Error 404: Doubting the Web

Stuart Moulthrop
School of Communications Design
University of Baltimore

The final version of this essay appears in Metaphor, Magic, and Power, Ed. A. Herman and T. Swiss. New York: Routledge, 2000, pp. 259-76.

1: Not found

If I say, "the Web" -- casual shorthand for "the World-Wide Web," itself a dubiously loose way of talking about certain things possible within Hypertext Transport and Internet Protocols -- you will probably form some immediate set of impressions. These may involve vast, trackless information spaces (the abstract or topological Web), or more likely, particular features in this indeterminate expanse ("pages," "sites," "channels," and more recently "portals"). Clearly the first alternative will not suffice. To speak of the World Wide Web as pure abstraction confers no more understanding than thinking about "the telephone" or "radio" or even "network television" in such imprecise terms. The many objects and interests caught up in the technologique of WWW/HTTP/IP can hardly be glossed so simply. They constitute something that is more event than object, more subjectivity than subject, more a network of bewildering particulars than a system of generalized content.

Yet the alternative strategy -- veering into the particular, as Joan Didion might say -- seems no more useful. What instance of Web production can serve as metaphor or even metonymy? Microsoft's vision of the Web fused into the Windows desktop might seem a logical candidate based on its audacity, if not its actual universality. To symbolize the Web in this way would identify the phenomenon mainly as a tool of productivity, the non-stick surface or myelin sheath for Mr. Gates' "frictionless economy." But even as its capital base expands to surreal proportions, the information economy has yet to demonstrate solid and reliable gains in efficiency or production (see Landauer). There is in fact good reason to suspect that digital networks promote substantially different models of commerce and even of value (see Kelly, Rules). Where else can we go today?

Perhaps to the city of Melbourne, where I used to see regularly a tram poster that said in large letters: STOP PLAYING WITH YOURSELF. Aimed at a certain kind of eternally adolescent male, the banner touted an Internet provider who offered access to multi-player, on-line games like Doom and Ultima II. The pitch seems revealing in both its obvious senses. The Internet and Web are often condemned as a great market of pornography (a function they inherited from earlier media like pulp printing and 8-millimeter film), and even when not about playing with yourself, the Web often panders to activities not permitted on company time. Maybe the Web and related technologies have less to do with business than pleasure -- or with entertainment, that always profitable combination of the two.

On the other hand, surely the World Wide Web has more to offer than fleshpots and gunsights. Reading the Web as an entertainment medium obscures its role as forum and fount of potentially important information. Consider on the one hand the millions of "home pages" featuring everything from vacation snaps to political philosophy, and on the other, the infamous "Drudge Report," the Web site whose scandal-mongering touched off one of the most significant political controversies in recent U.S. history. Each type of publication represents a watershed or possible turning point in the latter-day evolution of mass media, and in both one might see a turn away from centralized authority and rigid control of media markets. Consequences of these developments may be good (an end to the passive consumerism) as well as evil (misrule by fanatical elites); but in either case, some applications of the Web clearly go beyond mere amusement.

In fact, the Web's implications may be very large indeed. If it lives up to its currently dubious "World-Wide" status, we might expect the Web at least to complicate the effects of cultural imperialism and nation-state identity in the New World Order. Thought-experiment: choose ten Web pages that represent what you consider the most important developments in this medium at the moment. How many of those pages use a language other than English? What is to be said about Web sites that English-only speakers cannot read, or can only barely comprehend? Or to take the inquiry beyond language to cultural practice: what about Web publications that do not carry corporate advertising, or are not indexed by search engines and linked from "portal" sites?

To ask these questions is to raise issues of accessibility and access, which are always crucial in cases of technological innovation. Access can be mediated by language and culture, but material factors also come into play. To whom is the Web visible, after all? Or to turn this question the other way, even for those who have ready access to the Internet, how much of the Web is visible from moment to moment? What about the pages that are very hard to find -- or impossible to find at all? Perhaps this, after all, is the most representative aspect of anyone's Web experience:

Error 404:
The object you requested
could not be found on this server.

In fact, I have begun to think this error message may be the most profound thing one can say about the World Wide Web -- the best representative for all its shifting multiplicity. This notion leads to a very serious question: What if the Web as we think we know it does not really exist?

