Plenary Proposal

Looking Backward: Visual Culture and Virtual Aesthetics, 1984-1998

VRML Text Visualization by the Author

Matthew G. Kirschenbaum

Department of English and
Institute for Advanced Technology in the Humanities

University of Virginia

This is a paper about information theory, but not information theory in the traditional sense. Information theory as it is commonly discussed is associated with a group of philosophers, mathematicians, electrical engineers, and computer scientists acti ve in the late 1940s through the 1950s, who developed models for the transmission and reception of signals and messages based primarily on an understanding of information as a mode of difference. This work, much of it performed under the auspices of the M acy conferences on cybernetics, coincided with investigations of complex systems, positive and negative feedback, and the concept of homeostasis. In what follows, I propose that the last fifteen years or so -- from the early 1980s to the present day -- ha ve given rise to the emergence of a new phase of information "theory," equally significant though in some ways inverted from the abstract, schematic orientation of the earlier cybernetic strain. This latest phase of information theory is not pri marily technical or scientific in its origins, but rather cultural. (How could it be otherwise at a time when we are continually reminded that we are living under the appellation of an Information Age?) My basic contention is that information has now assu med visible and material form as a definable and even dateable set of aesthetic practices; a visible spectrum of tropes, icons, and graphic conventions that collectively convey the notion of "information" to the eye of the beholder. At stake is not whether any such conventions for representing information are accurate or correct to the formal ontology of information in an absolute sense, but rather the important fact that Western consumer culture has evolved sophisticated and compelling conceits for depicting information as an essence sufficient unto itself, or more properly, information as a synthetic, at times even haptic, commodity. That same idea can be succinctly conveyed via the shorthand term artificial information, which I use thr oughout this paper. Predictable though these latest permutations of the culture industry might seem, understanding information as a form of artifice -- and not as a token of difference -- carries with it implications that can be pursued across a broad arr ay of contemporary aesthetic, social, and technological developments, from new media artists who appropriate the look and feel of information to conduct their own kind of inquiries into the phenomenology of digital culture, to advertising and graphic des ign where the visible signs of information are consolidated and set in circulation through a variety of different media forms (printed matter not the least of them), to scientific research centers investigating new techniques in information design and vis ualization -- where information itself often becomes the explicit subject of representational technologies, as in attempts to "map" electronic data structures or the nodes and paths of computer networks.

To begin, a few words about methodology. Information theory as it was codified following the Second World War was primarily an academic phenomenon, though several of its most important proponents, such as Norbert Wiener, eventually achieved popular sta tus and notoriety for their work. But the critical process involved in studying this foundational era of information theory could only properly begin with attention to the institutional micro-histories comprised of its primary documents: the proceedings o f the Macy conferences (five of which have been edited and published by the Macy Foundation), the correspondence of key figures like John von Neuman, and the seminal papers and monographs, such as Wienerís Cybernetics: or Control and Communication in t he Animal and the Machine (1948), Claude Shannon and Warren Weaverís The Mathamatical Theory of Communication (1948), and Donald Mackayís various papers attempting to refute Shannonís foundational premise that information could be defined indep endently of meaning (eventually collected in a 1969 volume entitled Information, Mechanism, and Meaning). But the kind of information theory I am concerned with here and now, in our present day, does not manifest itself in the compressed and conde nsed form afforded by conference proceedings and other channels of scholarly communication. The primary locus of artificial information is without question the popular media: film, television, fiction and magazine publishing, and of course, the new digita l media. Consider William Gibsonís 1984 novel Neuromancer, which gives us what is surely the most famous prose evocation of cyberspace: "The matrix has its roots in primitive arcade games, early graphics programs and military experiments with cranial jacks. [. . .] Cyberspace. A consensual hallucination . . . A graphic representation of data abstracted from the banks of every computer in the human system. Unthinkable complexity. Lines of light ranged in the nonspace of the mind, clusters and c onstellations of data. Like city lights, receding . . ." (51). This sparse description (which, it may be of interest to note, is communicated not by Gibson as narrator but rather by the anonymous voice of a content-provider in a documentary sequence) has been frequently quoted, often as the de facto definition of "cyberspace" rather than the articulated fiction that it is. Though Sandy Stone is surely right to say that in the broadest sense the concept of cyberspace was "pulled from th e kinds of electronic networking [Gibson] saw already in use all around him" (varieties of networked experience that would also become the backdrop for Donna Harawayís "Manifesto for Cyborgs," first published only one year later and probabl y still the most influential cultural critique of the new information technologies ever written), there were also more specific points of departure for Gibson (33). These become evident when the fast- and forward-moving trajectory of the above passage is deflected laterally across the bright brittle terrain of the entertainment industry in the early 1980s. The arcade games mentioned in the first sentence, for example, were also the explicit inspiration for the computer graphics and virtual landscapes port rayed in the popular Walt Disney film Tron, released two years prior to Neuromancerís publication. Tronís special effects and animations, highly innovative for their day, are strikingly similar to the cyberspace envisioned by Gibson, as evidenced the following segment from Tronís screenplay: "We seem at first to be in the Electronic World still, flying over a vast circuit board lit by countless dots of light. . . .As we fly over the grid, descending, the image comes into c learer focus, and we realize that this is not a circuit board, but rather an actual landscape, a suburban grid at twilight." Whether or not Gibson was directly inspired by Tronís imagery seems secondary. (In interviews, he cites not the movie but rather his own experiences observing patrons of video arcades for the idea behind his version of cyberspace.) The primary question ought to consist in why it is that two of the most popular science fiction narratives of their day both deployed all bu t identical aesthetic conventions to evoke the otherwise enigmatic idea of an abstract electronic setting -- aesthetic conventions whose genealogies, it seems plain, originate much earlier in the century, with the widespread introduction of neon or outdoo r electrical lighting into urban environments (". . . like city lights, receding").

