof, the splintering of a fallen glass, the insolence of a clock ticking the night away could be represented by points. These natural events are then made into a visual form. The point captures the events that shed behind so many of their physical properties, such as noise and rhythm, or else a ghostly shadow in the case of the dead star. The point rules -- in it, nature finds a niche.

Also writing from the outside of the mimetic tradition, Paul Klee (1879-1940) would claim that images do not reproduce what is visible; images "render something visible" (Klee, 1973 p. 34).

Kandinsky (1970) wondered whether a point can produce a complete image and be a work in itself. It can, if we relate the point to another visual element, thus harmonizing point and plane. Kandinsky (1970, p. 45) presented the following example (Figure 1) made of the two most basic and simple forms -- the point and the plane -- as "the first image of any pictorial expression", and from which other images can derive.

Figure 1

Klee (1973) sees the matter in a very similar way. For him also, the point demands a plane; "before reaching the paper, the drop of ink was a point. After hitting the paper, the point expands in a spot, revealing what was hidden in the point, a field of forces just expressed through lines.

Figure 2

Time becomes a factor at the moment when a point moves and is made into a line. The same thing happens when a line engenders a surface, deploying itself" (p. 37).

In this image (Figure 2), the line travels through the surface of a page; it is the track left by the point, thus expressing time. New elements are introduced in image making. Slowly, the complexity required to imitate the outside world is put together. The conditions for mimetic representation spring from the interplay of non-representative forms. From the Modernist consciousness, we can discern the action of material and non-mimetic elements in the construction of mimetic representations; Renaissance mimetic images are perceived now in a different way.

As Meyer Schapiro (1969) mentioned: Historically, imitation had to come later, only after the initial visual and irregular fields of cave paintings were transformed into a continuous and neutral plane. In a flat and blank surface, imitation of the natural world becomes a possibility. An autonomous support, like a window from where we gaze at what is happening outward, is an indispensable condition for representation in three-dimensional depth. The visual field must be empty; nothing should interfere with the aim of displaying perfection in depicting nature. The flat and blank surface means more than an absolutely empty and idealized space. It is as though the spectator is left facing a window frame or a wall where images can arise. Images are certainly more than mere colors, shapes, and volumes. Colors, shapes, and volumes represent reality.

Representation of reality is not the first feature in image making. Images representing tangible events require consciousness. To see an image as a positive representation is to judge it. And what are we doing when we judge images as accurate representations of their objects? At this point, we leave mimesis as a visual problem, and transform it into a philosophical question -- the question of truth in representations.

Mimesis and Knowledge

As in many footnotes to what constitutes Western tradition, Plato (c. 429-347 B.C.) is the starting point. For Plato, image making must be understood as a part of the process of cognition. In this process, we find a perceived thing, its name, its definition, its representation, and finally, on a higher plane, understanding and true knowledge. There are multiple objects named by the word "circle," that can be drawn with a compass, and understood or else known as the concept circle. The different stages in the process relate to one another by mimesis. The object, the name, the definition, and the image represent mimetically the concept circle, in itself the full, the prior, the true, the immutable, the objective, the perfect, the timeless form (eidos).

Seen that way, mimesis is everywhere. Events in the natural world -- trees, tables, chairs, organisms, human beings and their behavior, just to name a few -- are no more than pictures of a divine and true order.

Thus, in the universe, there are three distinct hierarchical levels: 1) a peerless world of intellectual and perfect forms; 2) the world reached through the senses and that in its multiplicity copies and deforms the ideal world of forms; and 3) copies of copies, a world three times removed from the universe of forms. Copies of copies are mutant and false, as incorporeal as the glitter of sunlight dancing in the crest of waves, or in the surface of mirrors, both being simulacra of existence.

In Plato's argument, presented in Book X of The Republic, the carpenter builds multiple chairs through copying a unitary conceptual model -- chairs have a common form. On the other hand, the painter representing a chair in a painting does it from the viewpoint of a spectator gazing at the object. The painter copies a copy, produces a simulacrum, that is three times removed from the truth emanating from the ideal model. The painter is below the carpenter since he is hopelessly tied to the world of appearances The painter could never depict Truth; the painting is a phantasm of the essential reality.

Only the form, the perfect and unchanging essence, can be free from mimesis. The world of the senses is surely an imitation of the ideal world; all arts tend to be imitations of imitations. Even music is imitation. We have fragmentary and scant knowledge, most frequently based on textual sources, about the kind of musical works Plato had in mind when he said that music was mimetic. But he certainly felt that, being a principle affecting the universe as whole, mimesis would qualify both the musician and the painter -- "two names which you give to two other imitators" (Plato 1989, p. 458; Cratyllus 424). Or in The Republic (399a) when Plato (1989, p.644) refers to musical modes that "fittingly imitate the utterances and accents of a brave man who is engaged in warfare or in any enforced business."

