V. Radical Artifice
Critical writing about digital images tends to focus on what they are not: reliable, indelible, impartial, and impermeable. We look at digital images in this sidelong way because we have been conditioned by the knowledge that they enjoy no material existence outside of the electronic bits and bytes of data storage devices, and so can be easily and invisibly manipulated by desktop software packages like Adobe Photoshop. Consequently, anyone who knows anything about digital images knows that they are fundamentally untrustworthy unless accompanied by the most unimpeachable credentials. In what follows, I want to argue that our fixation with the unreliability and immateriality of digital images has eclipsed a number of other key components in our thinking about the medium. More particularly, and for reasons which I will begin to explain below, I want to argue that we have not yet evolved a fully adequate set of analytical and theoretical tools with which to discuss digital images with the same acuity one finds in the extensive critical literature on photography, film, and video.
We know a great deal about how digital images work of course -- there is no lack of documentation and literature from the technical communities involved in developing techniques, formats, and standards for computer graphics and digital imaging. But we -- by which I mean both the general public and more specialized academic communities -- know comparatively little about digital images as particularized modes of visual representation. Consider, for example, the commonplace notion that the 1991 Persian Gulf War resembled a "video game" or a "Nintendo" simulation -- almost always offered up in conjunction with widely-circulated images captured by military cameras on board of missiles and aircraft, depicting targets framed by crosshairs and data readouts and the like (figures 14 and 15). To the extent that such observations are meant to suggest something of the inhumanity and detachment that attends modern warfare, I have no disagreement. When ABCís Cokie Roberts opines that, "You see a building in a sight -- it looks more like a video game than anything else," we may be struck by the sinister cybernetics of the scene and so be moved to reiterate urgent questions of the sort Manual De Landa first posed in his prescient War in the Age of Intelligent Machines, published the same year as the events in the Gulf. But the conspicuous reference to video games -- repeated over and over again in both the popular media commentary and in the testimony of historians and other members of the intelligentsia -- is of interest to me here because it would seem to depend on a rather exacting visual correspondence between the images distributed by the military and the look and feel of flight simulators and arcade-style combat games. And in fact, I want to suggest that the Gulf War images look very little like most video games. It is certainly true that there are some superficial resemblances: for example, the Pentagon videos usually include a crosshairs or targeting apparatus, as do many arcade-style simulations; and it is perhaps also true that the cryptic data readouts sometimes visible in the images can be taken as something like a score or a point-tally. But to my eye, the differences far outweigh the similarities. The Pentagonís images are grainy and black-and-white. Explosions and bursts of light quickly overwhelm the infrared filters on the camera, washing across the screen as unprocessed white noise. By contrast, the aesthetic associated with video games, and digital images more generally, is firstly one of color and secondly one of crisp, clear, high-resolution detail (figure 16 depicts a relatively primitive combat game, and figure 17 a more sophisticated one -- but neither much resembles the Pentagon video images). Few of these attributes are evident in the crude images that are regularly called upon to situate the Gulf War within a horizon of digital artifice. Again, I want to stress that my point is not that observations about the push-button, channel-surfing "postmodern" conduct of the war are misguided in the broadest sense, but rather that the easy and even enthusiastic identification of a grainy black-and-white image with a "video game" suggests something about the low-definition lack of discrimination with which we habitually look at digital images. In this case, the most rudimentary of formal resemblances -- a crosshairs overlay -- suffices to seal the connection. To a large extent, the formal properties of electronic images are still invisible to the public eye, despite the rapidity with which their aesthetic is spreading through the cultural sphere.
Let me turn towards a second class of images by way of further example. These images are not representational in the conventional sense -- that is they do not correspond to any referent in the "real" word. Rather, they are first generation electronic artifacts, visualizations of electronic networks and digital file systems. Figure 18 is the most widely reprinted visualization of the Internet or "cyberspace." It was created in 1992, by researchers at the National Center for Supercomputing Applications and depicts the NSFNET, the backbone of the original ARPANET. The imageís aesthetic dimension is obvious. Clearly effort has been expended to make it visually appealing: the black background, for example, which enhances the slick cybernetic sheen of the image, or the three-dimensional geographical rendering of the United States when a two-dimensional outline would have been serviceable -- but much less striking. From these features, and from the brilliant luminosity of the lattice-work of data, it is obvious that there is more to this image than simply an objective visualization of the traffic on a layer of the network. (The color bar at the bottom indicates the intensity of information transmission.) What does this image really mean to tell us? Certainly not that data exchange is densest on the NSFNET backbone, a fact that would be common knowledge to those who care about such things. This image is an aesthetic event before it is an analytical device.
Figure 19 is a second example, from the same genre. This is a more recent visualization, from the University of Illinois, displaying server-loads on the World-Wide Web. One again there is a clear aesthetic element, most evident, I think in the dramatic perspective afforded by the birdís-eye view from atop the data towers jutting from the West Coast of the United States, a position that seems calculated to induce vertigo through the scale provided by the tops of other, lesser data stacks clinging to the landscape far below. On the one hand, it should come as no surprise that aesthetic elements have historically played an important role in cartography and information design. But if on one level these images are to be accepted as merely the concrete visible manifestations of the raw data and statistics from which they have been generated, then on another level they must also be accepted as imaginative structures (perhaps not altogether unlike the fanciful cartographic renderings of the New World created in the Sixteenth Century) -- premeditated compositions which are pre-mediated by the various contexts affecting our imagination of the imaginary space of cyberspace: contexts which include science-fiction, the popular media, the hacker subculture, and so on.
Figure 20 is a screenshot from MAPA, Java-based software under development by Paul Kahnís Dynamic Diagrams design firm. It is intended to provide users with maps of the file structure on a Web server. Hyperspace (figure 21) is an academic rather than a commercial project, from the Advanced Interaction Group at Birmingham in England. Once again its purpose is to provide a visible rendering of data in a file system (the molecular clusters represents folders and sub-directories and individual documents or objects). The contrast between this image and the MAPA softwareís more understated display -- in fact, Hyperspace closely resembles some of the renderings in Michael Benediktís Cyberpsace volume -- underscores what I would like to call the phenomenological materiality of electronic media. Not materiality in any real physical sense, although Nicholas Negroponte promises that "[m]ultimedia will someday be as subtle and rich as the feel of paper and the smell of leather." But our current technologies of human-computer interface are fundamentally graphical, what McLuhan would have called the electronic stained glass window. So electronic media, I am arguing, assume material form through the models and renderings and visualizations that we create to interact with it in various information states. This interaction occurs at the immediate phenomenological level of the user interface, and it is here that the visual materiality of information has the potential to become aestheticized -- even radically aestheticized, as in the case of the Hyperspace renderings.
Ergonomic postures dictate an optimum distance of eighteen to twenty-four inches between our eyes and the computer screen. We look into this visual field every day, but only rarely do we look at it -- perhaps because what we see there has already become all too familiar. But it is important that we begin refining the technical, analytical, and theoretical tools that will allow us to look at informationís phenomenological states more closely and more critically -- in essence demystifying "information" as the transcendental signifier of the Information Age. Computer-generated imagery will constitute an ever-expanding field of our vision -- as the data visualizations I have been discussing above amply demonstrate, for they subordinate information "itself" to the various conventions of representational artifice. Because of this I am convinced that aesthetics and visual form will be the most vital site for information and media studies in the coming years; no media, by definition, can exist without mediation, and the willful suppression of mediation cannot be achieved without a simultaneous surrender of knowledge concerning the relationships among various material technologies, their implementation, and ourselves.