2: Detour

Though wise to the dangers of Cartesian thinking, I have to ask indulgence at this point for something that might look like an attempted cogito. I can cheerfully enough deny existence of "the Web" in general, but I have a harder time calling into question one of the Web's fundamental features: hypertextuality. This is partly because I have spent more than fifteen years thinking about and tinkering with that concept. While I would not go so far as to say "I link, donc je suis," I do come back to hypertext as something not altogether dubious. What follows is a short and somewhat personal digression meant to explain how it is possible to maintain an interest in the Web even as one doubts or even denies its proper existence.

Two major influences helped solidify my emerging interest in hypertext long before anyone had heard of the Web. The first was Michael Joyce's experimental fiction afternoon, which convinced me both that hypertexts can be deeply frustrating and that this frustration, properly understood, yields a fresh approach to reading (see "Hyperreal"). Working through afternoon, I realized that the text was like the proverbial iceberg, or as I described it at the time, a miniature railroad controlled by some remote automaton. Both metaphors are meant to emphasize the importance of the unseen to any understanding of the text that is encountered. In the case of hypertext, what you see is only a small part of what you conceptually get. The text is not all there in a literal sense, and yet what is not visible or present matters very much.

The second early influence on my thinking came from a series of essays by Terry Harpold in which he argued on poststructuralist grounds that hypertext is a fundamentally perverse practice, a space of illusions and "detours" (Harpold, "Threnodies" and "Contingencies"). Like many people beginning to think about hypertext in those days, Harpold had read George Landow's eminently practical rhetoric of "arrivals and departures" (Landow), then headed in a different direction, as Landow himself would soon do. Instead of considering a link as a necessary joining of pre-ordained parts, Harpold insisted that no link ever runs true. Even when operating as intended, every link is phenomenologically a "detour," taking us someplace we did not anticipate. Building on Harpold's insight, Nancy Kaplan and I argued that links traverse a space of possibility that must be considered as much a part of the text as the visible expression itself (Kaplan and Moulthrop). Hypertext is always both seen and unseen, real and hypothetical.

When we came to the Web, or it came upon us, this notion of detours across semantic space gained new significance. Considering the things we began to see on the Web, it seemed that the space traversed by the link had material, social, and even economic implications that meant something important to a growing number of people, and not necessarily just the venture capitalists. It was at this point, around the end of 1994, that I began to see the importance of not-finding, or the deeper significance of Error 404. It was also then, mediated by my emerging understanding of hypertextual detours, that I started to wonder what we meant by "the World Wide Web"; but this was only the beginning of uncertainty.

3. Ask not what the Web means to you

I doubt anyone knows what she or he means by "the Web" -- but this unanswered question leads inevitably to others. To fully understand our situation we must turn the question the other way: What do we (present company intended) mean to the Web? It is important to recognize that as a community of scholars we belong, especially those of us trained from the 1960s through the 1980s, to to a communications regime that differs fundamentally from what may be emerging through the Internet. We were brought up on print and mass culture; and while I admit that these are broad generalizations whose meaning is even less precise than "the Web," it still seems true that for most of my present audience the discursive universe falls into two broad categories. At the center of this domain we find the stable and generally monologic productions of the "serious" intellectual disciplines (science, the law, and their aspiring ephebe, academic humanism). On the fringes of our attention, though perhaps far more present than we care to admit, come the ephemeral and rigidly traditional products of the entertainment industry, those overnight sensations that Pat Cadigan so usefully labels "porn:"

    Valjean had a screen for every porn channel, jammed together in the wall so that food porn overlapped med porn overlapped war porn overlapped sex porn overlapped news porn overlapped disaster porn overlapped tech-fantasy porn overlapped porn she had no idea how to identify. Maybe nobody did, maybe it just bypassed the stage where it would have been anything other than porn. Meta-porn, porn porn?

    I don't know what it is, but it makes me horny, and that's all that matters. (140; emphasis original)

Porn: to paraphrase the recent talk in Washington, it's not about sex, or not just about sex anymore. Something important has happened in the outer reaches of the sign system, a curious alignment of the spheres into more numerous and tightly defined micro-cultural orbits. The contents of the media waste land seem less like fragments shored against our ruins than like self-organizing structures. It may be that this perturbation of the system will send comets tumbling into the academic inner orbits, strange lights that cross our skies with portents of change. Maybe this time it really is the end of the world as we have known it.