With the exception of ancient Egyptian painting, Plato argues that representative arts are too close to deception. Music, painting, and dance imitate imitations. Image makers thrive in the mastery of illusion, and always produce contradictory images; therefore mimetic artists must be praised for their skills and banished from social life.

After Plato

As a direct consequence of his moral posture concerning images, Plato recognized the social import of art making. Society should control its images. Images ought to be judged in terms of unambiguous ethical principles and never be left solely in the hands of their makers and consumers.

Illusionary representations are definitely corrupting. The goal of having a stable and perfect society -- in Plato's conception, the ultimate goal of any society -- demands a similar mode of image making. Greek art during Plato's time did not do this, but Egyptian art followed rigid pictorial standards. In Egypt, "painters and practitioners of other arts of design were forbidden to innovate on these models or entertain any but the traditional standards" (Plato, 1989, p. 1255; Laws II 656e).

Concerning painting, Plato stated that what painters using perspective did was immoral. They distorted the proportions of reality, adjusting real properties of represented things to the demands of a gazing eye. How things truly were, how they corresponded to timeless models, receded in detriment of the task of representing the conditions of vision. Plato abhorred perspective. He preferred the modes of essence (objectivity, stability, permanence, eternity and truth) to the subjectivity of the point of view, from which representations in perspective were built. Plato recognized an exact parallel between the techniques of the painter, producing from afar the impression of reality, and the rhetorical tricks of the sophists (Schuhl, 1952, p. 52). Painters and sophists were content with imparting a partial and deceitful impression of reality. Truth was never their concern.

After Plato's arguments, it became impossible to deny that image making had a definite social impact. State censorship of the arts seemed justifiable. Yet no matter what he had to say about the pernicious social effects of perspective, later on, Renaissance painters chose to employ perspective as a dominant pictorial technique. To be an illusion, to be a copy, did not bother Paolo Ucello (1396-1475). Ucello would stay up late at night, exploring the mysteries and congeries of foreshortening, muttering to himself what a marvel perspective was (Vasari 1991).

Plato's critique of perspective had no lasting impact over painting, because his use of moral arguments to assess images is really a departure from the main issue of representation. For an image maker, the question is whether we should represent things and their properties as they actually are linked to each other or else how do we see them -- in other words, how do they appear to the eye.

Paradoxically, Plato had dispensed -- with a pejorative purpose that would be easily lost -- the idea that the ruling principle in nature is imitation. Then, why not imitate what the eyes see? And moreover, there is always the feeling that the objective representation supported by Plato is rather incomplete. If the image maker decides to represent things respecting the objective properties of things, the spectator is inevitably purged from the depicted scene. Is that fair? Certainly not, if we agree that images are always perceived by spectators: Truth requires witnesses. Could we call perspective an illusion? Or is perspective an alternative mode of truth? If perspective representations were in any form a valid and truthful representing truth, imitation would not qualify as necessarily inferior to objective truth. The ideal of perspective may not be to cheat the viewer, but rather to present a careful strategy of representation, simultaneously checking gaze against the understanding of the spectator, throwing vision against knowledge, thus conveying pleasure and awe to the viewer.

Furthermore, if we look at the history of painting, we realize that perspective became a ruling principle for image making. Plato's judgment of perspective was never taken seriously.To Renaissance painters, perspective was more than a common and technical figurative option. The intellectual importance of perspective is evident if we remember that, during the Renaissance, painters also wrote texts explaining the new pictorial procedure: Leon Battista Alberti (1404-1472) wrote Della pittura, Piero della Francesca (1410/20-1492) drafted De prospectiva pingendi, and Leonardo da Vinci (1452-1519) put together a Trattato de la pittura. So what is the meaning of such concerted effort?

In the footsteps of Cassirer's (1955a & 1955b) suggestions, Panofsky (1975) would claim that perspective is a symbolic form convening an intellectual content, central for a specific historical period, to a sensible mode of representation. Perspective cannot be reduced to the transposition of eternal conditions of gazing. Being a convention, perspective does not represent vision; it presents one of the possible representations of seeing. And through perspective, Western culture would provide a way of interpreting the world. Nature is figured as in a picture, where the gazing subject faces the object, and contemplates the world as if standing before a window. Made into an image, the world's traits are persistence, immutability, and permanence. It is the same conception of nature and knowledge that led René Descartes (1596-1650) and other post-Renaissance thinkers, all being part of what Martin Heidegger (1962) calls The Age of Representation.

