Along with David Summers’s Real Spaces: World Art History and the Rise of Western Modernism (New York: Phaidon, 2003) (click here for review), Whitney Davis’s A General Theory of Visual Culture is one of the most ambitious and potentially foundational books on art history in recent decades. It is unusually dense in logical argumentation, so it is more than a convention to say that it cannot helpfully be summarized. Because longer reviews will be needed to assess the book’s arguments, I want to use the generally shorter review length here in caa.reviews to raise two points about the book as a whole. But first I will evoke, as succinctly as possible, the book’s content, purpose, and significance.

Davis’s book ranges widely across the central examples of art-historical methodology, from Heinrich Wölfflin to Michael Baxandall, including discussions of writers as different as T. J. Clark, Arthur Danto, Ernst Cassirer, Nelson Goodman, and Giovanni Morelli. There are extended readings of texts by Erwin Panofsky, Richard Wollheim, E. H. Gombrich, and Ludwig Wittgenstein, and critiques of formal analysis (chapter 3), style analysis (chapter 4), and iconography (chapter 7). The book’s visual examples range from prehistory to Renaissance art to modernism and Warhol.

Davis’s principal purpose is to provide a “general theory of visual culture,” by which he means an account of the relation between what is cultural about vision, and what is visual about culture. He has many ways of putting this difference, and the variety is itself significant. (More on that later.) To ask about what is cultural about vision is to note that “styles of depiction . . . have materially affected human vision,” and to ask about what is visual about culture entails the possibility that “some things,” but not all, “are visual in culture, or visible as culture” (6; see also p. 8).

As a conceptual reorganization of art history’s fundamental terms of engagement with objects, the book is exemplary, and it is difficult to imagine a reader who is engaged with the discipline for whom this book is optional reading. I especially hope Davis’s reconceptualization will shift some of the conventional recurring assumptions of art history and visual studies, and I hope the rigor of his reconceptualization (as distinct from its content) will change the forms of argument in art history of visual studies. I hope so, but I am not sure it will happen, and I want to take this review to explore three issues that may be helpful for potential readers: the question of the book’s use-value; the question of what counts, in this book, as abstract argument; and the problem of assigning meaning.

1. Issues regarding potential readership

For philosophers of art, the framing of the book’s problems in terms of culture and visuality should pose no particular problems (I mean the framing, not the particular claims that Davis then goes on to develop); for art historians and graduate students of art-historical methodology, the framing should be recognizable as an attempt to provide a solid foundation for existing disciplinary habits. Art historians will recognize enough of Davis’s list of objects’ “aspects” (in one listing: “formality,” “representationality,” “stylisticality,” “pictoriality,” and “culturality”; 26, see also p. 42). But for people in visual studies, global art history, postcolonial studies, area studies, Bildwissenschaft, Bildvetenskap, and other fields, this book may appear to be concerned with a potentially pertinent but ultimately detached, and therefore optional, field of inquiry. That is a pity, because the field in which I am most engaged at the moment, visual studies, is in dire need of rethinking.

Davis has very little to say to visual studies directly, but what he does say is pointed. He scolds visual studies for implying “visuality is simply a cultural interpretation of what is seen,” for not considering “relevant . . . analogies” to the practices under study, for assuming “a cultural succession has a social matrix,” that it can be “simply a sociology of culture that happens to be visible,” and finally for treating viewers as “cultural servomechanisms” that “automatically adjust visually to visible worlds” (280, 316, 337, 8, 339). Yet I wonder if these passages are enough to catch the eye of visual studies scholars. The principal obstacle, I think, is the way politics is bracketed out of this book. I will give one example where what I want to point to is the manner of the exclusion, not the cogency of Davis’s arguments, which I do not often dispute.

Chapter 3, “What is Formalism?” introduces part 2 of the book by ruminating on the distinctions between formalism, style analysis, and iconography. Davis notes that “despite severe criticisms leveled over the years, formalism is a defensible hermeneutics” (47). It cannot be “dis-integrated” from style analysis, iconography, and other methods, and it can be understood as a “route” or “moment” in “integrated aspect-perception” (37, 47, 48). Formalists, Davis says, claim that formalism is “closer to visual perception than style or representationality,” but “this notion must be utterly wrong, and pernicious” because “no sensuously apprehended aspect of something is any closer to the object . . . than any other such aspect” (48). Davis then goes on to develop an account of “historical formalism” that complements modernist high formalism by proposing that previous visualities and previously visible configurative aspects might become visible to us. This expansion and clarification of formalism is interesting, but what I want to emphasize here is that the chapter may lose readers in visual studies, area studies, and other fields because of what Davis says after calling high modernist formalism’s claims “pernicious.” A reader in visual studies might well expect the next line to be something like, “because formalism is an escape from the object’s political context.” Davis hardly ever mentions political and social contexts, for good reasons: in relation to the arguments he develops here, those considerations are epiphenomenal or categorically misapplied. The politics, identity, gender construction, or socio-economic context of a work is provided through the kinds of “successions” and “recursions” that are under study. Davis is, of course, far from oblivious to claims regarding gender’s, or identity’s, primacy in cognition and apprehension. One argument about politics, of the sort that a reader in visual studies might expect, is buried in a brief summary of Art-Language’s response to Clark’s reading of Manet’s Olympia,¹ where Davis notes that “ideological misrepresentation in the Marxist sense” might characterize the entire history of bourgeois consciousness,” but “for that very reason this model of consciousness must be entirely uninformative” about the debate concerning interpretations of Manet’s painting, which needs to be seen “specifically as a slippage in the relays . . . of the iconographic succession as it moves . . . into depictive visibility” (228–29). In other words, questions of visibility, pictoriality, and “pictorial inadequacy” have to be sorted before debates like Art-Language’s can make sense. But my guess is almost no one interested in socio-economic, gender, identity, or other forms of visual histories will make it to page 229 to find that news. Davis’s other forthcoming books on gender will be read instead. Here, then, I just want to signal that omissions are not inadvertent.

2. The role of abstraction

Davis’s account is abstract, and not only in the sense that it is a “general theory,” a logical reduction, and a reconceptualization of the practices of art history. It is also abstract in two other salient senses. The book is abstract in the sense that it is a redescription of the sort that might be, effectively, a different discourse or language from the one it addresses. I have already made several remarks on that, regarding visual studies. And third, the book is abstract in that its argument relies on the stability of terms that are newly minted, or redefined from their vernacular usages. Such terms include “visuality,” initially defined as “the culturality of vision” (8); “visibilization” (10); “succession,” “iconographic succession,” and “cultural succession” (8, 21, 151, 255); “pictoriality” (22); “pictorialization” (150); “aspective interdetermination” (37); and “indiscernability” (22). These newly minted terms do several specifiably different kinds of work, and have specifiably different relations to the art-historical and philosophic discourses that inspired them. To use an analog that is central to Davis’s own arguments, the book’s specialized vocabulary comprises several partly disjunctive but analogous language games.

I want to close with an abstract meditation on this third sense of abstraction. What follows may not appear useful (or even comprehensible) if you have not yet read the book, but I hope it will be useful in negotiating Davis’s vocabulary.

Let me take “succession” and “recursion” as examples of the work that some of these newly or differently defined concepts are given. Succession first appears when Davis says that “vision is not inherently a visuality,” that is, the facts of human vision do not immediately correspond to the elements of visuality that appear in culture. “Rather,” he writes, “vision must succeed to visuality through a historical process” (8). The word is signaled as a technical term by the fact that it is set in italics. The first, partial definition is in the next sentence: “The recursions of this succession are not well understood.” Hence initially, succession is a process involving multiple, enchained steps called recursions. (More on them in a moment.) Davis also means succession to denote a complex process, including recursions, although sometimes recursions and successions are presented separately. Succession, he says in this passage, “occurs in complex relays and recursions of recognition” (9). On the same page, succession is glossed as a series of “relays, recursions, resistances, and reversions.” Elsewhere in the book viewers are said to “succeed” to “aspective horizons” such as formality and style, which gives the term a slightly different meaning, something more like “access” or “discern” or “become aware.” My point here is not at all that this is ill-defined for the work Davis wants it to do, but that the work is structurally open-ended, because Davis means to theorize the frame for such complexities, rather than the complexities themselves. In that sense, “succession” is like Wittgenstein’s “forms of life,” which Davis also employs: it is taken as descriptive rather than explanatory (373 n. 20; see also p. 287). That kind of decision about the deployment of abstract terms is, presumably, the reason that Davis writes, “vision must succeed to visuality through a historical process” and not, say, “vision must become visuality through a historical process” or “vision must be developed as visuality through a historical process” or “vision must be recognized as visuality through a historical process” or “vision must be assigned to visuality through a historical process.” (He does say these sorts of things in other passages; see also p. 322.)

Recursion, the sometime synonym for succession, is also subject to some latitude. At the beginning of chapter 6, Davis says a picture is “a visual recursion of nonpictorial stuff, of matter or marks. It transmutes their like-lookingness into looking-likeness. More exactly, it replicates . . . the visual event in which matter or mark is seen as something-or-other” into “looking-likeness, a visual event in which that something-or-other can be seen in the mark” (150). Here recursion is inversion, not repetition. In the process, a viewer could be said to “succeed” to a certain aspectivity, but that “succession” would not be an iterated, replicated, recursive sequence but an operative inversion. The difference matters, for example, when recursion is listed alongside reversion.