This then is one form that an investigation of artificial information might take: a survey of the more significant texts and artifacts involved in the process of synthesizing its unique aesthetic signatures. That the cyberspaces of both Neuromancer and Tron (as well as other cyberpunk productions such as the short stories in Gibsonís "Burning Chrome" anthology or even the 1981 animated feature Heavy Metal) are artificial alloys derived of complex cultural skeins may seem an obvious point, but I have seen that kind of comparative analysis developed only infrequently in the critical literature on cyberpunk and science fiction, not to mention studies of advanced information technology more generally. If we were to continue p ursuing the influence of Gibsonís work, we could move from popular entertainment to the speculative writings of the so-called digeratti, that loose clique of artists and public intellectuals who have emerged over the course of the last decade with some of the most widely-read accounts of the future of human-computer interaction. A touchstone here would be Michael Benediktís anthology entitled Cyberspace: First Steps, published in 1988 by the MIT Press. Prefaced by Gibson, the volume contains fiftee n essays, notable today, ten years later, for how literally they read his novels as starting points for actual research agendas in interface design and related fields. All of the essays oscillate between tacit recognition of the preliminary and tentative status of the actual technologies on the one hand, and a willingness to talk about cyberspace as though it were already an observable phenomenon on the other. Some contributors simply choose not to acknowledge this as an issue; Marcus Novak, for example, has no hesitation in informing us that, "The function of [a "cyberspace synthesizer"] is to receive a minimal description of the cyberspace, coded and compressed, and from it to render a visualization of that space for the user to navigate with" (233). The problem here, of course, is that in an essay filled with many similar gestures, Novak is assuming that cyberspace is subject matter evocative enough for the reader to suspend disbelief and benefit from a putatively sober description of a technology that does not in fact exist. Other contributors are more circumspect, such as David Tomas, who asserts the following: "Although cyberspace has been popularized by Gibsonís books, it is neither a pure Ďpopí phenomenon nor a simple tech nological artifact, but rather a powerful, collective, mnemonic technology that promises to have an important, if not revolutionary, impact on the future compositions of human identities and cultures" (31-2). This has the appearance of a balanced ass essment, yet it is clear that when Tomas talks about cyberspace as a "technology" he cannot mean technology in the sense of any specific hardware or software implementation -- a meaning he hastens to jettison by preceding his reference to " technological artifacts" with the qualifier "simple" and by placing the whole of the phrase in parallel with the equally ineffectual notion of cyberspace as a "pure Ďpopí phenomenon." Cyberspace, as it is invoked here, can only be a technology in the sense that the word itself, or more precisely, the idea of cyberspace mimics the behavior of certain material technologies, functioning as a "powerful, collective, mnemonic," or in other words, as a shorthand for a wh ole range of communicative agendas given depth and form by a shared aesthetic. Or to come to the point in different way, nowhere in Benediktís volume, which is replete with blueprints for cyberspace decks, lyrical evocations of virtual skylines, and attra ctive CAD-rendered illustrations and visualizations (which look a great deal like the sets in Tron), have I been able to locate a single sustained discussion of the then twenty-year-old ARPANET, the backbone of the Internet and graphical World-Wide Web as we know them today. From this it is perhaps not too tendentious to conclude that in 1988 the discourse of cyberspace (as recorded in the Benedikt collection and elsewhere) occupied a very different sphere from the telecommunications infrastructure of a global computer network that included electronic mail, electronic news, networked file transfers, remote systems logins, and a good deal more; while today, as anyone exposed to the popular cant of the Information Superhighway knows, terms such as &q uot;cyberspace," "surfing the Web," and "getting online" all arrive and depart from the same vague notion of universal electronic telepresence. The vocabulary that once belonged exclusively to an aesthetic conceit has been superi mposed, by both the popular media and the digeratti, over the hardware and software associated with client-server computing -- though of course the average Web site looks nothing at all like the fanciful renderings in the Benedikt book, and neither tabloi d talk of "Sexual predators stalking our children in cyberspace!" nor John Perry Barlowís libertarian "Declaration of the Independence of Cybersapce" have much resonance with Gibsonís original "fluid neon origami trick" and t he other epithets of the matrix.

