David G. Post(1)
Temple University Law School/Cyberspace Law Institute
Daniel W. Kops Freedom of the Press Lecture
September 16, 1996
* * * * * * * * * *
When Dean Altschuler invited me to give this talk this past June, there was fresh, exciting news about the topic on which he asked me to speak -- "free speech on the Internet." A federal court in Philadelphia had, just a few days before, decided the case of ACLU v. Reno. In that case, the court struck down the Communications Decency Act, the federal government's attempt to ban the transmission of "indecent" material over the Internet to persons under the age of 18, as an unconstitutional abridgement of the First Amendment's promise that Congress shall "make no law abridging the freedom of speech."
Now this was indeed an extraordinary event. I am certain that this decision deserves the appellation of "landmark," both in the age-old struggle to define the contours of what we, as a society, mean by 'the freedom of speech' that we have enshrined in the First Amendment, and in the fledgling struggle to define the relationship between the government and this new communications technology. After all, here was the first attempt by one of our primary law-making institutions -- Congress and the President, together -- specifically targeting the Internet; and the first encounter between this new medium and our other national law-making institution, the federal court system. This was, in a sense, the Scopes trial for the electronic age, with the Internet taking the place of the theory of evolution (or the scientific method itself); the net itself was, for the first but surely not the last time, on trial.
And if that were not alone sufficient to give this case instant landmark status, the court's written opinion should be enough to secure this decision a prominent place in legal history. The opinion is a truly remarkable document, and I urge any one of you even remotely interested in the Internet or the First Amendment -- a large category, I should hope -- to read it. It is first and perhaps foremost an extraordinarily readable description of what the Internet is and why it is unlike any other form of communication ever developed -- in the court's words, "the most participatory marketplace of mass speech that this country -- and indeed the world -- has yet seen," a place where "users . . . may speak or listen interchangeably, blurring the [very] distinction between 'speakers' and 'listeners'," a place where information, once it is made available to anyone, is "available to all other Internet users worldwide," a place that already contains material "as diverse as human thought," providing users with fulfillment of the Gutenbergian promise by providing "an easy and inexpensive way for a speaker to reach a large audience, potentially of millions."
And it is also a ringing and eloquent endorsement of the fundamental value of free and unfettered expression as a core component of human liberty:
"Cutting through the acronyms and argot that littered the hearing testimony . . . the Internet may fairly be regarded as a never-ending worldwide conversation. The Government may not, through the CDA, interrupt that conversation. As the most participatory form of mass speech yet developed, the Internet deserves the highest protection from governmental intrusion.. . . The absence of governmental regulation of Internet content has unquestionably produced a kind of chaos, but . . . [w]hat achieved success was the very chaos that the Internet is. The strength of the Internet is that chaos. [And so, too, does] the strength of our liberty depend upon the chaos and cacophony of the unfettered speech the First Amendment protects."But I come here tonight not to praise ACLU v. Reno but to bury it -- or, at least, to push it aside and out of view. I want to argue that the Internet calls for far more radical re-thinking of our basic notions of law and governance than even this court may have appreciated. I want to ask not "what do, or should, the laws governing the content of speech on the Internet look like?" but a more basic question: who makes those laws? Not "should pornography be banned on the Internet," or "is such a ban consistent with the First Amendment," but who decides whether pornography should be banned on the Internet? What is the source of the law that will govern our interactions in cyberspace, and who are the law-makers?
This is not an abstract question of political theory. If we are going to be communicating with one another in cyberspace, where do we look for the rules that can allow us to determine whether our speech is, or is not, wrongful? Questions about how the First Amendment applies to acts of the US Congress, however imaginatively or creatively we answer them, will simply not resolve for us the truly difficult questions that will determine the scope and the nature of free expression on the net, because the net, while it is not ungovernable or in a perpetual state of anarchy and chaos, as some would claim, is indeed ungovernable by traditional means.
It is ungovernable by traditional means because a core and irreplaceable component of our existing institutions for governance -- physical location itself -- is not, and cannot be made, operative on this network. We take for granted a world in which geographical borders--lines separating physical spaces--are of primary importance in determining the rules that define legal rights and responsibilities. Law-making sovereignty -- internationally recognized "statehood" -- itself is defined, at bottom, by control over a physical territory, and we look, in the first instance, to the law of "the place" where an event or transaction takes place for the source of the law governing that event or transaction -- the venerable principles of lex locus delicti and lex locus contractu, the law of the place of the wrong or the place of the contract.
