Last July, I wrote a piece “On Art and the Universal, I” and promised part two within a week. It’s almost 11 months later, so here it is. The first piece stands on its own as a critique of the political cynicism of the academic-gallery nexus. This second piece stands alone as well. Read part I, re-read it, or don’t bother.
As an art scholar and artist, I find the Greenbergian tradition invaluable. I studied for a year with Hal Foster in graduate school and was compelled by Rosalind Krauss’s essay on sculpture in the expanded field, as well as by Clement Greenberg’s efforts to find a trajectory for research within postwar painting. Briefly, Greenberg asserted that each art form should concentrate on its own unique properties or “the specificity of the medium“. Famously, Greenberg believed that the essence of modernism was to “use the characteristic methods of the discipline to criticize the discipline itself, not in order to subvert it but in order to entrench it more firmly in its area of competence.” To this end, painting, for Greenberg would best focus on the flatness of the canvas instead of imitating the three-dimensionality of sculpture. This was of great utility for the last generation of truly productive artists in the US, from Kenneth Noland to Donald Judd to James Turrell to my father, all of whom engaged with Greenberg—even when they disagreed with him. Disciplinary self-criticism and the specificity of the medium was a research project that embodied an Enlightenment ideal of a shared project of advancing human knowledge in a particular discipline. Krauss, who studied with Greenberg, reinterpreted his philosophy, moving away from the idea of medium specificity to propose art as an expanded field of practices and mediums, including conceptual, installation, and performance art. The object of interrogation ceased to be the medium and became the institution of art itself and with this, a greater element of political critique could be introduced. Foster took this further in his writings on the Pictures Generation, shifting to a postmodern exploration of the process of art making, originality and identity, and the nature of the sign itself.
Although I empathize with the Greenbergian search for politically progressive forces in art, this aspect of the project has run aground, even if is the only part of the project that remains popular. I detail this in my previous post, but in sum, the quest for the political in art has amounted to little more than a justification for guilty consciousness and the drive to affirm one’s virtue. Far from a place of resistance, the political in art is cynical in a Sloterdijkian sense: its proponents know that it has nothing to do with actual political progress, but they claim it nevertheless.
Perhaps not coincidentally, art lost the thread since the 1970s. Even as postmodernists deployed postmodernism as a totalizing concept, they claimed that totalization was obsolete (the classic boomer move of declaring itself the best and last generation at anything). For postmodernists, totalizing historical frameworks overgeneralize the intricacies and nuances of historical events and cultural phenomena, leading to oversimplification and inaccuracies, they overlook differences within a given time period, such as the experiences of marginalized groups, and they perpetuate existing power dynamics by privileging dominant cultural or social perspectives. But the price for rejecting totalizing narratives is that where art used to make clear, measured progress, after postmodernism, it is stuck in an endless loop of pluralism, sustained only by self-justifying statements about politics. Today, the relationship between theory and totality is fractured and postmodern thought, ironically, leans toward irrelevance. In his 1979 La condition postmoderne: rapport sur le savoir (translated as the Postmodern Condition: A Report on Knowledge), Jean-François Lyotard observed that knowledge—primarily science—was being fragmented into incommensurable discourse as an incredulity to metanarratives emerged. Today, the arts and humanities are also splintered into incommensurable discourses. But rather than being a position of greater strength and self-criticism, the fracture of narrative banally reflects our very existence, our selves intensely fragmented by the operations of media. Art practices and theories that exacerbate that fragmentation are merely accelerationist or, more likely, uncritical and reactive in nature. Lacking a metanarrative, however, there is little else they can do besides exacerbate fragmentation.
I contend that it’s time we breathe life back into the Greenbergian theoretical framework. This revival, however, should begin with a call for art to investigate itself again, not merely play to political activism for the sake of theater. The task at hand is to discern the proper object of knowledge for art, a fulcrum upon which we can rest our research. Or, if not the proper object, a proper object that would be suitable for investigation and productive of knowledge.
Except for the most feeble-minded of thinkers, the development of advanced levels of networked computation is the single biggest transformation in human existence in many decades. Our sense of what media is and our relationship to it has changed profoundly. Thus, although it is entirely possible for artists to pursue other, legitimate forms of research, my own work largely revolves around the role of technology in our lives. In the last year, I have specifically been compelled to explore the new generation of Artificial Intelligence software, particularly AI image generators.
What is specific to AI image generators is not the creation of the new, but rather their endless capacity to remix the history of art and imagery. We could see this as part of a dialectic, or more simply, as part of a back-and-forth process of art history since the late eighteenth-century loss of the absolute belief in the principles of classical art. After the archeological discovery that the ancient Greeks and Romans did not have a consistent system, art was set adrift with its terrifying newfound freedom. Nineteenth-century eclecticism followed: rules were treated flexibly and forms could be freely combined at will. The backlash came with modernism’s rejection of all past forms and its search for a new, universal language of form, a project refined in Greenberg’s late modernist turn toward the specificity of the medium. In response, Postmodernism critiqued the new and turned toward the semiotic recombination of past forms and/or imagery from popular culture and commercial art. Starting about 25 years ago, Network Culture or Metamodernism supplanted postmodernism, largely relying on a resurgence of interest in technical effects and their capacity to elicit sensation. Think of Anish Kapoor or Olafur Ellison, for example, or the emergence of the very large, technically flawless salon-painting-sized photographs by artists such as Andreas Gursky or Jeff Wall.
The era of AI creation is not, primarily, an era of the new. Architecture throws things into heightened relief. A furry, feathery building is not new. Nor is it interesting, except as a means of generating Instagram hits. Within a few years, AIs will be developed to effectively generate endless, plausible architectural models from a set of given parameters (site, area needed, programme, etc.), but even those are likely to remain endless permutations of the sort a follower of Frank Lloyd Wright or Mies van der Rohe might have done in their offices. For now, AIs are not yet capable of producing sophisticated three-dimensional models, but they are capable of producing imagery by remixing content. When something new emerges, it is through unusual juxtapositions thought up by the operator, but also through accidents. Malformed image generations can be interesting: for example, in my project on an alternative history of art in Vilnius, a series of glitched images appeared like the following image, which was supposed to be of a painting exhibit in a gallery. This process can be iterative since open-source AIs such as Stable Diffusion can be trained on specific datasets, so when accidents happen, artists can take those unusual results further.
AI image generation reveals that all art is already intertextual, that is, shaped by, and in turn shaping, other works through allusions, references, and influences. My father was a modernist but nevertheless spent his evenings looking at coffee table art books of Renaissance and Baroque masters for inspiration. Nor was this an uncommon practice among modern painters. We now have a different way of accessing that cultural subconscious. It does not reveal itself easily either. Working with AI image generators is, for the serious artist, as time-consuming as any other practice. The virtue of a Critical AI Art, however, is to explore how artworks are developed within a network of works, historical and recent, and the cultural contexts that surround them. A Critical AI Art expressly addresses intertextuality and its relation to the idea of originality, not merely because these are the issues raised by AI image generation, but because these are issues inherent to art itself.