But the visual aesthetics of information do not descend solely from William Gibson and early-eighties science fiction, nor is the process of cultural diffusion always one by which a fictive device becomes subsequently identified with functional technol ogies. Sometimes the reverse happens, as is the case with desktop publishing and current "post-alphabetic" trends in graphic design, which in large part stem from the technological event constituted by the mass-market release of the Apple Macint osh. Critics such as Richard Lanham have commented before on the implications of desktop publishing and digital typography, noting that the creative control afforded by the font libraries and clip art galleries at every userís fingertips contribute to the breakdown of traditional distinctions between reader and writer while dramatizing the malleability of words and images in a digital environment. Here I want to argue that graphic design is actually possessed of a deeper and much more specific import for critical observers of the new media: that it is in fact the single most important arena in which the public learns to recognize the look and feel of information qua information. Though it would be historically incorrect to equate the current post-a lphabetic style of graphic design with the Mac itself (significant experiments in non-linear design were being carried out in the seventies at such places as Detroitís Cranbrook Academy of Arts, to say nothing of loose but obvious associations with modern ist movements such as Dada and Futurism), it is correct to say that the Mac invigorated the nascent desktop publishing industry, made it a practical reality. Perhaps most importantly, the early limitations of the Macintosh (its low-resolution display) wer e quickly enlisted by type designers such as Emigreís Zuzana Licko (who began working seriously with the Mac within weeks of its debut) to provide the basic visual components of an electronic graphical identity. Licko says of the process, "I s tarted my venture with bitmap type designs, created for the course resolutions of the computer screen and dot matrix printer. The challenge was that because the early computers were so limited in what they could do you really had to design something speci al. . . . it was physically impossible to adapt 8-point Goudy Old Style to 72 dots to the inch. In the end you couldnít tell Goudy Old Style from Times New Roman or any other serif text face. . . . It is impossible to transfer typefaces between technologi es without alterations because each medium has its peculiar qualities and thus requires unique designs" (18, 23). What began as a material limitation in the medium and its underlying hardware was quickly accepted and adapted as an integral aspect of the mediumís native aesthetic identity, and that identity (think jaggies) has remained ironically intact and recognizable long after the technological base has shifted beyond its initial crudity (indeed, Emperor, Oakland, and other of Lickoís early Em igre bitmap fonts were to remain the basis for a number of her later PostScript designs). By contrast, advanced computer modeling and visualization technologies have only begun to replicate some of the more simplistic aspects of Gibsonís imagining of cyberspace, and functional virtual reality systems (helmets, goggles, data gloves and the like), which have been under investigation since the late sixties, may well prove to be a dead-end agenda. Yet the rational geometrics of Gibsonís "lines of lig ht" and the bitmapped jaggies of the early Emigre fonts both continue to function as among our most powerful agents of artificial information.

