Two Bits - The Cultural Significance of Free Software
Christopher M. Kelty (2008)

Part I the internet

1. Geeks and Recursive Publics

Since about 1997, I have been living with geeks online and off. I have been drawn from Boston to Bangalore to Berlin to Houston to Palo Alto, from conferences and workshops to launch parties, pubs, and Internet Relay Chats (IRCs). All along the way in my research questions of commitment and practice, of ideology and imagination have arisen, even as the exact nature of the connections between these people and ideas remained obscure to me: what binds geeks together? As my fieldwork pulled me from a Boston start-up company that worked with radiological images to media labs in Berlin to young entrepreneurial elites in Bangalore, my logistical question eventually developed into an analytical concept: geeks are bound together as a recursive public.

How did I come to understand geeks as a public constituted around the technical and moral ideas of order that allow them to associate with one another? Through this question, one can start to understand the larger narrative of Two Bits: that of Free Software [pg 28] as an exemplary instance of a recursive public and as a set of practices that allow such publics to expand and spread. In this chapter I describe, ethnographically, the diverse, dispersed, and as an exemplary instance of a recursive public and as a set of practices that allow such publics to expand and spread. In this chapter I describe, ethnographically, the diverse, dispersed, and novel forms of entanglements that bind geeks together, and I construct the concept of a recursive public in order to explain these entanglements.

A recursive public is a public that is constituted by a shared concern for maintaining the means of association through which they come together as a public. Geeks find affinity with one another because they share an abiding moral imagination of the technical infrastructure, the Internet, that has allowed them to develop and maintain this affinity in the first place. I elaborate the concept of recursive public (which is not a term used by geeks) in relation to theories of ideology, publics, and public spheres and social imaginaries. I illustrate the concept through ethnographic stories and examples that highlight geeks’ imaginations of the technical and moral order of the Internet. These stories include those of the fate of Amicas, a Boston-based healthcare start-up, between 1997 and 2003, of my participation with new media academics and activists in Berlin in 1999-2001, and of the activities of a group of largely Bangalore-based information technology (IT) professionals on and offline, especially concerning the events surrounding the peer-topeer file sharing application Napster in 2000-2001.

The phrase “moral and technical order” signals both technology—principally software, hardware, networks, and protocols—and an imagination of the proper order of collective political and commercial action, that is, how economy and society should be ordered collectively. Recursive publics are just as concerned with the moral order of markets as they are with that of commons; they are not anticommercial or antigovernment. They exist independent of, and as a check on, constituted forms of power, which include markets and corporations. Unlike other concepts of a public or of a public sphere, “recursive public” captures the fact that geeks’ principal mode of associating and acting is through the medium of the Internet, and it is through this medium that a recursive public can come into being in the first place. The Internet is not itself a public sphere, a public, or a recursive public, but a complex, heterogeneous infrastructure that constitutes and constrains geeks’ everyday practical commitments, their ability to “become public” or to compose a common world. As such, their participation qua recursive publics structures their identity as creative and autonomous [pg 29] individuals. The fact that the geeks described here have been brought together by mailing lists and e-mail, bulletin-board services and Web sites, books and modems, air travel and academia, and cross-talking and cross-posting in ways that were not possible before the Internet is at the core of their own reasoning about why they associate with each other. They are the builders and imaginers of this space, and the space is what allows them to build and imagine it.

Why recursive? I call such publics recursive for two reasons: first, in order to signal that this kind of public includes the activities of making, maintaining, and modifying software and networks, as well as the more conventional discourse that is thereby enabled; and second, in order to suggest the recursive “depth” of the public, the series of technical and legal layers—from applications to protocols to the physical infrastructures of waves and wires—that are the subject of this making, maintaining, and modifying. The first of these characteristics is evident in the fact that geeks use technology as a kind of argument, for a specific kind of order: they argue about technology, but they also argue through it. They express ideas, but they also express infrastructures through which ideas can be expressed (and circulated) in new ways. The second of these characteristics—regarding layers—is reflected in the ability of geeks to immediately see connections between, for example, Napster (a user application) and TCP/IP (a network protocol) and to draw out implications for both of them. By connecting these layers, Napster comes to represent the Internet in miniature. The question of where these layers stop (hardware? laws and regulations? physical constants? etc.) circumscribes the limits of the imagination of technical and moral order shared by geeks.

Above all, “recursive public” is a concept—not a thing. It is intended to make distinctions, allow comparison, highlight salient features, and relate two diverse kinds of things (the Internet and Free Software) in a particular historical context of changing relations of power and knowledge. The stories in this chapter (and throughout the book) give some sense of how geeks interact and what they do technically and legally, but the concept of a recursive public provides a way of explaining why geeks (or people involved in Free Software or its derivatives) associate with one another, as well as a way of testing whether other similar cases of contemporary, technologically mediated affinity are similarly structured. [pg 30]


Recursion (or “recursive”) is a mathematical concept, one which is a standard feature of any education in computer programming. The definition from the Oxford English Dictionary reads: “2. a. Involving or being a repeated procedure such that the required result at each step except the last is given in terms of the result(s) of the next step, until after a finite number of steps a terminus is reached with an outright evaluation of the result.” It should be distinguished from simple iteration or repetition. Recursion is always subject to a limit and is more like a process of repeated deferral, until the last step in the process, at which point all the deferred steps are calculated and the result given.

Recursion is powerful in programming because it allows for the definition of procedures in terms of themselves—something that seems at first counterintuitive. So, for example,

; otherwise return n times factorial of n-1; (defun (factorial n) ; This is the name of the function and its input n. (if (=n 1) ; This is the final limit, or recursive depth 1 ; if n=1, then return 1 (* n (factorial (- n 1))))) ; call the procedure from within itself, and ; calculate the next step of the result before ; giving an answer.1

In Two Bits a recursive public is one whose existence (which consists solely in address through discourse) is only possible through discursive and technical reference to the means of creating this public. Recursiveness is always contingent on a limit which determines the depth of a recursive procedure. So, for instance, a Free Software project may depend on some other kind of software or operating system, which may in turn depend on particular open protocols or a particular process, which in turn depend on certain kinds of hardware that implement them. The “depth” of recursion is determined by the openness necessary for the project itself.

James Boyle has also noted the recursive nature, in particular, of Free Software: “What’s more, and this is a truly fascinating twist, when the production process does need more centralized coordination, some governance that guides how the sticky modular bits are put together, it is at least theoretically possible that we can come up with the control system in exactly the same way. In this sense, distributed production is potentially recursive.”2

1. Abelson and Sussman, The Structure and Interpretation of Computer Programs, 30.

2. Boyle, “The Second Enclosure Movement and the Construction of the Public Domain,” 46. [pg 31]

From the Facts of Human Activity

Boston, May 2003. Starbucks. Sean and Adrian are on their way to pick me up for dinner. I’ve already had too much coffee, so I sit at the window reading the paper. Eventually Adrian calls to find out where I am, I tell him, and he promises to show up in fifteen minutes. I get bored and go outside to wait, watch the traffic go by. More or less right on time (only post-dotcom is Adrian ever on time), Sean’s new blue VW Beetle rolls into view. Adrian jumps out of the passenger seat and into the back, and I get in. Sean has been driving for a little over a year. He seems confident, cautious, but meanders through the streets of Cambridge. We are destined for Winchester, a township on the Charles River, in order to go to an Indian restaurant that one of Sean’s friends has recommended. When I ask how they are doing, they say, “Good, good.” Adrian offers, “Well, Sean’s better than he has been in two years.” “Really?” I say, impressed.

Sean says, “Well, happier than at least the last year. I, well, let me put it this way: forgive me father for I have sinned, I still have unclean thoughts about some of the upper management in the company, I occasionally think they are not doing things in the best interest of the company, and I see them as self-serving and sometimes wish them ill.” In this rolling blue confessional Sean describes some of the people who I am familiar with whom he now tries very hard not to think about. I look at him and say, “Ten Hail Marys and ten Our Fathers, and you will be absolved, my child.” Turning to Adrian, I ask, “And what about you?” Adrian continues the joke: “I, too, have sinned. I have reached the point where I can see absolutely nothing good coming of this company but that I can keep my investments in it long enough to pay for my children’s college tuition.” I say, “You, my son, I cannot help.” Sean says, “Well, funny thing about tainted money . . . there just taint enough of it.”

I am awestruck. When I met Sean and Adrian, in 1997, their start-up company, Amicas, was full of spit, with five employees working out of Adrian’s living room and big plans to revolutionize the medical-imaging world. They had connived to get Massachusetts General Hospital to install their rudimentary system and let it compete with the big corporate sloths that normally stalked back offices: General Electric, Agfa, Siemens. It was these behemoths, according to Sean and Adrian, that were bilking hospitals [pg 32] and healthcare providers with promises of cure-all technologies and horribly designed “silos,” “legacy systems,” and other closed-system monsters of corporate IT harkening back to the days of IBM mainframes. These beasts obviously did not belong to the gleaming future of Internet-enabled scalability. By June of 2000, Amicas had hired new “professional” management, moved to Watertown, and grown to about a hundred employees. They had achieved their goal of creating an alternative Picture Archiving and Communication System (PACS) for use in hospital radiology departments and based on Internet standards.