The arrangement of pictorial elements on a canvas is parallel to a distinctive conception of the world. In perspective, the depicted scene unfolds according to the idea that "all appears in proportion; nothing seems out of place; all of the parts fit nobly, over one to the other." (Holly 1996, p. 41). This principle defines how the world can be seen -- it goes beyond optics; it is a figurative principle; and through it, reality will be recognized as a unified scene. We are one step closer to the modern idea of space stated in the mathematical physics of Galileo Galilei (1564-1632).

Galileo (1953) pronounced that the book of the physical world is written in the alphabet of geometry, with circles, triangles, and squares. The idea of a common geometric language underlining the universe implies an infinite cosmos, ruled by universal and homogeneous laws. So, natural science conceives the world in a fashion analogous to what perspective had done before. The center from which the physical world is ordered is not the one from which a god-like entity looks at the world. The world is represented always from the point of view of an spectator. Movement, for instance, is not a property in itself; it is not like Aristotelian physics would say -- a moving body searches for its natural place in the hierarchy of the world. In the new image of the world put forward by Galilean physics, there is no hierarchy in the universe. Movement is shown in relation to a precise point of view; a body moves if it changes position in relation to a state of rest. Movement can also be reduced to a mathematical expression. There is no difference between this idea of space and the conception of spatial order that perspective puts forward. In both, as a direct consequence of the precedence given to the viewpoint, we find the idea of a homogeneous and infinite space, represented through mathematical, geometrical and linear means. Homogeneity of space reached from a spectators' unique point of view -- the beholder's gaze -- is the stipulation for a mimetic depiction of the natural world.

From Renaissance paintings to contemporary photographs the precepts of images remain the same. Images aim at being mimetic. Mimesis is an ideal representation putting the spectator at the center of image making whose goal is to render as faithfully as possible what the viewer is seeing. Perspective is a fair equation -- the image is directed toward the object that in its turn is made into a picture. Gombrich (1993) would then say -- perhaps the world may never look like an image, but an image may be made to look like the world.

Triumph and Crisis of the Mimetic Tradition

The mimetic ideal, grounded by the immense success of perspective as the main representational technique after the Renaissance, found its peak in the invention of photography. The historical debt is unquestionable; photography would not be feasible without the pictorial developments of the Renaissance, most specifically Leonardo's camera obscura. This statement can be regarded as almost a commonplace, but how could we forget that the principles of photography are akin to what guides representation in perspective -- the point of view is fixed, and the representing image is registered according to what is seen as coming from out there. With one momentous difference -- the painter would trace over the image projected from the camera obscura; whereas, in the case of photographs, a paper coated with a chemical emulsion sensitive to light fixes the projection coming from the outside.

Photography is then widespread, and mimetic simulation takes a radical turn. With photography, the mimetic accomplishment seems unsurpassable. No hand-made drawing can aspire to what photographic images attain. But, strangely enough, if photography with extreme realism simulates reality, it is never reality itself. The lesson to be drawn is lucid -- not even a mechanical device can achieve the mimetic ideal. Photographs make reality eerie. Photographs are the negative of a presence -- everyone knows that inside the camera, the photographed object is initially perceived as upside down -- resulting from formal decisions made through the handling of photographic equipment, therefore determined by the use of a certain lens, as well as the aperture of the camera's diaphragm, and the time required for exposing the film. These decisions are what establish the appearance of the photographed object: They are the image.

We must bear in mind that not far from now, the chemical technologies that were indispensable to photographs will be extinct. It is just a matter of time. Digital cameras still receive light coming from the outside world through lenses, but the radical difference is the digital mutation of light into bits, into information. Like never before, the transformation of chemical reactions into bits releases photographs from the strictures of representing what is a presence in nature. The possibilities of representation through images have become virtually without boundaries. Elements originally hauled from reality can be substituted and transformed to an unthinkable degree. Photographs will transcend the rigid role of witnesses and providers of evidence. Photographs will recast and subvert what was once regarded as reality.

But to see a photograph as the mere actualization of technical procedures can be truly disappointing. No wonder that the viewer of early portraits found that the traits of the person were captured with precision, yet the results were always "cold, schematic, and trivial" (Freund, 1974, p. 17). How would it be different if the technical obstacles to photograph were so great?

Nicéphore Niepce's (1765-1883) famous photograph of 1822, whose subject was a table with objects, took several hours to endow the subtle traces of a still life. In 1839, the time of exposure was considerably reduced -- it lasted 15 minutes. Two years later, the time was two minutes. In 1842, no more than 20 or 40 seconds were needed to capture a picture. The history of photography is also the story of how the instant was gradually seized. Photography reminded us that the plain reproduction of the conditions of vision did not suffice. A picture is rendered when it borrows from other visual solutions. The first photographic portraits had to be extensions of poses codified by traditional paintings.