At this point my basic methodological assumption should be clear: that any study of informationís artificial and aesthetic dimensions can only properly be a study in comparative media. Digital media cannot be isolated, functionally or theoretically, from the whole of the media spectrum, not least because digital production techniques have by now been so fully incorporated by other media -- the recording industry and Hollywood, to offer just two examples. Yet too many of those who h ave write critically about the new media seem to forget that the computer is not our only platform for digital data, and that computer networks are not our only conduits for "information." Likewise, the fact that digital media operate as multime dia -- or more properly meta-media, in that they accommodate and subsume other kinds of media content (audio, video, etc.) -- is beside the point. The processes by which different media platforms interface with the culture at large (their oft-heral ded convergence into a single black-box receiver notwithstanding) differ in important ways -- the apparent failure of WebTV being a case in point. Or else consider Wired magazine, whose first issue was published in January of 1993. Though tame by c omparison to much of the innovative design work of the late eighties and early nineties, Wired, like no other publication I know, seems clearly intended to look like information, to look like it through and through and cover to cover. This v isual flamboyance has been frequently remarked upon by its critics, and the magazine often performs as a synecdoche for the "Information Age" -- despite the fact (or as I would be inclined to argue, precisely because of the fact) that it is a pr inted and not an electronic artifact. Yet, from its inception in 1993 (when it was initially published only bi-monthly), Wired has, like any corporate entity, evolved into a multivalent media presence that now cultivates a variety of distribution n etworks, among them: an imprint for popular books about cyberculture, HardWired; an online forum and Web center, HotWired, as well as separate Web sites for the magazine itself and other services (WiredNews, LiveWired); and two allied search engines, HotBot and NewsBot. To remark upon Wired the magazine independently of this broader media identity seems tendentious at best. In this context, Michele Pierson, who has written about the aesthetics of Wired in relation to the Hollywood special effects industry, observes: "The computer generated imagery that glosses the pages of Wired magazine are [sic] visually more spectacular than anything that Hotwired -- the magazineís on-line count erpart -- has to offer. By comparison, Hotwiredís clunky frames, day-glo wallpaper and occasional animated graphics offer a computer graphics environment lacking in visual effects" (46). To this I would add that Hotwiredís relative imp otence in the very medium its collective name-brand identity serves iconify dramatizes the extent to which informationís aesthetic conventions bear no necessary or organic relation to the technological base of any one media form (a lesson also apparent fr om our current, and ultimately arbitrary, association of "cyberspace" with the HTTPD layer of the World-Wide Web, or from PostScript fonts that self-reflexively retain their bitmapped edges). We also begin to see, in keeping with Lickoís observa tions above, that it is the differences between media platforms -- manifest at the deep ontological levels of data standards, transmission protocols, and the material functionality of the hardware itself -- that produce some of our most compelling instanc es of artificial information. The printed version of Wired, in other words, is a powerful instrument for distilling the aesthetic artifice of information precisely because the magazine and its putative subject matter occupy two separate media, ther eby necessitating a good deal of careful thought and decision-making about how best to capture and consolidate representations of digital media in their most potent and accessible form. That Wired (the magazine) is often received as the avatar of t he Information Age testifies not only to the self-consciousness its designers and art director have had in this regard, but also to the power and resiliency of the traditional publishing industry in this, the late age of print -- and let us also not forge t that laptops, PDAs, wearable computers, and digital ink notwithstanding, print still enjoys a more intimate and pervasive network of mechanical reproduction than anything in the online world.