At that point, in the spring of 2000, Sean could still cheerfully introduce me to his new boss—the same man he would come to hate, inasmuch as Sean hates anyone. But by 2002 he was frustrated by the extraordinary variety of corner-cutting and, more particularly, by the complacency with which management ignored his recommendations and released software that was almost certainly going to fail later, if not sooner. Sean, who is sort of permanently callow about things corporate, could find no other explanation than that the new management was evil.

But by 2003 the company had succeeded, having grown to more than 200 employees and established steady revenue and a stable presence throughout the healthcare world. Both Sean and Adrian were made rich—not wildly rich, but rich enough—by its success. In the process, however, it also morphed into exactly what Sean and Adrian had created it in order to fight: a slothlike corporate purveyor of promises and broken software. Promises Adrian had made and software Sean had built. The failure of Amicas to transform healthcare was a failure too complex and technical for most of America to understand, but it rested atop the success of Amicas in terms more readily comprehensible: a growing company making profit. Adrian and Sean had started the company not to make money, but in order to fix a broken healthcare system; yet the system stayed broken while they made money.

In the rolling confessional, Sean and Adrian did in fact see me, however jokingly, as a kind of redeemer, a priest (albeit of an order with no flock) whose judgment of the affairs past was essential to their narration of their venture as a success, a failure, or as an unsatisfying and complicated mixture of both. I thought about this strange moment of confession, of the combination of recognition and denial, of Adrian’s new objectification of the company as an [pg 33] investment opportunity, and of Sean’s continuing struggle to make his life and his work harmonize in order to produce good in the world. Only the promise of the next project, the next mission (and the ostensible reason for our dinner meeting) could possibly have mitigated the emotional disaster that their enterprise might otherwise be. Sean’s and Adrian’s endless, arcane fervor for the promise of new technologies did not cease, even given the quotidian calamities these technologies leave in their wake. Their faith was strong, and continuously tested.

Adrian’s and Sean’s passion was not for money—though money was a powerful drug—it was for the Internet: for the ways in which the Internet could replace the existing infrastructure of hospitals and healthcare providers, deliver on old promises of telemedicine and teleradiology, and, above all, level a playing field systematically distorted and angled by corporate and government institutions that sought secrecy and private control, and stymied progress. In healthcare, as Adrian repeatedly explained to me, this skewed playing field was not only unfair but malicious and irresponsible. It was costing lives. It slowed the creation and deployment of technologies and solutions that could lower costs and thus provide more healthcare for more people. The Internet was not part of the problem; it was part of the solution to the problems that ailed 1990s healthcare.

At the end of our car trip, at the Indian restaurant in Winchester, I learned about their next scheme, a project called MedCommons, which would build on the ideals of Free Software and give individuals a way to securely control and manage their own healthcare data. The rhetoric of commons and the promise of the Internet as an infrastructure dominated our conversation, but the realities of funding and the question of whether MedCommons could be pursued without starting another company remained unsettled. I tried to imagine what form a future confession might take.

Geeks and Their Internets

Sean and Adrian are geeks. They are entrepreneurs and idealists in different ways, a sometimes paradoxical combination. They are certainly [pg 34] obsessed with technology, but especially with the Internet, and they clearly distinguish themselves from others who are obsessed with technology of just any sort. They aren’t quite representative—they do not stand in for all geeks—but the way they think about the Internet and its possibilities might be. Among the rich story of their successes and failures, one might glimpse the outlines of a question: where do their sympathies lie? Who are they with? Who do they recognize as being like them? What might draw them together with other geeks if not a corporation, a nation, a language, or a cause? What binds these two geeks to any others?

Sean worked for the Federal Reserve in the 1980s, where he was introduced to UNIX, C programming, EMACS, Usenet, Free Software, and the Free Software Foundation. But he was not a Free Software hacker; indeed, he resisted my attempts to call him a hacker at all. Nevertheless, he started a series of projects and companies with Adrian that drew on the repertoire of practices and ideas familiar from Free Software, including their MedCommons project, which was based more or less explicitly in the ideals of Free Software. Adrian has a degree in medicine and in engineering, and is a serial entrepreneur, with Amicas being his biggest success—and throughout the last ten years has attended all manner of conferences and meetings devoted to Free Software, Open Source, open standards, and so on, almost always as the lone representative from healthcare. Both graduated from the MIT (Sean in economics, Adrian in engineering), one of the more heated cauldrons of the Internet and the storied home of hackerdom, but neither were MIT hackers, nor even computer-science majors.

Their goals in creating a start-up rested on their understanding of the Internet as an infrastructure: as a standardized infrastructure with certain extremely powerful properties, not the least of which was its flexibility. Sean and Adrian talked endlessly about open systems, open standards, and the need for the Internet to remain open and standardized. Adrian spoke in general terms about how it would revolutionize healthcare; Sean spoke in specific terms about how it structured the way Amicas’s software was being designed and written. Both participated in standards committees and in the online and offline discussions that are tantamount to policymaking in the Internet world. The company they created was a “virtual” company, that is, built on tools that depended on the Internet and allowed employees to manage and work from a variety of locations, though not without frustration, of course: Sean waited years for broadband access in his home, and the hospitals they served [pg 35] hemmed themselves in with virtual private networks, intranets, and security firewalls that betrayed the promises of openness that Sean and Adrian heralded.

The Internet was not the object of their work and lives, but it did represent in detail a kind of moral or social order embodied in a technical system and available to everyone to use as a platform whereby they might compete to improve and innovate in any realm. To be sure, although not all Internet entrepreneurs of the 1990s saw the Internet in the same way, Sean and Adrian were hardly alone in their vision. Something about the particular way in which they understood the Internet as representing a moral order—simultaneously a network, a market, a public, and a technology—was shared by a large group of people, those who I now refer to simply as geeks.

The term geek is meant to be inclusive and to index the problematic of a recursive public. Other terms may be equally useful, but perhaps semantically overdetermined, most notably hacker, which regardless of its definitional range, tends to connote someone subversive and/or criminal and to exclude geek-sympathetic entrepreneurs and lawyers and activists. 23 Geek is meant to signal, like the public in “recursive public,” that geeks stand outside power, at least in some aspects, and that they are not capitalists or technocrats, even if they start businesses or work in government or industry. 24 Geek is meant to signal a mode of thinking and working, not an identity; it is a mode or quality that allows people to find each other, for reasons other than the fact that they share an office, a degree, a language, or a nation.

Until the mid-1990s, hacker, geek, and computer nerd designated a very specific type: programmers and lurkers on relatively underground networks, usually college students, computer scientists, and “amateurs” or “hobbyists.” A classic mock self-diagnostic called the Geek Code, by Robert Hayden, accurately and humorously detailed the various ways in which one could be a geek in 1996—UNIX/ Linux skills, love/hate of Star Trek, particular eating and clothing habits—but as Hayden himself points out, the geeks of the early 1990s exist no longer. The elite subcultural, relatively homogenous group it once was has been overrun: “The Internet of 1996 was still a wild untamed virgin paradise of geeks and eggheads unpopulated by script kiddies, and the denizens of AOL. When things changed, I seriously lost my way. I mean, all the ‘geek’ that was the Internet [pg 36] was gone and replaced by Xfiles buzzwords and politicians passing laws about a technology they refused to comprehend.” 25

For the purists like Hayden, geeks were there first, and they understood something, lived in a way, that simply cannot be comprehended by “script kiddies” (i.e., teenagers who perform the hacking equivalent of spray painting or cow tipping), crackers, or AOL users, all of whom are despised by Hayden-style geeks as unskilled users who parade around the Internet as if they own it. While certainly elitist, Hayden captures the distinction between those who are legitimately allowed to call themselves geeks (or hackers) and those who aren’t, a distinction that is often formulated recursively, of course: “You are a hacker when another hacker calls you a hacker.”

However, since the explosive growth of the Internet, geek has become more common a designation, and my use of the term thus suggests a role that is larger than programmer/hacker, but not as large as “all Internet users.” Despite Hayden’s frustration, geeks are still bound together as an elite and can be easily distinguished from “AOL users.” Some of the people I discuss would not call themselves geeks, and some would. Not all are engineers or programmers: I have met businessmen, lawyers, activists, bloggers, gastroenterologists, anthropologists, lesbians, schizophrenics, scientists, poets, people suffering from malaria, sea captains, drug dealers, and people who keep lemurs, many of whom refer to themselves as geeks, some of the time. 26 There are also lawyers, politicians, sociologists, and economists who may not refer to themselves as geeks, but who care about the Internet just as other geeks do. By contrast “users” of the Internet, even those who use it eighteen out of twenty-four hours in a day to ship goods and play games, are not necessarily geeks by this characterization.

Operating Systems and Social Systems

Berlin, November 1999. I am in a very hip club in Mitte called WMF. It’s about eight o’clock—five hours too early for me to be a hipster, but the context is extremely cool. WMF is in a hard-to-find, abandoned building in the former East; it is partially converted, filled with a mixture of new and old furnishings, video projectors, speakers, makeshift bars, and dance-floor lighting. A crowd of around fifty people lingers amid smoke and Beck’s beer bottles, [pg 37] sitting on stools and chairs and sofas and the floor. We are listening to an academic read a paper about Claude Shannon, the MIT engineer credited with the creation of information theory. The author is smoking and reading in German while the audience politely listens. He speaks for about seventy minutes. There are questions and some perfunctory discussion. As the crowd breaks up, I find myself, in halting German that quickly converts to English, having a series of animated conversations about the GNU General Public License, the Debian Linux Distribution, open standards in net radio, and a variety of things for which Claude Shannon is the perfect ghostly technopaterfamilias, even if his seventy-minute invocation has clashed heavily with the surroundings.