Given such a cataclysmic outlook, perhaps a print-based academic can say nothing useful about the Web. Maybe we should consign its strange productions to the cultural Oort cloud along with pop songs, TV shows, comic books, pro sports, and other excremental spectacles. Maybe the smartest strategy is to step from skepticism to denial. Word: What Web? Is there anything out there that really matters? What if the whole thing is just more Rupert Murdoch and Ted Turner propaganda, pure Silicon Alley and Hollywood hype? "Hypertext," after all, begins with that nasty four-letter word. Nor are other formulations more convincing. Gibson's original gloss on cyberspace may be tiresomely familiar but it remains accurate. He called the technology a "consensual hallucination" -- illusion, mirage, mass-mediated phantasm (51). This definition is worth remembering. Any denial can be carried too far, and I would be the last to defend a purely reactionary retreat to print; but at the same time I think we must approach the Web with an attitude of unbelief.

This is not to say, however, that all varieties of unbelief are equal. Some must clearly be excluded as trivial or tautological: for instance, the assertion that the Web, as precursor to the grand vision of cyberspace, somehow falls outside our temporal jurisdiction, vested in a future evermore about to be. The head of a prominent university press once told me, with evident bitterness, that the prospect of electronic publishing made her glad she would soon be retiring. While it is nice not to have to worry about the future, this sort of skepticism is no longer defensible with respect to the Internet. "Cyberspace" may still be science fiction, but the World Wide Web is not -- and anyway, as Bruce Sterling has famously said, we live in a science-fictional world (xi). It is increasingly hard to separate fiction from reality these days. Cyberspace may be hallucinatory (or hallucinogenic), but "consensual" it certainly is not, especially if we understand that word not in Gibson's unusual sense of sensory input but with its more familiar sense of community or polity. Taken this way, the unreal image of the Web seems a shadow or projection of very real social and economic concerns. Things may matter even though -- or as -- they do not exist.

So the Web does not exist; but it fails or refuses to exist in a particular way of which we must take note. If we must give up the pleasure of historical nihilism, we should probably also forego any easy anti-essentialism or reflex affirmation that "the" Web, monologic singular, cannot exist because Web space is of course a multicosm, a heterogeneous network of complex and dynamic regimes. The truth of this statement equals its banality. Heterogeneity matters, as we said in beginning, but that is a given. What things in life are not complex and dynamic? These are fundamental qualities of most if not all social activity, at least insofar as that activity is reflected in discourse. True, the infamous "404" often indicates mutability and instability, and the Web is indeed a strange and shifty set of appearances; but these qualities must be the beginning, not the end of doubt.

We should know, or at least suspect, that reflex anti-essentialism can lead to more serious abuse. Dr. Johnson knew what he was doing when he attacked Berkeley's idealism by kicking the nearest rock. The gesture had an Archimedean meaning. To kick the essence out of something demands leverage. One must plant the other foot securely somewhere. Where do I stand when I point my toe at the Web? What can we say about post-print technologies from the standpoint of the classically industrial, mass-communication regime? From what venue do we issue our announcements that the Web is hallucinatory, chimerical, not quite there? Look in the library, the bookstore, or these days out along the Interstate, in those remainder shops that populate the outlet malls. The Web may leave us in uncertainty, but we know where we are with print.

4: Monster

Enter the devil's advocate, or super-skeptic. Perhaps we can know where we are in relation to the Web as well -- and perhaps, as I suggested earlier, that position will not seem tenable. What if the Web is not just hallucination but pernicious delusion? Let us assume not simply that the Web does not exist, but that surely nothing like it ever could exist. Consider the World Wide Web as (in every sense of the word) a monster. No one will be admitted to the theater during the terrifying death-of-reading scene.

You may have heard that the Web is inherently hypertextual (Haraway 125), even though many chief exponents of that more general technology, from Ted Nelson to Michael Joyce, deeply resist the claim. Authorities may differ about their nature and function, but hypertext links do provide a foundation for Web discourse. Links bring with them an important element of intertextual relatedness, and even a kind of hypertextuality, so it seems safe to assert that hypertext and the Web have something important in common.