Suddenly, we discern how images depend upon other images. Nude pictures echo the poses of Jean Auguste Dominique Ingres (1780-1867). Che Guevara dead on a table, surrounded by his captors, reflects Rembrandt's (106-1669) Anatomy Lesson. In these cases, and in many others, the reference of an image is not its referent, but other representations.

It is difficult to believe that nowadays anyone would seriously uphold the mimetic ideal; the experience of being surrounded by proliferating images clearly hints that representations refer to representations, not to the outside world. The obsessive growth of images drove us away from referents. Actual experience was purified to the brink of hallucination.

How can we accept the idea that images should match reality, and so be dependent upon a correspondence with the natural world, when images have become more realistic than reality? Nature is now just the appearance of appearances; and the represented object matures into a pretext that we are urged to forget. In contrast with the traditional mimetic premise that images have primarily something to do with either fidelity or truth, we are immersed in a world of images that could not care less for the loss of referents.

Images do not aspire to capture reality -- they have become the normative pattern bestowing reality to events in the world out there. A bride enters the church to be married. The nuptial march is performed. The guests oversee an event unfolding primarily to flashes and cameras. The event will be later resurrected as the photograph album is handed from person to person, or when the video tape is projected onto a screen. The resulting impression is that we do not live, we pose. What matters is the transmutation of the event into images. Experience is turned into representations. Living experience is pale in comparison to the simulation that images offer.

Concerning images, the natural world does not rule completely anymore. Presently, images are forceful, autonomous, and unquestioned representations. Images are turned into enchantment and enigma.

It is also surprising that mimetism can be found almost everywhere in daily experience, in holograms, in computerized simulation, in machines of virtual reality, while being at the same time almost absent from several authors interpreting contemporary image making (Bryson, 1983; Moxey 1994; Bryson, Holly and Moxey 1994). Images can act as surrogates of reality, without being reality itself; instead they are signs.

In such intellectual landscape dawns the idea that semiotics -- the general theory of signs -- can furnish the interpretative tools to deal with images. Even if we accept without challenge this argument, there is an important question to be asked. What kind of theory of signs is needed to deal with interpretation of images in an age where theory refuses mimetic explanations in all its forms? A specific brand of semiotics, the conventionalist interpretation, becomes an alternative to the mimetic paradigm. This approach has been itself entangled in what appears to be a hopeless predicament. I will present and critically examine the conventionalist hypothesis and will then argue in favor of another semiotic approach to the interpretation of images free from the strictures of rigid conventionalism.

Roots of Conventionalism

Rigid conventionalism asserts that the communicative power and immediacy of images come from the fact that their program of production is shared by makers and consumers of visual creations. The plain implication of this assumption is the belief in the existence of a productive force in visuality, generating and determining its possibilities. From that view, strict conventionalism fosters a deterministic conception of history and creation. That would appear to be enough to discard it as a productive theoretical option in interpreting images.

But how could we argue against conventionalist interpretations? Conventionalism presides over a distinguished intellectual legacy, cutting across several fields of the humanities. And, furthermore, conventionalist theories present a plausible answer to why images have a communicative dimension. The result is that there seems to be no other account for the social import of images as messages.

We have gotten so used to conventionalist theories that they are supposed to be immune to criticism. However, conventionalism deploys a theoretical strategy that naturalizes a distorted and fragmentary theory of signs, in its claim that images are just composed of two distinct elements, form and matter, made into one by the powerful action of rules shared socially. Conventionalism relies upon a sort of linguistic and anthropological imperialism -- bonded as the idea of culture -- excluding such a plethora of signs that it may qualify as unsatisfactory reductionism.

Conventions are a belated and precarious force in sign production, being no more than unfulfilled conditionality. As unfulfilled conditionalities, conventions cannot be deemed the exclusive, the central, or the ruling element in the construction and the interpretation of images. So why is this tenet so omnipresent in contemporary humanities?

With any paradigm a theoretical frame conducts knowledge and interpretation at a certain historical moment. The mimetic paradigm was left behind and displaced by conventionalism. Therefore, conventionalism has to be assessed as the height of a revolutionary conscience that subverts a previous disciplinary matrix dominating the tradition of interpreting images (cf. Kuhn, 1970). There is no doubt that one important moment in this theoretical revolution is Gombrich's (1961) redefinition of mimesis, in Art and illusion. Gombrich transformed mimesis into the historical unfolding of conventions, generated to match nature. At this moment, it is evident that, in ideas about images, nature is transformed into a secondary element. Conventions are regarded as what predominantly defines image making.

However, the origin of conventionalism is not in Gombrich. It goes back to Kantian aesthetics. And certainly, in its debt to German tradition in art history, Gombrich's analysis of mimetic illusion is only possible if mimesis is denoted as a normative criterion. And this is the working of Immanuel Kant (1724-1804); for Kant buried the dominance of nature in aesthetic considerations, displacing nature in favor of genius.

