The following essay is adapted from “A Spider’s Web,” the tenth chapter of my book The Big Switch.
It’s natural to think of the Internet as a technology of emancipation. It gives us unprecedented freedom to express ourselves, to share our ideas and passions, to find and collaborate with soul mates, and to discover information on almost any topic imaginable. For many people, going online has felt like a passage into a new and radically different kind of democratic state, one freed of the physical and social demarcations and constraints that can hobble us in the real world.
The sense of the Web as personally “empowering,” to use the common buzzword, remains strong, even among those who rue its commercialization or decry the crassness of much of its content. In early 2006, the editors of the Cato Institute’s online journal Cato Unbound published a special issue on the state of the Internet. They reported that the “collection of visionaries” contributing to the issue appeared to be “in unanimous agreement that the Internet is, and will continue to be, a force for liberation.” The Net is “the world’s largest ungoverned space,” declare Google’s Eric Schmidt and Jared Cohen in their new book The New Digital Age. “Never before in history have so many people, from so many places, had so much power at their fingertips.” David Weinberger, in Small Pieces Loosely Joined, summed up the Internet’s liberation mythology in simple terms: “The Web is a world we’ve made for one another.”
It’s a stirring thought, but like most myths it’s at best a half-truth and at worst a delusion. Computer systems in general and the Internet in particular put enormous power into the hands of individuals, but they put even greater power into the hands of companies, governments, and other institutions whose business it is to control individuals. Computer systems are not at their core technologies of emancipation. They are technologies of control. They were designed as tools for monitoring and influencing human behavior, for controlling what people do and how they do it. As we spend more time online, filling databases with the details of our lives and desires, software programs will grow ever more capable of discovering and exploiting subtle patterns in our behavior. The people or organizations using the programs will be able to discern what we want, what motivates us, and how we’re likely to react to various stimuli. They will, to use a cliché that happens in this case to be true, know more about us than we know about ourselves.
Even as the latest Net-related technological developments—the cloud, social media and networks, powerful handheld computers—grant us new opportunities and tools for self-expression and self-fulfillment, they are also giving others an unprecedented ability to influence how we think and what we do, to funnel our attention and actions toward their own ends. The technology’s ultimate social and personal consequences will be determined in large measure by how the tension between the two sides of its nature—liberating and controlling—comes to be resolved.
All living systems, from amoebas to nation-states, sustain themselves through the processing of matter, energy, and information. They take in materials from their surroundings, and they use energy to transform those materials into various useful substances, discarding the waste. This continuous turning of inputs into outputs is controlled through the collection, interpretation, and manipulation of information. The process of control itself has two thrusts. It involves measurement—the comparison of the current state of a system to its desired state. And it involves two-way communication—the transmission of instructions and the collection of feedback on results. The processing of information for the purpose of control may result in the release of a hormone into the bloodstream, the expansion of a factory’s production capacity, or the launch of a missile from a warship, but it works in essentially the same way in any living system.
When, in the 1880s, Herman Hollerith created the punch-card tabulator, the antecedent of today’s digital computer, he wasn’t just pursuing his native curiosity as an engineer and an inventor. He was responding to an imbalance between, on the one hand, the technologies for processing matter and energy and, on the other, the technologies for processing information. He was trying to help resolve what James R. Beniger, in The Control Revolution, calls a “crisis of control,” a crisis that was threatening to undermine the stability of markets and bring economic and technological progress to a halt.
Throughout the first two centuries of the Industrial Revolution, the processing of matter and energy had advanced far more rapidly than the processing of information. The steam engine, used to power ships and trains and industrial machines, allowed factories, transportation carriers, retailers, and other businesses to expand their operations and their markets far beyond what was possible when production and distribution were restricted by the limitations of muscle power. Business owners, who had previously been able to observe their operations in their entirety and control them directly, now had to rely on information from many different sources to manage their companies. But they found that they lacked the means to collect and analyze the information fast enough to make timely decisions. Measurement and communication both began to break down, hamstringing management and impeding the further growth of businesses. As the sociologist Emile Durkheim observed in 1893, “The producer can no longer embrace the market in a glance, nor even in thought. He can no longer see limits, since it is, so to speak, limitless. Accordingly production becomes unbridled and unregulated.” Government officials found themselves in a similar predicament, unable to assemble and analyze the information required to regulate commerce. The processing of materials and energy had progressed so rapidly that it had gone, quite literally, out of control.