Despite my lame German, I still manage to jump deeply into issues that seem extremely familiar: Internet standards and open systems and licensing issues and namespaces and patent law and so on. These are not businesspeople, this is not a start-up company. As I would eventually learn, there was even a certain disdain for die Krawattenfaktor, the suit-and-tie factor, at these occasional, hybrid events hosted by Mikro e.V., a nonprofit collective of journalists, academics, activists, artists, and others interested in new media, the Internet, and related issues. Mikro’s constituency included people from Germany, Holland, Austria, and points eastward. They took some pride in describing Berlin as “the farthest East the West gets” and arranged for a group photo in which, facing West, they stood behind the statue of Marx and Lenin, who face East and look eternally at the iconic East German radio tower (Funkturm) in Alexanderplatz. Mikro’s members are resolutely activist and see the issues around the Internet-as-infrastructure not in terms of its potential for business opportunities, but in urgently political and unrepentantly aesthetic terms—terms that are nonetheless similar to those of Sean and Adrian, from whom I learned the language that allows me to mingle with the Mikro crowd at WMF. I am now a geek.

Before long, I am talking with Volker Grassmuck, founding member of Mikro and organizer of the successful “Wizards of OS” conference, held earlier in the year, which had the very intriguing subtitle “Operating Systems and Social Systems.” Grassmuck is inviting me to participate in a planning session for the next WOS, held at the Chaos Computer Congress, a hacker gathering that occurs each year in December in Berlin. In the following months I will meet a huge number of people who seem, uncharacteristically for artists [pg 38] and activists, strangely obsessed with configuring their Linux distributions or hacking the http protocol or attending German Parliament hearings on copyright reform. The political lives of these folks have indeed mixed up operating systems and social systems in ways that are more than metaphorical.

The Idea of Order at the Keyboard

If intuition can lead one from geek to geek, from start-up to nightclub, and across countries, languages, and professional orientations, it can only be due to a shared set of ideas of how things fit together in the world. These ideas might be “cultural” in the traditional sense of finding expression among a community of people who share backgrounds, homes, nations, languages, idioms, ethnos, norms, or other designators of belonging and co-presence. But because the Internet—like colonialism, satellite broadcasting, and air travel, among other things—crosses all these lines with abandon that the shared idea of order is better understood as part of a public, or public sphere, a vast republic of letters and media and ideas circulating in and through our thoughts and papers and letters and conversations, at a planetary scope and scale.

“Public sphere” is an odd kind of thing, however. It is at once a concept—intended to make sense of a space that is not the here and now, but one made up of writings, ideas, and discussions—and a set of ideas that people have about themselves and their own participation in such a space. I must be able to imagine myself speaking and being spoken to in such a space and to imagine a great number of other people also doing so according to unwritten rules we share. I don’t need a complete theory, and I don’t need to call it a public sphere, but I must somehow share an idea of order with all those other people who also imagine themselves participating in and subjecting themselves to that order. In fact, if the public sphere exists as more than just a theory, then it has no other basis than just such a shared imagination of order, an imagination which provides a guide against which to make judgments and a map for changing or achieving that order. Without such a shared imagination, a public sphere is otherwise nothing more than a cacophony of voices and information, nothing more than a stream of data, structured and formatted by and for machines, whether paper or electronic. [pg 39]

Charles Taylor, building on the work of Jürgen Habermas and Michael Warner, suggests that the public sphere (both idea and thing) that emerged in the eighteenth century was created through practices of communication and association that reflected a moral order in which the public stands outside power and guides or checks its operation through shared discourse and enlightened discussion. Contrary to the experience of bodies coming together into a common space (Taylor calls them “topical spaces,” such as conversation, ritual, assembly), the crucial component is that the public sphere “transcends such topical spaces. We might say that it knits a plurality of spaces into one larger space of non-assembly. The same public discussion is deemed to pass through our debate today, and someone else’s earnest conversation tomorrow, and the newspaper interview Thursday and so on. . . . The public sphere that emerges in the eighteenth century is a meta-topical common space.” 27

Because of this, Taylor refers to his version of a public as a “social imaginary,” a way of capturing a phenomena that wavers between having concrete existence “out there” and imagined rational existence “in here.” There are a handful of other such imagined spaces—the economy, the self-governing people, civil society—and in Taylor’s philosophical history they are related to each through the “ideas of moral and social order” that have developed in the West and around the world. 28

Taylor’s social imaginary is intended to do something specific: to resist the “spectre of idealism,” the distinction between ideas and practices, between “ideologies” and the so-called material world as “rival causal agents.” Taylor suggests, “Because human practices are the kind of thing that makes sense, certain ideas are internal to them; one cannot distinguish the two in order to ask the question Which causes which?” 29 Even if materialist explanations of cause are satisfying, as they often are, Taylor suggests that they are so “at the cost of being implausible as a universal principle,” and he offers instead an analysis of the rise of the modern imaginaries of moral order. 30

The concept of recursive public, like that of Taylor’s public sphere, is understood here as a kind of social imaginary. The primary reason is to bypass the dichotomy between ideas and material practice. Because the creation of software, networks, and legal documents are precisely the kinds of activities that trouble this distinction—they are at once ideas and things that have material effects in the [pg 40] world, both expressive and performative—it is extremely difficult to identify the properly material materiality (source code? computer chips? semiconductor manufacturing plants?). This is the first of the reasons why a recursive public is to be distinguished from the classic formulae of the public sphere, that is, that it requires a kind of imagination that includes the writing and publishing and speaking and arguing we are familiar with, as well as the making of new kinds of software infrastructures for the circulation, archiving, movement, and modifiability of our enunciations.

The concept of a social imaginary also avoids the conundrums created by the concept of “ideology” and its distinction from material practice. Ideology in its technical usage has been slowly and surely overwhelmed by its pejorative meaning: “The ideological is never one’s own position; it is always the stance of someone else, always their ideology.” 31 If one were to attempt an explanation of any particular ideology in nonpejorative terms, there is seemingly nothing that might rescue the explanation from itself becoming ideological.

The problem is an old one. Clifford Geertz noted it in “Ideology as a Cultural System,” as did Karl Mannheim before him in Ideology and Utopia: it is the difficulty of employing a non-evaluative concept of ideology. 32 Of all the versions of struggle over the concept of a scientific or objective sociology, it is the claim of exploring ideology objectively that most rankles. As Geertz put it, “Men do not care to have beliefs to which they attach great moral significance examined dispassionately, no matter for how pure a purpose; and if they are themselves highly ideologized, they may find it simply impossible to believe that a disinterested approach to critical matters of social and political conviction can be other than a scholastic sham.” 33

Mannheim offered one response: a version of epistemological relativism in which the analysis of ideology included the ideological position of the analyst. Geertz offered another: a science of “symbolic action” based in Kenneth Burke’s work and drawing on a host of philosophers and literary critics. 34 Neither the concept of ideology, nor the methods of cultural anthropology have been the same since. “Ideology” has become one of the most widely deployed (some might say, most diffuse) tools of critique, where critique is understood as the analysis of cultural patterns given in language and symbolic structures, for the purposes of bringing [pg 41] to light systems of hegemony, domination, authority, resistance, and/or misrecognition. 35 However, the practices of critique are just as (if not more) likely to be turned on critical scholars themselves, to show how the processes of analysis, hidden assumptions, latent functions of the university, or other unrecognized features the material, non-ideological real world cause the analyst to fall into an ideological trap.

The concept of ideology takes a turn toward “social imaginary” in Paul Ricoeur’s Lectures on Ideology and Utopia, where he proposes ideological and utopian thought as two components of “social and cultural imagination.” Ricoeur’s overview divides approaches to the concept of ideology into three basic types—the distorting, the integrating, and the legitimating—according to how actors deal with reality through (symbolic) imagination. Does the imagination distort reality, integrate it, or legitimate it vis-à-vis the state? Ricoeur defends the second, Geertzian flavor: ideologies integrate the symbolic structure of the world into a meaningful whole, and “only because the structure of social life is already symbolic can it be distorted.” 36

For Ricoeur, the very substance of life begins in the interpretation of reality, and therefore ideologies (as well as utopias—and perhaps conspiracies) could well be treated as systems that integrate those interpretations into the meaningful wholes of political life. Ricoeur’s analysis of the integration of reality though social imagination, however, does not explicitly address how imagination functions: what exactly is the nature of this symbolic action or interpretation, or imagination? Can one know it from the outside, and does it resist the distinction between ideology and material practice? Both Ricoeur and Geertz harbor hope that ideology can be made scientific, that the integration of reality through symbolic action requires only the development of concepts adequate to the job.

Re-enter Charles Taylor. In Modern Social Imaginaries the concept of social imaginary is distinctive in that it attempts to capture the specific integrative imaginations of modern moral and social order. Taylor stresses that they are imaginations—not necessarily theories—of modern moral and social order: “By social imaginary, I mean something much broader and deeper than the intellectual schemes people may entertain when they think about social reality in a disengaged mode. I am thinking, rather, of the ways in [pg 42] which people imagine their social existence, how they fit together with others, how things go on between them and their fellows, the expectations that are normally met, and the deeper normative notions and images that underlie these expectations.” 37 Social imaginaries develop historically and result in both new institutions and new subjectivities; the concepts of public, market, and civil society (among others) are located in the imaginative faculties of actors who recognize the shared, common existence of these ideas, even if they differ on the details, and the practices of those actors reflect a commitment to working out these shared concepts.