Since Kant, we observe a progressive retraction from the idea that visual creation is caged in the normative precept and that the imitation of nature is the task of image making. There is a distinction between image making (a product of a different nature, namely human nature) and 'ordinary nature' per se (Kamal, 1986, p. 18). Images convey the impression of being nature, but they are a supreme and unobtrusive creation. Rules include nature and the tendency to regard them as self-contained.

In the same way that Kantian ethics and its categorical imperative are an ecumenical obligation, or, in other words, a universal law ruling individual purposes, genius is not an absolute free agency. Genius is more than the action of an individual. In Critique of Judgement, Kant (1952, p. 169) asserts that the genius does not know how the ideas for creation "have entered his head, nor has he it in his power to invent the like at pleasure." Guided by gifts bequeathed personally, genius creates rules "to be followed by another genius" (Kant 1952, p. 181). A rule is created and matures into a collective triumph.

Now, Beauty obviously does not descend from nature; it comes from its value as representation. Representations ascend to the foreground and nature recedes to a status of an unprovable conjecture. Useful conjecture, but unprovable, as Kant (1966) argued in Critique of Pure Reason.

The things-in-themselves (noumena) cannot be known, except as phenomena, which means representations. But we can think about noumena, even if it leads us to antinomies, in other words, unresolvable contradictions. Being outside the reach of the human mind, the things-in-themselves are beyond rational knowledge. Kant (1966, p. 292-293) states that it is impossible to "positively extend the sphere of the objects of our thought beyond the conditions of our sensibility, and assume besides appearances objects of pure knowledge, that is noumena, since such objects have no assignable positive meaning."

With Kant, we are led into an intellectual system where the dominance of rules is unquestioned. The utmost creation, the achievements of aesthetic genius, begins with the breaking of rules and ends up with the constitution of rules. The grounds for conventionalism are beginning to be paved.

Slowly, the idea of culture ultimately emerges, and representations in given surroundings will be conceived as part of an organic whole. It is with Ernest Cassirer (1955a & 1955b) that the Kantian idea of an a priori conducting experience is redefined as a set of collective representations known as symbolic forms that comprise culture. Now culture is the dominant factor in human experience; it precedes possible experience. Nature is entirely omitted from this theoretical viewpoint. An image is seen as determined by the totality of its culture. A condition to interpret a message circulating in a society becomes the unveiling of its generative rules, conventionally shared.

Culture functions as an a priori, molding individuals actions. The system of cultural institutions becomes similar to a unitary subject, omnipresent in historical phenomena. Culture creates a symbolic net, a filter, allowing us to move among other human beings, as well as our natural environment.

The symbiotic nature of culture, codes, and symbols is nowadays almost a platitude. Cultural codes are deemed vital for filtering and repositioning sensorial stimulus. To sustain this view, Roman Jakobson (1971, p. 339-340) quotes the observations of M. Aronson who, in a radio broadcast, tried to transfer natural noises of a train arriving to a station. The effect was disastrous. The informational value of natural noises was next to zero. There was no way of discriminating them clearly without conventions. Even to copy reality we must use a mediating schema. The schema organizes our perceptual experiences and vindicates the necessity of considering codes which are operative in a cultural milieu. It is the same idea presented by structural linguistics. For post-Saussurian linguistics, a prior system of rules is a condition for any utterance. Conventions, fixed as rules, abate the arbitrariness of the linguistic sign.

Conventionalists could then profess a rupture with the mimetic tradition. And rightly so. But I will argue that conventionalists do not step aside from the fundamental metaphysical tenet of mimetic theories.

Mimetism and conventionalism are intellectual solutions that present different replies to the same problem; they deal with the same elements. Mimetism and conventionalism surmise the existence of two distinct natures -- form and matter, mind and matter, or conceptual and corporeal elements -- fastened to one another by a third ingredient, a mode of relation, effective enough to bridge this gap. Mimetism alleges that similarity is what links matter and form, whereas conventionalism professes that a rule joins what is different; they are the extremes of a long-standing tradition.

At the origin of this tradition, we meet, once again, Plato's contention that mimesis projects a world of conceptual forms over matter. The stage is set with three distinct actors, hierarchically related to one another. There are two worlds; one intelligible -- a realm of timeless forms, a domain of archetypes -- and the always changing, imperfect, and moving world as revealed by the senses. Thanks to mimesis, mind and matter grow into a couple that will outlast this long-standing tradition. So, however critical Aristotle was of so many of Plato's solutions, he kept the same basic idea -- the objects are a union of matter and form, and form is superior to matter. Form is active, whereas the sensible side of a thing is passive. In this tradition stemming from Plato's essentialism, form imposes continuity to the natural world. We classify objects of the natural world, because we recognize in them the same active form. Besides the duality of form and matter, we will find a third element at work. Again, this third element is a mimetic relationship.

