Friday, April 01, 2011

The Hugo Black Lecture, Part V


[On March 23, 2011, I delivered the 20th annual Hugo L. Black lecture on freedom of expression at Wesleyan University. I'm publishing the prepared text of the lecture in installments this week on Balkinization. Part One, Part Two, and Part Three, and Part Four have already appeared.]

VII. The Difference Infrastructure Makes

The story of Wikileaks, like the story of the Egyptian protests, is about the infrastructure of free expression and how it helps or hinders the activities of democracy.

The government did not seek an injunction against Wikileaks largely because the digital infrastructure makes it futile. Assange did not have to rely on the facilities of a major newspaper to publish his revelations. He created mirror sites in multiple countries around the world that made it impossible to block all of his copies. He worked with newspapers for a different reason: to give himself political cover.

Professor Benkler has pointed out another important feature of the new digital infrastructure: Once the leaker (we assume Private Manning) uploaded the materials on the Wikileaks website, Assange could not be co-opted in the same way that traditional media organizations could. Assange picked newspapers in different countries and promised them a scoop in their counties in return for helping him sort through the materials. Because the papers knew that someone else in their country would get the scoop if they refused, they had incentives to cooperate. And because Assange worked with multiple newspapers in different countries, his disclosures would not be prevented if one or two of them were coopted by their governments.

Compare this with the New York Times’ revelation of the Bush Administration’s secret domestic surveillance program. President Bush and his associates had probably violated the law. Nevertheless, the Bush Administration convinced the Times to delay publication until well after the 2004 election, possibly helping George W. Bush win a second term as President. Why did the Times agree to delay publication? It was probably a complicated set of reasons: a personal request from the President, a sense of patriotism, and a desire to maintain the access and contacts with government that contemporary journalists crave. Today major news organizations depend on a series of leaks and background information from government officials, and they do not want to bite the hand that feeds them too often or too hard.

The Obama Administration had no such leverage over Assange and Wikileaks. Even if the Administration co-opted the New York Times, there were plenty of other papers to take its place, and Assange could publish the materials himself without anyone’s permission. Indeed, the Obama Administration refused to deal with Assange directly, although it was apparently willing to talk to its acquaintances at traditional papers like the Times.

Daniel Ellsberg, the leaker in the Pentagon Papers case, faced a very different infrastructure of free expression in 1971. The key technological innovation in his day was the copying machine. Ellsberg made about 18 copies of the Pentagon Papers. He first tried to get the Pentagon Papers read on the Senate floor, where they would be protected by the Constitution’s Speech or Debate Clause. When that failed he went to the New York Times. After a federal district court issued an injunction against the Times, he sent copies to the Washington Post and then to about 16 other papers, which began publishing the papers one after the other.

This was actually an early version of Wikileaks' strategy. Each newspaper had incentives to participate because each circulated in a different geographical area. Ellsberg hoped that the Nixon Administration wouldn’t or couldn’t go after all of the country’s major newspapers at once.

Even so, it was a dangerous game in the pre-Internet world. First, all of the newspapers were located in one country. If the Supreme Court ruled against any one paper that was the end of the game. Second, there were a finite number of copies, and making new ones—and distributing them to new locations—took time. If the government could enjoin the papers and round the copies up quickly enough, Ellsberg might not be able to create more and place them in secure locations. Third, unlike Assange, the publisher in the Wikileaks case, Ellsberg did not own his own newspaper or broadcasting facility. He relied on the mass media distribution of newspapers.

Wikileaks is the beginning of a new model of journalism that uses digital networks to obtain sensitive information anonymously, secure it in multiple sites, and publish it in defiance of territorial governments. Wikileaks-style organizations have been springing up around the Internet, and later innovators will no doubt improve on the techniques pioneered by Wikileaks.

Traditional newspapers will eventually join this trend: they will create and install their own anonymous file depositories and form cooperatives with other newspapers around the world. Conversely, NGOs like the ACLU will do an increasing amount of what is now called investigative journalism, obtaining documents, processing them, and releasing them to the press. In the 20th century professional journalists, NGOs and ordinary citizens had different and clearly demarcated functions and activities; that’s giving way to a new decentralized system in which other actors become more like traditional journalism organizations and traditional journalism organizations become more like Wikileaks and NGOs. As this happens, professional identities and professional norms will change.

Monroe Price, who teaches at the Annenberg School and Cardozo Law School, has pointed out that most countries not only have internal policies for regulating their own media, they also have policies for regulating the media of other countries. Countries want to affect the knowledge and information circulating in other nations. During the Cold War, for example, the United States invested in Radio Free Europe and Radio Liberty; and Radio Marti to influence Eastern Europe, Central Asia, the Middle East, and Cuba; Secretary of State Clinton’s speech on global Internet policy shows that the United States believes that it is our interest to promote Internet freedom and the dissemination of information in other countries, especially countries that the U.S. disagrees with.

Taking Price’s insight one step further, we might say that the Internet allows each individual, or each NGO, to have its own media policy, giving people in other countries information that their governments don’t want them to see. That includes the United States of America. The U.S., naturally, doesn’t like it one bit.

In discussing the Middle East protests, I noted that governments and protestors are in an arms race or an innovation cycle. Just as China has worked to design its infrastructure to undermine protest, governments will find new ways to seek to undermine Wikileaks and its successors.

