Many terrorist groups share a common goal with mainstream organizations and institutions: the search for greater efficiency through the Internet. This pursuit of on-line efficiency has spawned a First Amendment dilemma. The Internet’s ability to link geographically dispersed individuals to changing data without the filtering provided by traditional media is a substantial asset for domestic and transnational networks, from violent white supremacist groups to Al Qaeda. However, much of the Web’s terrorism-related content, including the abstract advocacy of violence, has manifest value as an exercise of free speech. Modern First Amendment jurisprudence protects extreme speech as a form of engagement in the polity, and responds to fears that unleashing the government on speakers will permit the targeting of groups outside the perceived mainstream. Nevertheless, some Internet communications intended for the mobilization and implementation of violence should forfeit protection. The difficulty lies in drawing lines that reach Internet communication the state can legitimately prohibit without chilling protected speech. Theorists of the Internet would be the logical candidates for resolving this dilemma. Unfortunately, most Internet theorists have touched on terrorism in ways that are perfunctory or incomplete. Even more seriously, many theorists argue for a descriptive model of Internet Exceptionalism, premised on qualitative distinctions between the Internet and earlier media, that would result in either over- or under-regulation of Internet communications. The Internet Exceptionalist model has produced two groups that draw sharply different normative conclusions: the celebratory and cautionary schools. Celebratory and cautionary approaches both invoke two attributes of the Internet: simultaneity and resistance to mediation.
Simultaneity leverages the Internet’s speed and flexibility to allow people and groups to communicate readily and rapidly with others around the globe. Lack of mediation allows Internet users to articulate their perspectives and plans without the filters that shape messages in traditional media. Celebratory commentators laud simultaneity and resistance to mediation as virtues that boost creativity, enhance the flow of information, and promote interaction and dialog. Cautionary scholars argue that the Internet can polarize populations and replace democratic discourse with the mere aggregation of consumer preferences.
The invocation of simultaneity and absence of mediation has historically been a staple of the debate about new media, newcomers to America, and the application of the First Amendment in times of crisis. The World War I cases such as Schenck, Abrams, and Frohwerk applied variants of the “bad tendency” test, which allowed the government to punish speech that tended to entice illegal acts, to limit the spread of dissent over America’s involvement in World War I. Government repression stemmed from concern about the simultaneity with which dissent could proliferate, aided by then relatively new technologies such as telegraph cable, motion pictures, and direct mail. Doubts about the loyalty of America’s immigrant and working class population, perceived as simultaneously physically present in America but linked to enemy nations abroad, also played a crucial role.
Holmes’s clear and present danger elaboration on the bad tendency test, with its compelling metaphor of fire in a crowded theatre, illustrates the speed of dissent in a new technological age, and its immunity from tempering influences. Subsequent use of the clear and present danger test in McCarthy Era cases such as Dennis v. United States suggests an increasing concern over technological innovation and simultaneous loyalties. Even in the modern free speech area shaped by Brandenburg v. Ohio [insert cite], supposedly “neutral” legislation like the statute criminalizing the burning of draft cards upheld by the Court in United States v. O’Brien seemed driven by distaste for the mass-media spectacle of draft-eligible young people burning their draft cards in opposition to the Vietnam War.
In the Internet context, the cautionary view depicts on-line communication as a volatile, combustible space resembling Holmes’ fire in a crowded theatre. This view could trigger the return of the repressive bad tendency test through the interpretation of neutral statutory provisions, such as the prohibition of material support of terrorist organizations or conspiracies. It would also legitimize government use of such provisions to marginalize particular groups, such as Arabs, South Asians, or Muslims, much as the World War I prosecutions targeted immigrants from Eastern and Central Europe. In addition, a cautionary view would also permit suppression of on-line materials dealing with terrorist tactics or scientific knowledge, such as the ingredients for poison gas, that serve First Amendment values precisely because of their troubling content. On the other hand, accepting the celebratory view, with its pervasive skepticism about government regulation, could facilitate use of the Internet to build capabilities for collective violence. It could also permit domestic networks to use the Internet to threaten private persons, as in the case of the “Nuremberg Files” where individual doctors performing abortions were labeled as war criminals.
Responding to the inadequacies of Internet Exceptionalism, this article offers a participant-centered analysis of terrorism and the Internet under the First Amendment. The participant-centered view draws on the work of two thinkers, Louis Brandeis and Hannah Arendt, who argued both that engagement in civic discourse is crucial to democracy and that new technology poses particular challenges for self-governance. To promote participation, the participation-centered approach examines how proposed measures for regulating terrorist material on the Internet affect the level of civic engagement of both speaker and audience.
For the participant-centered view, civic engagement and the right to privacy are complementary. Efforts to promote participation should include protection of privacy as a refuge from what Arendt describes as the harsh and sometimes threatening “glare” of the public realm. Under this view law should address the Internet’s power to both connect and intrude.
To this end, the participant-centered account acknowledges that the Internet’s reach holds unprecedented potential for global conversations and connections. In times of crisis, governments often regard this potential as a threat, and target the participation of newcomers or subordinated groups in new media. In response, courts should interpret statutes, including purportedly “neutral” enactments that impose incidental burdens on speech, to preserve participation and guard against targeting of perceived outsiders. Courts should also scrutinize attempts by government and corporations to limit participation by monopolizing knowledge deemed too risky for distribution on the Internet. However, some Internet regulation is necessary to preserve participation from the chill of threats and to reach concerted illegality outside the realm of civic engagement. Law should curb terrorist networks’ use of the Internet for communication about pending operations and acquisition of new resources for violence. Law should also limit the use of the Internet to intrude on privacy for purposes of intimidation.
The article is in six Parts. Part I discusses the history of new media, newcomers, and the First Amendment. Part II discusses the development by courts and commentators of cautionary and celebratory perspectives on Internet Exceptionalism. Part III analyzes the Internet’s implications for terrorist organizations and government policy after September 11, and discusses the perils of both over- and under-regulation of the Internet. Part IV presents the participant-centered view as an alternative to Internet Exceptionalism’s over- or under-regulation. Part V applies the model to specific examples, including the prohibition of Internet-based “material support” to terrorist organizations and conspiracies, the publication on the Web of terrorist manuals or dangerous scientific processes, and the use of the Internet to circulate “true threats” against groups such as abortion providers. Part VI considers alternatives to the participant-centered model.
The history of restrictions on free speech, like the history of free speech itself, is a story of legal attempts to cope with the rush of the “new”. The “new” includes new technology, and also new demographic developments promoted by immigration. Courts, legislators, and the executive branch, as well as commentators, have focused on two overlapping phenomena linked with the new: simultaneity and absence of mediation.
Simultaneity involves the cultural ability to identify, empathize, and coordinate with other persons sharing cultural, national, or ideological backgrounds despite geographic dispersion. For the scholar Benedict Anderson, this conception of simultaneity emerged after the discovery of the “New World” of the Americas, fueled by “an accumulation of technological innovations” in shipbuilding, navigation, and printing. Emigration from Europe to the Americas accelerated this notion of simultaneity, as settlers in the New World cultivated parallels with their countries of origin.
Absence of mediation refers to the ability to bypass institutions that influence and temper thinking, feeling, and acting. Philosophers and social commentators confronting modernization in the late 19th Century were concerned that traditional sources of mediation would break down as more people concentrated in urban centers to take advantage of employment generated by technological innovation. Mobility of people was disconcertingly matched by mobility in ideas, goods, and capital, which could move more quickly than the mediating structures that earlier had sufficed to contain them. Absence of mediation also reflects demographic movements. As Northern Europeans established their dominion in the Americas, they became concerned that immigration of poor people and people from Southern and Eastern Europe would result in resistance to established cultural and national traditions. Regulation of new media occurred at the juncture of technological and demographic change.
To see the impact of these conceptions of simultaneity and absence of mediation on the development of doctrine and attitudes toward the First Amendment, it is useful to consider the history of government regulation of media and technology in the decades leading up to World War I. Many commentators of the period saw the need for greater government regulation to prevent abuses of an untrammeled market affected by new technology. This trend affected books and printed matter, such as newspapers, which became cheaper because of falling paper prices and new printing technology, and more methodical and “scientific” marketing. Devices such as the telegraph and the telephone enhanced simultaneity and resistance to mediation, and the railroads facilitation in the delivery of mail meant that media were “able to move... messages more quickly than in the past, breaking down barriers of distance and tying markets and other institutions more closely together.”
