For those of you wondering where my Generator.X paper would be available online, here it is, complete with hyperlinks to relevant references and images…
Unframing Perspective: Towards a Definition of the Generative Sublime
Boredom Research: Ornamental Bug Garden
In the Renaissance, Brunelleschi’s invention of one-point perspective foregrounded the inextricable connection between representational systems and the technologies that are used to generate these systems. The metric system of Renaissance perspective, and later systems like quaduratura, the system used to paint Baroque ceiling frescoes, (see for example Andreas Pozzo) are pre-digital manifestations of the way in which a representational “code” or algorithm comes to shape what we see as “realistic” or “true to life.” The system of Renaissance perspective creates the sense of looking through a planar “window” onto the world by following a rigorous geometrical model of rectilinear grid lines and converging orthogonals. In this paper, I take up Heidegger’s notion of the world picture to postulate that both the computational and aesthetic properties of generative art destroy the definition of the window-gazing subject and instead indicate the development of a new, dynamically fluctuating “ground” from which we perceive, interpret, and communicate about the world.
In theoretical writings discussing the advent of generative art within a broader context of digital media production, authors often emphasize the relationship between code as material and the visual world generated by code. If all language creates a set of possibility conditions that simultaneously afford and constrain our range of experience, how does generative code reconfigure our notions about representation? In this paper, I will address the question of whether generative art facilitates the reshaping of the metaphors we use to construct a particular world picture.
In “The Age of the World Picture,” Martin Heidegger argues that we can only ever know our world as a representation. And because representation is never complete, but always a more or less rough “map” of a given territory, any world picture will always be oriented to a small subset of all possible knowable data. The philosophical ramifications of the invention of a metric model of space are multifold, but the primary issue to consider here is how a “system” like Renaissance perspective, associated with the metaphor of the window, creates a particular world picture, and then how an alternate system, like generative art, might provoke the formation of new world picture.
In “What is Generative Art” Philip Galanter defines generative art as “any art practice where the artist uses a system, such as a set of natural language rules, a computer program, a machine, or other procedural invention, which is set into motion with some degree of autonomy contributing to or resulting in a completed work of art. The traditionally dominant representational system instantiated by Albertian perspective demands that the image, the “completed” artwork, be viewed from a fixed perspective. The single point of view places the viewer in a relatively passive position from which the image is observed with disinterested and contemplative absorption. This viewing position depends on a notion of a human subject that is reducible to a single, monocular and disembodied point matched by an equal and opposite vanishing point in the discrete world mapped out within the boundaries of the frame.
Generativity, on the other hand, calls attention not to the static object, nor to the static viewer, but to evolution through time. I contend that this temporal extension draws generative art away from older representational systems, which derive their aesthetic success from fixing the body and attention of the viewer into an ideal “point of view”, and also from detemporalizing space. The disinterested contemplation in Kantian aesthetics is often critiqued for disembodying the viewer by reducing perception to an abstract intellectual process divorced from sensorially integrated lived experience. As Henri Lefebvre has argued, human beings to do not move passively through an undifferentiated, voidlike envelope of space. Instead, space is created by action, movement, and so on. Linear perspective instantiates a world picture imagined as a rigid system of geometrically converging lines and points, or a rationalized grid, whereas generative art and design uncovers the action of structuration as a spatio-temporally extended process.
Code, in generative art, works performatively to shake the notion of “image” loose from what Henri Bergson would call a “spatialized” fixity. For Bergson, spatialization is synonymous with reification, with a forgetfulness of the fact that no object is ever truly static and unchanging. Coding has been recognized as a powerful force in integrating process into art practice, so that the viewer is forced to confront duration itself. Whereas film, video, and live performance are all time-based media, all three are subject to narrative and material constraints that enframe their temporality, and thus reclaim them to spatialized/fixed modes of art production.