Divide these techniques into “old school” and “new school” forms of censorship. “Old school” censorship means enjoining publications, taking control of newspapers and television stations, and rounding people up, either to prevent them from speaking, or to teach others a lesson that the government is not to be messed with. As I said, I think this is what the U.S. government is doing in the case of Bradley Manning.

“New school” censorship tries to control the digital infrastructure of free expression; it leverages privately owned networks and employs public-private cooperation. It may prove just as effective in the long run.

First, states can use their power over information infrastructure to insert government controls and surveillance technologies into the infrastructure. They can order businesses who control elements of the infrastructure to hinder, delay, block or censor content and speakers. These parties include the owners of telecommunications facilities like Verizon, technology companies like Cisco, domain name registrants like GoDaddy, site hosting services like Amazon, institutions of electronic commerce like Mastercard and Paypal, platform owners like Blogger, Facebook or Twitter, and search engines like Google. These are all potential private censors, and thus they are all potential targets of government control.

Second, instead of direct orders, the government can coax, persuade, or signal to private owners of the information infrastructure to hinder or block offending publications and speakers. This allows the government to assert that private parties are doing the censoring, not the state itself. And because it is private censorship, we should respect it, first because the market will check any abuses, and second, because corporations have free speech rights to own and control the infrastructure.

What is new here is the use of the various elements of the digital infrastructure—telecommunications conduits, servers, domain name registars, and payment systems—to censor. The strategy of government officials coaxing or signaling private parties to censor, by contrast, is not new at all. It was used, for example, during the McCarthy period, when private parties blacklisted people for fear of being thought communist sympathizers.

Both direct control of infrastructure and public-private cooperation were used in the Middle East and in China. Both are fully available to the United States, and, moreover, the United States is currently employing them. The strategies of the new digital censorship in other parts of the world are not as dissimilar from what we do in the United States as you might think. The greatest threat to freedom of speech today is not simply that of public power or private power. It is their potent combination.

The Hugo Black of the Associated Press case in 1941—who was worried about the power of private combinations to suppress speech—turns out to be just as important to understanding free speech on the Internet today as the Hugo Black of the Pentagon Papers case in 1971, who was worried about government prosecution.

We criticize Yahoo when it capitulates to China. We should also criticize Amazon and Paypal when they capitulate to Joe Lieberman, just as we should criticize Senator Lieberman himself for making appeals that he would never direct at the New York Times, the Guardian Online or Der Spiegel. If Senator Lieberman had suggested that Mastercard stop processing the New York Times’ subscriptions or that Amazon stop hosting its content on its servers, people would have thought he was a lunatic.

Despite the pressure, however, Wikileaks was able to find new intermediaries to work with. The resiliance of Wikileaks suggests how crucial certain features of digital infrastructure are to freedom of speech. The network as currently organized does not respect national boundaries, it is decentralized, redundant, flexible, plastic, and allows for many competitors and services. But that free speech-friendly design is not guaranteed in the future.

VIII. Conclusion

Right now there is enormous pressure in the United States to build back doors to allow surveillance on Internet networks and digital platforms in the United States, and to implement technologies that will make it easy for governments and corporations to filter content and block access to disfavored content.

The pressure on the United States to do these things is not coming from authoritarian strongmen in the Middle East. It is not coming from the People’s Republic of China. It is coming from our government, and from the content industry. The government is worried about potential criminal activities and terrorist networks that use Skype, Facebook, and Gmail; the content industry is worried about file sharing and intellectual property. Meanwhile, telecommunications and broadband companies, which oppose some of these proposals, have their own shopping list: they want to protect the right to block and filter traffic that interferes with their business models or to favor traffic by their business partners; that’s why the industry vigorously opposes network neutrality.

The danger in these proposals is that they threaten the American infrastructure of free expression. Building networks that allow you to filter for intellectual property also allows you to filter for anticompetitive reasons, or even for ideological reasons. Implementing broadband technologies to slow and block traffic that your business partners don’t like allows slowing and blocking traffic for other reasons as well.

Moreover, building a back door into online communications means building a surveillance system into every aspect of our lives that uses digital communications systems, ranging from e-mail to Facebook to gaming software to Google Docs. If the government required that building contractors install bugs and hidden cameras in every home or apartment, people would object strenuously even if the government assured them that the bugs and cameras would only be turned on if the government obtained a warrant.

Building a backdoor greatly lowers the costs of routine surveillance. In the pre-digital world the government had to decide whether the cost of a wiretap or a surveillance stakeout was worth the manpower and the expense. When government builds surveillance into digital communications systems, the cost of surveillance, including unnecessary surveillance, declines rapidly, so it’s reasonable to expect that there will be more of it.

The same is true of proposals to require that broadband providers install filtering systems or deep packet inspection systems to look for contraband intellectual property. Once these facilities are built into a system, they greatly reduce the costs of blocking, filtering, and censoring. Designing an infrastructure in this way shifts the cost of surveillance and censorship away from government and onto citizens.

The First Amendment is an information policy for democracy, but it is only one information policy among many. It needs the assistance of an infrastructure to make good on its promises. The fight over free speech today, around the world, is a fight over how that infrastructure is designed and implemented. If we want to preserve a free Internet, we must have networks that cannot easily be abused in the future. We must design democratic values into the infrastructure of free expression if we want an infrastructure that protects democracy.

Older Posts
Newer Posts