For progressives, decisions to leave ownership of the telegraph and the telephone, as opposed to the mails, in private hands, created concerns about overreaching and actions against the public interest. Supporting greater regulation of new technology through the common law, Brandeis and Warren wrote about the role of these technological developments in making individual and private information readily available with a pace and absence of context hitherto unprecedented, thereby threatening privacy, dignity, and attributes of personhood. In sum, for many in the Progressive Era, simultaneity and the erosion of mediating structures in the new media created not only opportunities, but also risks requiring regulation.
In the early years of the 20th Century, the burgeoning popularity of motion pictures triggered concern because their makers and distributors, as well as their audience, included many immigrants. The Supreme Court’s insistence in Mutual Films Corp. v. Industrial Com. of Ohio that movies were not entitled to First Amendment protection exemplified this trend. The Court cited what it viewed as the potential corruption of entire families, both adults and children, through the movies as a basis for this susceptibility to regulation. For the Court, motion pictures were unmediated “spectacles... representations of events.” However, moviemakers intended them not as expressions of mediated opinion, but instead as “a business pure and simple, originated and conducted for profit.”
The lack of mediation the courts perceived in movie production was echoed in the effect of movies on their audiences: motion pictures were “capable of evil... [and] insidious... corruption” because of their “attractiveness and manner of exhibition.” The unmediated character of the movies’ content and audience, with children receiving messages directly, without the tempering influence of parental translation, exacerbated the risk of simultaneity. Movies could quickly and irreversibly inculcate audiences with undesirable sentiments and habits, much as diseases might spread through tightly packed immigrant communities.
Concern over the pernicious combination of new media and immigration attained currency not because of proof of causation, but because of the power of metaphor. When future Justice of the Supreme Court John Hessin Clarke noted in 1901 that he detested the “‘philosophy’ of the ‘reds’... [that] should find no room for culture or spread” in American society, the analogy to newly discovered bacteria was clear. The metaphor of combustibility was also pervasive. Delivering explosives in the mail and in the public square was a tactic of violent political groups of the time. Progressives had long perceived cities, with their substantial immigrant communities, as incubators of corruption. Just as an explosion occurs because an explosive agent comes in contact with susceptible material, extreme speech under this metaphor reacts when it comes into contact with congested immigrant urban communities lacking personal, social, or institutional inhibitions. Explosive violence results. Perhaps the Progressives’ interest in science and technology made them more willing to employ images drawn from science, such as injection of poison and the chain reaction, to justify greater government authority over the media and speech.
Progressive anxiety about the interaction of new media and immigration came to a head during the American involvement in World War I. The Wilson Administration feared the role of modern technology such as submarine cable in spreading information abroad about American dissent, and so encouraging wartime opponents. For Wilson, the technologies of cable, inexpensive printing, and efficient mail delivery provided the means for “hyphenated groups” linked in real time with their ethnic brothers and sisters in enemy nations. Opponents of the war, who believed that American involvement aided imperialist regimes and risked American lives, sometimes used technology that was sophisticated for its day. One group used such a technology – targeted direct mail – to send leaflets opposing American involvement in World War I to 15,000 inductees. In response, Wilson invoked images of contagion and toxicity, warning against those who would “inject the poison of disloyalty into our own most critical affairs.” To deal with the threat, the Wilson Administration developed a comprehensive program of monitoring, surveillance, and censorship.
The Wilson Administration also wished to limit the impact of “propaganda” spread by Germany and alleged German sympathizers in the United States. To counter this propaganda effort, George Creel, who ran Wilson’s wartime information department, the Committee on Public Information, started his own propaganda machine. Creel took an active interest in the very new medium of motion pictures, and even provided scripts sending appropriate wartime messages to the motion picture industry.
At the same time, the bad tendency test that dominated judicial treatment of First Amendment issues reflected the scientific rhetoric of simultaneity and unmediated risks. Here, too, perceptions of the risks of technology and immigration influenced outcomes. The court’s basic premise stemmed from the law of seditious and criminal libel, which allowed government to punish speech that possessed a tendency to encourage illegal acts. As interpreted by the courts, the bad tendency test relied on theories of group psychology that viewed crowds as easily susceptible to suggestion. The Court was willing to defer to legislative views that the proliferation of media advocating offensive positions created a tendency to disobey the law. In fact, concern over the tendency of immigrants to exploit the volatile moods of concentrated urban populations helped push the Court to uphold the conviction and deportation of John Turner, an English anarchist, under the Alien Immigration Act of 1903, which excluded from admission to the United States “anarchists, or persons who believe in or advocate the overthrow by force or violence of the Government of the United States or of all government or of all forms of law.” For the Court, Turner’s deportation had no implications for free speech, but rested instead on Congress’s ability to limit “undesirable additions to our population.”
The dissent of political radicals, many of them immigrants, to American involvement in World War I heightened judicial reliance on images of simultaneity and unmediated risk. In Frohwerk v. United States, Justice Holmes wrote for the Court upholding a conviction under the Espionage Act based on distribution of a circular protesting the war, deferring to the government’s view that “circulation of the paper was in quarters where a little breath would be enough to kindle a flame.” Most famously, in Schenck v. United States, Holmes upheld a conviction under the Espionage Act for the targeted direct mail campaign described above, even though the defendants never urged the inductees to refuse to report. Concerned that the bad tendency test was unduly restrictive of free speech, Holmes nevertheless argued that in war-time the stakes might be so high that dissent, permissible at other times, would pose a “clear and present danger” of interference with the war effort. While the “clear and present danger” test stressed the simultaneity of the speech and the evil it would produce, it also captured the unmediated nature of the risk that communities of immigrants and others open to suggestion would respond to dissent not with deliberation, reflection, or debate, but instead with lawlessness. Illustrating his point, Holmes again invoked the combustibility metaphor, arguing that the First Amendment would not protect a person in “falsely shouting fire in a theatre and causing a panic.”
However, as Holmes argued a few months later, the tropes of simultaneity and absence of mediation also pose risks because of their ability to justify government overreaching. In Abrams v. United States, a case dealing with Jewish immigrants who circulated pamphlets, one written in Yiddish, criticizing America’s efforts to send troops to Russia after the Russian Revolution, the majority followed the bad tendency test, but Holmes dissented. Holmes’ view here, influenced by Prof. Chafee of Harvard, Judge Learned Hand, and Justice Brandeis, was that ideas, including “opinions that we loathe,” would compete for public acceptance. Action would not necessarily follow thought mechanically, but would result from a more mediated, deliberative process. For more extreme ideas, dissipation and dilution would be the most likely outcome of prolonged exposure.
Unfortunately, new crises after World War II again precipitated resort to Holmes’s “clear and present danger” test as a vehicle for addressing the simultaneity and unmediated risks posed by new technology and immigration. In Dennis v. United States, a case upholding the Smith Act, which prohibited membership in the Communist Party, the Court again resorted to the combustibility image, justifying its decision by alluding to the “inflammable nature of world conditions” and noting that, “[i]f the ingredients of the reaction are present, we cannot bind the government to wait until the catalyst is added.” The Court’s language invoked not just a standard explosion, but the nuclear variety that had become all too familiar since the conclusion of the war.
Picking up the absence of mediation theme, Justice Frankfurter in his Dennis concurrence also noted, in an echo of the Court’s earlier Mutual Film case, that government could regulate new technology such as radio far more readily than traditional hard-copy sources, quoting a First Amendment scholar’s argument that the radio “is not engaged in the task of enlarging and enriching human communication.... [but] in making money.” Focusing also on the descendants of immigrants who purportedly formed the backbone of the Communist Party in the United States, Frankfurter noted the threat of simultaneity by citing the Chinese Exclusion cases that held that such immigration could constitute a particularly insidious form of “foreign aggression and encroachment.” The opinion of the Court echoed this theme, noting the danger posed by “countries with whom petitioners were... ideologically attuned.”