But here is a problem. To anyone brought up in an ideology of textual mastery -- which is, after all, what many of us were taught in graduate school -- hypertext is apt to inspire horror. Not only does this sort of writing expand the scope of the textual universe by allowing links from one body of signs to others, it also invites users to complicate and exfoliate their textual productions. There is more and more text all the time and more discursive volume within the component texts. The burden on critics and editors, to say nothing of ordinary readers, expands exponentially. As a literary or literate practice -- viewed from the comfortable parameters of the print regime -- hypertext does not seem to fit into our world of discourse. On practical if not intellectual grounds, then, we might wonder how there could ever be a World Wide Web, at least in the sense of an enterprise we mean to take seriously. The devil's advocate rests.

However, this is cannot be the end of our skepticism. Michel Foucault defined the "author function," our practice of delimiting bodies of literary work under proper names, as the "thrifty" principle that prevents signification from proliferating out of control (158-59). The economic metaphor was well chosen, for there seems to be a deep connection between textual and material economies. This is quite evident today. Now that we have undone the limits on text, the old rules of business practice seem also to have disappeared. We have entered a period of explosion, if not inflation, in finance as well as texts. A recent article in the corporate journal InfoWorld observes that the seven most prominent "all-Internet" businesses, companies like Netscape, Excite, Infoseek, and Amazon, have registered roughly a billion dollars in losses after two to three years in operation (Reed). Most are still losing money and have no clear plan to generate profits except by dominating some inestimable market. Yet at the time of this report these unprofitable companies had capitalizations ranging from $1 billion to $17 billion, figures equal to or greater than those of old-line concerns like American Airlines and Sears Roebuck.

While the vast, silent majority of info-business startups have already failed, and though massive retrenchment may always be just around the corner, there has been as yet no sustained retreat from Internet or Web commerce. The boom continues to defy logic. The gamble on information technologies continues to prove irresistible -- witness the notorious case of Microsoft, which may have been undone, in a judicial if not an economic sense, by its desire to rule the Internet. And yet the economics of the Web, even in a fundamentally sound business like Amazon (of which more later), seem to make little sense. Amazon may be profitable, but it seems unlikely to succeed at the titanic levels expected. Surely the Web cannot sustain itself as a business proposition, if by such we imply anything beyond pure symbolism.

Having turned to the economics of the sign, however, we may find that our skepticism expands in unsettling directions. Perhaps it is not the Internet or the Web that is untenable, but the entire social order from which these technological boondoggles arise. For the same bizarrely self-contradictory pattern that seems to obtain in textual production and Internet business plans -- a mad expansion that threatens to outrun the capacity of any underlying market -- can be seen at the most general level of social relations. Technologies likle the Web seem to promise a general devolution, at least in discursive or symbolic production. Every computer-equipped, reasonably wired person can be a publisher or "content provider." Let a million startups bloom! And yet even as this vision unfolds, the logic of mergers and acquisitions consolidates ultimate control of communications channels, to say nothing of finance and industrial production, in the hands of a vanishingly few: the Gateses, the Murdochs, the Turners, the equity holders of Time-Warner, Bertelsmann, Wolters-Klüwer. Are we living through an expansion or a huge contraction -- the apotheosis of free markets or their final implosion? How can the answer be both at once?

Surely no society such as this can really exist. By reading the Web as illusion, we may have discovered that it is indeed a creature of fantasy, a very scary monster indeed -- but the creature is no stalker of the Hollywood night, but rather the Red King from Through the Looking-Glass. Which would make us the tenuous people of his dreaming.

5: What use?

Maybe the revery of postindustrial capitalism will not last much longer; sleepers eventually awaken. If the Web does not exist, it may be because it is, like the society that engenders it, an artifact of transition, a blur in the slow-motion film of history, or to try a less anachronistic metaphor, a file that can no longer be found. 404 indeed. Perhaps the world is truly about to end, or change utterly. As McLuhan claimed, one may identify certain social formations or identities with evolutions in technology (1967, 68-69). He argued that "the public," the body of rational individuals inculcated by the Enlightenment, could be traced in large measure to the influences of print, while "the mass," the next social identity to evolve, owed its emergence largely to broadcast media. If the Web and other forms of Internet communication represent nascent forms of something yet to be fully defined, will they usher in a third form of humanity? Who or what succeeds the mass?