Social imaginaries are an extension of “background” in the philosophical sense: “a wider grasp of our whole predicament.” 38 The example Taylor uses is that of marching in a demonstration: the action is in our imaginative repertory and has a meaning that cannot be reduced to the local context: “We know how to assemble, pick up banners and march. . . . [W]e understand the ritual. . . . [T]he immediate sense of what we are doing, getting the message to our government and our fellow citizens that the cuts must stop, say, makes sense in a wider context, in which we see ourselves standing in a continuing relation with others, in which it is appropriate to address them in this manner.” 39 But we also stand “internationally” and “in history” against a background of stories, images, legends, symbols, and theories. “The background that makes sense of any given act is wide and deep. It doesn’t include everything in our world, but the relevant sense-giving features can’t be circumscribed. . . . [It] draws on our whole world, that is, our sense of our whole predicament in time and space, among others and in history.” 40

The social imaginary is not simply the norms that structure our actions; it is also a sense of what makes norms achievable or “realizable,” as Taylor says. This is the idea of a “moral order,” one that we expect to exist, and if it doesn’t, one that provides a plan for achieving it. For Taylor, there is such a thing as a “modern idea of order,” which includes, among other things, ideas of what it means to be an individual, ideas of how individual passions and desires are related to collective association, and, most important, ideas about living in time together (he stresses a radically secular conception of time—secular in a sense that means more than simply “outside religion”). He by no means insists that this is the only such definition of modernity (the door is wide open to understanding alternative modernities), but that the modern idea of moral order is [pg 43] one that dominates and structures a very wide array of institutions and individuals around the world.

The “modern idea of moral order” is a good place to return to the question of geeks and their recursive publics. Are the ideas of order shared by geeks different from those Taylor outlines? Do geeks like Sean and Adrian, or activists in Berlin, possess a distinctive social imaginary? Or do they (despite their planetary dispersal) participate in this common modern idea of moral order? Do the stories and narratives, the tools and technologies, the theories and imaginations they follow and build on have something distinctive about them? Sean’s and Adrian’s commitment to transforming healthcare seems to be, for instance, motivated by a notion of moral order in which the means of allocation of healthcare might become more just, but it is also shot through with technical ideas about the role of standards, the Internet, and the problems with current technical solutions; so while they may seem to be simply advocating for better healthcare, they do so through a technical language and practice that are probably quite alien to policymakers, upper management, and healthcare advocacy groups that might otherwise be in complete sympathy.

The affinity of geeks for each other is processed through and by ideas of order that are both moral and technical—ideas of order that do indeed mix up “operating systems and social systems.” These systems include the technical means (the infrastructure) through which geeks meet, assemble, collaborate, and plan, as well as how they talk and think about those activities. The infrastructure—the Internet—allows for a remarkably wide and diverse array of people to encounter and engage with each other. That is to say, the idea of order shared by geeks is shared because they are geeks, because they “get it,” because the Internet’s structure and software have taken a particular form through which geeks come to understand the moral order that gives the fabric of their political lives warp and weft.

Internet Silk Road

Bangalore, March 2000. I am at another bar, this time on one of Bangalore’s trendiest streets. The bar is called Purple Haze, and I have been taken there, the day after my arrival, by Udhay Shankar [pg 44] N. Inside it is dark and smoky, purple, filled with men between eighteen and thirty, and decorated with posters of Jimi Hendrix, Black Sabbath, Jim Morrison (Udhay: “I hate that band”), Led Zeppelin, and a somewhat out of place Frank Zappa (Udhay: “One of my political and musical heroes”). All of the men, it appears, are singing along with the music, which is almost without exception heavy metal.

I engage in some stilted conversation with Udhay and his cousin Kirti about the difference between Karnatic music and rock-androll, which seems to boil down to the following: Karnatic music decreases metabolism and heart rate, leading to a relaxed state of mind; rock music does the opposite. Given my aim of focusing on the Internet and questions of openness, I have already decided not to pay attention to this talk of music. In retrospect, I understand this to have been a grave methodological error: I underestimated the extent to which the subject of music has been one of the primary routes into precisely the questions about the “reorientation of knowledge and power” I was interested in. Over the course of the evening and the following days, Udhay introduced me, as promised, to a range of people he either knew or worked with in some capacity. Almost all of the people I met appeared to sincerely love heavy-metal music.

I met Udhay Shankar N. in 1999 through a newsletter, distributed via e-mail, called Tasty Bits from the Technology Front. It was one of a handful of sources I watched closely while in Berlin, looking for such connections to geek culture. The newsletter described a start-up company in Bangalore, one that was devoted to creating a gateway between the Internet and mobile phones, and which was, according to the newsletter, an entirely Indian operation, though presumably with U.S. venture funds. I wanted to find a company to compare to Amicas: a start-up, run by geeks, with a similar approach to the Internet, but halfway around the world and in a “culture” that might be presumed to occupy a very different kind of moral order. Udhay invited me to visit and promised to introduce me to everyone he knew. He described himself as a “random networker”; he was not really a programmer or a designer or a Free Software geek, despite his extensive knowledge of software, devices, operating systems, and so on, including Free and Open Source Software. Neither was he a businessman, but rather described himself as the guy who “translates between the suits and the techs.” [pg 45]

Udhay “collects interesting people,” and it was primarily through his zest for collecting that I met all the people I did. I met cosmopolitan activists and elite lawyers and venture capitalists and engineers and cousins and brothers and sisters of engineers. I met advertising executives and airline flight attendants and consultants in Bombay. I met journalists and gastroenterologists, computer-science professors and musicians, and one mother of a robot scientist in Bangalore. Among them were Muslims, Hindus, Jains, Jews, Parsis, and Christians, but most of them considered themselves more secular and scientific than religious. Many were self-educated, or like their U.S. counterparts, had dropped out of university at some point, but continued to teach themselves about computers and networks. Some were graduates or employees of the Indian Institute of Science in Bangalore, an institution that was among the most important for Indian geeks (as Stanford University is to Silicon Valley, many would say). Among the geeks to whom Udhay introduced me, there were only two commonalities: the geeks were, for the most part, male, and they all loved heavy-metal music. 41

While I was in Bangalore, I was invited to join a mailing list run by Udhay called Silk-list, an irregular, unmoderated list devoted to “intelligent conversation.” The list has no particular focus: long, meandering conversations about Indian politics, religion, economics, and history erupt regularly; topics range from food to science fiction to movie reviews to discussions on Kashmir, Harry Potter, the singularity, or nanotechnology. Udhay started Silk-list in 1997 with Bharath Chari and Ram Sundaram, and the recipients have included hundreds of people around the world, some very well-known ones, programmers, lawyers, a Bombay advertising executive, science-fiction authors, entrepreneurs, one member of the start-up Amicas, at least two transhumanists, one (diagnosed) schizophrenic, and myself. Active participants usually numbered about ten to fifteen, while many more lurked in the background.

Silk-list is an excellent index of the relationship between the network of people in Bangalore and their connection to a worldwide community on the Internet—a fascinating story of the power of heterogeneously connected networks and media. Udhay explained that in the early 1990s he first participated in and then taught himself to configure and run a modem-based networking system known as a Bulletin Board Service (BBS) in Bangalore. In 1994 he heard about a book by Howard Rheingold called The Virtual [pg 46] Community, which was his first introduction to the Internet. A couple of years later when he finally had access to the Internet, he immediately e-mailed John Perry Barlow, whose work he knew from Wired magazine, to ask for Rheingold’s e-mail address in order to connect with him. Rheingold and Barlow exist, in some ways, at the center of a certain kind of geek world: Rheingold’s books are widely read popular accounts of the social and community aspects of new technologies that have often had considerable impact internationally; Barlow helped found the Electronic Frontier Foundation and is responsible for popularizing the phrase “information wants to be free.” 42 Both men had a profound influence on Udhay and ultimately provided him with the ideas central to running an online community. A series of other connections of similar sorts—some personal, some precipitated out of other media and other channels, some entirely random—are what make up the membership of Silk-list. 43

Like many similar communities of “digerati” during and after the dot.com boom, Silk-list constituted itself more or less organically around people who “got it,” that is, people who claimed to understand the Internet, its transformative potential, and who had the technical skills to participate in its expansion. Silk-list was not the only list of its kind. Others such as the Tasty Bits newsletter, the FoRK (Friends of Rohit Khare) mailing list (both based in Boston), and the Nettime and Syndicate mailing lists (both based in the Netherlands) ostensibly had different reasons for existence, but many had the same subscribers and overlapping communities of geeks. Subscription was open to anyone, and occasionally someone would stumble on the list and join in, but most were either invited by members or friends of friends, or they were connected by virtue of cross-posting from any number of other mailing lists to which members were subscribed.


Silk-list is public in many senses of the word. Practically speaking, one need not be invited to join, and the material that passes through the list is publicly archived and can be found easily on the Internet. Udhay does his best to encourage everyone to speak and to participate, and to discourage forms of discourse that he thinks [pg 47] might silence participants into lurking. Silk-list is not a government, corporate, or nongovernmental list, but is constituted only through the activity of geeks finding each other and speaking to each other on this list (which can happen in all manner of ways: through work, through school, through conferences, through fame, through random association, etc.). Recall Charles Taylor’s distinction between a topical and a metatopical space. Silk-list is not a conventionally topical space: at no point do all of its members meet face-to-face (though there are regular meet-ups in cities around the world), and they are not all online at the same time (though the volume and tempo of messages often reflect who is online “speaking” to each other at any given moment). It is a topical space, however, if one considers it from the perspective of the machine: the list of names on the mailing list are all assembled together in a database, or in a file, on the server that manages the mailing list. It is a stretch to call this an “assembly,” however, because it assembles only the avatars of the mailing-list readers, many of whom probably ignore or delete most of the messages.

Silk-list is certainly, on the other hand, a “metatopical” public. It “knits together” a variety of topical spaces: my discussion with friends in Houston, and other members’ discussions with people around the world, as well as the sources of multiple discussions like newspaper and magazine articles, films, events, and so on that are reported and discussed online. But Silk-list is not “The” public—it is far from being the only forum in which the public sphere is knitted together. Many, many such lists exist.

In Publics and Counterpublics Michael Warner offers a further distinction. “The” public is a social imaginary, one operative in the terms laid out by Taylor: as a kind of vision of order evidenced through stories, images, narratives, and so on that constitute the imagination of what it means to be part of the public, as well as plans necessary for creating the public, if necessary. Warner distinguishes, however, between a concrete, embodied audience, like that at a play, a demonstration, or a riot (a topical public in Taylor’s terms), and an audience brought into being by discourse and its circulation, an audience that is not metatopical so much as it is a public that is concrete in a different way; it is concrete not in the face-to-face temporality of the speech act, but in the sense of calling a public into being through an address that has a different temporality. It is a public that is concrete in a media-specific [pg 48] manner: it depends on the structures of creation, circulation, use, performance, and reuse of particular kinds of discourse, particular objects or instances of discourse.

Warner’s distinction has a number of implications. The first, as Warner is careful to note, is that the existence of particular media is not sufficient for a public to come into existence. Just because a book is printed does not mean that a public exists; it requires also that the public take corresponding action, that is, that they read it. To be part of a particular public is to choose to pay attention to those who choose to address those who choose to pay attention . . . and so on. Or as Warner puts it, “The circularity is essential to the phenomenon. A public might be real and efficacious, but its reality lies in just this reflexivity by which an addressable object is conjured into being in order to enable the very discourse that gives it existence.” 44

This “autotelic” feature of a public is crucial if one is to understand the function of a public as standing outside of power. It simply cannot be organized by the state, by a corporation, or by any other social totality if it is to have the legitimacy of an independently functioning public. As Warner puts it, “A public organizes itself independently of state institutions, law, formal frameworks of citizenship, or preexisting institutions such as the church. If it were not possible to think of the public as organized independently of the state or other frameworks, the public could not be sovereign with respect to the state. . . . Speaking, writing, and thinking involve us—actively and immediately—in a public, and thus in the being of the sovereign.” 45

Warner’s description makes no claim that any public or even The Public actually takes this form in the present: it is a description of a social imaginary or a “faith” that allows individuals to make sense of their actions according to a modern idea of social order. As Warner (and Habermas before him) suggests, the existence of such autonomous publics—and certainly the idea of “public opinion”— does not always conform to this idea of order. Often such publics turn out to have been controlled all along by states, corporations, capitalism, and other forms of social totality that determine the nature of discourse in insidious ways. A public whose participants have no faith that it is autotelic and autonomous is little more than a charade meant to assuage opposition to authority, to transform [pg 49] political power and equality into the negotiation between unequal parties.

Is Silk-list a public? More important, is it a sovereign one? Warner’s distinction between different media-specific forms of assembly is crucial to answering this question. If one wants to know whether a mailing list on the Internet is more or less likely to be a sovereign public than a book-reading public or the nightly-news-hearing one, then one needs to approach it from the specificity of the form of discourse. This specificity not only includes whether the form is text or video and audio, or whether the text is ASCII or Unicode, or the video PAL or NTSC, but it also includes the means of creation, circulation, and reuse of that discourse as well.

The on-demand, Internet-mediated book, by contrast, will have a much different temporality of circulation: it might languish in obscurity due to lack of marketing or reputable authority, or it might get mentioned somewhere like the New York Times and suddenly become a sensation. For such a book, copyright law (in the form of a copyleft license) might allow a much wider range of uses and reuses, but it will restrict certain forms of commercialization of the text. The two publics might therefore end up looking quite different, overlapping, to be sure, but varying in terms of their control [pg 50] and the terms of admittance. What is at stake is the power of one or the other such public to appear as an independent and sovereign entity—free from suspect constraints and control—whose function is to argue with other constituted forms of power.

The conventionally published book may well satisfy all the criteria of being a public, at least in the colloquial sense of making a set of ideas and a discourse widely available and expecting to influence, or receive a response from, constituted forms of sovereign power. However, it is only the latter “on-demand” scheme for publishing that satisfies the criteria of being a recursive public. The differences in this example offer a crude indication of why the Internet is so crucially important to geeks, so important that it draws them together, in its defense, as an infrastructure that enables the creation of publics that are thought to be autonomous, independent, and autotelic. Geeks share an idea of moral and technical order when it comes to the Internet; not only this, but they share a commitment to maintaining that order because it is what allows them to associate as a recursive public in the first place. They discover, or rediscover, through their association, the power and possibility of occupying the position of independent public—one not controlled by states, corporations, or other organizations, but open (they claim) through and through—and develop a desire to defend it from encroachment, destruction, or refeudalization (to use Habermas’s term for the fragmentation of the public sphere).

The recursive public is thus not only the book and the discourse around the book. It is not even “content” expanded to include all kinds of media. It is also the technical structure of the Internet as well: its software, its protocols and standards, its applications and software, its legal status and the licenses and regulations that govern it. This captures both of the reasons why recursive publics are distinctive: (1) they include not only the discourses of a public, but the ability to make, maintain, and manipulate the infrastructures of those discourses as well; and (2) they are “layered” and include both discourses and infrastructures, to a specific technical extent (i.e., not all the way down). The meaning of which layers are important develops more or less immediately from direct engagement with the medium. In the following example, for instance, Napster represents the potential of the Internet in miniature—as an application—but it also connects immediately to concerns about the core protocols that govern the Internet and the process of standardization [pg 51] that governs the development of these protocols: hence recursion through the layers of an infrastructure.

These two aspects of the recursive public also relate to a concern about the fragmentation or refeudalization of the public sphere: there is only one Internet. Its singularity is not technically determined or by any means necessary, but it is what makes the Internet so valuable to geeks. It is a contest, the goal of which is to maintain the Internet as an infrastructure for autonomous and autotelic publics to emerge as part of The Public, understood as part of an imaginary of moral and technical order: operating systems and social systems.

From Napster to the Internet

On 27 July 2000 Eugen Leitl cross-posted to Silk-list a message with the subject line “Prelude to the Singularity.” The message’s original author, Jeff Bone (not at the time a member of Silk-list), had posted the “op-ed piece” initially to the FoRK mailing list as a response to the Recording Industry Association of America’s (RIAA) actions against Napster. The RIAA had just succeeded in getting U.S. district judge Marilyn Hall Patel, Ninth Circuit Court of Appeals, to issue an injunction to Napster to stop downloads of copyrighted music. Bone’s op-ed said,

Popular folklore has it that the Internet was designed with decentralized routing protocols in order to withstand a nuclear attack. That is, the Internet “senses damage” and “routes around it.” It has been said that, on the ’Net, censorship is perceived as damage and is subsequently routed around. The RIAA, in a sense, has cast itself in a censor’s role. Consequently, the music industry will be perceived as damage—and it will be routed around. There is no doubt that this will happen, and that technology will evolve more quickly than businesses and social institutions can; there are numerous highly-visible projects already underway that attempt to create technology that is invulnerable to legal challenges of various kinds. Julian Morrison, the originator of a project (called Fling) to build a fully anonymous/untraceable suite of network protocols, expresses this particularly eloquently. 46

Bone’s message is replete with details that illustrate the meaning and value of the Internet to geeks, and that help clarify the concept [pg 52] of a recursive public. While it is only one message, it nonetheless condenses and expresses a variety of stories, images, folklore, and technical details that I elaborate herein.

The Napster shutdown in 2000 soured music fans and geeks alike, and it didn’t really help the record labels who perpetrated it either. For many geeks, Napster represented the Internet in miniature, an innovation that both demonstrated something on a scope and scale never seen before, and that also connected people around something they cared deeply about—their shared interest in music. Napster raised interesting questions about its own success: Was it successful because it allowed people to develop new musical interests on a scope and scale they had never experienced before? Or was it successful because it gave people with already existing musical interests a way to share music on a scope and scale they had never experienced before? That is to say, was it an innovation in marketing or in distribution? The music industry experienced it as the latter and hence as direct competition with their own means of distribution. Many music fans experienced it as the former, what Cory Doctorow nicely labeled “risk-free grazing,” meaning the ability to try out an almost unimaginable diversity of music before choosing what to invest one’s interests (and money) in. To a large extent, Napster was therefore a recapitulation of what the Internet already meant to geeks.

Bone’s message, the event of the Napster shutdown, and the various responses to it nicely illustrate the two key aspects of the recursive public: first, the way in which geeks argue not only about rights and ideas (e.g., is it legal to share music?) but also about the infrastructures that allow such arguing and sharing; second, the “layers” of a recursive public are evidenced in the immediate connection of Napster (an application familiar to millions) to the “decentralized routing protocols” (TCP/IP, DNS, and others) that made it possible for Napster to work the way it did.

Bone’s message contains four interrelated points. The first concerns the concept of autonomous technical progress. The title “Prelude to the Singularity” refers to a 1993 article by Vernor Vinge about the notion of a “singularity,” a point in time when the speed of autonomous technological development outstrips the human capacity to control it. 47 The notion of singularity has the status of a kind of colloquial “law” similar to Moore’s Law or Metcalfe’s Law, as well as signaling links to a more general literature with roots in [pg 53] libertarian or classically liberal ideas of social order ranging from John Locke and John Stuart Mill to Ayn Rand and David Brin. 48

Bone’s affinity for transhumanist stories of evolutionary theory, economic theory, and rapid innovation sets the stage for the rest of his message. The crucial rhetorical gambit here is the appeal to inevitability (as in the emphatic “there is no doubt that this will happen”): Bone establishes that he is speaking to an audience that is accustomed to hearing about the inevitability of technical progress and the impossibility of legal maneuvering to change it, but his audience may not necessarily agree with these assumptions. Geeks occupy a spectrum from “polymath” to “transhumanist,” a spectrum that includes their understandings of technological progress and its relation to human intervention. Bone’s message clearly lands on the far transhumanist side.

A second point concerns censorship and the locus of power: according to Bone, power does not primarily reside with the government or the church, but comes instead from the private sector, in this case the coalition of corporations represented by the RIAA. The significance of this has to do with the fact that a “public” is expected to be its own sovereign entity, distinct from church, state, or corporation, and while censorship by the church or the state is a familiar form of aggression against publics, censorship by corporations (or consortia representing them), as it strikes Bone and others, is a novel development. Whether the blocking of file-sharing can legitimately be called censorship is also controversial, and many Silk-list respondents found the accusation of censorship untenable.

Proving Bone’s contention, over the course of the subsequent years and court cases, the RIAA and the Motion Picture Association of America (MPAA) have been given considerably more police authority than even many federal agencies—especially with regard to policing networks themselves (an issue which, given its technical abstruseness, has rarely been mentioned in the mainstream mass media). Both organizations have not only sought to prosecute filesharers but have been granted rights to obtain information from Internet Service Providers about customer activities and have consistently sought the right to secretly disable (hack into, disable, or destroy) private computers suspected of illegal activity. Even if these practices may not be defined as censorship per se, they are nonetheless fine examples of the issues that most exercise geeks: the use of legal means by a few (in this case, private corporations) to [pg 54] suppress or transform technologies in wide use by the many. They also index the problems of monopoly, antitrust, and technical control that are not obvious and often find expression, for example, in allegories of reformation and the control of the music-sharing laity by papal authorities.

Third, Bone’s message can itself be understood in terms of the reorientation of knowledge and power. Although what it means to call his message an “op-ed” piece may seem obvious, Bone’s message was not published anywhere in any conventional sense. It doesn’t appear to have been widely cited or linked to. However, for one day at least, it was a heated discussion topic on three mailing lists, including Silk-list. “Publication” in this instance is a different kind of event than getting an op-ed in the New York Times.

The material on Silk-list rests somewhere between private conversation (in a public place, perhaps) and published opinion. No editor made a decision to “publish” the message—Bone just clicked “send.” However, as with any print publication, his piece was theoretically accessible by anyone, and what’s more, a potentially huge number of copies may be archived in many different places (the computers of all the participants, the server that hosts the list, the Yahoo! Groups servers that archive it, Google’s search databases, etc.). Bone’s message exemplifies the recursive nature of the recursive public: it is a public statement about the openness of the Internet, and it is an example of the new forms of publicness it makes possible through its openness.

The constraints on who speaks in a public sphere (such as the power of printers and publishers, the requirements of licensing, or issues of cost and accessibility) are much looser in the Internet era than in any previous one. The Internet gives a previously unknown Jeff Bone the power to dash off a manifesto without so much as a second thought. On the other hand, the ease of distribution belies the difficulty of actually being heard: the multitudes of other Jeff Bones make it much harder to get an audience. In terms of publics, Bone’s message can constitute a public in the same sense that a New York Times op-ed can, but its impact and meaning will be different. His message is openly and freely available for as long as there are geeks and laws and machines that maintain it, but the New York Times piece will have more authority, will be less accessible, and, most important, will not be available to just anyone. Geeks imagine a space where anyone can speak with similar reach and staying [pg 55] power—even if that does not automatically imply authority—and they imagine that it should remain open at all costs. Bone is therefore interested precisely in a technical infrastructure that ensures his right to speak about that infrastructure and offer critique and guidance concerning it.

The ability to create and to maintain such a recursive public, however, raises the fourth and most substantial point that Bone’s message makes clear. The leap to speaking about the “decentralized routing protocols” represents clearly the shared moral and technical order of geeks, derived in this case from the specific details of the Internet. Bone’s post begins with a series of statements that are part of the common repertoire of technical stories and images among geeks. Bone begins by making reference to the “folklore” of the Internet, in which routing protocols are commonly believed to have been created to withstand a nuclear attack. In calling it folklore he suggests that this is not a precise description of the Internet, but an image that captures its design goals. Bone collapses it into a more recent bit of folklore: “The Internet treats censorship as damage and routes around it.” 49 Both bits of folklore are widely circulated and cited; they encapsulate one of the core intellectual ideas about the architecture of the Internet, that is, its open and distributed interconnectivity. There is certainly a specific technical backdrop for this suggestion: the TCP/IP “internetting” protocols were designed to link up multiple networks without making them sacrifice their autonomy and control. However, Bone uses this technical argument more in the manner of a social imaginary than of a theory, that is, as a way of thinking about the technical (and moral) order of the Internet, of what the Internet is supposed to be like.

In the early 1990s this version of the technical order of the Internet was part of a vibrant libertarian dogma asserting that the Internet simply could not be governed by any land-based sovereign and that it was fundamentally a place of liberty and freedom. This was the central message of people such as John Perry Barlow, John Gilmore, Howard Rheingold, Esther Dyson, and a host of others who populated both the pre-1993 Internet (that is, before the World Wide Web became widely available) and the pages of magazines such as Wired and Mondo 2000—the same group of people, incidentally, whose ideas were visible and meaningful to Udhay Shankar and his friends in India even prior to Internet access there, not to mention to Sean and Adrian in Boston, and artists and activists in [pg 56] Europe, all of whom often reacted more strongly against this libertarian aesthetic.

For Jeff Bone (and a great many geeks), the folkloric notion that “the net treats censorship as damage” is a very powerful one: it suggests that censorship is impossible because there is no central point of control. A related and oft-cited sentiment is that “trying to take something off of the Internet is like trying to take pee out of a pool.” This is perceived by geeks as a virtue, not a drawback, of the Internet.

For Jeff Bone (and a great many geeks), the folkloric notion that “the net treats censorship as damage” is a very powerful one: it suggests that censorship is impossible because there is no central point of control. A related and oft-cited sentiment is that “trying to take something off of the Internet is like trying to take pee out of a pool.” This is perceived by geeks as a virtue, not a drawback, of the Internet.

On the other side of the spectrum, however, this view of the unregulatable nature of the Internet has been roundly criticized, most prominently by Lawrence Lessig, who is otherwise often in sympathy with geek culture. Lessig suggests that just because the Internet has a particular structure does not mean that it must always be that way. 50 His argument has two prongs: first, that the Internet is structured the way it is because it is made of code that people write, and thus it could have been and will be otherwise, given that there are changes and innovations occurring all the time; second, that the particular structure of the Internet therefore governs or regulates behavior in particular ways: Code is Law. So while it may be true that no one can make the Internet “closed” by passing a law, it is also true that the Internet could become closed if the technology were to be altered for that purpose, a process that may well be nudged and guided by laws, regulations, and norms.

Lessig’s critique is actually at the heart of Bone’s concern, and the concern of recursive publics generally: the Internet is a contest and one that needs to be repeatedly and constantly replayed in order to maintain it as the legitimate infrastructure through which geeks associate with one another. Geeks argue in detail about what distinguishes technical factors from legal or social ones. Openness on the Internet is complexly intertwined with issues of availability, price, legal restriction, usability, elegance of design, censorship, trade secrecy, and so on. [pg 57]

However, even where openness is presented as a natural tendency for technology (in oft-made analogies with reproductive fitness and biodiversity, for example), it is only a partial claim in that it represents only one of the “layers” of a recursive public. For instance, when Bone suggests that the net is “invulnerable to legal attack” because “technology will evolve more quickly than businesses and social institutions can,” he is not only referring to the fact that the Internet’s novel technical configuration has few central points of control, which makes it difficult for a single institution to control it, but also talking about the distributed, loosely connected networks of people who have the right to write and rewrite software and deal regularly with the underlying protocols of the Internet—in other words, of geeks themselves.

Many geeks, perhaps including Bone, discover the nature of this order by coming to understand how the Internet works—how it works technically, but also who created it and how. Some have come to this understanding through participation in Free Software (an exemplary “recursive public”), others through stories and technologies and projects and histories that illuminate the process of creating, growing, and evolving the Internet. The story of the process by which the Internet is standardized is perhaps the most well known: it is the story of the Internet Engineering Task Force and its Requests for Comments system.

Requests for Comments

For many geeks, the Internet Engineering Task Force (IETF) and its Requests for Comments (RFC) system exemplify key features of the moral and technical order they share, the “stories and practices” that make up a social imaginary, according to Charles Taylor. The IETF is a longstanding association of Internet engineers who try to help disseminate some of the core standards of the Internet through [pg 58] the RFC process. Membership is open to individuals, and the association has very little real control over the structure or growth of the Internet—only over the key process of Internet standardization. Its standards rarely have the kind of political legitimacy that one associates with international treaties and the standards bodies of Geneva, but they are nonetheless de facto legitimate. The RFC process is an unusual standards process that allows modifications to existing technologies to be made before the standard is finalized. Together Internet standards and the RFC process form the background of the Napster debate and of Jeff Bone’s claims about “internet routing protocols.”

A famous bit of Internet-governance folklore expresses succinctly the combination of moral and technical order that geeks share (attributed to IETF member David Clark): “We reject kings, presidents, and voting. We believe in rough consensus and running code.” 51 This quote emphasizes the necessity of arguing with and through technology, the first aspect of a recursive public; the only argument that convinces is working code. If it works, then it can be implemented; if it is implemented, it will “route around” the legal damage done by the RIAA. The notion of “running code” is central to an understanding of the relationship between argumentby- technology and argument-by-talk for geeks. Very commonly, the response by geeks to people who argued about Napster that summer—and the courts’ decisions regarding it—was to dismiss their complaints as mere talk. Many suggested that if Napster were shut down, thousands more programs like it would spring up in its wake. As one mailing-list participant, Ashish “Hash” Gulhati, put it, “It is precisely these totally unenforceable and mindless judicial decisions that will start to look like self-satisfied wanking when there’s code out there which will make the laws worth less than the paper they’re written on. When it comes to fighting this shit in a way that counts, everything that isn’t code is just talk.” 52

Such powerful rhetoric often collapses the process itself, for someone has to write the code. It can even be somewhat paradoxical: there is a need to talk forcefully about the need for less talk and more code, as demonstrated by Eugen Leitl when I objected that Silk-listers were “just talking”: “Of course we should talk. Did my last post consist of some kickass Python code adding sore-missed functionality to Mojonation? Nope. Just more meta-level waffle about the importance of waffling less, coding more. I lack the [pg 59] proper mental equipment upstairs for being a good coder, hence I attempt to corrupt young impressionable innocents into contributing to the cause. Unashamedly so. So sue me.” 53

Eugen’s flippancy reveals a recognition that there is a political component to coding, even if, in the end, talk disappears and only code remains. Though Eugen and others might like to adopt a rhetoric that suggests “it will just happen,” in practice none of them really act that way. Rather, the activities of coding, writing software, or improving and diversifying the software that exists are not inevitable or automatic but have specific characteristics. They require time and “the proper mental equipment.” The inevitability they refer to consists not in some fantasy of machine intelligence, but in a social imaginary shared by many people in loosely connected networks who spend all their free time building, downloading, hacking, testing, installing, patching, coding, arguing, blogging, and proselytizing—in short, creating a recursive public enabled by the Internet.

Jeff Bone’s op-ed piece, which is typically enthusiastic about the inevitability of new technologies, still takes time to reference one of thousands (perhaps tens of thousands) of projects as worthy of attention and support, a project called Fling, which is an attempt to rewrite the core protocols of the Internet. 54 The goal of the project is to write a software implementation of these protocols with the explicit goal of making them “anonymous, untraceable, and untappable.” Fling is not a corporation, a start-up, or a university research project (though some such projects are); it is only a Web site. The core protocols of the Internet, contained in the RFCs, are little more than documents describing how computers should interact with each other. They are standards, but of an unusual kind. 55 Bone’s leap from a discussion about Napster to one about the core protocols of the Internet is not unusual. It represents the second aspect of a recursive public: the importance of understanding the Internet as a set of “layers,” each enabling the next and each requiring an openness that both prevents central control and leads to maximum creativity.

RFCs have developed from an informal system of memos into a formal standardization process over the life of the Internet, as the IETF and the Internet Society (ISOC) have become more bureaucratic entities. The process of writing and maintaining these documents is particular to the Internet, precisely because the Internet [pg 60] is the kind of network experiment that facilitates the sharing of resources across administratively bounded networks. It is a process that has allowed all the experimenters to both share the network and to propose changes to it, in a common space. RFCs are primarily suggestions, not demands. They are “public domain” documents and thus available to everyone with access to the Internet. As David Clark’s reference to “consensus and running code” demonstrates, the essential component of setting Internet standards is a good, working implementation of the protocols. Someone must write software that behaves in the ways specified by the RFC, which is, after all, only a document, not a piece of software. Different implementations of, for example, the TCP/IP protocol or the File Transfer Protocol (ftp) depend initially on individuals, groups, and/or corporations building them into an operating-system kernel or a piece of user software and subsequently on the existence of a large number of people using the same operating system or application.

In many cases, subsequent to an implementation that has been disseminated and adopted, the RFCs have been amended to reflect these working implementations and to ordain them as standards. So the current standards are actually bootstrapped, through a process of writing RFCs, followed by a process of creating implementations that adhere loosely to the rules in the RFC, then observing the progress of implementations, and then rewriting RFCs so that the process begins all over again. The fact that geeks can have a discussion via e-mail depends on the very existence of both an RFC to define the e-mail protocol and implementations of software to send the e-mails.

This standardization process essentially inverts the process of planning. Instead of planning a system, which is then standardized, refined, and finally built according to specification, the RFC process allows plans to be proposed, implemented, refined, reproposed, rebuilt, and so on until they are adopted by users and become the standard approved of by the IETF. The implication for most geeks is that this process is permanently and fundamentally open: changes to it can be proposed, implemented, and adopted without end, and the better a technology becomes, the more difficult it becomes to improve on it, and therefore the less reason there is to subvert it or reinvent it. Counterexamples, in which a standard emerges but no one adopts it, are also plentiful, and they suggest that the standardization process extends beyond the proposal-implementation-proposal-standard [pg 61] circle to include the problem of actually convincing users to switch from one working technology to a better one. However, such failures of adoption are also seen as a kind of confirmation of the quality or ease of use of the current solution, and they are all the more likely to be resisted when some organization or political entity tries to force users to switch to the new standard—something the IETF has refrained from doing for the most part.

Conclusion: Recursive Public

Napster was a familiar and widely discussed instance of the “reorientation of power and knowledge” (or in this case, power and music) wrought by the Internet and the practices of geeks. Napster was not, however, a recursive public or a Free Software project, but a dot-com-inspired business plan in which proprietary software was given away for free in the hopes that revenue would flow from the stock market, from advertising, or from enhanced versions of the software. Therefore, geeks did not defend Napster as much as they experienced its legal restriction as a wake-up call: the Internet enables Napster and will enable many other things, but laws, corporations, lobbyists, money, and governments can destroy all of it.

I started this chapter by asking what draws geeks together: what constitutes the chain that binds geeks like Sean and Adrian to hipsters in Berlin and to entrepreneurs and programmers in Bangalore? What constitutes their affinity if it is not any of the conventional candidates like culture, nation, corporation, or language? A colloquial answer might be that it is simply the Internet that brings them together: cyberspace, virtual communities, online culture. But this doesn’t answer the question of why? Because they can? Because Community Is Good? If mere association is the goal, why not AOL or a vast private network provided by Microsoft?

My answer, by contrast, is that geeks’ affinity with one another is structured by shared moral and technical understandings of order. They are a public, an independent public that has the ability to build, maintain, and modify itself, that is not restricted to the activities of speaking, writing, arguing, or protesting. Recursive publics form through their experience with the Internet precisely because the Internet is the kind of thing they can inhabit and transform. Two [pg 62] things make recursive publics distinctive: the ability to include the practice of creating this infrastructure as part of the activity of being public or contesting control; and the ability to “recurse” through the layers of that infrastructure, maintaining its publicness at each level without making it into an unchanging, static, unmodifiable thing.

The affinity constituted by a recursive public, through the medium of the Internet, creates geeks who understand clearly what association through the Internet means. This affinity structures their imagination of what the Internet is and enables: creation, distribution, modification of knowledge, music, science, software. The infrastructure—this-infrastructure-here, the Internet—must be understood as part of this imaginary (in addition to being a pulsating tangle of computers, wires, waves, and electrons).

The Internet is not the only medium for such association. A corporation, for example, is also based on a shared imaginary of the economy, of how markets, exchanges, and business cycles are supposed to work; it is the creation of a concrete set of relations and practices, one that is generally inflexible—even in this age of socalled flexible capitalism—because it requires a commitment of time, humans, and capital. Even in fast capitalism one needs to rent office space, buy toilet paper, install payroll software, and so on.

The Internet is not the only medium for such association. A corporation, for example, is also based on a shared imaginary of the economy, of how markets, exchanges, and business cycles are supposed to work; it is the creation of a concrete set of relations and practices, one that is generally inflexible—even in this age of socalled flexible capitalism—because it requires a commitment of time, humans, and capital. Even in fast capitalism one needs to rent office space, buy toilet paper, install payroll software, and so on.

The urgency evidenced in the case of Napster (and repeated in numerous other instances, such as the debate over net neutrality) is linked to a moral idea of order in which there is a shared imaginary[pg 63] of The Public, and not only a vast multiplicity of competing publics. It is an urgency linked directly to the fact that the Internet provides geeks with a platform, an environment, an infrastructure through which they not only associate, but create, and do so in a manner that is widely felt to be autonomous, autotelic, and independent of at least the most conventional forms of power: states and corporations—independent enough, in fact, that both states and corporations can make widespread use of this infrastructure (can become geeks themselves) without necessarily endangering its independence.

 23. For the canonical story, see Levy, Hackers. Hack referred to (and still does) a clever use of technology, usually unintended by the maker, to achieve some task in an elegant manner. The term has been successfully redefined by the mass media to refer to computer users who break into and commit criminal acts on corporate or government or personal computers connected to a network. Many self-identified hackers insist that the criminal element be referred to as crackers (see, in particular, the entries on “Hackers,” “Geeks” and “Crackers” in The Jargon File, http://www.catb.org/~esr/jargon/, also published as Raymond, The New Hackers’ Dictionary). On the subject of definitions and the cultural and ethical characteristics of hackers, see Coleman, “The Social Construction of Freedom,” chap. 2.

 24. One example of the usage of geek is in Star, The Cultures of Computing. Various denunciations (e.g., Barbrook and Cameron, “The California Ideology”; Borsook, Technolibertarianism) tend to focus on journalistic accounts of an ideology that has little to do with what hackers, geeks, and entrepreneurs actually make. A more relevant categorical distinction than that between hackers and geeks is that between geeks and technocrats; in the case of technocrats, the “anthropology of technocracy” is proposed as the study of the limits of technical rationality, in particular the forms through which “planning” creates “gaps in the form that serve as ‘targets of intervention’” (Riles, “Real Time,” 393). Riles’s “technocrats” are certainly not the “geeks” I portray here (or at least, if they are, it is only in their frustrating day jobs). Geeks do have libertarian, specifically Hayekian or Feyerabendian leanings, but are more likely to see technical failures not as failures of planning, but as bugs, inefficiencies, or occasionally as the products of human hubris or stupidity that is born of a faith in planning.

 25. See The Geek Code, http://www.geekcode.com/.

 26. Geeks are also identified often by the playfulness and agility with which they manipulate these labels and characterizations. See Michael M. J. Fischer, “Worlding Cyberspace” for an example.

 27. Taylor, Modern Social Imaginaries, 86.

 28. On the subject of imagined communities and the role of information technologies in imagined networks, see Green, Harvey, and Knox, “Scales of Place and Networks”; and Flichy, The Internet Imaginaire.

 29. Taylor, Modern Social Imaginaries, 32.

 30. Ibid., 33-48. Taylor’s history of the transition from feudal nobility to civil society to the rise of republican democracies (however incomplete) is comparable to Foucault’s history of the birth of biopolitics, in La naissance de la biopolitique, as an attempt to historicize governance with respect to its theories and systems, as well as within the material forms it takes.

 31. Ricoeur, Lectures on Ideology and Utopia, 2.

 32. Geertz, “Ideology as a Cultural System”; Mannheim, Ideology and Utopia. Both, of course, also signal the origin of the scientific use of the term proximately with Karl Marx’s “German Ideology” and more distantly in the Enlightenment writings of Destutt de Tracy.

 33. Geertz, “Ideology as a Cultural System,” 195.

 34. Ibid., 208-13.

 35. The depth and the extent of this issue is obviously huge. Ricoeur’s Lectures on Ideology and Utopia is an excellent analysis to the problem of ideology prior to 1975. Terry Eagleton’s books The Ideology of the Aesthetic and Ideology: An Introduction are Marxist explorations that include discussions of hegemony and resistance in the context of artistic and literary theory in the 1980s. Slavoj Žižek creates a Lacanian-inspired algebraic system of analysis that combines Marxism and psychoanalysis in novel ways (see Žižek, Mapping Ideology). There is even an attempt to replace the concept of ideology with a metaphor of “software” and “memes” (see Balkin, Cultural Software). The core of the issue of ideology as a practice (and the vicissitudes of materialism that trouble it) are also at the heart of works by Pierre Bourdieu and his followers (on the relationship of ideology and hegemony, see Laclau and Mouffe, Hegemony and Socialist Strategy). In anthropology, see Comaroff and Comaroff, Ethnography and the Historical Imagination.

 36. Ricoeur, Lectures on Ideology and Utopia, 10.

 37. Taylor, Modern Social Imaginaries, 23.

 38. Ibid., 25.

 39. Ibid., 26-27.

 40. Ibid., 28.

 41. The question of gender plagues the topic of computer culture. The gendering of hackers and geeks and the more general exclusion of women in computing have been widely observed by academics. I can do no more here than direct readers to the increasingly large and sophisticated literature on the topic. See especially Light, “When Computers Were Women”; Turkle, The Second Self and Life on the Screen. With respect to Free Software, see Nafus, Krieger, Leach, “Patches Don’t Have Gender.” More generally, see Kirkup et al., The Gendered Cyborg; Downey, The Machine in Me; Faulkner, “Dualisms, Hierarchies and Gender in Engineering”; Grint and Gill, The Gender-Technology Relation; Helmreich, Silicon Second Nature; Herring, “Gender and Democracy in Computer-Mediated Communication”; Kendall, “‘Oh No! I’m a NERD!’”; Margolis and Fisher, Unlocking the Clubhouse; Green and Adam, Virtual Gender; P. Hopkins, Sex/Machine; Wajcman, Feminism Confronts Technology and “Reflections on Gender and Technology Studies”; and Fiona Wilson, “Can’t Compute, Won’t Compute.” Also see the novels and stories of Ellen Ullman, including Close to the Machine and The Bug: A Novel.

 42. Originally coined by Steward Brand, the phrase was widely cited after it appeared in Barlow’s 1994 article “The Economy of Ideas.”

 43. On the genesis of “virtual communities” and the role of Steward Brand, see Turner, “Where the Counterculture Met the New Economy.”

 44. Warner, “Publics and Counterpublics,” 51.

 45. Ibid., 51-52. See also Warner, Publics and Counterpublics, 69.

 46. The rest of this message can be found in the Silk-list archives at http://groups.yahoo.com/group/silk-list/message/2869 (accessed 18 August 2006). The reference to “Fling” is to a project now available at http://fling.sourceforge.net/ (accessed 18 August 2006). The full archives of Silk-list can be found at http://groups.yahoo.com/group/silk-list/ and the full archives of the FoRK list can be found at http://www.xent.com/mailman/listinfo/fork/.

 47. Vinge, “The Coming Technological Singularity.”

 48. Moore’s Law—named for Gordon Moore, former head of Intel—states that the speed and capacity of computer central processing units (CPUs) doubles every eighteen months, which it has done since roughly 1970. Metcalfe’s Law—named for Robert Metcalfe, inventor of Ethernet—states that the utility of a network equals the square of the number of users, suggesting that the number of things one can do with a network increases exponentially as members are added linearly.

 49. This quotation from the 1990s is attributed to Electronic Frontier Foundation’s founder and “cyber-libertarian” John Gilmore. Whether there [pg 319] is any truth to this widespread belief expressed in the statement is not clear. On the one hand, the protocol to which this folklore refers—the general system of “message switching” and, later, “packet switching” invented by Paul Baran at RAND Corporation—does seem to lend itself to robustness (on this history, see Abbate, Inventing the Internet). However, it is not clear that nuclear threats were the only reason such robustness was a design goal; simply to ensure communication in a distributed network was necessary in itself. Nonetheless, the story has great currency as a myth of the nature and structure of the Internet. Paul Edwards suggests that both stories are true (“Infrastructure and Modernity,” 216-20, 225n13).

 50. Lessig, Code and Other Laws of Cyberspace. See also Gillespie, “Engineering a Principle” on the related history of the “end to end” design principle.

 51. This is constantly repeated on the Internet and attributed to David Clark, but no one really knows where or when he stated it. It appears in a 1997 interview of David Clark by Jonathan Zittrain, the transcript of which is available at http://cyber.law.harvard.edu/jzfallsem//trans/clark/ (accessed 18 August 2006).

 52. Ashish “Hash” Gulhati, e-mail to Silk-list mailing list, 9 September 2000, http://groups.yahoo.com/group/silk-list/message/3125.

 53. Eugen Leitl, e-mail to Silk-list mailing list, 9 September 2000, http://groups.yahoo.com/group/silk-list/message/3127. Python is a programming language. Mojonation was a very promising peer-to-peer application in 2000 that has since ceased to exist.

 54. In particular, this project focuses on the Transmission Control Protocol (TCP), the User Datagram Protocol (UDP), and the Domain Name System (DNS). The first two have remained largely stable over the last thirty years, but the DNS system has been highly politicized (see Mueller, Ruling the Root).

 55. On Internet standards, see Schmidt and Werle, Coordinating Technology; Abbate and Kahin, Standards Policy for Information Infrastructure.

License: Licensed under the Creative Commons Attribution-NonCommercial-Share Alike License, available at http://creativecommons.org/licenses/by-nc-sa/3.0/ or by mail from Creative Commons, 559 Nathan Abbott Way, Stanford, Calif. 94305, U.S.A. "NonCommercial" as defined in this license specifically excludes any sale of this work or any portion thereof for money, even if sale does not result in a profit by the seller or if the sale is by a 501(c)(3) nonprofit or NGO.
Duke University Press gratefully acknowledges the support of HASTAC (Humanities, Arts, Science, and Technology Advanced Collaboratory), which provided funds to help support the electronic interface of this book.
Two Bits is accessible on the Web at twobits.net.

≅ SiSU Spine ፨ (object numbering & object search)

(web 1993, object numbering 1997, object search 2002 ...) 2024