In 1969, the Court finally drained the vestiges of the bad tendency test from the standard for incitement, joining a demanding version of Holmes’s clear and present danger test with a subjective element involving the speaker’s intent. Under this “modern” understanding of the First Amendment, statements preceding acts of violence can constitute incitement only if the speaker intended violence to result and violence was an imminent and likely consequence of her statements. Yet, Brandenburg reflects the Court’s understanding that in times of crisis government will seek to exaggerate simultaneity and minimize the existence of mediation.
However, the consensus surrounding the modern First Amendment does not herald the end of history for the jurisprudence of free speech. Doctrine has carved out a number of areas where the government can regulate communication, making Brandenburg less central. Moreover, in the shadow of September 11, law arising from new media such as the Internet retains the potential to disrupt the modern equilibrium.
As a new medium, the Internet inherits the concerns about simultaneity and unmediated risk raised by earlier generations of media innovations. At the same time, the growth in concern over terrorist networks since September 11 magnifies apprehensions about exploitation of the Internet’s capabilities for violent purposes. This convergence of concern requires a nuanced treatment of the interaction between the Internet’s capabilities and the nature of transnational terrorist networks. An unrestrained celebratory view of the Internet may discount synergies with terrorism, leading to under-regulation of the Internet. However, an overly cautionary approach may neglect the substantial benefits of the Internet for democracy, and lead to over-regulation, including the targeting of immigrants characteristic of the World War I cases. This section analyzes theoretical and judicial conceptions of the Internet. In the process, it prepares the ground for a nuanced examination of the relationship between Internet capabilities and terrorist violence.
Many prominent commentators embrace a view we can call “Internet Exceptionalism,” which stresses distinctions between the Internet and earlier communications media such as books, newspapers, and broadcasts. Internet Exceptionalists cite a variety of the Internet’s attributes, centering on the same simultaneity and absence of mediation that preoccupied courts and commentators with regard to previous technological innovations. For example, Internet Exceptionalists note how the Internet enhances consumers’ ability to assemble an individualized collage of information from a variety of specialized and partisan sources, without the intercession of an intermediary, such as an editor, who may offer a broader perspective.
This descriptive view of the Internet does not lead to a uniform normative outlook. Within the Internet Excepionalist school we can identify two principal points of view. The first we can call the celebratory approach. Commentators taking this view celebrate the attributes of the Internet and argue that the technological and user interface attributes of the medium create a new imperative of user freedom. For these commentators, government regulation stands in the way of the full celebration of the Internet’s advantages for the dissemination of information and the cultivation of innovation. Lawrence Lessig, for example, argues that the application of copyright and other conceptions of intellectual property to the Internet can foster unhealthy government and corporate control. Indeed, for celebratory commentators, government regulation is not only normatively inappropriate, it is futile, given the relatively free-flowing nature of current Internet architecture.
In contrast, another, perhaps less populous school of Internet Exceptionalists takes a cautionary view that accepts many of the premises of the celebratory school but draws far more dire normative conclusions. This school would argue that a greater government role is necessary to curb the potential of the Internet for polarization and intellectual and civic fragmentation. One theorist has acknowledged that, “[p]eer-to-peer networks... don’t do much to build new communities... few systems allow fans to deliberate about music or make connections and decisions collaboratively... [o]ften you find what you already know about.” On this cautionary view, mediation, although occasionally oppressive and inconvenient, can also allow consumers of information the opportunity to try something they might otherwise have rejected as inconsistent with their preconceptions. In the process, people’s minds can change, their perspectives can evolve, and group-think and polarization become a little more difficult. This is the goal of core mediating institutions in a liberal society, such as universities. Serendipity – the benefits of a surprise encounter with the unexpected or underestimated – is one of the virtues of this idea. Unfortunately, since many Internet searches reflect the predispositions of the user, cyberspace in practice offers “little opportunity for serendipity.”
The celebratory and cautionary Internet Exceptionalists focus on different bugbears: government and corporate control for the celebratory scholars, and polarization for the cautionary school. The literature is rich, but an integrated approach addressing each of these perils has been elusive. As the next subsection demonstrates, courts confronting individual cases have similarly produced sharply varying results.
Courts have been mixed in their treatment of the themes of simultaneity and absence of mediation on the Internet. Some decisions have taken a cautionary view, stressing the need to cabin the power of Internet to protect property rights or safeguard the public. Other decisions have veered toward the celebratory view, asserting that the speed and flexibility of the Internet make regulation futile and sometimes unfair. Decisions regarding intellectual property, on-line pornography, and criminal law all reveal this split.
For a classic example of the cautionary view, consider Universal City Studios v. Corley, in which the court upheld as an incidental burden on speech a statutory prohibition on distributing software for the primary purpose of circumventing restrictions on viewing digital versatile disks (DVD’s). The court asserted that codes distributed in this manner “instantly cause a computer to accomplish tasks and instantly render the results...available throughout the world via the Internet.” Another court, in DVD Copy Control Ass’n v. Bunner disagreed, arguing that the instantaneous nature of Internet transmission meant that the value of the plaintiffs’ asserted property right had already been compromised many times over prior to the court’s decision. Therefore, according to the court, there was no irreparable harm, making an injunction unnecessary and punitive.
In the on-line pornography area, the Supreme Court has adopted a celebratory view when dealing with content-based regulation. For example, consider the Supreme Court’s recent decision in Ashcroft v. ACLU. In this case, the Court held that a restriction on Internet pornography harmful to minors would not survive First Amendment scrutiny. The majority observed that the sheer number of hosts worldwide made content-based regulation at the source difficult, if not impossible. The Court also declined to view children’s exposure to pornography on-line as unmediated, citing the potential for evolving technology such as parentally-installed filters to provide checks at the recipient end without chilling speech at the source.
However, the cautionary view is dominant when the Court can classify the measure at issue as outside the realm of content-regulation. In United States v. American Library Association, for example, the Court upheld a measure requiring libraries to install filters to block pornography on the Internet as a condition of receiving federal financial assistance. Here, the Court was so concerned about the prospect of unmediated access to the Internet that it downplayed the overbroad character of the filters required, which adversely affected access to a range of non-pornographic content. The Court also dismissed plausible alternatives to the filter requirement.
In the criminal law area, courts sometimes recognize that the accelerated and unmediated world of the Internet also creates a far greater risk of inadvertent or ephemeral actions leading to criminal liability. A hard copy or even speech in real-time requires far greater deliberation, increasing the likelihood that acts considered orders, solicitations to perform criminal acts, incitements, or true threats will be premeditated. In contrast, the Internet’s absence of mediation can become a trap for the unwary. For example, in United States v. Alkhabaz, a student posted a graphic rape and abduction fantasy about a person bearing the name of a female dorm-mate to a message board, and was subsequently charged with making a threat of violence using instrumentalities of interstate commerce. Holding that the intent to make the female student aware was a crucial element of the charge, the Sixth Circuit dismissed the charges. As the Sixth Circuit noted, there was no evidence that the defendant wished to make the female student aware of his fantasy. Before the Internet, the defendant may have circulated his story in hard-copy form to persons sharing his admittedly troubling interests. He would have understood that sending a copy to the female student could have yielded tort or even criminal liability, and would have conformed his conduct accordingly. In this way, the absence of mediation on the Internet – the breaking down of barriers between intended and unintended audiences – can lead to overreaching in the regulation of Internet speech and unfairness to defendants, unless courts pay careful attention to context.
Here, too, however, courts sometimes embrace a pro-regulation view as a means of controlling the Internet’s new fora. For example, a court has held that a government agent need not have probable cause to view conversations in a chat room and subsequently pose as an individual interested in child pornography to secure evidence against another chat room visitor. A court could instead view a chat room as a forum whose openness some participants may not fully comprehend, leading to confused expectations. By declining to suppress the agent’s chat room surveillance, the Court may have encouraged intrusive law enforcement tactics.
In sum, courts move from a celebratory to a cautionary view with little consistency or overall analytical framework. As the Internet matures, judicial consistency may similarly develop. However, the risk is that the evolution of case law will suffer permanently from this patchwork beginning.