At this point we might think about tempering somewhat our sustained skepticism. If space may produce new worlds and cyberspace brave new people, then maybe there is something to look forward to after all. Maybe the Red King can remain in his dream state a good while longer; long enough, at least, to consider a more constructive set of questions. Why should the Web exist? Assuming the manifold contradictions of information culture could be worked through, what would be its likely social effects? Or at the risk of a certain banality: What is the World Wide Web good for?

One canonical answer, perhaps the standard Enlightenment answer, looks to augmentation or prosthesis. The Internet is the distributed human nervous system, Oversoul, nošsphere, or as Don DeLillo once wickedly named it, "Space Brain" (DeLillo, 45). So who are these Web people even now climbing out of their learning pods and Skinner boxes, ready to re-invent the digital economy, or at least start processing all that hypertext? Whoever they are, I suspect they are the millennial equivalent of Fitzgerald's rich -- not the same as you and me. They are, one hopes, smarter.

The dream of augmenting or "bootstrapping" human intelligence can be a good thing, less perhaps in the self-serving rhetoric of Gates or Negroponte than in the visionary thinking of Douglas Engelbart, the great designer who invented many fundamental technologies on which our current dreams are based (Engelbart). Still, there is always a gap between vision and reality. Writing about a particularly bizarre piece of fiction from William Burroughs, the critic Charles Newman called it "an aesthetic experience recommended for a species which has yet to appear on earth" (93). Generalizing from literature to culture more broadly, we might find in Newman's critique an important corrective. Biologists differ about the speed at which speciation occurs, but there can be little doubt that it takes longer than the release cycle of most Web browsers -- a fact that may explain much.

The true children of the Internet may be already among us, or they may be much longer in coming. It is probably impossible to characterize their arrival except in retrospect. In any event, the apocalyptic notion of a technological Great Awakening or Childhood's End seems to have little bearing on the present status, or not-quite-entity, of the World Wide Web.

Hopelessly addicted to the McLuhan Channel, I have always preferred another crackpot explanation for our current predicament. Admittedly, this thinking probably works better as myth or fiction than cultural analysis, but I offer it nonetheless. McLuhan noted that technologies "reverse" as they approach some limit of development or expression (1964, 35). Carried to its extreme in the 20th century, industrial mass communication reversed from the linear and perspectival medium of print into the "cool" immediacy of broadcasting. Scaling up this rationale, we can derive the origin of the Internet and its curious illusion the Web from the reversal of that supreme signifier, the thermonuclear bomb (as Susan Sontag quipped, "cogito ergo boom"). Having produced doomsday weapons, we turned from technologies that reduce discursive potential to one (I win) or zero (everyone loses) to technologies that ramify discourse beyond any dream of control, as Kevin Kelly points out. This was the creative leap of technoculture, the grand postmodern swerve from the path of mutual assured destruction. In the words of Harold Bloom: "Discontinuity is freedom" (39). Or in this case, survival.

Though not intended as legitimate historical analysis, this account at least registers (though it cannot explain) the irrationality of our current condition. To continue in a Bloomian vein: As we fell out of the modern nightmare we swerved, and now we lie in a postmodern hell improved by our own making. Welcome to the Information Age. But before I suggest that the end of the Cold War was engineered by the Trilateral Commission in order to spur commercial development of the Internet, let me confess that the World Wide Web makes no more sense as a bridge across the gulf of Apocalypse than as an agency of bootstrapping or Steigerrung.

Neither scenario helps explain the fundamental problem with which we began: though everywhere in evidence, the existence of the Web cannot be accounted for in any satisfactory way. What after all would the Web exist for?-- and note that the answer to this question must be framed in terms of practice, not theory. Thomas Landauer, system designer and former research chief at Bellcore, puts this question most effectively in his trenchant study of information technology and its business culture (13-14). After noting that the massive computerization of the 1980s and 1990s yielded no appreciable gain in productivity, and that business would on the whole have done better to put its technology dollars into the bond market, Landauer raises a crucial and uncomfortable question. He asks of information technology, what's the use?