In contrast, generative art asserts the temporal quality of experience by adding elements of the unknown or alien. We are forced to question here what kind of space-time we are seeing in the proliferating images, and to ask what kind of world, is being not just reflected but even possibly created ex nihilo by the building blocks of generative code. This mode of art making shakes loose our habitual viewing practices and our assumptions about naturalism and realism. One point perspective cannot be maintained precisely because the image is not centered around a single, fixed point but rather a complex set of interacting parts. This leaves us with a sense that we are witnessing the growth of some varietal of “nature”, a system that is in many ways much more “natural” than the geometrical “code” of linear perspective. In viewing this unfurling of another nature, it is easy to experience a simultaneous fear of and fascination with this alien organicity. Along with a loss of the familiar perspectival standpoint comes the need to re-evaluate the aesthetic categories we typically apply to art objects, most notably in this case the category of the sublime.
Kant defines the sublime as that which exceeds our perceptual capacity, but this is rarely considered in terms of temporal perception. In the experience of the sublime, there remains an implicit fixed position that is engulfed by the terror of limitlessness. The observer is again reduced to a point instead of existing as a stereoscopic, multisensory being that is always pulsing, vibrating, and dynamically resonant at the most basic physiological level.
While Peter Bentley jokes that coding is literal, unambiguous, and tends to produce literal, unambiguous thinking in human programmers, it is also true that the methodical nature of code restructures the way we conceive of and perceive the world by reaching to the bare bones of structure itself. If generative code is that which is always in the process of mycelial dividing, unfolding, and becoming, its sublimity lies precisely in its temporal extension, the algorithmically unfurling flow and flux that so often gives generative art the appearance of organicity.
In the press release for BoredomResearch Lab’s Ornamental Bug Garden, Vicky Isley and Paul Smith write that “the complexities of the overall sound composition are the result of emergence within the systems” and later that “sound…carries the time signature of emergence”. Whereas traditional musical pieces become enframed by a quasi-narrative arc, as we observe in Martin Wattenburg’s Shape of Song, the time signature of emergence in the Bug Garden brings us precisely to a temporal sublime, and the possibility of apprehending the visual object not according to the usual categories of perspective, point of view, or spatialized fixity, but instead the sublimity that comes from an endless process of what Rich Doyle calls the “becoming-visible of the future,” a “becoming…whose ‘ontology’ is algorithmic and complex” instead of a product of 18th century classificatory systems that are stratified and linear.
For Kant the descriptive category of the sublime is reserved for phenomena that which exceeds the perceptual capacities of the human mind, and it is the goal of romantic artists to find a way to express spatial limitlessness within the constraints of visual representation. In his article on the “Anti-Sublime Ideal in Data Art,” Lev Manovich argues that “data visualization artists aim at precisely the opposite: to map such phenomena into a representation whose scale is comparable to the scales of human perception and cognition”. We can conceive of mapping, as it is commonly understood, as exhibiting the same drive towards simplification of vast information sets into a manageable image.
But the more aesthetically charged examples of generative art take a different direction. In software art, the code itself is foregrounded as the aesthetic (or anti-aesthetic) object. Generative art uses code to relinquish authorial control and to allow emergent patterns to unfold with some degree of autonomy, within basic constraints. But while, as Cox, McLean, and Ward point out, “to separate the code and the resultant action would simply limit the aesthetic experience”, they argue simultaneously that “the aesthetic value of code lies in its execution.” In generative art, emergence carries an aesthetic resonance that makes the object seem simultaneously like life and mysteriously other from it.
My overall argument is that generative art breaks down and reconfigures normative models of spatialization, so that space must now be understood as temporally extended and fluid and not subject to the same framing procedures as traditional pictorial representation. Perception, position and perspective can no longer be imagined as issuing from a monocular, unisensory viewpoint. Another piece that instantiates this theoretical perspective is Pierre Proske’s Synchronized Swamp project. This project shows how the phenomenon of synchronization occurs across nature, so that two pendulum clocks, the flashing of a swarm of fireflies, the chirping of insects, and the human respiratory and nervous system, all tend to synchronize. The piece combines movement through space and multi-sensory perception to demonstrate the phenomenon of biorhythmic sympathy. Because the sounds in the piece are distributed through the exhibition space, the listener/observer can “experience the synchronization from different perspectives”. This effectively destabilizes the immobilized, atemporal “viewing position” of the ideal observer of a perspectival construction.