Analysis of the interaction of terrorism and the Internet also reflects the perils of dichotomies. The Internet offers significant synergies to entrepreneurs of violence. However, it also holds out the hope of enhanced access for subordinated groups. Neglecting either dimension of the Internet’s capability is dangerous. To establish the synergies between the Internet and terrorist activity, this Part turns first to the absence of mediation in cyberspace, and secondly to the impact of simultaneity. It then examines the countervailing potential of the Internet for building democracy through dialog. Finally, this Part sketches some challenges at the post-September 11 intersection of the Internet and terrorism.
The absence of mediation on the Internet can promote polarization and permit consumers to avoid the unexpected teachable moment. On a technical level, the absence of entry barriers to Internet communication also facilitates the operations of terrorist groups. I discuss each in turn.
Lack of mediation is a key ingredient in the production of polarization and concerted violence against innocents to achieve political, cultural, or social aims. This concerted violence, typically, if perhaps too glibly called “terrorism,” can result from the actions of states or private groups of domestic or transnational origin. From the genocide of Rwanda and the Sudan to the efforts of white supremacists and anti-abortion extremists in the United States, governments or other organizations practicing concerted violence against innocents often emerge from shared sentiments of inequity or displacement. Exploiting these sentiments, entrepreneurs promote the appeal of exclusionary images of “authenticity.” Substantively, these exclusionary images thrive through comparisons with “inauthentic” others, defined through traits such as race, ethnicity, nationality, or even occupation. Procedurally, this stark view of authenticity entails the rejection of introspection, dialog, and debate as means for peacefully coming to terms with difference. By stigmatizing others and discrediting dialog, authenticity entrepreneurs designate violence as the tactic of choice.
The unmediated nature of the Internet exacerbates polarization. Most media, including broadcast and paper sources, have some form of mediation mechanism between their content and their audience. In a newspaper, for example, the editor reviews the content, often seeking a range of stories and a range of views. This may not be the case for a web-site with minimal physical infrastructure, investment, or stake in the community. Websites face even fewer constraints than most traditional media to the propagation of ideologically homogenous content, and often lack the legal staff or institutional culture to curb rumor, innuendo, and libel. Web sites catering to particular niches tend to attract an audience of persons who already agree with the extreme opinions featured on the site. Homogeneous groups are likely to perceive both identity and grievances in a far more polarized fashion, thereby promoting recruitment for violent acts, and impeding the interchange and deliberation crucial to democracy.
The Internet’s lack of mediation also facilitates rapid revision of web-site content in a fashion that can assist terrorist organizations. A site that contains specific or personal information about potential targets of terrorist attacks, including buildings, installations, or people, can readily update that site as new information becomes available. While people or groups can alter patterns of behavior to deal with other kinds of public threats, the flexibility and modifiability of Internet communications mean that those behavior alterations can be quickly passed on to persons who might be committed to executing attacks. In addition, the flexibility of the Internet makes it easy to shut down web-sites and set up new ones to avoid detection.
The speed of Internet communication offers the prospect of connections across the globe, enhancing ability to communicate operational information regarding terrorist activities. Many violent networks are geographically dispersed. Transnational groups such as Kach, Hamas, and Al Qaeda raise money and recruit operatives on an international scale. Domestic networks such as extremist anti-abortion or white supremacist organizations also have members in far-flung locations. The qualities of the internet make it perfect for such dispersed communication. Participants in a network can log in from locations all over the map, and gain access to information more efficiently than is possible with traditional media.
Complementing this notion of simultaneity is the Internet’s facilitation of asynchronous interactions. Visitors to a web-site need not communicate in real-time; instead they can send messages that will be read and responded to at the convenience of others, but without the transmission delays of older modes like “snail mail.” The Internet also facilitates the automatic and user-accessible archiving of material awaiting the visitor’s attention. Coupled with the Internet’s geographic reach, the temporal versatility of the Internet is a boon to terrorists.
The simultaneity and absence of mediation on the Internet also have salutary consequences. For example, the Internet’s speed and geographic reach can enable the connection of diasporated communities on a regional and global basis. Consider here Edward Said’s account of Palestinian refugees in a camp on the West Bank setting up the “Across Borders Project,” which used the Internet to connect residents of the camp with residents of other Palestinian refugee camps in Lebanon, Jordan, Syria, and Gaza. The Project allowed residents to exchange views on important issues of relevance to the Palestinian community, such as the future of the peace process. One might not agree with all, or indeed any of the opinions articulated in such an exchange. One can hardly gainsay, however, the value to ideals of political participation represented by this capability.
Similarly, the Internet can provide intellectuals and activists with avenues to bypass the status quo-centered mediation offered by most mainstream media. When major media offer privileged access to Bush Administration officials promoting an intervention in Iraq justified by faulty intelligence, intellectuals and activists can turn to the Internet to communicate an alternative vision. Some countries, including Myanmar, Saudi Arabia, and the People’s Republic of China, restrict Internet access for their citizens out of fear of such alternatives. The United States, however, should model a more robust commitment to democratic debate.
In an uncertain world, the interaction between fear of new technology like the Internet, and apprehension about newcomers to the United States, labeled as suspected terrorists, prompts distortions in First Amendment doctrine. A cautionary approach to the Internet might increase the risk of a return to the bad tendency test, perhaps camouflaged in the rhetoric of content-neutrality. A prosecution that resulted in the deportation of a Saudi national with a student visa, Sami Al-Hussayen, for coordinating extremist web-sites raises this concern, as do subsequent cases. Similarly, the post-September 11 climate may further incline courts and legislatures to curb the publication of information about terrorist tactics or scientific processes, even when this information contributes to public debate. However, in this climate the use of the Internet by home-grown hate groups to intimidate opponents may not receive comparable attention, both because such groups do not fit the terrorist “profile” and because celebratory scholars may view the ubiquity of hate on the Internet as neutralizing the intimidation conveyed. The potential result is the worst of all worlds, as skewed perceptions generate both over- and under-regulation.
A contextual view that will both respect free speech and limit the use of the Internet to cause collective violence can help resolve this dilemma. Meeting this need, a participant-centered view echoes the position of thinkers and activists that participation in public discourse is a central value in a democracy. This view, developed here through the work of Louis Brandeis and Hannah Arendt, preserves participation against the totalizing force of the state, and the state’s tendency to stigmatize outsiders. It identifies the practice of concerted violence not as an aspect of participation, but as an activity that undermines deliberation and discourse. It also preserves private space as a refuge and respite from the rigors of participation, in the face of the Internet’s invasion of the private realm.
Brandeis and Arendt reflect the focus of the participant-centered view on both engagement with and sanctuary from the public sphere. Brandeis argued in his famous dissent in Whitney v. California that “the final end of the state was to make men free to develop their faculties; and that in its government the deliberative forces should prevail over the arbitrary... the greatest menace to freedom is an inert people.” Similarly, for a civic republican theorist like Arendt, involvement in public discourse is crucial to the realization of human potential. Through sharing stories in the public realm, human beings open themselves to the realm of the unexpected, acknowledging the contingency of their own prejudices and preconceptions.
While engagement in human governance is the highest goal or pursuit for a participation-centered account, government exists in an ambivalent relationship with participation. Government, properly understood, needs participation to develop new ideas and challenge old habits. However, government officials often seek ways to domesticate or manage participation, robbing it of the alliance with the unexpected that makes it a central expression of what it means to be human. Seeking to manipulate public opinion, governments frequently alter facts and massage the truth. Governments also seek to stigmatize dissenters, casting them as outsiders of questionable loyalty. Brandeis, for example, feared that in times of crisis the simultaneity and unmediated nature of government repression – fostered by “hysterical, unintelligent fear” – would distort democratic deliberation and chill or punish participation.
To reinvigorate civic engagement, both Brandeis and Arendt looked to outsiders. Both thinkers believed that immigrants strengthened democracy by bringing new ideas and renewed commitments. Each identified the labor movement as a central engine of participation for the hitherto excluded and viewed repression of the movement as a danger to participation-centered ideals. Arendt also praised the role of the anti-war and civil rights movements of the 1960's in making the government accountable.
Although civic republican theorists’ insist on the need for dissent, they tend to view organized violence against others as undermining engagement. For Arendt, violence reflects a homogenized viewpoint, certain of its conclusions, and focused too often on the mechanics of death and pain. Such a view corrodes the commitment to transparency – to letting a concept be seen from all perspectives – that animates participation. Violence inevitably skews discourse among both perpetrators and survivors, creating habits of insularity and fear that inspire further destruction.