This question is significantly hard to address (though Landauer's answers are well worth considering); but even in confronting the question we might begin to reach a better understanding of the deeply dubious Web. If the Web does not yet exist as a fully formed communications regime, or indeed as an economic proposition, perhaps it is because we have not yet understood, recognized, or even formulated the uses to which it should be put. We have yet to understand the parameters of our fallen state. It may be that the World Wide Web requires a thorough re-thinking of what we mean by use. Much of that process is still before us. We may wish, following Kevin Kelly and Sherry Turkle, to think of the Web as an "emergent" phenomenon whose nature will be revealed as it unfolds. Taking this line in a pragmatic direction, we may search with Jay David Bolter for a process of "remediation" in which the ecology of media settles itself into new arrangements, self-motivated, autopoetic, and deeply recursive (Bolter and Grusin, 4-5). Though Bolter recognizes remediation as something of a sideshow trick that demands critical inspection, there is a palpable change here from his earlier line, which began by arguing that "this [hypertext] will destroy that [print]" (Writing Space, 1). Perhaps Bolter is on to something. Is there remediation (if not remedy) for our doubts? Might it be possible to stop worrying and love the Web?

6: Credo

So the Web is all in your head, pure illusion, not so much consensual hallucination as special digital effect. That doesn't mean you can't learn to appreciate the thing, or perhaps even cherish it, especially if you happen to hold a chair in a humanities department somewhere in the wired world. Bolter's new pragmatism -- and it is mine as well, since I come from a working-class college balanced on the ax-edge of budget cuts -- can have strong appeal, considering the career prospects of majors in literature and the fine arts. Could the Web be the great salvation of the humanities at century's end, a decent fallback option for talented people left wanting by feeble academic job markets? If so, the Web could be the most important development for the humanities since interlibrary loan.

But what if one still refuses, perversely and adamantly, to accept this miracle? Some people cannot bring themselves to embrace this sort of unbridled (and perhaps unprincipled) pragmatism. They are not necessarily wrong. Cautious engagement seems the best course, a position that neither dismisses the possibilities for emergence nor takes as read what is not yet written. There is undeniably a danger in the Web mirage -- perhaps especially, to strike closer to home, in the conceit of Web design as a cure-all for a moribund academic culture still yoked to the printing press. Palliatives may conceal symptoms of more serious disease. If we choose to believe pragmatically in the Web, we should remember that it may be a diversion meant to hold the attention of intelligent people while the masters of capital lock down the gates of oligopoly control.

What, after all, do Web designers design? To adapt Landauer's line, what is the use of textual production on the Web? There may of course be valid answers to these questions. Perhaps there are things for weavers of Webs to do besides creating user interfaces for PC banking software, corporate intranets, or banner ads for the latest Hollywood disaster epic. If we can lift our eyes from the print-derived metaphor of Web "pages," if we imagine that working on the Web might mean something other than creating discrete, marketable commodities, then we might be able to open some space for change. Even now the "hacktivists" are among us (Dominguez et al.). Their program may offer firmer ground for action, and thus for belief.

At the same time, there may also be possibilities for improvement within the oligopolized space of late-late capitalism. Consider this story about a funny thing that happened on my way to the virtual cash register. Seeking a course text for my class in Hypermedia Production, I paid a visit to and carried out name searches for several authors whose books I had used with satisfaction in the past. One of these searches turned up a book whose title contained the phrase "Communication Design." Among other positive indications, this phrase closely resembles the name of my academic program, so I was ready to add the book to my virtual shopping cart, with thoughts about making it a course requirement; but these thoughts vanished as I read further down the Web page.

The descriptions of books at Amazon prominently display brief, unsolicited reviews by readers. The very first comment I encountered for the text in question (which in fairness should remain unnamed) advised that this book, billed as a collaboration between a respected senior writer and a relative unknown, was actually written almost entirely by the unknown. The comment concluded with the simple prescription: "Avoid." After looking at the negative review in more detail and remembering problems with similar books in the past, I followed the reviewer's advice. There was no sale.

