James Elkins identifies mediation with the phenomenological account of embodied experience, associates my project with an undeclared embrace and exploration of a phenomenological tradition that art history more often takes for granted than engages with directly, and equates my attempt to think beyond social art history with phenomenology's relative autonomy from social analysis. I recognize the many resonances with phenomenology's concerns and am happy to acknowledge the debt; I would even agree that phenomenology is the necessary "substrate" of art historical analysis. Nonetheless, I am far from seeing it as providing my project with such coherence as it possesses. For one thing, my intervention analyzed the mediations of the artwork in terms of communication, which is not phenomenology's focus. For another, what I discussed as individuation is not, as Elkins has it, a matter of phenomenological being-in-the-world; it is a process in which different kinds of work of cultural, social, and ecological mediation converge. A further difference is that while phenomenology goes a long way toward mitigating the opposition of subject and object, it does not ultimately break with these categories. David Summers, for his part, characterizes my interest in mediation as the outgrowth of a semiotic approach. I think I was careful not to make a generalized use of the sender-receiver model that underpins semiotics, the approach against (but not outside) which our most radical art historical formalisms have defined their own focus on the suspension of definite meaning.
I take the concept of mediation specifically from the systems theory of Niklas Luhmann; it provides us, I believe, with a necessary complement to the concept of medium on which modern art history is founded. One way of defining that necessity would be to say that if phenomenological insights into the work performed by art are to have purchase today, they need the frame of a functionalist reconsideration of the artwork. Very schematically, for Luhmann, communication is not transmission but instead the making of distinctions; in this sense his view of artistic communication has more in common with communication at the level of cells than with semiotics. (1) Art entails "the triggering effect of a specific difference," which, when it succeeds as form, "sets in motion a special kind of communication that draws upon the capacity to perceive or on the imagination and yet cannot be mistaken for the world we ordinarily perceive." Art also has a specific relation to verbal communication. It avoids and circumvents language, establishing a different type of structural coupling between consciousness and communication. It does this "by using perceptions contrary to their primary purpose" in order to enforce the distinction between information (or hetero-reference) and the operation that produces information--utterance (or self-reference)--thereby producing meaning. (2) Art repeats the artificial "cut" distinguishing information and utterance in new forms throughout any given artwork. "What is at stake in art," writes Luhmann, "is not a problem to be solved once and for all but a provocation--the provocation of a search for meaning that is constrained by the work of art without necessarily being determined in its results." (3) From a systems-theory point of view, therefore, the artwork produces meaning through the recurrent distinction it makes between information and utterance, with the operational logic of this mediation of communication having primacy over the artwork's objecthood and unity.
At this very general level, the relevance to art historical practice is hardly obvious. It is not until the specificities of artistic structure come into play that Luhmann's approach reveals its interest. It can't be said that he makes it easy. One formidable obstacle to any straightforward art historical use of Luhmann's work lies in the fact that information/utterance distinctions are necessarily enforced across another fundamental distinction, between medium and form; this distinction, too, repeats itself at every level of the artwork. The apparent familiarity of the terms medium and form is utterly misleading. Luhmann means by medium something quite different from the materiality associated with this term in art history. The medium of systems theory is a liquid and evasive potentiality, involving a loose coupling of elements, which can be apprehended only indirectly through its actualizations in form. Form is understood to be a tight coupling brought about by recursivity and the establishment of a perceptible boundary between marked and unmarked space. Confusingly for the art historian, one is forced to conclude that art history's concept of medium is an example of systems theory's concept of form. In order to adapt this alien but ultimately helpful schema to the needs of art historical analysis, in my essay I had to set aside Luhmann's often rebarbative vocabulary in favor of terms that can more easily inhabit art historical discourse. The distinction between the potentiality and loose coupling of systems-theory medium and its actualization in the recursivity and tight coupling of systems-theory form emerges at two levels of my analysis. It appears first when I contrast the subimage and image levels of the artwork under the names of "ground" and "scape" (as in "imagescape") respectively. And it appears again in my distinction between a major mode of mediation such as depiction, which is shared by different forms of pictorial art, and its actualization in the material medium of ink painting. Avoiding entirely Luhmann's special use of the word "medium," I try in the intervention to sketch out the beginnings of a properly art historical theory of the artwork at the intersection of mediation and medium as art history normally understands it.
In order to avoid all confusion with phenomenology, it is worth briefly parsing the mediation-medium relation in terms of loose and tight coupling of elements. Ground is in effect the operational mode of mediation, the way in which mediation inhabits and informs the artifact. Relative to mediation, ground is the corresponding visual form of tight coupling, but in its relation to the image level of the artwork, which I am calling scape, ground instead takes on a function of loose coupling, with scape then becoming tight coupling in its turn. The same mutability of function can be seen in the case of scape. Scape functions as tight coupling only in relation to ground; elsewhere, it functions as a form of loose coupling. The relation of imagescape, for example, to its particular possibilities in a specific material medium, be it ink painting or woodblock prints, is again one of loose to tight coupling. Thus, the relation between mediation and medium passes through a series of more particular relations--mediation-ground, ground-scape, and scape-medium--each of which reproduces the logic of loose/tight coupling.
Why should art historians care? To give one example of the art historical interest of this chain of relations, it has the immediate advantage of providing us with a way of dealing with the fact that many (most? all?) artworks incorporate more than one set of aesthetic possibilities. Here I am responding to two justified criticisms made by Summers: first, that my intervention privileges the visual over the material, and second, that it underemphasizes the importance of medium. In the study of Chinese art, by convention we treat A Solitary Temple below Brightening Peaks as part of the history of the medium of ink painting, and we tend to consider secondary, or supplementary, the fact that any display of the painting would, as display, have brought it into the realm of interior decoration. True, we may note resonances between A Solitary Temple as a landscape painting making the most restrained use of color, on the one hand, and the Northern Song interest in muted colors and texture for ceramic glazes, lacquers, and some textiles, on the other. We may also note that in the decorative arts, such restraint was often deliberately set against the opulence of precious metals, gilding, or mineral pigments, just as landscape paintings might have roller ends of jade or gold. When such questions are discussed, it is often under the heading of...