The participant-centered account also argues that private forces can undermine democracy and engagement as severely as a repressive government. Indeed, under the participant-centered account, there is often a continuum between repressive private and public forces. For Brandeis, of course, corporate power was a profound threat to civic engagement and accountability. Arendt also lamented the rise of corporate power as displacing the public realm. In addition, Arendt devoted substantial attention to charting the rise of polarized private groups, such as the Nazis and Fascists, who used violence and intimidation to take over governments, converting them into repressive and sometimes genocidal regimes.
As a bastion against oppression, civic republicans also value a zone of privacy. For Brandeis, a space “to be let alone” ensured the flourishing of each person’s “spiritual nature, of his feelings and intellect.” It also ensured the vitality of the public sphere, since the “intensity and complexity” of engagement in public matters required periodic opportunities for “retreat from the world.” Arendt for her part wrote powerfully about the need for a “place of one’s own.” For Arendt, this space “offers the only reliable hiding place from the common public world.” Without this sanctuary, public engagement becomes “shallow” and trivial.
The participant-centered view argues that modern media such as the Internet can frustrate participation by destroying privacy. Brandeis was concerned that the virtually “instantaneous” speed of modern “inventions... business methods... [and] mechanical devices” would “invade[ ] the sacred precincts of private and domestic life.” Arendt described the corrosive effects of the exposure of private persons to the ministrations of mass media, observing that, “[f]ame penetrates the four walls, invading their private space, bringing with it... the merciless glare of the public realm, which floods everything in the lives of those concerned.” In addition to these direct effects, the marketing of unmediated trivialities about personal affairs would crowd out matters of genuine public importance, undermining more difficult but necessary discourse and confirming people’s pre-packaged opinions.
Even more ominously, for Brandeis and Arendt, emerging technology spawns more intrusive government methods of surveillance, such as the wire-tapping that Brandeis criticized as an unreasonable search and seizure in his dissent in Olmstead v. United States. Technological advances might ultimately offer the government clandestine access to “the most intimate occurrences of the home,” and to “unexpressed beliefs, thoughts, and emotions,” enabling government to more tightly manage the people’s unruly urge to participate. Technology could enable the government to more effectively dominate discourse, leveraging new media and techniques of persuasion to obscure or erase inconvenient facts about the world.
However, the participant-centered view would also value the countervailing capability that technology offers for dissenters. In this vein, Arendt noted that governmental attempts to erase the past will fail so long as government cannot “wield power over the libraries and archives of all countries of the earth.” The virtually limitless archives of the Internet thus mediate the government’s attempts to achieve dominion over data, sustaining alternative resources for participation. In addition, the decentralized spaces created by the Internet have the potential to generate new kinds of civic interaction and fresh perspectives on perennial problems.
The participant-centered model also captures the two central elements of concern about the Internet: lack of mediation and simultaneity. Those acting as agents of terrorist organizations or operations are not only far less susceptible to mediation, but more committed to preventing others with whom they interact from obtaining access to mediated views. This is the danger of conspiracy and agreement that has typically made such operational terrorist “networks” unprotected under the First Amendment. Simultaneity is important in the privacy context, where non-public persons can be intimidated by the Internet transmission of continuously updated information accessible to network members that have a track record of violence directed against such private persons. The power of this threat potentially serves to intimidate people into silence, disabling the “more speech” antidote outlined by Brandeis in Whitney.
The participant-centered view protects participants in public discourse, including those participants testing or exceeding the limits of polite discussion. However, the participant-centered view denies protection to willful participation in agreements or conspiracies that set the stage for escalating violence. The absence of a “more speech” antidote shuts down mediating mechanisms for such participants. At the same time, the participant-centered view protects individuals engaged in high-risk activities deemed important to democratic life, including witnesses in high-profile criminal trials and persons engaged in vindicating reproductive rights. By attaching special protection to certain roles important for democratic participation, and simultaneously depriving protection to roles that chill the participation of others, a participant-centered view emphasizes the social nature of both protected and unprotected activities.
An initial step in a participant-centered approach is to identify three different ways in which speech can facilitate violence by terrorist groups or institutions in the Internet context. This analysis sets out three modes of crime-facilitating speech or communication: information and affiliation, mobilization, and operation. It acknowledges, however, that these categories are unstable. Moreover, certain mainstays of typical First Amendment analysis, such as the requirement that harm be imminent to criminalize speech that is otherwise protected and the general view that First Amendment protection attaches to a greater degree to public rather than private speech, do not necessarily hold true in the context of the Internet and terrorism.
A participant-centered view accords affiliational and informational speech the highest level of protection. This speech involves a statement of political support for a particular group or cause, no matter how abhorrent to the majority, or a description of a group, event, or thing in the world. The modern First Amendment generally protects abstract calls for violence captured in either the affiliational or informational mode, as in, “I support the overthrow of the United States government,” or, “X supports the use of force to effectuate the revolution.” Such statements, however extreme, are tied to participation in public discourse, as long as they retain a general focus. Speech of this kind would include on-line discussions of the status of violent jihad in Islamic doctrine.
Extreme speech warrants protection because of its strong ontological and pragmatic links to ideals of participation. Such speech may be an element of a learning process for participants, who may test extreme rhetoric, and then determine that it does not meet their needs. Extreme rhetoric can also serve as an outlet for dissent, and as an indication of vulnerabilities for regimes that may then undertake reform. Similarly, the participant-centered approach would generally protect reports about world events, such as newspaper accounts of speeches by leaders of alleged terrorist groups, as serving similar functions. Moreover, a government with the power to deter or punish such statements of affiliation or information may use its power broadly to stifle all disagreement and target disfavored groups, chilling civic involvement. As indicated previously, this is the theory underlying the “bad tendency” test and the broad interpretation of “clear and present danger,” as endorsed by the Dennis Court.
In some cases, however, speech couched in the language of information or opinion may be operational in intent or effect. Operational communications involve commands, agreements, solicitations, or threats to perform illegal acts. Such acts, particularly those that involve violence, reflect a shift away from participation in civic discourse and a subsequent move toward conduct that is either largely self-interested or focused on undermining the participation of others.
History reveals that leaders of violent organizations often use the language of opinion or information to authorize violence. Sometimes an authorization may come in the form of a question, as when Henry II of England asked, “Will no one rid me from this turbulent priest?” Henry’s select audience of nobles responded with the murder of Thomas à Becket. More recent leaders of violent organizations have cast extortion and intimidation as the simple supply of information, tendering data about an offer their interlocutor “cannot refuse.” For example, an alleged Mafia boss on trial in New York resorted to the language of opinion, averring that he “would like to leave a receipt” for targets of his displeasure; those targets soon found themselves migrating to settings of higher (or lower) altitudes.
While most of the above examples concern individual targets, leaders of organizations with the capacity and inclination to practice violence may also have designs on collective targets. For example, leaders of such networks may target members of particular ethnic, religious, national, or even occupational groups, such as Palestinians on the West Bank, Jews in Israel, Tutsis in Rwanda, or doctors performing abortions in the United States. Here, too, leaders may couch their directives in terms of opinion or metaphor, noting, for example, that “the gates of resistance are open totally.” What separates such expressions from mere statements of opinion or sentiment is their context: the organizational structure of the group and its pattern or practice of violent acts suggests that the audience for the remarks consists of members of the group who view the expressions not as opinions, but as operational instructions.
In the context of Internet-driven collective violence, the public expression of the authorization and the occasional time lag between the authorization and subsequent violence do not vitiate the operational character of the remarks. The asynchronous nature of Internet time allows operatives to readily gain access to material on web-sites at their convenience. An operative who is inaccessible during the “real-time” delivery of the authorization of violence can readily pick up the thread at a subsequent point. Imminence also seems to be irrelevant where, as in the case of September 11, terrorist operations can take months or years of planning. From an anti-terrorist policy point of view, the relevant issue is the ongoing capability of the terrorist organization to plan and execute violence once authorized, not the time period between the authorization and the completed act.