My tale from the cybermall may say something important about the reforming potential of the World Wide Web. It has been some time since I walked into a bookstore and picked a title from the shelves only to be talked out of the purchase -- although this did happen more than once, in another place and time, when I was dealing not with multinational chains but with a dedicated, independent bookseller. I wonder if this bookseller is still in the business; things are different now. I do not recall ever choosing a book at my local Borders or Barnes and Noble or B. Dalton only to find stuck to the cover a warning to "Avoid." This sort of thing does not and cannot happen in a regime dominated by inventory costs, hyper-competition, and the demand for ever higher profits.

Yet the value of books transcends their commodity status -- a reason we still have lending libraries and (for the moment) first-use rights. Amazon's hallucinatory business model, in which it holds only a nominal inventory and can afford to un-sell the books that line its virtual shelves, represents a very interesting revision of commodity capitalism, albeit in a limited, local instance. Perhaps it merely corrects a perverse mistreatment of books, which were never meant to be sold like hamburgers; and perhaps the Amazon effect will not transfer or, in that most ominous requirement of e-business, "scale." But it does seem possible that Amazon's approach indicates fundamental and eligible changes in the way vendors define their relation to consumers -- changes in which Web designers as well as Internet radicals might find common ground.

If this seems an extravagant suggestion, consider that the notable success of Amazon as a retailer of books, and lately music and videos, may represent only the first stage in the development of a new market for textual goods. Amazon has successfully separated its trade from traditional channels of inventory and distribution, but this transformation can be taken further. Since the value of a book, music CD, or videotape inheres mainly in its content and not in the material substrate, why not eliminate the object altogether? Why print books? Relatively cheap and lightweight display devices now on the market can store hundreds of titles. Price and performance of these "electronic books" seem likely to improve markedly over the next few years. Amazon could easily deliver texts for these devices as bitstreams transmitted through the Internet, as several vendors of electronic books are doing already. For those who still cherish the physical object, local service outlets could return to the ancient practice of booksellers and print and bind on demand. Even physical bookstores might survive this change. Redefined as marketing and browsing places, they might come to resemble lounges and cafes even more than they do now.

Would these differences make a difference in the larger scheme of things? Much depends, of course, on unpredictable social and political articulations. In concept, however, sale by download could allow providers of textual goods to bypass and the large industrial concerns that now control production and distribution. depends entirely on the News Corporation, Time/Warner, Macmillan, and a few other major interests that provide its stock in trade. But this might not always be the case. If publishing no longer meant expensive production and delivery of physical objects, content providers might find new outlets for their work.

There would of course be further complications. The choke point in publishing might shift from production to evaluation and publicity, with capital interests arguing, as they already do in Web publications like Salon and Slate, that they are the only proper arbiters of textual value. The capitalists might then shift their arguments for heavy investment and high profit margins to the demands of taste-making, or advertising. However, these functions depend on tight control of product lines. It is relatively easy to shape the public's desire for movies at the cineplex or paperbacks at the airport, where consumer options are limited to a handful of products only briefly available. Would the same reasoning apply to a market where the shelves or marquee are replaced by a hypertextual catalog, and where no title ever goes out of print?

While this logic has yet to penetrate the relatively backward book trade, there has been movement in this direction in the popular music market, spurred by the advent of MPEG-3 recorders and the rapid growth of music download sites on the Web. Television programmers may be waking up as well. Thomas Rosenstiel, director of the Project for Excellence in Journalism, recently said: "Our mass media depends [sic] on an audience that no longer exists -- a mass audience which is now fragmented" (Barringer, C1). Might we reach a point at which the monolithic mass market, for some commodities at least, becomes as chimerical as the Web seems today?

Probably not, if the current owners of the media have any say in the matter. It is worth noting that even as MPEG-3 and e-books make their appearance, the U.S. Patent Office has begun to award alarmingly broad protections for basic business practices -- a development which apparently spurred the Microsoft Corporation to apply for a patent on sale of electronic magazines by subscription over the World Wide Web. That such an application would even be considered seems instructive. Oligopoly capital continues to call the tune, in this country at least, and will continue to do so as long as political campaigns are paid for by corporate subvention. Any major shift in markets is bound to arouse opposition.