Following the spirit of the mimetic tradition, Cartesian dualism retains the cleavage between mind (res cogitatio) and matter (res extensio), qualified as corporeal nature expanding in space. Therefore, in terms of Cartesian philosophy, mind and matter are irreducible limits tied by adequatio.

It must not be forgotten that Descartes's arguments were presented in his Meditations, whose purpose was to ascertain the existence of God. God is logically necessary as a bridge to the exclusion of mind and matter; if there is a steady and structural split of substances, there ought to be a primal substance where the differences do not exist, and out of which the cleft was engendered. Being divine, the primal substance is prior to all creation. In it, there is no differentiation of mind and matter. It is prior to the duality perceived in the natural world. Adequatio itself is possible because an underlying substance accommodates mind and matter.

The door is opened to Kant's defense of an a priori. Even if the Kantian argument does not call for God or any divine substance, in fact, without an a priori, how is it possible to have adequatio of radically distinct substances? Conventionalism retains the mind and matter distinction, introducing, after Kant, and as a binding factor, a rule or set or rules preceding mind and matter. The function of the a priori is to fuse a concept and a material component as if they were two sides of a sheet of paper.

Kantian conceptions ripple across many disciplines in the humanities. Kant is at the core of the linguistics proposed by Ferdinand de Saussure (1857-1913) -- in itself synchronic, thus concerned with the structural state of a language rather than its historical evolution. The ties of Kant and Saussure are more than clear after pondering Claude Lévi-Strauss's (1964) attempt to apply structural linguistics to non-verbal elements of the social world like kinship. It is no surprise that Paul Ricoeur (1969, p. 55) noticed promptly that Saussurian applications lead to a distinct philosophical stand -- "a Kantianism without transcendental subject". The validity of the interpretation is obvious from the fact that Lévi-Strauss himself never denied the accuracy of this remark.

Saussurian linguistics aims at isolating logic and psychological relations that unite material sounds and concepts, matter and mind, in the guise of a system. The linguistic system looks like a notation preceding any actual performance or utterance. The system is an a priori; two distinct natures, one material, the other conceptual, will be made into one. So, in linguistics, and in all of his semiological project, Ferdinand de Saussure presumes the existence and the predating of an a priori system of rules welding signifiant and signifié, signifier and signified.

The ideas held by Saussurian linguistics would be eventually absorbed in art history that will give precedence to synchronic analytical methods in detriment of strictly diachronic approaches. Heinrich Wölfflin (1950) and Alois Riegl (1931) would then emphasize that art making is ruled by a system -- akin to the linguistic contract -- through which conventions are shared in a community of social actors. The ancestry of both assumptions -- either in linguistics or in art history -- is clearly Kantian. The similarity between historical interpretation of images and linguistic analysis is striking.

In Kant's philosophy, in structural linguistics and in historical interpretation of images, we find the same metaphysical furniture assembled with new screws.

Signs and images beyond the separation of mind from matter

I decline the separation of mind and matter. The theory of signs I defend supposes a continuity of mind and matter, and repudiates rigid conventionalism, whose most immediate effect is the domestication of chance and uncertainty, precisely the most persistent attributes of historical processes. The patterns that may be cast as time progresses do not result from the action of determined regularities. These regularities, so dear to conventionalist theories of sign, such as Saussurian semiology, are a belated discovery. There is a lasting state of indetermination in the endowments of nature, arising from the spontaneity of origins. Chance is never fully controlled: "An element of pure chance survives and will remain until the world becomes an absolutely perfect, rational, and symmetrical system, in which mind is at last crystallized in the infinitely distant future" (Peirce, 1992, p. 297).

In a semiotics derived from such cosmogony there is no room for the notion that an antedating pattern rules signs. We can have nothing but "approximate regularities" (Peirce, 1992, p. 302). The regularities themselves are always precarious. Signs can only be conceived as functions acting as representations, and moved over by endless chance and change. This is a radical departure that overcomes the duality of mind and matter embraced by conventionalism. A renewed theory of representation must start with the admission that representations are manifest through three classes, three stages, of sign production. Signs should not be envisaged as if they were things. Signs are functions of relatives. The logic of signs is the logic of possibilities.

Any representation is, at first, an hypothesis expressing logical possibilities. Representations are nothing but three modes of possibilities: "a may-be, a can-be, and a would-be" (Peirce 1976, p. 868). The three manners of existence intertwine in a subtle way. The path does not go from possibility to reality, and finally crowning in necessity. Necessity can perfectly interpret and mediate the link between possibility and actuality.

The triadic structure of representations is even clearer when we ponder the logical procedures involved in the act of knowing. In his writings on logic, Peirce (1976, 1992) identified three basic modes of reasoning: abduction (also hypothesis or even retroduction), induction, and deduction. Abductive reasoning invents or puts forward a hypothesis from available evidence; it is the origin of knowing. However precariously, induction yields the evaluation and the establishment of the hypothesis conducting any act of reflection. Deduction explains hypothesis, extracting the necessary consequences through which conjectures would be tested. Abduction, induction, and deduction are logical mechanisms organizing the diversity of experience and accounting for our cognitive schema.

Therefore, representations should be patterns structured as a triadic model. Again, representations are functions, not things. And by functions it is meant operations in which if a change is made in one variable, the result is an altogether different representation.

In a sign-function, we would have a material ground of the representation, in other words a first representing a second, therefore its object, and then these two elements suffer the action of a third, in itself an interpretation reducing the arbitrary chasm separating the ground from its object. The third element in the triadic schema is called the interpretant.

This model escapes duality. It is completely dissimilar from all conventionalist models that postulate an a priori system of rules fusing two radically diverse natures -- one intelligible, the other sensible. The triadic model is the condition for refusing the mingling with any kind of a priori assumptions; the best way to qualify the triadic model is to see it as an interminable interaction, as a dialogue (Fisch, 1986, p. 318). It is sheer action, with no hierarchy excluding what is inseparable and continuous. It is a dialogue between equals, an action of signs granting no privilege to anything, not even to conventions.

Signs are more than mere actualization of conventions. They are activities and processes. In them, conventions do little more than indicate possibilities.

The resort to a triadic and evolutionary action of signs, installed by the force of interpretants, is incompatible with the Saussurian semiological conception of signs and its Kantian reliance on both an a priori and the duality of mind and matter. This last sentence could be seen as inconsistent with Kant's undeniable influence over Peirce's philosophical system. It is quite true that, from the viewpoint of interpreting Peirce's monumental oeuvre, Kant is more than a driving force. As early as 1859, and when still a student, Peirce wrote a paper named "The Axioms of Intuition; After Kant", now classified as Ms. 50. He also delivered a Harvard Lecture on Kant in 1865. Furthermore, Peirce's (1992, p. 1-10) fundamental text on the categories, "On a new list on categories", published in 1868, is certainly Kantian in scope. Yet, Peirce's progressive acknowledgement of the importance of biological thought drives him away from strict Kantianism. The first recognized reference to Darwin in his writings is casual, and can be found in the IX Lowell Lecture of 30 November 1866. The text reads: "The great disputes of science have been usually between those who ask for causes and those who ask for classification; and the Darwinian controversy is an instance in point" (Pierce 1982, p. 488). But later on, Peirce (1992, p. 218) would state that "evolution is a postulate of logic", and that his evolutionary cosmology "is only Darwinism analyzed, generalized, and brought into the realm of Ontology" (Peirce 1992, p. 222). In the latter phase of his writings, the triadic model is not anymore the imposition of an a priori grid over experience, but the an illustration of the actual living flux. In that case, Kant and Darwin are divergent thrusts in Peirce's semiotic project. The triadic model implies necessarily acts of endless sign mediation. It is worlds away from correspondence or similarity between distinctively dual constituents.

The principle that applies to all signs -- and thus to images -- is approximation. Signs, including images, are hypothetical constructs. In fact, there is no way that consciousness could have a direct access to the external world. Images, even the ones we regard mimetic are a result of signs acting upon signs. They are an intricate cluster of representations, emancipated from the immediate ideal of corresponding to the outside world. As is always the case, a sign substitutes another sign, and mediated by them, the cognitive mind is led in the direction of whatever exists out of consciousness. With the external world, images entertain a relation of clash, not one of correspondence. The possible autonomy of images accounts for the fact that they can act as normative patterns to perceptions in daily life.

Images are not subservient copies; they are additions succeeding from a fundamentally creative agency. This redefinition of image is the first step to produce the insight of a complex theory of sign production, triadic and not dicotomic, and fully qualified to overcome the exclusion of mind and matter present in mimetic and conventionalist doctrines afflicting the interpretation of images.

References

Baudrillard, J. (1981). Simulacres et simulation [Simulacra and simulation]. Paris: Gallé.

Barker, A. (Ed.). (1984a). Greek musical writings; The musician and his art.(Vol. 1). Cambridge: Cambridge University Press.

Barker, A. (Ed). (1984b). Greek musical writings; Harmonic and acoustic theory. (Vol. 2). Cambridge: Cambridge University Press.

Boorstin, Daniel J. (1964). The image: A guide to pseudo-events in America. New York: Harper & Row.

Bryson, N. (1983). Vision and painting: The logic of gaze. New Haven: Yale University Press.

Bryson, N., Holly, M. A., & Moxey, K. (1994). Visual culture: Images and interpretations. Hanover: Wesleyan University Press, University Press of New England.

Descartes, R. (1993). Meditations on first philosophy: In which the existence of God and the distinction of the Soul from the Body are demonstrated. Indianapolis: Hackett.

Cassirer, E. (1955a). The philosophy of symbolic forms: Language. (Vol. 1). New Haven: Yale University Press.

Cassirer, E. (1955b). The philosophy of symbolic forms: Mythical thought. (Vol. 2). New Haven: Yale University Press.

Eco, U. (1976). A theory of signs. Bloomington: Indiana University Press.

Fisch, M. (1986). Peirce, semeiotics, and pragmatism. Bloomington: Indiana University Press.

Freund, G. (1974). Photographie et société [Photography and society]. Paris: Seuil.

Galilei, G. (1953). Dialogue concerning the two chief world systems, Ptolemaic and Copernican. Berkeley: University of California Press.

Gombrich, E. H. (1961). Art and illusion: A study in the psychology of pictorial representation. New York: Pantheon.

Gombrich, E. H. (1993). Aditional thoughts on perspective. The Journal of Aesthetics and Art Criticism 51, 69.

Heidegger, M. (1962). Chemins qui mènent nulle part [Roads that lead to nowhere]. Paris: Gallimard.

Hirsch, E. D., Jr. (1986). Beyond convention. New Literary History 14, 389-397.

Hochberg, J. & Brooks, V. (1962). Visual recognition as an unlearned ability: A study of one child's performance. American Journal of Psychological 73, 624-628.

Holly, M. A. (1996). Past looking: Historical imagination and the rhetoric of the image. Ithaca: Cornell University Press.

Jakobson, R. (1971). Selected writings II: Word and language. The Hague: Mouton.

Kandinsky, W. (1970). Point, ligne, plan: Contribuition a l'analyse des élements picturaux [Point, line, plane: Contribution to the analysis of pictorial elements. Paris: Denöel/Gonthier.

Kamal, S. (1986). Kant and fine art: An essay on Kant and the philosophy of fine art and culture. Oxford: Clarendon Press.

Kant, I. (1952). Critique of judgement (James Creed Meredith, trans.). Oxford: Clarendon.

Kant, I. (1966). Critique of pure reason. (Norman Kemp Smith, trans.). Garden City, NY: Doubleday.

Klee, P. (1973). Theorie de l'art moderne [Theory of modern art]. Paris: Denöel/Gonthier.

Kuhn, T. S. (1970). The structure of scientific revolutions. Chicago: University of Chicago Press.

Lévi-Strauss, C. (1964). Le cru et le cuit [The Raw and the cooked]. Paris: Plon.

Moriarty, S. E. (1996). Abduction: A theory of visual interpretation. Communication Theory 6, 167-187.

Moxey, K. (1994). The practice of theory: Poststructuralism, cultural politics, and art history. Ithaca: Cornell University Press.

Panofsky, E. (1970). Meaning in the visual arts. Harmondsworth: Penguin.

Panofsky, E. (1975). La perspective comme forme symbolique et autres essais [Perspective as a symbolic form and other essays]. Paris: Minuit.

Plato (1989). The collected dialogues, including the letters. Princeton: Princeton University Press.

Peirce, C. S. (1985). Writings of Charles Sanders Peirce: A chronological edition. (Vol. 1) Bloomington: Indiana University Press.

Peirce, C. S. (1992). Houser, N., & Kloessel C. (Eds.). The essential Peirce: Selected philosophical writings.. Bloomington: Indiana University Press.

Peirce, C. S. (1976). The new elements of mathematics. (Vol. 3). The Hague: Mouton.

Riegl, A. (1931). Das holländische Grüppenporträt [The Dutch group portrait]. Vienna: Österreichische Staatsdr.

Ricoeur, P. (1969). Le conflit des interprétations [The conflict of interpretations]. Paris: Seuil.

Saussure, F. de (1982). Cours de linguistique générale [Course on general lingustics]. Paris: Payot.

Schapiro, M. (1969). On some problems in the semiotic of visual art: Field and vehicle in image-signs. Semiotica 1, 223- 242.

Schuhl, P.-M. (1952). Platon et L'art de son temps [Plato and the art of his time]. Paris: Presses Universitaires de France.

Summers, D. (1994). Form and gender. In N. Bryson, M. A. Holly, & K. Moxey (Eds.). Visual culture: Images and interpretations (pp. 384-412). Hanover: Wesleyan University Press, University Press of New England.

Vasari, G. (1991). The lives of the artists. Oxford: Oxford University Press.

Wölfflin, H. (1950). Principles of art history: The problem of style in later art. New York: Dover.

Notes

ENSAIOS