But of course, in comparative terms, even this offers only a very partial perspective. Let us not forget that the aesthetic of Wired (or that of lesser-known but more visionary publications such as RayGun or Mondo 2000) is also, fo r example, the mixed-media aesthetic of MTV, which first went on the air in 1981 with a pop jingle by the Buggles that is itself a titular lesson in comparative media: "Video Killed the Radio Star." And throughout the eighties, music videos were important showcases for computer-generated animation and effects, notably in work by performers such as Peter Gabriel and Todd Rundgren. Yet MTV cannot be understood apart from the much broader late twentieth century phenomenon the critic Larry McCaffery has dubbed Avant Pop, which circumscribes the increasingly commonplace fusion of popular media imagery to the techniques of the modern and postmodern avant garde -- a fusion that is often practically accomplished by means of digital technology. And so fo rth. The media ecology of the past twenty year is, as hypertext guru Ted Nelson is sometimes wont to say, deeply intertwingled.

After developing the concept of artificial information along the broad lines sketched above, I will turn toward three specific case studies: David Blairís extraordinary electronic film and hypermedia environment Wax, or the Discovery of Television A mong the Bees, representing the new media arts; David Carsonís graphic design, representing the post-alphabetic aesthetic in contemporary typography; and various examples of information mapping and scientific visualization, which provide us with some of the most vivid displays of information as visual artifice. I close this proposal with some more specific words about each of these topics and their relevance to my project of a visual genealogy of virtual culture.


Comprised of over 2000 nodes and 25,000 links, Blairís WAXweb is by far the single largest creative hypermedia environment extant today, dwarfing such works as Michael Joyce's landmark hypertext novel afternoon (which is comprised of fewer than 600 nod es and under 1000 links), and Stuart Moulthrop's Victory Garden (approximately 1000 nodes and 3000 links). But the most compelling reasons for study arise from WAXweb's synchronic and diachronic complexity. As a hypermedia environment, WAXweb occupies sev eral parallel media planes, which currently include or have at one time have included: Storyspace webs, MOO objects, HTML documents, JPEG images files (approximately 4800 of them), MPEG video files (560 of them), WAV audio files (2200 of them), and VRML o bjects and scenes (approximately 300 separate renderings). This complex media ecology has evolved over some four years of online development: as new technologies and data standards have become available, the project has been ported and migrated across sev eral platforms and file systems, discarding some of its earlier medias, while absorbing and layering others. (The present incarnation of WAXweb is publicly identified as version 3.0, Alpha -- its current focus appear to be shifting from the MOO to VRML sp aces.) Moreover, WAXweb is an extension of Blair's 1991 film Wax, or the Discovery of Television Among the Bees (85 min.), itself constituted enitirely by electronic images and graphics -- these form an antediluvian layer in the environment that is WAXweb today. (The current VRML objects, for example, were derived directly from the original 3D renderings used in the film.) In 1993, one year before the first version of the WAXweb hypermedia was brought online, Wax the film was broadcast across the I nternet's high-bandwidth multimedia backbone, the m-bone, becoming the first full-length film to be distributed in the medium. The WAXweb, which in fact recognizes the artifice of "information" as one of its central thematic touchstones, provid es a rich environment for examining the extent to which the aesthetic properties of information are in fact residual traces of material differences among platforms and media forms, as manifest at the deep ontological level of data standards, transmission protocols, and the interface between hardware and software.

The New Typography

That there is a material aesthetic associated with the virtual is plain to see, not only in digital media but also printed publications. It is reflected in its most bastardized and least interesting form in Wiredís teflon sheen, but its visible spectrum extends from Wired to the grunge fonts and multilayered letter-forms which have emerged as the signature styles of the graphic design programs at such places Detroitís Cranbrook Academy of Arts and CalArts in Valencia. Since 1984, the same year as the mass-market release of Appleís Macintosh, much of this design work has been associated with Emigre magazine, founded by Rudy VanderLans and Zuzana Licko to showcase the Mac-generated fonts created at their digital type foundry in Berke ley. Influential designers such as Edward Fella, Neville Brody, Anne Burdick, Steve Tomasula, Susan LaPorte, and Michael Worthington, many of whom have Cranbrook or CalArts affiliations, all gained early exposure through Emigre, which offers itself as an alternative to the more mainstream trade journals EYE and Print. Since the early nineties, the best-known practitioner of innovative visual and graphic design -- whose work is often described as the new typography, deconstructive typ ography, or digital typography -- has been David Carson. A former surf-celebrity with little formal design training, Carson has attracted an international following for the layouts and experimental fonts which first appeared in the six-issue run of a maga zine called Beach Culture; in 1992 Carson became the art director at the music magazine Ray Gun, founded by publisher Marvin Jarrett as an underground competitor to Rolling Stone, Creem, and Spin. Today Carson commutes t o art-school workshops and seminars around the globe, while designing dissonance for the likes of Coca-Cola, Swatch, and Hardees. I shall be discussing the dynamics of his career and cult-following in a good deal more detail. For now suffice to say that C arson, as the most closely-watched designer of the nineties, has done as much as Template Gothic to consolidate the look of the decade.