The structure of violent networks also undermines the rationale for the protection of public speech. In the collective violence context, victims are fungible, targeted because of group characteristics such as race or ethnicity rather than individual identity. Authenticity entrepreneurs can deliver instructions to commence targeting such fungible victims more efficiently in public than in private. Committed or prospective perpetrators may also view directives on a web-site as more authoritative than dissemination in a more private medium, whether someone’s living room or a chat room on the Internet. Moreover, in an area such as Rwanda, which in the 1990's approached a “tipping point” of violence as a result of authorizations and collective appeals by “authenticity entrepreneurs,” perpetrators may also be fungible, deciding to commit violence because of opportunity rather than a long pedigree of commitment to the cause. Finally, when operatives are committed but dispersed and compartmentalized to promote secrecy, a broad public authorization is a useful starting point for planning. Once an authenticity entrepreneur issues a directive, planning can proceed in a more covert fashion, with the secret and homogenous structure of the organization shutting out mediating views.
Threats offer yet another context in which abstract justifications of violence can combine with specific, personal information about possible targets of violence to generate a message that, viewed as a whole, is operational. In such a case, a participant-centered approach would consider the impact of such hybrid speech on the participation of targets, as well as speakers. The participant-centered approach, therefore, is inherently reciprocal, protecting the communication of those who do not materially interfere with the ability of others to speak.
Moving beyond these categories, communication can also entail the mobilization of infrastructure for future illegal activity. Mobilizing activity stops short of operational action’s direct link to violence, instead building an organization’s resources and capabilities. For example, a person may seek to collect money in order to hand it over to a transnational network such as Al Qaeda, whose track record of violence has resulted in its designation as a terrorist organization. Such an individual might also knowingly manage a web-site on Al-Qaeda’s behalf, or contribute to a violent network’s inventory of information about security measures and vulnerabilities of possible terrorist targets, knowing that such information will aid the network in making final target selections.
Courts generally regard such mobilization as appropriately prohibited by government, as long as the limits are carefully tailored to avoid direct regulation of protected speech. Under this analysis, Congress can prohibit raising cash for a transnational organization with a recent track record of violence, because the transnational and horizontally integrated nature of the group makes it difficult to ensure that money goes only to nonviolent activities. Congress cannot, however, prohibit pure affiliational or informational speech in support of the group’s ends or means. Nevertheless, as we shall see in the next section, distinguishing between protected affiliational or informational statements and mobilizing conduct is not always easy, creating the risk that regulating mobilization will become a back-channel method for restricting content and singling out marginalized groups such as immigrants.
In sum, each of these categories, including affiliational, informational, operational, and mobilizational speech, bleed into each other. While each category has paradigm cases where consensus is plausible, at the margins overlaps and ambiguities persist. To consider how the participant-centered approach assists in sorting out those ambiguities, consider the examples in Part V.
Ambiguities between state power and civil liberties may be a permanent legacy of September 11. Cyberspace has more than its share of such uncertain boundaries. Legal issues surround the prohibition of Internet-based “material support” to terrorist organizations and conspiracies, the publication on the Web of terrorist manuals or scientific processes, and the use of the Internet to circulate “true threats” against groups such as abortion providers.
The participant-centered approach helps to clarify some problems associated with conduct on the Internet that might violate the prohibition on “material support” of a designated terrorist organization (DFTO) or terrorist conspiracy. The material support prohibitions are an example of government regulation that imposes incidental burdens on speech in the interest of regulating conduct. Too broad an interpretation of “material support” on the Internet could recapitulate the sorry history of the bad tendency test in a more neutral guise. Applying the participant-centered view, however, offers an exit from this difficulty.
1. Incidental Burdens Generally
Permitting government to impose incidental burdens on speech in the course of vindicating substantial government interests is both necessary and dangerous. As a general matter, incidental burdens analysis holds that the government can regulate conduct related to free speech when: (1) such regulation is content-neutral; (2) the regulation serves an important governmental interest; and (3) the regulation is narrowly tailored to serve the interest. Without some ability to impose incidental burdens, government itself would become impossible. However, the incidental burdens test, which is less demanding on government than the test for content-based speech, leaves open a back-door for government efforts to control speech.
This problem with the incidental burdens test is emphasized in the leading Supreme Court decision, United States v. O’Brien. In O’Brien, the Court upheld a statute passed at the height of protests against the Vietnam War which criminalized burning one’s draft card. The Court asserted that the government was merely trying to vindicate its interest in an orderly draft system. For the Court, the system would become disorderly if registrants for the draft or draft-eligible individuals could destroy their draft cards. For the Court, therefore, the statute was a narrowly tailored vehicle for serving a significant government interest. According to the Court, since the statute was indifferent to any political view that an individual might seek to express in burning his draft card, simply penalizing the conduct itself, the statute was content-neutral.
The same analysis figures in the more recent case of Universal City Studios v. Corley. In Corley, the Second Circuit upheld the provisions of the Digital Millennium Copyright Act (DMCA) that prohibited the distribution of software that had the ability to defeat use and reproduction restrictions on digital products licensed for sale. The court reasoned that the DMCA does not directly impinge on speech, since it does not restrict persons from criticizing the restrictions contained in licensing agreements, but simply promotes enforcement of those agreements. Thus, according to the Corley court, the DMCA is a reasonable and narrowly tailored vehicle for promoting intellectual property rights.
Commentators have vigorously criticized both O’Brien and Corley as requiring far less tailoring than the incidental burdens test seems to contemplate. In O’Brien, for example, commentators have noted that Congress had already enacted measures requiring that registrants for the draft have their draft card in their possession. The government did not demonstrate that these pre-existing provisions were inadequate to safeguard the government’s interests. Moreover, commentators assert that the timing of the statute’s enactment, after the start of protests against the Vietnam War, strongly suggests that the statute was designed to suppress speech, and was therefore not content-neutral.
Similarly, commentators have argued that Corley did not take into account that the use restrictions at issue had little relevance to the goal of preventing unauthorized duplication. Moreover, these commentators have in effect argued that the DMCA was not content-neutral, since it took sides between those arguing for corporate intellectual property rights, and consumers arguing for preserving fair use rights. In addition, the Court’s invocations of simultaneity proved too much – if the software was already widely available, enforcing the statute against the hackers would not effectively serve a government interest and would merely suppress hackers’ efforts to encourage corporations to build a better, more responsive business model.
2. Material Support, the Internet, and the First Amendment
Transported to the law and terrorism arena, material support concerns frame the analysis of the recent prosecution of Sami Omar Al-Hussayen, a Saudi national studying computer science at the University of Idaho. Al-Hussayen was arrested with great fanfare in early 2003 and charged with providing and conspiring to provide material support to terrorist groups. According to the government, the defendant provided “expert advice or assistance” by using his computer and Internet skills to set up a web-site for a group later designated as a terrorist organization, which also included a link to a web-site tied to the Palestinian DFTO organization Hamas. Al-Hussayen also posted accounts and audio files of speeches by radical Saudi clerics to his site, to chat rooms, and to e-mail lists. The government did not allege and presented no evidence that Al-Hussayen actually collected money for the purpose of conveying funds to Hamas, or held himself out as an agent of Hamas for this purpose. Indeed, the government never alleged that Al-Hussayen worked for Hamas at all. However, the government alleged that Al-Hussayen specifically intended to provide material support to Hamas and other terrorist organizations.
While a jury ultimately acquitted Al-Hussayen of the material support charges, Al-Hussayen’s victory was of limited value to the defendant. Al-Hussayen’s family left the country, and Al-Hussayen was himself imprisoned for over a year pending trial (during which time he continued to work on his degree from his jail cell), before he was subsequently deported. The government was able to proceed to trial in the case, confining Al-Hussayen for a year, because specific intent is virtually always a triable issue. Because the government continues to bring other material support cases involving the Internet, now is an appropriate time for considering whether the specific intent standard in Internet material support cases is sufficiently speech-protective.
The Al-Hussayen prosecution illustrates the risk that the material support prohibition will become the contemporary equivalent of the content-based regulation the courts permitted in the World War I cases like Schenck. One of the government’s recent theories was in effect the mirror image of Schenck: while Schenck sent direct mail to inductees, allegedly with the intent of persuading them not to serve, Al-Hussayen sent material on-line to persons interested in Islamic issues, allegedly with the specific intent of encouraging them to participate in terrorist training camps or contribute funds to terrorist organizations.