What I am pointing to here is the presence of specifiably different uses of abstract or “general” terms, entailing different senses of abstraction. It is this issue, I think, that will have to be judged by readers as they assess the book’s applicability to their own interests. For me, the form of Davis’s imagination is disassemblative (if I can coin a word in keeping with the terms in the book): he often finds the most cogent way forward is to disassemble a putative whole (visuality, iconography, representation, denotation) into its combinatoric units. The great virtues of this are the crystalline clarity it produces and the open-endedness it permits him. Davis can be exact about terms such as “formality,” “representationality,” “stylisticality,” “pictoriality,” and “culturality,” and that allows him to propose multiple, shifting, indefinitely intricate combinations of “successions” and “recursions.” The potential drawbacks are symmetrical to those strengths: the coherence of the account depends on the stability of the framing terms, and the utility of the account depends on the reader’s capacity to locate particular configurations within the open-ended combinations of “succession.” One of the questions that a disassemblative approach opens, and intentionally declines, is the limit of structural fine detail in any given practice: why stop, for example, at “the seeing of each human agent”? (319) What is to prevent a more particulate approach to context, meaning, explanation, and intention? And, on the other end of the scale, what concepts can adjudicate and gather the largest categories of “formality,” “representationality,” “stylisticality,” “pictoriality,” and “culturality”? These are not questions begged by the book, which is in full control of its apparatus, and is aimed at what I take it Davis understands as plausible and rewarding sorts of cultural interpretation. But in an abstract accounting, there is no delimiting frame.

3. The problem of meaning

I want to end with a note on the book’s philosophic strengths, which should recommend it to specialists in Wittgenstein and language philosophy. Chapter 9, the book’s most abstract chapter, is an exemplification of some consequences of language games posited near the beginning of Wittgenstein’s Philosophical Investigations. Here Davis begins by developing more terms for “successions,” including “networks” and “analogies” (281–82). He adumbrates a philosophic position on analogy, in which it appears as a more fundamental synonym for “succession,” especially when he declares that “successions . . . are the accessions of analogy in visibilizing configuration” (283). This non-ontological position is comprehensible in relation to the chapters that have preceded it, but the appearance of analogy is surprising; for me at least, it is not entirely clear why analogy has not borne the weight of the concept of “succession” from the outset. From a philosophical standpoint, these opening pages of chapter 9, which also include the fullest discussion of “forms of life,” are fundamental; I recommend readers begin with these pages, and then return to the opening of the book. Most of chapter 9 is a daringly abstract and literal realization of some of Wittgenstein’s ideas about language games, using somewhat hokey-looking computer graphics. And yet it is here that Davis makes his strongest claims about culture and the visual. The third of his three examples is a study of “super-language games” in which the participants employ analogies between language games, which in turn permits the creation of meaning. (In Wittgenstein’s opening example, a builder just calls for blocks of stone, and another builder responds. In that scenario there is no room for questions about the “meaning” of the building they produce.) The example is intended to show that visually indiscernible practices can result from very different language games. In the next and final chapter, Davis draws a deeper or more properly philosophic consequence from that, observing that “there could be cases in which a form of life that looks to us like our own is in fact unimaginably different,” so that, for example, “we readily overlook the cases in which different forms of life seem to be morphologically similar” (324). On the other hand, when we observe a cultural practice that appears entirely meaningless (devoid of plausible analogies in our own lives), it “must seem absurd to us.” What would such a person’s form of life “possibly be like?” he asks, and answers, “it cannot be like anything at all” (326).

This is incrementally close to Wittgenstein’s own concerns, as they have been explored by a number of commentators. It also seems like a pitch of abstraction in the second sense that I mentioned above, because it appears entirely devoid of application to art history; but this is exactly the moment when the deepest charges can be laid against art history, visual studies, and related fields, because the double possibility of indiscernibly similar practices that result from very different language games, on the one hand, and of practices that appear not to result from any knowable language game, on the other hand, point directly at unconceptualized voids in scholarship and critical understanding. It is easy to miss dissimilarities in the cultural practices that produce apparently similar works (for example, contemporary installation practices in different parts of the world, which can seem similar), and to afford them inappropriate, projected meanings; and it is just as easy to fail to assign difference, and therefore meaning, because practices seem to lack analogies (for instance, avant-garde practices in Theodor Adorno’s sense, which at first may appear blank, meaningless, or, in Davis’s preferred term, “wooden”) (325).

In Davis’s account, meaning is saying what something is like, and we do that far too loosely (294, 333). Among the many things for which A General Theory of Visual Culture is a corrective, perhaps this is the most important. I hope that Davis’s book succeeds in making some scholars in art history and visual studies ponder why their analyses encounter so few obstructions, why meaning and visuality seem to flow so effortlessly from cultural practices.