During the second half of the nineteenth century, a series of technological advances in information processing helped administrators, in both business and government, begin to re-impose control over commerce and society, bringing order to chaos and opening the way for even larger organizations. The construction of the telegraph system, begun by Samuel F.B. Morse in 1845, allowed information to be communicated instantaneously across great distances. The establishment of time zones in 1883 allowed for more precise measurement of the flows of goods. The most important of the new control technologies, however, was bureaucracy—the organization of people into hierarchical information-processing systems. Bureaucracies had, of course, been around as long as civilization itself, but, as Beniger writes, “bureaucratic administration did not begin to achieve anything approximating its modern form until the late Industrial Revolution.” Just as the division of labor in factories provided for the more efficient processing of matter, so the division of labor in government and business offices allowed for the more efficient processing of information.
But bureaucrats alone could not keep up with the flood of data that needed to be processed—the measurement and communication requirements went beyond the capacities of even large groups of human beings. Just like their counterparts on factory floors, information workers needed new tools to do their jobs. That requirement became embarrassingly obvious inside the U.S. Census Bureau at the end of the century. During the 1870s, the federal government, struggling to administer a country and an economy that were growing rapidly in size and complexity, had demanded that the Bureau greatly expand the scope of its data collection, particularly in the areas of business and transport. The 1870 census had spanned just five subjects; the 1880 round was expanded to cover 215. But the new census turned into a disaster for the government. Even though many professional managers and clerks had been hired by the Bureau, the volume of data overwhelmed their ability to process it. By 1887, the agency found itself in the uncomfortable position of having to begin preparations for the next census even as it was still laboring to tabulate the results of the last one. It was in that context that Hollerith, who had worked on the 1880 census, rushed to invent his information-processing machine. He judged, correctly, that it would prove invaluable not only to the Census Bureau but to large companies across the nation. (Hollerith’s company would become IBM.)
The arrival of the punch-card tabulator was a seminal event in a new revolution—a “Control Revolution,” as Beniger terms it—that followed and was made necessary and inevitable by the Industrial Revolution. Through the Control Revolution, the technologies for processing information finally caught up with the technologies for processing matter and energy, bringing the living system of society back into equilibrium. The entire history of automated data processing, from Hollerith’s tabulator through the mainframe computer and on to the modern computer network, is best understood as part of that ongoing process of reestablishing and maintaining control. “Microprocessor and computer technologies, contrary to currently fashionable opinion, are not new forces only recently unleashed upon an unprepared society,” writes Beniger, “but merely the latest installment in the continuing development of the Control Revolution.”
It should come as no surprise, then, that most of the major advances in computing and networking, from Hollerith’s time to the present, have been spurred not by a desire to liberate the masses but by a need for greater control on the part of commercial and governmental bureaucrats, often ones associated with military operations and national defense. Indeed, the very structure of a bureaucracy is replicated in the functions of a computer. A computer gathers information through its input devices, records information as files in its memory, imposes formal rules and procedures on its users through its programs, and communicates information through its output devices. It is a tool for dispensing instructions, for gathering feedback on how well those instructions are carried out, and for measuring progress toward some specified goal. In using a computer, a person becomes part of the control mechanism. He turns into a component of what the Internet pioneer J. C. R. Licklider, in the seminal 1960 paper “Man-Computer Symbiosis,” described as a system integrating man and machine into a single, programmable unit.
But while computer systems played a major role in helping businesses and governments reestablish central control over workers and citizens in the wake of the Industrial Revolution, the other side of their nature—as tools for personal empowerment—has also helped shape modern society, particularly in recent years. By shifting power from institutions to individuals, information-processing machines can disturb and disrupt control as well as reinforce it. Such disturbances tend to be short-lived, however. Institutions have proven adept at reestablishing control through the development of ever more powerful information technologies. As Beniger explains, “information processing and flows need themselves to be controlled, so that informational technologies continue to be applied at higher and higher levels of control.”
The arrival of the personal computer in the 1980s, for example, posed a sudden and unexpected threat to centralized power. It initiated a new, if much more limited, crisis of control. Pioneered by countercultural hackers and hobbyists, the PC was infused from the start with a libertarian ideology. As memorably portrayed in Apple Computer’s dramatic “1984” television advertisement, the personal computer was to be a weapon against central control, a tool for destroying the Big Brother-like hegemony of the corporate mainframe. Office workers began buying PCs with their own money, bringing them to their offices, and setting them up on their desks. Bypassing corporate systems altogether, PC-empowered employees gained personal control over the data and programs they used. They gained freedom, but in the process they weakened the ability of bureaucracies to monitor and steer their work. Business executives and the IT managers that served them viewed the flood of PCs into the workplace as “a Biblical plague,” in the words of computer historian Paul Ceruzzi.