Despite the premise of judicial decisions upholding the material support statute as being content-neutral, the government’s theories nevertheless focused on the content of the material conveyed on-line by the defendant. A statement such as, “I support Hamas and I urge everyone on this list to support it monetarily and otherwise,” reflects an opinion (praise of Hamas) and advocacy of an abstract course of action (financial or other support) that should receive First Amendment protection. Prosecution for Internet speech under the material support statute in this context, far from merely imposing incidental burdens on speech, constitutes a de facto content-based classification. Permitting such law enforcement action also encourages the profiling of particular groups the government associates with terrorism, including immigrants of the Muslim faith. Moreover, prosecutions of this kind unduly discount the efficacy of mediation, such as the deterrent value of prohibiting the actual provision of material support in the form of financial assistance or other resources. Unfortunately, the court in the Al-Hussayen case failed to rule on the issue, thereby posing no obstacles to future government use of the material support statute for this purpose.
To avoid the risk of a rebirth of the bad tendency test, courts presiding over material support prosecutions involving Internet communications should require a showing that the defendant acted as an agent for a specific terrorist organization, or as a participant in a conspiracy to commit a specific terrorist act. To be an agent in this context, a defendant charged with material support of a designated terrorist organization would have to reach an express or implied agreement with decisionmakers within the organization to seek resources on the organization’s behalf. Agency could involve a defendant holding himself out as collecting money or other resources for the organization, offering advice to prospective contributors on laundering contributions to avoid detection, or providing resources such as website management directly to the organization – actions that can subsidize or facilitate violence. Proof of such activity moves the speaker’s conduct squarely into the operational or mobilizational realm, minimizing the chances that the material support prohibition will be used to chill affiliation or information-seeking.
Courts have often used a comparable approach to determine liability in cases involving speech facilitating violation of the tax laws. These cases mirror material support cases like Sami Al-Hussayen’s – in the tax context, defendants allegedly seek to deprive the government of lawful revenue, while in the material support context, defendants allegedly seek to provide revenue to unlawful organizations. In tax cases, courts have held that the First Amendment protects defendants who support the position that tax laws are unfair and that persons therefore are morally justified in violating them. Courts have often looked for some indicia of collusive activity between the speaker and the audience, including concrete assistance in the filing of fraudulent tax returns.
In the tax context, such collusive behavior lends an operational tone to the speaker’s conduct, removing it from the realm of affiliation or information. Specific collaboration between the speaker and the audience raises particular concerns about simultaneity and the absence of mediation. A visitor to a DFTO or tax-evasion website that receives concrete advice is far more likely to quickly make the decision to go forward with a specific violation, having been enabled with operational guidance facilitating such conduct. Similarly, the agreement between the speaker and the audience member makes it far less likely that other countervailing influences, including the general deterrence provided by law, will interfere with the audience member’s plans, or that the government will learn of the illegal activity. Furthermore, such concerted activity gives the speaker a stake in the transaction, making it more likely that he or she will take further actions to ensure the success of the particular mode of illegal conduct.
The above analysis would have resulted in the dismissal of the charges against Al-Hussayen, but would have permitted prosecution to go forward in United States v. Sattar. In Sattar, the government charged three individuals with violating the material support prohibition by seeking to secure the approval of an incarcerated terrorist leader, Sheikh Abdel-Rahman, for the Internet posting of a decree urging members of the Sheikh’s organization to kill Americans and Jews. The organization had been designated by the government as a terrorist organization after the leader was convicted of participating in a conspiracy to commit numerous acts of violence within the United States. The group later declared a “cease-fire.” The co-defendants allegedly sought to procure the jailed leader’s approval of the fatwah and its subsequent posting on the Internet to disrupt the cease-fire and turn the group back toward violence.
If it is true, as the government alleges, that the Sattar defendants attempted to place the incarcerated leader of the organization back in the “loop” in order to direct future violent acts, the defendants went beyond the realm of information and statements of affiliation into the operational realm of conspiracy. Although the fatwah was couched in the semantic frame of opinion, the recent history and the statement’s particular wording suggested otherwise. The evidence that the Sattar defendants acted in concert with the organization’s leader on the organization’s behalf and with knowledge of the organization’s structure and practice distinguishes the Sattar case from the independent informational activities engaged in by Al-Hussayen. The participant-centered approach’s focus on context thus holds decisionmakers in organizations accountable for statements that are the operational predicates for violent acts, while simultaneously protecting independent participants in civic discourse who merely offer information or express their affiliation, no matter how extreme.
The participation-centered approach also argues for greater protection for the on-line publication of terrorist techniques and scientific methods. Disclosure of such material may lead to greater citizen involvement and more vigorous debate. Moreover, the arguments against such participation are likely to be highly skewed variants of the simultaneity and unmediated risk claims that our jurisprudence has rightly rejected in the core area of political speech.
To explore this argument in the Internet and terrorism setting, it is useful to return to Universal City Studios v. Corley, in which the court enjoined a website’s posting of computer code that enabled website visitors to play DVDs on Linux computers. Although the code had been written by hackers without knowledge of or participation in the work that produced the use restriction, the court validated the recording industry’s vague argument that the use restriction also discouraged piracy. The Corley court’s deference to the recording industry was particularly troubling in light of the First Amendment value of the code in question.
In the DVD cases, the on-line availability of techniques that may aid illegal conduct is important to alert society that such knowledge is available, and therefore to encourage business organizations seeking to benefit from copyright-protective technologies to better equip themselves to compete in the marketplace of ideas. One way to deal with music or movie piracy, for example, would be to create a user-friendly alternative to peer-to-peer networks. Corporations that avoid devising such alternatives, because of reliance on a technological lock-down, are relying on short-term fixes, instead of considering long-term values.
A consideration of the First Amendment value of the DVD-hacking code in the Internet context also makes short work of the mediation and simultaneity claims. While a code that penetrated copyright restrictions on software would allow individuals to make unauthorized copies, software manufacturers have a range of mediative strategies at their disposal. As suggested in the preceding paragraph, mediation may be indirect, taking the form of attractive alternative technologies devised and marketed by the manufacturers themselves. In addition, if manufacturers wish to retain the option of coercive measures, they can institute legal actions against end-users to deter piracy. Indeed, manufacturers have pursued a combination of these two mediative strategies because the simultaneity argument, irrespective of legal results, cuts so clearly against them as a matter of business planning. Once software of such general interest is on the Internet, putting the genie back in the bottle is all but impossible. If the goal of suppressing speech is to bar access to information, legal efforts to cope with the simultaneity of the Internet will literally fail before they begin. The global reach of the Internet compounds the futility of legal intervention in such cases.
Armed with analysis from the software piracy context, we can investigate how these issues play out regarding two areas important for law and terrorism: disclosure of terrorist tactics and scientific processes for the formation of weapons of mass destruction.
Internet publication of detailed instructions for terrorists on assassinations, extortion, and the like, inevitably give rise to justified public apprehensions about the impact of these violent acts. Yet here, as in the context of extremist speech, current legal tests based on specific intent may not provide adequate protection for speech. Consider the example of “Hit Man,” a detailed guide for hired killers published in hard-copy form as well as on the Internet. Because the publisher stipulated for purposes of summary judgment that it intended that the materials be used by actual or aspiring hit-men, a court awarded damages to victims of killings where the killer apparently used the book. It was clear, however, that there was no actual or knowing participation in the killings by the book’s author or publishers.
While the “Hit-Man” decision has its proponents, it is troubling because of the guide’s First Amendment value as well as the court’s failure to think through the issues of simultaneity and mediation. The book’s value, like that of a Jihad manual for terrorists-in-training, is in illustrating the vulnerabilities of our current system, and the ease with which a lawbreaker can operate with impunity. The Hit-Man and Jihad materials also offer useful insights into the psychological dynamics of the individuals who embark on such a violent course of action, thus allowing participants in civic debates to learn about the homogeneity of thought and selective moral concern of terrorist groups as well as democratic counter-examples that value diversity.
