Spam NPR coverage of Spam: A Shadow History of the Internet by Finn Brunton. News, author interviews, critics' picks and more.
NPR logo Spam


A Shadow History of the Internet

by Finn Brunton

Hardcover, 270 pages, Mit Pr, List Price: $27.95 |


Buy Featured Book

A Shadow History of the Internet
Finn Brunton

Your purchase helps support NPR programming. How?

NPR Summary

Spam emails — those commonplace nuisances in inboxes around the world — are a surprisingly complex phenomenon, part of a global criminal infrastructure in which we are all, unwittingly, entangled.

Read an excerpt of this book

Note: Book excerpts are provided by the publisher and may contain language some find offensive.

Excerpt: Spam



Pitcairn Island in the South Pacific is the smallest and least populated jurisdiction in the world, described by the first mariner to observe it as a "great rock rising out of the sea ... about a thousand leagues westward to the continent of America." It's a lonely place; when the Bounty mutineers needed a refuge from the global empire whose laws they'd broken, they sailed for Pitcairn and vanished for almost two decades. As of last year's electoral rolls, there are 45 people on Pitcairn, most of them descended from the mutineers. There is a government-subsidized satellite Internet connection for the island's houses. Per capita, Pitcairn Island is the world's number-one source of spam.

How is this possible? Can this island, whose major export industries are handicrafts for passing ships and stamps for philatelists, host a cabal of spammers bombarding the world with phishing messages and Viagra ads by satellite connection? In the list of spam production per capita, Pitcairn is followed by Niue and Tokelau, also in the South Pacific; the Principality of Monaco, whose population overwhelmingly consists of rich expatriates dodging taxes in the world's second smallest country; and the Principality of Andorra, a country one-fifth the size of Rhode Island. Are there really that many Catalan-speaking Andorran spam barons, accumulating stolen credit card data by the ski slopes of the Pyrenees?

The answer is no, of course. The Polynesians who live on Niue, like the wealthy Europeans overlooking the Mediterranean, have been unknowingly conscripted into the business of spam. Virtually all of us on the Internet have been, on one side or another. Spam has changed laws and communities at the points of friction where the network's capacities rub against prior orders of work and governance. It has changed our language, economics, and culture and exerted a profound effect on our technologies. It has subtly — and not so subtly — deformed the shape of life online, pulling it into new arrangements that make no more sense than the movement of the galaxies unless you allow for the mass of all the dark matter. Only at scale, in time and in space, can we begin to bring this shape into focus. It stretches from embryonic computer networks in the 1960s and 1970s to the global social graph of 2010, from the basements of MIT to the cybercafes of Lagos, Rotterdam, and Tallinn. It encompasses points across the network from covert chat channels to Google's server farms to " ghost number blocks" to anonymous banks of airport pay phones. The spam most of us experience every day as a minor and inexplicable irritant is like the lonely rock that sailor sighted, merely the visible peak of vast and submerged infrastructures with much to say about the networked world.

The word "spam" means very different things to different people at different times. It is a noun, collective and singular, as "this spam" can mean "all these messages I've received" or "this particular email." It is a verb, as in "they spam me," and an adjective, as in "this is spammy." It refers to many varieties of exploitation, malfeasance, and bad behavior, and spam terminology has branched out into specific subdomains, from "phishing spam" and "419 spam" to splogs, linkfarms, floodbots, content farms. (All of these new terms and forms will be defined and described in the following chapters.) It can easily slide into what philosophers call the "sorites paradox" ( "sorites," appropriately for talking about spam, comes from the Greek word for "heap" or "pile" ) to describe the linguistic confusion between grains of sand and "sand" in dunes, the moment at which individual bits become a single big pile. When "spam" is discussed in journalism or casual conversation, the word is often meant as I have used it in the previous paragraph, as a unified substance or a continuous event, like smog or "a mass or pulp," as Susanna Paasonen puts it. But spam begins to make sense only when we get specific and separate out the different types, motives, actors, and groups.

Spam is not a force of nature but the product of particular populations distributed through all the world's countries: programmers, con artists, cops, lawyers, bots and their botmasters, scientists, pill merchants, social media entrepreneurs, marketers, hackers, identity thieves, sysadmins, victims, pornographers, do-it-yourself vigilantes, government officials, and stock touts. Long as it is, this list mentions only those who live there more or less full-time, because everyone online participates in the system of spam, however indirectly. We fund and enable it with choices we make and trade-offs we are willing to accept because the benefits to the network outweigh the costs. We generate areas of relevance and attention whose discovery and exploitation is the heart of the business. We alter how spam works and break up its current order with choices, refusals, and purchases whose consequences we may not understand.

Those houses on Pitcairn, for example, connected to their satellite link, do not shelter spammers hard at work. At some point, a few of the computers on Pitcairn were taken over by a piece of malware, probably arriving as the misleading payload of a spam message that appeared to come from a friend, taking advantage of unpatched or insecure software that can be daunting for the user to maintain. This malware quietly commandeered the computers without their owners ever noticing (perhaps they merely thought that the machine, and the Internet connection, had become a little slower) and enrolled them along with many thousands of other captured computers in homes and offices around the world into a system called a "botnet." One of the functions of the botnet is to use spare computer power and Internet connection bandwidth to send out spam messages on behalf of the botnet's controller, who can thus generate hundreds of millions of messages at effectively no cost. A desktop computer, manufactured in Shenzhen or Xiamen or Chennai, powered by a diesel generator and online through a satellite over the Tropic of Capricorn, on a remote island in the Melvillean Pacific, has become part of a distributed machine controlled by a remote group of botmasters in Denver or St. Petersburg. This is a system unprecedented in human history, a vision out of science fiction, that writes constantly repeating messages of crushing banality: "YOU HAVE WON!!!/Congratulations!!! You have won £ 250,000.00POUNDS from Pepsi Company award 2010, Please provide your Full name, Age, Sex, Country, Occupation" ; "Permanent En1arged- Penis/She prefers your lovestick bigger" ; "Listen up. I must let you in on a few insider *secrets*: Instead of waiting months to generate sales on your site, you can start gaining the hits you want right now. "


This is a book about spam for anyone who wants to understand what spam is, how it works, and what it means, from the earliest computer networks to the present day. To understand spam means understanding what spam is not, because — as you will see — the history of spam is always a history of shifting definitions of what it is that spam harms and the wrong that it produces. The history of spam is the negative shape of the history of people gathering on computer networks, as people are the target of spam's stratagems. It is defined in opposition to the equally shifting and vague value of "community." (In fact, many of the early cases of spam provoke groups of people on computers into the task of self-definition and self-organization as communities.) To put this history and this book into a single sentence: spamming is the project of leveraging information technology to exploit existing gatherings of attention.

Attention, the scarce resource of human notice, is what makes a community on the network, and the creation of communities, the invention of "we" on the Internet, is an act of attention. Communities and spam as a whole are projects in the allocation of attention, and spam is the difference — the shear — between what we as humans are capable of evaluating and giving our attention, and the volume of material our machines are capable of generating and distributing when taken to their functional extremes. 4 Over four decades of networked computing, spammers have worked in that gulf between our human capacities and our machinic capabilities, often by directly exploiting the same technologies and beneficial effects that enable the communities on which they predate. These two forces build, provoke, and attack each other, and the history of one cannot be understood without accounting for its nemesis.

This co-constitutive feedback loop between communities and spammers is a major event in the technological drama that is the Internet and its sister networks. This term, "technological drama," is from the work of Bryan Pfaffenberger and provides framing and context for the story to come; it is the life of a technology, from conception and design through adoption, adaptation, and obsolescence. 5 Why a drama ? Because technologies are statements about the distribution of needful things such as power, status, access, wealth, and privilege, for good and for ill. They are stages on which social and political arguments and counterarguments are made. These arguments may be not merely irrational but also deeply cultural and ritualistic, expressing convictions that lie well beyond the immediate physical constraints of engineering. A technological drama is suspenseful: it is not played out to foregone conclusions but is rife with the possibility of subversion, takeover, and unexpected captures. It is a drama of escalation and feedback loops, for technologies are never merely passive vessels for holding ideas and ideologies but active things in the world that open new possibilities and capacities. They change the communities that created them and those that take them up.

The inciting incident that frames any technological drama and gets it moving is the gathering of a "design constituency," in Pfaffenberger's phrase. "Constituency" is well-chosen, because we are not simply referring to the engineers, inventors, scientists, or designers who actually work the lathe, draw out the blueprints, or blow the glass but to the entire body of people who participate in the design and who stand to directly benefit from the technology's success. It is to their benefit if it sells, of course, assuming that it is the kind of technology that can be commoditized into widgets and not, for instance, a civil engineering project. More important, however, is that the values embedded in the technology, intentionally or unintentionally, become dominant. Those values reflect an arrangement of power, control, and prestige that the design constituency would like to see in the world, whether centralized and privatized, open and egalitarian, or otherwise. This design constituency can include the engineers and applied designers themselves, as well as managers and shareholders in firms, but also politicians, experts, theorists, and elites. What is complex and important here is to be able to view technologies in two ways at once: seeing both their physical capacities and functions, and their social and political assertions — the moves they make in the allocation of resources and power. We will explore computer networks and the people building and using them with this perspective.

To get some perspective on how technological dramas work as an analytic tool, consider the case of heavier-than-air aviation. No one would argue against the profound benefits delivered by the development of powered flight, but to really understand the adoption and adaptation of the technology we must account for the acts of political stake-planting and the renegotiations of influence and control that went with it. Aviation's roots included the powerful early idea of "air-mindedness," for which planes were not just powered, winged craft for flight but objects whose existence was a direct expression of a rational, modern, and global mindset that would simultaneously bring about world peace and subdue colonized populations through bombing. H. G. Wells, for instance, in his role as a public intellectual and policy advocate, framed the work of getting from one place to another by air as nothing less than the formation of a new generation of intellectual pilot-samurai who would enforce a technocratic world state. This coming class would be necessarily scientific, cosmopolitan, and forward thinking, because they used planes. Airplanes, Le Corbusier asserted, were "an indictment, an accusation, a summons" to architects and city planners, shaming their retrograde ideas: planes were pure futurity, the avatars of the machine age. At the same moment, figures such as Gabriele D'Annunzio, aviator and Fascist literary icon, and Italo Balbo, commander of Italian North Africa and leader of the famous Century of Progress Rome – Chicago flight, were using both the glamour and threat of flight as an assertion of Fascism's fusion of futuristic dynamism with nationalist and archaic agendas. In the United States, flight included the democratic vision of self-taught tinkerers hacking on planes in barns and garages and potent assertions of military superiority and the projection of power abroad in an abruptly smaller world. And so on. This many-layered complexity of motives, ideas, fantasies, and goals obtains across the technological domain. To understand any technical event in depth, we need to be able to describe the full range of interests in the work of the design constituency.

The other side of the drama is the "impact constituency," those whose situation is being changed by what the design constituency produces. Neil Postman simply called them the "losers," because the rearrangement of the world, although not necessarily one in which they lose out explicitly, is one over which they have little initial control. Examples from Pfaffenberger ' s research are particularly well suited for this purpose. Consider the politics of irrigation projects in Sri Lanka, a particularly bald-faced attempt at social engineering via technological deployment. Landed, powerful, and deeply anti-industrial local elites sought to manage the threat of dispossessed, dissatisfied, and mobile peasants by creating a neatly controllable system of rural settlements. It was a project of getting them back on the farm, to put it simply, where they would stay ignorant of modernity and easier to tax and manage, forestalling the arrival of an industrial order in which the landholding "brown sahibs" stood to lose. The landholders did the designing, and peasants felt the impact. James C. Scott's work provides several cases along these lines — of compulsory villagization, collectivization of agriculture, and high modernist urban planning — in which the capture and redistribution of value is exerted through the production of artifacts, places, and systems of living, always framed as "for their own good." We could also speak on the far smaller and more intimate scale of something like the Victorian household's hallway bench, with its ornate mirrors and carved hooks, for the household's masters to view as they walk by and hang their hats, and the strikingly uncomfortable bare low bench on which the servants, messengers, and peddlers are to sit and wait.

What makes these relationships of design and impact into a drama is the back-and-forth of technological statements and counterstatements. After the original assertion made in the design and deployment of a technology come the responses, as the impact constituencies take things up, change them, and accept them or fight back. The design constituency cooks up methods and language for using technologies to arrange, distribute, or contain power, status, and wealth, and impact constituencies have countermoves available. They can produce arguments to get their hands on the technology, for instance, and reconstitute it, which does not merely ameliorate the painful setbacks produced by the deployment but actually generates a new technology that builds on the old for their own purposes.

The most obvious and canonical instance of reconstitution in a technological drama, producing a "counterartifact" out of the existing technology, is the personal computer. Decades of development in computing had been the product of the military, academia, and contractors and corporations such as IBM. The computer was "the government machine," sitting in the ballistics lab and running models for building nuclear weapons and game theory analyses of Cold War outcomes. Computing had become virtually synonymous with a bureaucratic, administered society in which people were subsumed as mere components, in what Lewis Mumford termed the "megamachine." Against this concept came the first hackers and subversive tinkerers, activists, and artists, with Ted Nelson (who coined the term "hypertext" ) asserting, "You can and must understand computers now," "countercomputers" for the counterculture, the Homebrew Computer Club, the Apple I and II, and so on. The drama never stops, and Fortune's wheel keeps turning, casting down and raising up and demanding new technical statements and counterstatements.

Getting a significant new technology that is instilled with cultural values and political goals to "work" is an act of assembly, of mobilizing different groups, ideas, forces, and stakeholders together. There needs to be a flexibility to the idea, enough that you can pull heterogeneous and sometimes even opposed powers into an alliance behind it. A technology too clearly and narrowly defined may not be able to develop all the alliances it needs. For the Sri Lankan irrigation project, it meant creating a complicated alliance of ethnic and religious chauvinism, paternalistic philanthropy, opposition to the old British colonial order (by an elite whose power derived from the restructuring of society under the British), and so on. A similar multiplex set of alliances could be seen in the free/libre open source software (FOSS) movement, with businesses, individual charismatic activists, developers, and political radicals of many stripes and very diverse agendas trying to gather under the same banner. The epic semantic fork between the models of open source and free software in 1998 captures one of the moments when the ambiguities became unsustainable and had to be reformulated. "Movement," writes Christopher Kelty, "is an awkward word; not all participants would define their participation this way. ... [They] are neither corporations nor organizations nor consortia (for there are no organizations to consort); they are neither national, sub-national, nor international; they are not 'collectives' because no membership is required or assumed. ... They are not an 'informal' organization, because there is no formal equivalent to mimic or annul. Nor are they quite a crowd, for a crowd can attract participants who have no idea what the goal of the crowd is." This complex mesh, sharing practices and debating ideology, is a design constituency, gathering around the technology and trying to marshal support for different and sometimes conflicting visions to push the project into the world.

This process of gathering and marshaling is strengthened by some founding ambiguities. These let the designers cast a wide net and make it easier to argue that this technological project speaks for you, too. Some of these ambiguities are invoked of necessity. The story that a design constituency builds to draw in stakeholders from different domains to support a new technology must draw on what Victor Turner terms "root paradigms," the deep organizing principles of a culture and an epoch that provide a rationale and a motive for action. Root paradigms aren't exact and precise, and they are never simply true or false. Whether they be submission to the free market, the sanctity of human life, the vital and cleansing power of war, or the unquestionable role of the dynastic king, root paradigms are dynamic, messy, and enormously powerful concepts built on internal oppositions. They draw their energy and vitality from their unsettled condition of irreconcilable struggle within which new technologies, political initiatives, and movements can be placed and contextualized. At major turning points in the development of the Internet and spam, struggles between constituencies were played out that drew on far older root paradigms such as absolute freedom of speech, communal self-defense and self-organization, the technological autonomy of the capable individual, the inevitability of destructive anarchy without governance, and the centrality of commerce to society. The presence of these paradigms gives technological deployments the thrilling, and often later inexplicable, attraction of social movements (because that is, in fact, what they are). They draw their strength from roots sunk deep into the earth, where the bones of prior orders and the ruins of previous civilizations underlie the present.

These foundational ambiguities in a technology's design are a crucial resource for the impact constituencies and others to exploit. Sally Moore describes how a reworking of the arrangements introduced by a technology is made possible, by "exploiting the indeterminacies of the situation, or by generating such indeterminacies ... areas of inconsistency, contradiction, conflict, ambiguity, or open areas that are normatively indeterminate." The indeterminate space is the place for trade-offs and concessions, ways to get many diverse parties working together and pointed in the general direction envisioned by the design constituency. It is also leaves apertures and affordances in the plan for the manipulation, escapes, and exploitation of others, from innovations and improvements to exploits and deliberate sabotage — and thus for things like spam.

This complex indeterminacy obtained at every stage of the Internet's development. As will be discussed later in this book, there was deep uncertainty and widely varying understandings as to what this thing was to be and how it should be used by people and by hardware and software. This uncertainty was an enormous boon for innovators and inventors, for the strange frontiers of network culture, and for both hackers and criminals, whose somewhat blurry relationship and ambiguous legal status recurs in this history. Spam survived and prospered by operating in the edge cases around these big ideas, in the friction between technical facts and the root paradigms that are expressed in them where powerful concepts like trust, anonymity, liberty, and community online were reinvented, modified, and sometimes discarded. In following spam, we will explore how these ideas evolved and, above all, how human attention online, and the strategies for capturing it, changed over time.


Appropriately for a technological drama, the history of spam has three distinct acts, which are reflected in this book's three sections. The first, from the early 1970s to 1995, begins with conversations among the architects of the earliest computer networks, who were trying to work out acceptable rules, mores, and enforcement tools for online communication. It closes with the privatization of the Internet and the end of the ban on commercial activity, the arrival of the Web, and the explosion of spam that followed the Green Card Lottery message on Usenet in May 1994. It is a quartercentury of polylogue concerning the fate and the purpose of this astonishing thing the online population was fashioning and sharing ( "polylogue " being a term from an early computer network for this new form of asynchronous and many-voiced conversation on screens). It includes a remarkable cast of postnational anarchists, baronial system administrators, visionary protocol designers, community-building "process queens," technolibertarian engineers, and a distributed mob of angry antispam activists. It is an era of friction between concepts of communal utility, free speech, and selfgovernance, all of which were shaped in a negative way by spam. "Spam" here is still used in its precommercial meaning of undesirable text, whether repetitive, excessive, or interfering. The imminent metamorphosis of these ideas as the values and vision of the network changed in the mid-1990s was partially signaled and partially led by changes in spam's significance and means of production.

The next phase lasts about eight years, from 1995 to 2003, or from the privatization of the network through the collapse of the dot-com boom and the passage of the CAN-SPAM Act in the United States. It is about money and the balance between law and collective social action. Those years are filled with the diversification of spam into a huge range of methods and markets, following and sometimes driving innovations on the Internet and the Web, from search engine manipulation and stock market "pump-and-dump" schemes to fake password requests and "Nigerian prince" fraud campaigns. During this time, a strange class of magnates and hustlers is born, arbitraging the speed of new technological developments against the gradual pace of projects to regulate them. Their nemeses start here as well, with antispam projects ranging from message-filtering programs to laws and coordinated campaigns of surveillance, research, and harassment. This period is fraught with uncertainty about the role of nations and territorial boundaries online, the ambiguous rights and responsibilities of "users," and the relationship between what algorithms can process and what humans can read.

The most recent phase, from 2003 to the present day, turns on these questions of algorithms and human attention. A constellation of events is dramatically changing the economics of the spam business: the enforcement of laws, the widespread adoption of powerful spam filters, and the creation of user-produced content tools. To keep the business profitable, those spammers who survive the transition will develop systems of automation and distributed computing that span the globe and lead to military applications — building nothing less than a criminal infrastructure. In turn, antispammers will rely on sophisticated algorithms and big data to minimize the labor of catching and blocking spam messages. As spam prefigured and provoked crises in community and governance on the Internet, it now shows us an imminent crisis of attention — in the most abject and extreme form, as always. After four decades of interrupting conversations, grabbing clicks, demanding replies, and muddling search results, spam has much to teach us about the technologies that capture our attention, now and to come.

From Spam, by Finn Brunton. Copyright 2013 by Finn Brunton. Excerpted by permission of MIT Press.