The breakdown of control proved fleeting. The client-server system, which tied all the previously autonomous PCs together into a single network connected to a central store of corporate information and software, was the means by which the bureaucrats reasserted their control over information and its processing. Together with an expansion in the size and power of IT departments, client-server systems enabled companies to restrict access to data and to limit the use of software to a set of prescribed programs. Ironically, once they were networked into a corporate system, PCs actually enabled companies to monitor, structure, and guide the work of employees more tightly than was ever possible before. “Local networking took the ‘personal’ out of personal computing,” explains Ceruzzi in A History of Modern Computing. “PC users in the workplace accepted this Faustian bargain. The more computer-savvy among them resisted, but the majority of office workers hardly even noticed how much this represented a shift away from the forces that drove the invention of the personal computer in the first place. The ease with which this transition took place shows that those who believed in truly autonomous, personal computing were perhaps naïve.”
The popularization of the Internet, through the World Wide Web and its browser, brought another and very similar control crisis. Although the construction of the Internet was spearheaded by the Department of Defense, a paragon of centralized power, it was designed, paradoxically, to be a highly dispersed, loosely organized network. Since the overriding goal was to build as reliable a system as possible—one that could withstand the failure of any of its parts—it was given a radically decentralized structure. Every computer, or node, operates autonomously, and communications between computers don’t have to pass through any central clearinghouse. The Net’s “internal protocols,” as New York University professor Alexander Galloway writes in his book Protocol, “are the enemy of bureaucracy, of rigid hierarchy, and of centralization.” If a corporate computer network was akin to a railroad, with tightly scheduled and monitored traffic, the Internet was more like the highway system, with largely free-flowing and unsupervised traffic.
At work and at home, people found they could use the Web to once again bypass established centers of control, whether corporate bureaucracies, government agencies, retailing empires, or media conglomerates. Seemingly uncontrolled and uncontrollable, the Web was routinely portrayed as a new frontier, a Rousseauian wilderness in which we, as autonomous agents, were free to redefine society on our own terms. “Governments of the Industrial World,” proclaimed John Perry Barlow in his 1996 Declaration of the Independence of Cyberspace, “you are not welcome among us. You have no sovereignty where we gather.” But, as with the arrival of the PC, it didn’t take long for governments, and corporations, to begin reasserting and even extending their dominion.
The error that Barlow and many others have made is to assume that the Net’s decentralized structure is necessarily resistant to social control. They’ve turned a technical characteristic into a metaphor for personal freedom. But, as Galloway explains, the connection of previously untethered computers into a network governed by strict protocols has actually created “a new apparatus of control.” Indeed, he writes, “the founding principle of the Net is control, not freedom—control has existed from the beginning.” As the fragmented pages of the World Wide Web turn into the unified and programmable database of the cloud, moreover, powerful new forms of surveillance and control become possible. What is programming, after all, but a method of control? Even though the Internet still has no center, technically speaking, control can now be wielded, through data-mining algorithms and other software code, from anywhere. What’s different, in comparison to the physical world, is that acts of control, even as they grow much larger in scale, become harder to detect and the wielders of control become more difficult to discern.
EPILOGUE: Shortly after I posted this piece, word broke about the National Security Agency’s up-to-now secret PRISM program, in which the spy agency accesses vast amounts of data from major Internet-service companies, including Google, Facebook, Apple, and Microsoft, to track and otherwise gather intelligence on individuals. Here we see the basic pattern of the Control Revolution again playing out, with a disruption of control, caused by a flood of unregulated information, followed by the development of new information technologies that enable the reestablishment of control at a higher level. A report in the Times seems pertinent:
Today, a revolution in software technology that allows for the highly automated and instantaneous analysis of enormous volumes of digital information has transformed the N.S.A., turning it into the virtual landlord of the digital assets of Americans and foreigners alike. The new technology has, for the first time, given America’s spies the ability to track the activities and movements of people almost anywhere in the world without actually watching them or listening to their conversations. …
While once the flow of data across the Internet appeared too overwhelming for N.S.A. to keep up with, the recent revelations suggest that the agency’s capabilities are now far greater than most outsiders believed. “Five years ago, I would have said they don’t have the capability to monitor a significant amount of Internet traffic,” said Herbert S. Lin, an expert in computer science and telecommunications at the National Research Council. Now, he said, it appears “that they are getting close to that goal.”
As does this, from the Journal:
Key advances in computing and software in recent years opened the door for the National Security Agency to analyze far larger volumes of phone, Internet and financial data to search for terrorist attacks, paving the way for the programs now generating controversy. … The NSA’s advances have come in the form of programs developed on the West Coast—a central one was known by the quirky name Hadoop—that enable intelligence agencies to cheaply amplify computing power, U.S. and industry officials said. The new capabilities allowed officials to shift from being overwhelmed by data to being able to make sense of large chunks of it to predict events, the officials said.