In his forward to a hardbound Ray Gun retrospective entitled Out of Control, William Gibson describes the work collected there, much of which belongs to Carson, as "The event horizon of futurity, as close as any windshield, its textu res mapped in channel-zap and the sequential decay of images faxed and refaxed into illegibility . . . brave new worlds abraded onto the concrete of the now. . . . This is design pushing back against the onslaught of an unthinkable present" (13). In what follows, I want to suggest that this post-alphabetic aesthetic is one that appears precisely at the point of print mediaís imperative to formalize a representation of its own putative demise. That is, it is an aesthetic that is intens ely self-reflexive in its attempt to depict, and at some level iconify, the material conditions of printís communicative exhaustion. The body of graphic design work associated with Carson, Ray Gun, Emigre, Cranbrook and CalArts ther efore bears close scrutiny by students of the new medias, for it dramatizes that aspect of the relationship between print and electronic textualities driven by the need of the former to assimilate and contain the ruptures of the latter.

Information Landscaping

The finally area of the paper will be directed largely toward the convergence of information and aesthetics in scientific research, particularly in such fields as computer-assisted visualization and information mapping. Perhaps nowhere is this phenomen on so dramatically illustrated as in the online documentation for the Language Visualization and Multilayer Text Analysis project by members of the Visualization Group at Cornell University's Theory Center. This is a project which aims to develop: ". . . a tool which allows a researcher to explore interactively the structures and typologies of discursive formation in large samples of textual data and develop new techniques for reading and interpreting text space." In other words, the researchers are attempting to find ways to effectively visualize and present textual structures as information. What's most striking about the images they've produced, however, is that they are, to my eyes, simply gorgeous to look at -- the screenshots of the actual i nterface no less than the more deliberately aestheticized compositions in the documentation's "palimpsest gallery." A second striking example comes out of Sandia National Labs, where researchers are now in the process of fine-tuning algorithms which will analyze relationships among various fields of scientific research -- chemistry, biology, physics, etc. -- and then render those relationships as a three-dimensional landscape, complete with mountains, valleys, continents, and oceans: an Imago Mundi for the next millennium. Such undertakings may be grouped under the umbrella name of "information landscaping." With origins in Muriel Cooper's Visible Language Workshop at MIT's Media Lab, information landscaping attempts to visualize and render the "d eep" contours of information structures. In this same context I also explore computer and virtual reality interface design, and information mapping as it is currently carried out on the desktop, including the recent flurry of Web-based visualization aids, such as Web View (software which maps the distribution of documents across a server) and Alta Vista's LiveTopics prototype, a tool which spatializes the results of a search query. It's also worth noting that many of the same principles and heuristics of visual design evident Carson's graphic design, are also employed by those involved in interface design and data visualization. The result, therefore, is that relationships between information and aesthetics in these contexts quickly become thoroughly imbr icated, partly because aesthetic qualities are recognized as important design elements, but also because the finished products, as with the Cornell project, are often quite striking to look at. This last point is not to belabor the obvious, but a s urvey of the critical literature on electronic media and related topics reveals surprisingly little attention devoted to visual aesthetics (as opposed to visual rhetoric) as a relevant category of experience for medium. Yet, it is my contention that the l uminous ray-tracings of the virtual can be as evocative as some of the finest examples of Renaissance cartography.

Vita and Contact Information available at: