Artbyte (New York) 2, no. 2 (Summer 1999), pp. 26-27
Do we want convergence?
Cable-modem pitchmen boast downloads ten or a hundred times faster than ordinary copper-wire modems. WebTV advocates talk of connecting billions of people who can't afford computers to the Internet via their televisions. Microsoft is pitching its scaled-down version of Windows, known as CE, as the operating system of choice for everything from palmtops to automobile navigation systems to toasters. Cisco Systems and other routers are at work on "Internet 2.0," a protocol whereby the same phone line would carry voice, fax, voicemail, and e-mail simultaneously.
I don't know about you, but this kind of talk scares the hell out of me. It's not that I have a particular affection for glacial downloads or dumb appliances; it's just that the "advantages" offered by these particular models of convergence are asymmetric. Cable download speeds *are* blistering; cable upload speeds are not. Bringing the Web to TV *would* encourage more poor people to surf the Internet, but in migrating from desk to den the cyberspace portal would move from a place reserved for work to a place reserved for passive consumption. Having my toaster hooked up to the Internet would tell me nothing about GE--but it could tell them quite a bit about me. And unlike the old, democratic IP switches, which couldn't distinguish between an AT&T commercial and e-mail from your brother-in-law, the current designs for Internet 2.0 routers would prioritize some traffic over others, so that companies can bill them separately at different rates. All of this means a lower percentage of netizens making Web pages and sending e-mail, and a higher percentage channel-surfing a push-driven Internet on the same box as NBC and CNN. Let's face it: the average couch potato, if left on the couch, will choose a TV dinner over a keyboard any day.
So what, say convergence advocates. People who are really dying to hang out in chat rooms and upload pictures of their kittens will still go through the effort, and everyone else will have access to information on an unprecedented scale. Besides, a big problem with the Web right now is that you can't trust the information you find there. People rely on the Time Warners and MSNBCs of the world because the familiar logos reassure them. If the ratio of people who use the Internet as a participatory, two-way medium drops, that's just the price we'll have to pay to get more people online.
For the short term, this sounds like a reasonable compromise; for the long term, it's a disaster. As I argued in the December-January 1998-99 Cross Talk ("Is Cyberspace Really a Space?"), the architecture we call cyberspace is still under construction, and the short-term decisions we make now about its character will very likely be interpreted as eternal laws by future netizens. So let's assume asymmetric architecture prevails and fast-forward into the future to imagine the kind of cyberspace that might result. The asymmetric geometry touted by a certain faction of convergence advocates--in which it's easier to travel one way than another--reminds me of nothing more than the Panopticon, Jeremy Bentham's notorious 1843 design for a maximum-security prison. Bentham's plans called for a central guard tower whose windows look out on an annular building divided into prison cells. Although impenetrable walls divide the inmates from each other, each cell has a window facing back at the tower as well as one on the opposite wall facing out into the world. The function of this window to the outside is not to give inmates a room with a view, but to backlight them against the light of day to make them all the more visible to the inspectors in the tower. For their part, the inspectors employ Venetian blinds to ensure that inmates cannot tell at any given moment if they are being watched. Although Bentham had in mind the limited application of his design to prison architecture, Michel Foucault called the Panopticon "the diagram of a mechanism of power reduced to its ideal form," and went on to describe its influence in the nineteenth and twentieth centuries on the architecture of hospitals, workshops, and schools--in short, "whenever one is dealing with a multiplicity of individuals on whom a task or a particular form of behavior must be imposed."
Of course, in Bentham's Panopticon, the elite get information from the population but not vice versa; in the network TV model, the population gets information from the elite (if you call sitcoms and soap commercials "information"). Since in both cases it is the centralized elite that *controls* the information, this difference doesn't mean much. Nevertheless there are more subtle and important ways that a twenty-first-century Panopticon would be different from its nineteenth-century counterpart. The original Panopticon, as its name implies, depended on light as a means of gathering information. But the way light illuminates a perspectival space is different from the way electromagnetic signals--even if traveling through fiber optic cables--spread knowledge through network. Perspectival light radiates. An illuminated object reflects light outward in all directions from its position. Conversely, a central eye, by swiveling in different directions, can monitor incoming light from every quarter. For this reason, perspectival space is continuous: a guardian of the Panopticon could do a quick count of the entire prison population simply by panning his glance 360 degrees across the rows of cells in the peripheric ring.
Cyberspace, however, is built not from bricks and glass but from computer code. And no matter how much computer memory and processor time programmers devote to simulating perspectival space, even the most advanced forms of virtual reality are fundamentally composed of multiple, intersecting loops--and often show it. A good example of this fundamental discontinuity is reflected in Jenny Holzer's virtual World Two (1993). Viewers who don the headset of this virtual world can steer into the darkened portal of a dwelling, but just as they pass through they will notice a momentary pause in the program before they find themselves in an empty room. This delay is caused not by a bug in the virtual world's computer code, but by the very structure of the code itself. To take advantage of the computer's knack for doing a lot of dumb calculations very quickly, programmers write procedures that repeat the same simple calculation over and over in a recursive loop. For complicated applications, programmers nest these small program loops inside larger loops. When the computer has to switch between two very large loops, however, even a sophisticated machine often halts output for a second or two. To some extent, then, programmers are fighting the natural tendencies of the computer when they try to smooth over the seam between two loops in a program. No matter how well camouflaged for the user, the threshold between two loops in a program--such as the doorway leading from a landscape into a house--may require the machine's processor to perform a prodigious leap across pages of computer code.
Hoping to be seduced by the virtual world they have paid to experience, most viewers will ignore this bump in the road of perspective and assume the exterior space of the landscape is continuous with the interior space of the room. Holzer, however, doesn't let them swallow the illusion quite so easily. Say when they entered the portal the sky was blue and the sand a brilliant yellow. Once they steer their way out again, they may find scrubby green tundra against a coal-gray horizon. It's not really that the textures and colors of the landscape completely changed while they were inside--there was no landscape waiting outside when they were in the room, just a switch to connect them to another loop in the program. And just because they entered through a certain door is no guarantee that leaving through that door will return them to the same place. Architects of an asymmetric cyberspace can "build in" a preferred direction: no matter which door you take, it might lead you ineluctably toward the same destination.
Along with continuity, cyberspace does away with nineteenth-century assumptions about contiguity as well. If an inmate were missing from a given cell in an ordinary prison, the authorities would interrogate the next-door neighbors in the same cellblock, figuring they might be in on a conspiracy. Precisely to prevent such collaboration, Bentham designed walls between adjacent cells so that contiguous prisoners couldn't see each other. But in cyberspace, information's tendency to suffuse the local neighborhood is replaced by a tendency for information to shoot off in one or more discrete vectors. Contiguity used to be geographic, and a message was associated most strongly with its origin: "This letter came from Yokohama, which is near Tokyo." If there is any contiguity in cyberspace, however, it is linear, and a message is associated most strongly with its pathway. To track down cyberconspiracies, overlords of the network would interrogate not neighbors but nodes. The next time you get one of those e-mail chain letters, take a closer look the return-path header of the message:
means the message traveled from a university in Yokohama to the Food and Drug Administration in Washington to a pharmaceutical company in Philadelphia to a private mailbox in San Francisco. While the contacts between individuals revealed by such data trails are linear rather than geographic, they can still be incriminating in the right circumstance. Microsoft recently trotted out e-mail by disgruntled Netscape employees to suggest that its domination of the market was due to its competitors' incompetence rather than unfair monopolistic practices. Should the courts admit e-mail dug out of archives as legal evidence, it will discourage the public from using this otherwise private, reciprocal medium and further encourage the architecture of asymmetry.
Along with continuity and contiguity, the third assumption of the Panopticon that won't survive the transition from enlightenment to connectedness is the horizon. In the past, the horizon was where enemies came from, so you built lookout towers to see barbarian invaders before they got to you. The horizon was also where you escaped to, so the guardians of the Panopticon controlled the sight lines through which escaped inmates would try to flee. Now that the backlit horizon of the Panopticon has been replaced with a backlit screen, we no longer see an invader before it is upon us. Paul Virilio has noted that in previous eras military control hinged on keeping the target always in sight. In the future, however, the means of control may not depend on visual contact. Reconnaissance can be replaced by a global Search, strikes by Search and Destroy, invasion by Search and Replace, colonization by Search and Replicate. (These new tactics can be just as prone to "collateral damage": witness the breast-cancer survivor groups deleted by America On Line administrators, who indiscriminately eradicated every chat that included the word "breast.") As the offense abandons the Panopticon, so the defense must abandon the medieval tower. In Panoptic space, the higher the lookout tower, the better the control of information. In cyberspace, however, control is maintained not at a high level but at a low one: back doors into Windows programs can be opened by those who know DOS, and back doors into DOS can be opened by those who know machine language. The lowest level of computer code is the most powerful. So perhaps the best defense would be to abandon the lookout tower and go underground, tunneling from one domain name and e-mail address to another, screening sensitive communications with deep encryption. Only then could netizens go against the preferred direction established by the architects of asymmetry.
As in the Panopticon, the disempowered in an asymmetric Internet will still be, in Foucault's words, "the object of information, never a subject in communication." In a footnote to his discussion of Bentham's prototype, Foucault mentions that Bentham's first plan for the Panopticon had included an acoustic surveillance system with pipes leading from the cells to the central tower. The problem was that the prisoners could also hear what the inspectors were doing. It is tempting to see Bentham's acoustic network as a prototype for today's Internet, especially since Marshall McLuhan once claimed that electronic networks are based on an acoustic world-view. It is also tempting to conclude from Bentham's rejection of the idea that free access to information is automatically empowering. But we must not mistake information flow for information control; the mere ability to receive information does not guarantee that you can use it to better your situation or other people's. Empowerment, whether physical or electronic, isn't just a matter of receiving a commodity, but of taking it home ("downloading"), modifying it, and passing it on ("uploading"). Only by encouraging a cyberspace that allows reciprocal exchange on as many levels as possible can we escape the network equivalent of the Panopticon.