As usual, those who would enter this contest on the side of change must subsist largely on illusions -- radical economic models, faith in individual enterprise, and anachronistic notions of a public good. To this list of illusions we might now add the World Wide Web and some of the possibilities it may hold for electronic commerce. To be sure, it would be foolish to place in these imaginings anything but the most conditional belief. Like all technologies, the Web and the Internet in themselves make little difference. Visions do not change the world, except as they inform real work. But work without vision leads nowhere.

Works Cited

Barringer, Felicity. "In Washington, Is There News After Scandal?" New York Times 15 February 1999. C1, C6.

Bloom, Harold. The Anxiety of Influence: A Theory of Poetry. 2nd Ed. New York: Oxford UP, 1997 [orig. 1973].

Bolter, Jay David. Writing Space: The Computer, Hypertext, and the History of Writing. Hillsdale, NJ: Lawrence Erlbaum, 1991.

Bolter, Jay David and Richard Grusin. Remediation: Understanding New Media. Cambridge, MA: MIT Press, 1999.

Cadigan, Pat. Synners. New York: Bantam, 1991.

DeLillo, Don. Ratner's Star. New York: Vintage, 1977.

Dominguez, Ricardo, Carmin Karasic, Brett Stalbaum, and Stefan Wray. "Zapatista Floodnet Public Version Release." On-line: World-Wide Web. 27 January, 1999.

Engelbart, Douglas. "Boosting Collective IQ, and an Open Hyperdocument System." On-line document. World Wide Web, 27 January, 1999.

Foucault, Michel. "What Is an Author?" In Josué Harari, ed. Textual Strategies: Perspectives in Post-Structuralist Criticism. Ithaca: Cornell UP, 1979, 141-60.

Gibson, William. Neuromancer. New York: Ace, 1984.

Haraway, Donna J. Modest_Witness@Second_Millennium. FemaleManİ_Meets_OncoMouseŞ: Feminism and Technoscience. New York: Routledge, 1997.

Harpold, Terry. "Threnody: Psychoanalytic Digressions on the Subject of Hypertexts." Hypermedia and Literary Studies. Ed. P. Delany and G. Landow. Cambridge, MA: MIT Press, 1991. 171-84.

Harpold, Terry. "The Contingencies of the Hypertext Link." Writing on the Edge 2(2) [1991]: 126-38.

Joyce, Michael. Afternoon: A Story. Cambridge, MA: Eastgate Systems, 1990.

Kaplan, Nancy, and Stuart Moulthrop. "Where No Mind Has Gone Before: Ontological Design for Virtual Spaces." Proceedings of the ACM Hypertext Conference. Edinburgh, 1994. 206-12.

Landow, George P. "Relationally Encoded Links and the Rhetoric of Hypertext." Hypertext '87 Papers. Chapel Hill: Association for Computing Machinery, 1987. 331-44.

Moulthrop, Stuart. "Hypertext and 'the Hyperreal.'" Hypertext '89 Papers. Pittsburgh: Association for Computing Machinery, 1989. 259-68.

Kelly, Kevin. Out of Control: The Rise of Neo-Biological Civilization. Cambridge, MA: Addison Wesley, 1994.

Kelly, Kevin. New Rules for the New Economy: Ten Radical Strategies for a Connected World. New York: Viking, 1998.

Landauer, Thomas K. The Trouble with Computers: Usefulness, Usability, and Productivity. Cambridge, MA: MIT Press, 1996.

McLuhan, H. Marshall. Understanding Media: The Extensions of Man. Cambridge, MA: MIT Press, 1994 [orig. 1964].

McLuhan, H. Marshall, Quentin Fiore, and Jerome Agel. The Medium is the Massage: An Inventory of Effects. San Francisco: Hardwired, 1996 [orig. 1967].

Newman, Charles. The Post-Modern Aura: The Act of Fiction in an Age of Inflation. Evanston: Northwestern UP, 1985.

Reed, Sandy. "Internet Companies are Rewriting the Accepted Rules of the 'Old' Economy." InfoWorld 2 November 1998. 67.

Sterling, Bruce. Mirrorshades: The Cyberpunk Anthology. New York: Ace, 1988.

Turkle, Sherry. Life on the Screen: Identity in the Age of the Internet. Cambridge, MA: MIT Press, 1997.