For over a thousand years in the history of European art, perhaps the principal relationship between "word" and "image" was an asymmetric one in which "word"—the sacred literature of the Bible and the consensus of its theological interpretation—is the major term. "Image," in this tradition, refers to objects as varied as liturgical vestments, sacred vessels, illuminated manuscripts, altarpieces; even whole churches complete with their programs of sculpture and stained glass. These "images," different as they may be one from another, have a common status defined by their relationship to thoughts and concepts expressed in narrative and argument, commentary and interpretation. Even as late as the beginning of the seventeenth century, such images were normally considered to be the product of crafts or mechanical arts and, as such, secondary expressions of theology and the liberal arts.|
In principle, and to a great extent in practice, the craftsmen responsible for making religious images were considered to be something like visual engineers; their knowledge of materials and techniques enabled them to give visual expression to someone else's theological concepts. The images they made are intellectual images in the sense that they are meant to diffuse ideas. They are not representations of ordinary visual experience; they are visual equivalents of a consensus of textual interpretation and commentary, so whenever ordinary visual experience and a theological concept conflict, the theological concept takes precedence. Images in this tradition diffuse the work of scholarship far beyond the faculties of theology. They reach even beyond the circle of the literate, and address anyone who can see. They are a substitute for books.
As early as the Second Council of Nicea (787), "image"—icons or sacred pictures—was formally described as subordinate to "word." The council held that everything about a religious image concerned with meaning belongs to the theologians, the execution alone belongs to the painters and sculptors. The distinction is something like the one that exists today between architects, who design buildings, and the construction crews who actually build them. The architects who designed medieval churches were themselves considered craftsmen; the intellectual layer of such design belonged to the theologians.
It may be a little difficult today to think of Saint-Etienne de Bourges or the Ghent Altarpiece as examples of "the Bible of the poor," but that is how they were seen by almost all literate intellectuals before 1789. Literate intellectuals today often forget the importance this relationship between word and image especially in the case of painters such as Jan van Eyck and Rogier van der Weyden, who worked entirely within this tradition but whose paintings continued to be valued in the context of later conceptions of "visual art."
Since the theological ideas that inform the works of Jan and Rogier are unlikely to be of much interest to people who encounter their paintings today, the painters' attempts to represent these ideas often go unrecognized. As early as the sixteenth century, collectors began to turn these pictures into "curiosities;" by the late eighteenth century, they had become documents in the history of painting, often a history that understood them as "primitive" attempts to do something that later painting does in a more developed way. Today, they are beginning to be treated in the context of theories about images—theories as foreign to the masters who made them as were the cabinets d'amateur and museums that began the process of turning religious images into "works of art."
To see religious images as they were seen by the unlettered of times past requires an unusual and erudite form of literacy today. Contemporary art historians consider the relationship of religious art to the texts that inform them mainly when it is a question of glossing what have become esoteric symbols or seemingly arbitrary conventions. When we turn to the modern masters of medieval iconography, it is because we are interested in the pictures, not in the ideas they express. Weirdly enough, as the audience for these pictures moved from the illiterate many to the most self-consciously literate members of society, "image" supplanted "word" as the major term in this relationship.
Among its consequences, this reversal has obscured the distinctive relationship of an important tradition of religious images to natural language. The tradition of representing sacred literature and its theological interpretation in pictures rests on the assumption that the concepts expressed in natural language can have equivalent expression in pictures, that words can be "translated" into images. There was no serious and sustained argument against this position for approximately a thousand years. To what extent is it true? Even after the wholesale destruction of religious images in Northern Europe early in the sixteenth century and again in the years just after l789, there remains a vast body of work to test this assumption and an exceptional opportunity to learn something about language.
I want to consider two areas where metaphors operate differently in natural language and in iconography. The first, I will consider very briefly. The commonplace notion that essence can be deduced from appearance operates both in language and in pictures, but it has greater power and range in pictures than in language. Neither physical appearance nor clothing, for example, are ordinarily described in New Testament texts. Yet there is a tradition rarely broken in Christian iconography until the sixteenth century that requires the saints and prophets to be beautiful in body and countenance and to be attractively dressed. Their dignified appearance (something that can can be seen) is mapped onto their spiritual dignity (something that cannot be seen) and is a visual expression of the commonplace notion that essence can be deduced from appearance combined with the basic metaphor PEOPLE ARE CONTAINERS.
In this metaphor, psychological, moral or spiritual states are understood in terms of the physical interior of a container. Gestures, facial expressions, and speech, all of which can be perceived, are understood in terms of the exterior of the container. This is an example of the Great Chain metaphor in which attributes of human character and psychological, moral or spiritual attributes are understood in terms of a physical object with part-whole structure. To understand the relationship between the physical and the psychological attributes of a person in terms of a physical object with a part-whole structure is to understand the relationship metaphorically. The conceptual metaphor can be combined with either of two commonplace views of the relationship between the exterior and the interior. The first is that we can know the interior part by knowing the exterior part. When we adopt this view, we frequently elaborate the metaphor so that the physical exterior is understood as transparent; that is why we can know the interior. The other commonplace view is that you cannot know the interior part by knowing the exterior part. When we adopt this view, we frequently elaborate the metaphor so that the exterior is opaque. I will give the name APPEARANCE REVEALS ESSENCE to the combination of the commonplace view that essence can be deduced from appearance and the conceptual metaphor that the exterior is transparent.
APPEARANCE REVEALS ESSENCE is a well-established convention in Christian iconography partly because in pictures, it is always necessary to deal with appearance. In language, it isn't necessary to deal with appearance at all.
In the Gospel of Matthew, the figure of John the Baptist is an exceptional case. His clothing is described, and the literary description conflicts with the iconographic convention APPEARANCE REVEALS ESSENCE. John the Baptist's clothing is described as a "rough coat of camel's hair." (Mt. 3:4). The description raises a problem for iconography because to contradict the text is impossible and to suspend the convention of clothing standing for spiritual dignity in the single case of John the Baptist creates an unwanted ambiguity. One iconographic solution can be seen in the landmark Ghent Altarpiece of Jan van Eyck (1432). John is dressed in a gorgeous robe, hemmed with jewels, but open and worn over an animal skin so that both are visible.
Pictures have a physical immediacy that natural language does not, but pictures cannot refer to a person's abstract moral qualities or interior states of a person except by relating appearance to essence. A picture's range of direct address is limited to the visible world; its subject must be represented in visual terms. If a picture represents moral qualities or inner transformations, painters have no recourse except to map images from a visible domain onto their subject. Throughout the middle ages, John the Baptist was considered to be the greatest of all the male saints. In images of the Last Judgment, he takes precedence over all of the Old Testament prophets and all of the New Testament apostles. This precedence is represented by spatial arrangement—he is closer to the Christ of Judgment than the apostles and the canonical Hebrew prophets—and size—he is larger than the other male saints. But he is also a figure whose appearance expresses his dignity. When APPEARANCE REVEALS ESSENCE is conventional, he cannot be pictured as bald or overweight or homely or unimpressive in countenance. The literary description of John—the textual authority for placing him immediately to the right of the image of the Christ of Judgment—is a phrase from the Gospel of Matthew spoken by the Christ himself (11: 11): "… never has there appeared on earth a mother's son greater than John the Baptist … ." The text says nothing at all about his physical appearance. It is quite possible in natural language to address a person's moral qualities, even as here in metaphoric language ("greater than" is an example of the conceptual metaphor SIZE IS IMPORTANCE, a metaphor so completely conventional that we hardly recognize it as a metaphor at all) without giving any physical description of the person, without having to be specific about what the person is wearing. In a picture, it is necessary to be specific about what someone looks like. If the picture is not concerned with the person's literal appearance but is very much concerned with portraying invisible qualities, the conceptual metaphor, APPEARANCE REVEALS ESSENCE is a fundamental resource.
I also want to consider a kind of ambiguity in language that is easily overlooked because it is so deeply conventional: the ambiguity between the so-called literal and the metaphoric use of words. I am thinking of the way we commonly understand that although we seem to be talking about one kind of thing, physical movement, for example, we are referring to an entirely different domain of human experience, an intellectual or moral one, for example. When we talk about Caravaggio's "followers," we are hardly aware of any ambiguity in our language. When we use conceptual metaphors such as CHANGE OF STATE IS CHANGE OF PLACE the source domain can belong to ordinary visual experience, when the subject—what Lakoff and Turner, following Lakoff and Johnson, call the target domain—does not. When conceptual metaphors become conventional resources of everyday language, they affect the lexicon. In such cases, the lexicon allows us to do things that have no equivalent in pictures.
Consider this passage from the Gospel of Matthew. "As he passed on from there Jesus saw a man named Matthew at his seat in the custom-house, and said to him, 'Follow me'; and Matthew rose and followed him." [The New English Bible, Mt. 9:9]. The sentence belongs to a textual unit that begins at 8: 18 and includes two other short passages about following Jesus, two passages about Jesus' powers over physical forces and over spiritual forces, and another about a cripple getting up from his bed and going home. This last is explicitly a "sign" of a change in the cripple's spiritual state. Jesus first tells him his sins are forgiven. When this scandalizes some of the lawyers present, Jesus restores the cripple to health by telling him to get up, take his bed and go home. He offers this physical change of state as a visual proof of his power to change the man's spiritual state. The sentence about Matthew follows immediately and is glossed in the four following verses, which complete the unit. These final verses explain why Jesus associates with the demented, the possessed, the sinful and the worldly. He eats with bad characters—of whom tax-gatherers are a type—because "It is not the healthy that need a doctor, but the sick." So as doctors associate with the sick (and presumably bring them to health), Jesus associates with the sinful and worldly and brings them to a state of grace and spiritual dignity.