We can take the example from this sound installation to see how visually generative pieces also elicit biorhythmic sympathy. In observing Ryoji Ikeda’s data.spectra, we see that looking at abstract flow of visual information is an experience of sheer duration. In reference to data.spectra, curator Mark Stubbs writes that “we ‘feel’ an endless flow of data” when looking at this piece. We respond to the aesthetic properties of data itself, not just to the picture as mimetic representation of a known world structure. Generative art facilitates a “feel” for data by giving it form. Ben Fry has also remarked in regard to his Genomic Cartography project that to “feel data” is to orient oneself to it. And to orient oneself is literally to adjust one’s position, to work out one’s direction and facingness in relation to the object. The surface, or phenotype, in generative art affords a terrain: a quasi-geographical realization of code. Works like the Synchronized Swamp, or Wilfried Hou je Bek’s Generative Psychogeography project invoke humanness and even anthrophmorphism, but on a different register from the model of the human given by one-point perspective. Hou je Bek writes about generative psychogeography that it seems chaotic and disorienting. This is the case because the project works to dehabituate spatial practice, and to force the walker to experience at a fundamental physiological level an alternate relationship to the structure of urban space/time. The generative sublime emerges from the shift away from the representational metaphor of looking through a window toward a mode of interfacing with the object through a phenomenological sympathy.
If, following Manovich, modern art abandons its traditional function of depicting the human being (or representing the world mimetically) in favor of presenting abstraction, mechanical reproduction, and, crucially, data, it also demands that humans assess what it is, finally, that a body can do and why it is that we choose to represent or conjure images of the phenomenal world. Visualization in generative art is less about framing a world picture than unframing and extending it, or dismantling the window frame and showing space not as a rigid grid but as a field of morphogenic possibilities.
The Romantics treated the sublime as a spatial phenomenenon. Bierstadt, a member of the American Hudson River School, attempted to represent the unrepresentable by enframing rugged mountainous terrain. He and artists like him continued to operate within the paradigm of the enframed window on the world. The computational sublime, however, must be defined according to a different set of rules, which suggests the existence of an alternate aesthetic paradigm.
Leonel Moura’s robot paintings exhibit the new modality of the computational sublime insofar as they can be seen as “the mapping of some sort of deterministic chaos”. Moura’s own writings about his work emphasize precisely the unfolding of the image in time and space, which calls attention less to the conceptual art historical ontology of generative art than to a magical utopianism in which generative works lead to a restructuration of nature. In works like these, our fascination develops from a sense that the “behavior” we witness both mirrors natural patterns and confounds them. Generative art has the capacity to make us nervous because it is not subject to the same representational “syntax” or dominant metaphors we witness in traditional modes of art production. The “liveliness” of the image may not be life per se, but it suggests the possibility of another nature, not just conceptually but also perceptually.
While code supplies the possibility conditions for sublimity to arise, the sublime experience is not located in the code itself, but rather in the surficial “organism” it produces. As Paul Harrison discusses with regard to his Ghost Diagrams, algorithmically generated objects expand or morph the edges of our world-picture-frame by playing with our capacity to process the world around us through our intuitive “feel” for natural phenomena. Computationally generated objects astound us because they confound our assumptions about the way nature works. In some ways, algorithmically-generated art goes hand in hand with the modern scientific world picture criticized by Heidegger, in which we believe that if we break down the world into more and more infinitesimal parts, we will ultimately arrive at the fundamental truth of our nature and the nature of the world around us. But as Harrison and others point out, generated art objects also seem to take on qualities of their own “There is no upper limit to their capacity to surprise us. Furthermore, tiles have an intuitive quality that other forms of computation lack. You can see how they fit together.”
An organism is more than the sum of its organs. When the organs are fitted together, the organism becomes something more. This surprising something more we call “spirit” or “ghost.” The most effective generative images infuse us with a sense of the hauntedness of the structure. The simultaneous fear and fascination that is proper to the sublime lies in the semi-autonomy of the organismic entity we see onscreen. But, the excitement of dislocation produced by the generative sublime also reorients us to the potential for compossible world pictures. Ultimately, the most successful works of generative art act to map a new terrain and to shift the Heideggarian concept of the frame. The notion of the subject, here, one of the most powerful rhetorical figures of late 20th century philosophy, expands beyond the single monocular point described by the geometrical algorithm of Renaissance perspective. A durational model of emergence shows the body of the subject never to remain self-same, but always in interaction with environmental stimuli to which we are startlingly other and uncannily familiar.