In the Hit-Man and Jihad manual context, our assessment of simultaneity and mediation is likely to be skewed by our apprehension of the events described in the materials. First, consider mediation. If short-term punitive fixes are available against those who distribute such information, we may unduly discount the impact of longer-term mediative strategies, such as policy alternatives that promote global equality and thereby blunt the recruiting of new terrorist operatives. Such alternatives will not always work, particularly given the polarized discourse within violent networks and the small number of people required to create catastrophic damage. However, legal rules should nonetheless encourage such options.
Analysis of simultaneity leads to the same conclusion. As the World War I cases demonstrate, combustibility is a powerful metaphor for the language of incitement, shaping not only views of an event itself, but also assessments of communications perceived as related to the event. People often attribute the same speed and sudden impact characteristic of a catastrophic event, such as an explosion, to abstract speech advocating or describing illegal conduct perceived as related to the catastrophe. These connections, however, are far less clear in practice. Putting together a terrorist operation or a hit requires elaborate planning, as the September 11 hijackers demonstrated. Groups with the structure and control over their members required for such planning also generally have the resources for developing and distributing their own proprietary tactical materials. Even when, as apparently occurred in the “Hit Man” case, a perpetrator draws inspiration and data from published materials, these materials often duplicate other materials already available. Given the simultaneity of the Internet, attempts to stop the flow of information qua information will merely inspire an endless virtual fun-house of mirror sites.
While on the surface, information about scientific processes may be more technical than information regarding terrorist tactics, the same analysis ultimately applies. Benefiting from a formula for poison gas or a nuclear device, for example, requires the cultivation of a body of knowledge and professional judgment, as well as the resources to build and maintain a physical plant for the manufacture and distribution of the weapon. Developing the expertise and the infrastructure to exploit that formula thus “demands a significant investment of time and money.” Persons with sufficient skills and resources to exploit the formula are unlikely to need the published formula to do their work. Similarly, persons lacking such expertise cannot benefit from the formula even it is published on the Internet or elsewhere.
Moreover, similar to the Hit-Man and Jihad manual circumstances, democratic governments have mediative measures available in the scientific process context as well. Persons with scientific expertise who act as agents of terrorist groups to enhance those groups’ destructive capacities can be prosecuted under a mobilization rationale. So can persons who donate substantial sums of money to terrorist organizations – money that can be used to purchase the services of persons with such expertise. Similarly, a participant-centered rationale would allow legal recourse against an individual who discloses information gained through participation in proprietary research, or those who induce such a disclosure, since these actions undermine participation in the research process. Absent evidence of such participation, the legal system should not criminalize or enjoin on-line publication of tactics, techniques, or scientific methods.
While the participation-centered approach offers heightened protection in dealing with incidental burdens and disclosure of terrorist tactics and scientific methods, it argues for minimal protection of the provision of personal information in a threatening context over the Internet. Here too, however, categories are murky. Material that intimidates can also inform. The participant-centered approach resolves issues by focusing on participation not only by speakers, but also by the targets of speech.
In a democratic society, threats are punishable not because they necessarily lead to physical violence, but because they deter participation in a broad range of activities. Despite the focus on immigrants that has historically tainted antiterror efforts, many of the most serious threats to participation in American political discourse are home-grown. The extortion practiced by organized crime is punishable because it affects people engaged in a broad range of commercial activities beneficial to society, either forcing people out of these activities or “offering” them the possibility of staying in business if they pay a tax to the offeror. Practitioners of collective violence such as the Ku Klux Klan also use fear to stifle participation. If victims of such intimidation seek to inform the authorities, the organization in question retaliates. In this fashion, threats deter the important civic duty of holding wrongdoers accountable. Participation withers, and alienation takes its place. The result is the corruption of ordinary democratic processes by practitioners of violence.
Threats occur not merely expressly, but also implicitly, sometimes in conjunction with proffers of information. Context, including justifications for violence, a recent pattern of violence directed at the target group, and specific information about the group, are important. For example, symbolic speech such as cross-burning on the property of a member of the target group, combined with the Klan’s historic justifications for and pattern of violence, sends a powerful message about that target’s vulnerability. In perhaps a more controversial case, the Ninth Circuit held that the “Nuremberg Files” website that identified individual doctors performing abortions as mass-murderers, labeleded them as “wanted,” and placed a line through the names of doctors that had been murdered, constituted a “true threat” which was subject to injunction. Thus, a website that fails to express an overt threat may nonetheless tender a true threat if the context demonstrates that the designers of the website intend to intimidate a person or persons.
The Internet is a singularly useful medium for such intimidation. Traditional media will typically decline to carry such information because of safety or taste concerns. In addition, the simultaneity of the Internet, which dispenses with the “lead time” of any other medium, presents a greater intrusion on people targeted and gives them less opportunity to make adjustments to promote safety. Updates and revisions of rapidly-changing information are routine on the Internet. If targets change their schedules, addresses, or other personal information, the site’s webmaster can post the new information on the site virtually immediately. Because threats constitute attempts to intimidate others, courts should view them as inherently operational, and not subject to the constraints on regulation of information and affiliation under the First Amendment. A hate group’s “death sentence” can reach intending and aspiring “executioners” most efficiently through the Internet. The capabilities of the Internet support the argument that a punishable threat should not require distinct proof of collaboration with persons who will attempt to carry out the threat because violence need not be a certainty in order to deter listeners’ participation. Given a context of violence, potential victims should not be required to guess whether a person making a threat of violence that seems clear in its intent to intimidate has the connections or resources to carry out that threat. Indeed, even in cases where there is no express interaction between the speaker and those individuals with the means and propensity to carry out the speaker’s threat, the fear sparked by uttering the threat may be amplified when related violent actions have already occurred. Leveraging violence in this fashion is not a form of political participation, but rather a rejection of the premises of the participatory model. If the threat, in itself, would intimidate a reasonable person, such a threat is a sufficient basis for culpability.
In addition, persons making on-line threats should not be able to cite the presence of other protected material on a site as a means for evading accountability. Some of the material on the “Nuremberg Files” site, such as the general characterization of doctors performing abortions as murderers and war criminals, clearly reflects protected social and political views. However, anti-abortion extremists have ample means available for expressing such sentiments that do not contain personal information about specific doctors. Because of the wide availability of alternatives, a rule imposing criminal or civil liability for implicit threats does not unreasonably burden the participation of the speaker.
To illustrate how true threat doctrine might apply to the Internet, consider a recent series of events involving a government witness in the trial of prominent white supremacist Matthew Hale for soliciting the murder of a federal judge. Witness Tony Evola’s testimony led to Hale’s conviction. A website sympathetic to Hale excoriated the witness, publishing the home phone and address of an individual who happened to share the witness’s name, but was in fact an entirely different person. Previously, one of Hale’s followers had gone on a homicidal rampage in connection with an earlier legal defeat for Hale, killing two people. Meanwhile, a radio host of an avowedly “eugenics-based” broadcast dedicated to propagating conceptions of racial superiority gave out information about the website, “In case anyone wants to say hi.”
In the climate of violence engendered by the Hale case, intimidation of the kind noted above goes beyond threats of social ostracism, and leverages fear of physical harm in order to chill participation in the legal system. Responding to the threat, the participant-centered account would have supported an injunction and a claim for damages against the offending website and against the radio host, had the host continued to encourage his listeners to commit violence. Criminal liability under a carefully drafted statute that bars intimidation of witnesses may also have been appropriate, along with tailored intervention with Internet Service Providers (ISP’s) to limit the publication of the material.
Legal authority to regulate threats should not simultaneously regulate the independent disclosure of lawfully obtained personal information. The presence of the intent to intimidate distinguishes the personal information revealed in the above discussion from the personal information revealed in cases such as the Kobe Bryant prosecution. In that case, the Colorado Supreme Court enjoined news media outlets from publishing information that they had received in an erroneous electronic transmission containing information about the complainant and recent motions concerning the scope of Colorado’s “rape shield” law. In the order issuing the injunction, the Colorado Supreme Court sought to distinguish two significant United States Supreme Court cases that had struck down measures prohibiting or creating liability for disclosure of personal information about parties in criminal cases.
Despite the Colorado Supreme Court’s efforts in the Kobe Bryant case, the information about the complainant, obtained without wrongdoing on the media outlet’s part, seems to fall squarely within the First Amendment ambit est