The sentence about Matthew, then, is not about what can be seen, a man getting up from his seat and following another man. It is about a spiritual change. A member of a notorious profession leaves a settled state of worldly corruption for the active pursuit of spiritual perfection. It is a normal and conventional resource of language to refer to such invisible spiritual change by speaking of visible physical change.
This conventional resource of language has no direct equivalent in iconography. The spiritual transformation of the man whom Jesus encountered at his seat in the custom house cannot be represented by a picture showing a man who has evidently left his seat and is now walking behind Jesus because a picture of one man walking behind another will not be understood to be about spiritual transformation. The problem of representing such a transformation in a static picture points to a general asymmetry in word and image, natural language and iconography. Pictures of physical change may be given a conventional meaning so that they can be understood to refer to a different domain, but these conventional meanings will be esoteric because they are essentially arbitrary. In language, the lexicon itself is affected by the use of such conceptual metaphors. The force of words such as "rise" and "follow" can hardly be said to be primarily literal since their metaphoric use is so common and has become so entirely conventional.
The actual bodily motions of rising and following, however, do not maintain the same ambiguity between literal and metaphoric meaning. Neither then will a static picture of any point in such a motion. How then can a static picture translate Matthew's spiritual transformation? If just one point of either physical movement (rising or following) is depicted, the picture cannot convey the subject of the narrative I want to consider Caravaggio's "Calling of Saint Matthew" (Rome: Capella Contarelli, San Luigi dei Francese) as a solution to this problem.
This picture takes careful account of both of the metaphors in Matthew 9:9. In that text, Matthew starts off seated, and then stands. CHANGE OF POSITION IS CHANGE OF STATE. Once standing, he has new capacities. CHANGE OF POSTURE IS CHANGE OF CAPACITIES. He leaves a settled worldly state, and this gives him new capacities. Having left worldly values behind, he can now follow Jesus.
Caravaggio cannot use dynamic metaphors, so he translates the first part of the text into a picture by switching to spatial metaphors involving geometric figures, (a broken circle and a line), gestures, and arrangement.
Matthew is seated at a table, but he is not alone. There are four other figures arranged two on each side of him around a table. Jesus is standing to the right of this group, but he is not alone either. The figure standing (partly) in front of him is Peter, who is not following Jesus in the sense of walking behind him; he is instead imitating his posture and his gesture.
Matthew is not in the act of standing. But he is looking at Jesus while the figure directly across from Jesus is looking down at the money on the table. The figure bent over beside Matthew is also ignoring Jesus, but Matthew is looking at him and is responding to Jesus' gesture with one of his own. The five figures around the table form a rough circle with Matthew in the middle. The circle is a broken one, opened at the right, broken open by the light and Jesus' gesture. Matthew isn't in the act of changing his (physical) state in the picture, instead the arrangement allows him to be contrasted to the other figures who do not look at Jesus or do not respond to him. Caravaggio can make use of spatial relations as a source domain because they are static. He cannot, however, translate the dynamic concept of following into a static image. His solution here is to substitute a citation from another picture for the metaphor in the text. Jesus' gesture substitutes for Matthew's metaphorically understood dynamic action of following. Matthew becomes a disciple in the text because of something he does. He follows Jesus. But Caravaggio makes a crucial change. What makes Matthew a disciple in the picture is a creative act on Jesus' part. He points to Matthew and his gesture cites a passage from the creation of Adam in Michelangelo's painting from the Sistine Chapel—as famous, at least in Rome, when Caravaggio painted as it is today.
The inability of pictures to use dynamic source domains and the necessity for them to depict physical features even when they are concerned only with psychological or spiritual states suggest limits on the translation of texts into pictures. But this asymmetric relationship points to a interesting feature of natural language too. The lexicon incorporates an ambiguity in such words as "rise" and "follow" that the actions themselves if pictured do not have. The postures that can be represented in static pictures do not share the lexical ambiguity between so-called literal and metaphoric understanding.
Beyond this important asymmetry between word and image, images have properties of their own not shared by natural language; static pictures incorporate an inevitable static bias. It is not a considered theological choice that led Caravaggio to give Matthew a passive role in the representation of his change of spiritual state. It is a general bias of images, one that exercises a powerful influence in commonplace thinking—massively affected as it is by the tacit iconographic education offered by even the most iconoclastic cultures.