But cyberspace does not merely weaken the significance of physical location, it destroys it altogether. The cost and speed of message transmission on the net is entirely independent of physical location: messages can be transmitted from any physical location to any other location without any distance- or location-based degradation, decay, or delay. There are no physical cues or barriers that might otherwise keep certain geographically remote places and people separate from one another. The net enables simultaneous transactions between large numbers of people who do not know, and in many cases cannot know, the physical location of the other party. A World Wide Web server in Frankfurt Germany looks for all intents and purposes exactly like a server in Frankfort, Kentucky. The net has no territorially based boundaries at all. One can meaningfully talk of the "location" of events and transactions in cyberspace, but only in reference to a virtual space consisting of the "addresses" of the machines between which messages and information are routed; the system is entirely indifferent to the physical address or location of those machines.
Events in cyberspace thus, in a sense, take place "nowhere," at least nowhere in the physical world, a challenging prospect for a global legal system that relies on physical location to determine the source of applicable law and the identity of the relevant law-makers. We have long since abandoned a strict and inflexible application of the doctrine of the "law of the place." In many contexts, we take a broader view, and look not to the "location" of events to determine the law that is applicable to particular conduct but rather to the jurisdiction where the effects of that conduct are felt. But in cyberspace, this brings us no closer to a solution to our problem, because the effects of online activities are felt simultaneously in every corner of the global network. Our World Wide Web server in Frankfurt can be accessed just as easily by users in Frankfort, Kentucky -- or anywhere with a net connection -- as by those in Berlin; all jurisdictions simultaneously feel the effects of the information posted there, and thus all would appear to have equal claims to make the law governing the content of this site, surely a recipe for international chaos.
In short, as my colleague Larry Lessig has put it, events in cyberspace take place nowhere, yet everywhere, at the same time. There is simply no room for this phenomenon in our existing legal infrastructure, no way to ask the most basic question of all: who makes the law? If material placed on the net is equally accessible across the globe regardless of the geographical location from which it originates, how can the U.S. Congress determine whether or not the material is pornographic, or infringing of an author's copyright? If it does so, how can it enforce its determination when the source of the information is not within its borders? And if Congress points to the within-border effects of these online actions to justify a right to make law applicable to, say, a World Wide Web server physically located in Germany, or Mexico, or Brazil, what is to prevent the Germans, or Mexicans, or Brazilians from asserting the corollary right to make law applicable to Web servers in Des Moines or Dallas? As I suggested above, this is not international law, it is international chaos.
Let me give you a concrete example. Suppose that the contents of this speech that I am now giving violate the law of some existing jurisdiction -- that Singapore, say, has a law making it a crime to deny the governability of the net. This is not as far-fetched as it might sound; Germany, after all, has a law prohibiting denial that the Holocaust occurred, and, more generally, there are wide differences among jurisdictions in terms of what is or is not considered libelous, defamatory, or unlawfully privacy-invading in printed material. And suppose that this speech is published in the print version of the Ithaca Journal. We have little problem asserting that the law of Ithaca New York, subject to the law of New York State, subject to the law (and Constitution) of the United States, must be consulted to determine whether, for example, I have said anything defamatory, or libelous, or criminally inciteful. It is preposterous to suggest that Singaporean officials can prevent publication of the paper or punish me, or the editors of the paper, for violating its law. It is preposterous not just because Singapore does not have the power to compel compliance with its laws in Ithaca, but because it is illegitimate for it to do so; the people of Singapore, sovereign on their side of an internationally-recognized boundary, have an insufficient stake in this activity to assert any legitimate claim to control it. It is the people on this side of the boundary -- citizens of New York and of the United States -- who most directly feel the effects of that publication and are the legitimate sources of the law governing it.
But suppose now that the Ithaca Journal also places the text of this speech on its World Wide Web server. The people of Singapore now have access to this text that is as direct as the residents of Ithaca, and the effects of this speech are now felt as palpably in Singapore as in Ithaca. Now the people of Singapore do have a stake in what I'm saying, a claim that they feel the effects of my words no less strongly than do the people of New York. But giving credence to that claim is untenable, for every country on the globe has precisely the same claim that the Singaporeans have to regulate this behavior. Their citizens, too, have the same instantaneous access to my words. And it simply cannot possibly be the case that the law of all of these individual countries can be applied to the words I am speaking, for that would subject me to an incomprehensible morass of conflicting obligations.
This is, I think, obvious and irrefutable -- so obvious and irrefutable that you might even wonder why I bring it up. To appreciate its impact, reverse the context: a Singaporean Web server that contains material that is criminal in this country -- advice on how to assassinate the President, say, or, to use a less inflammatory example, copies of software programs, available for free downloading, that are protected under US Copyright law but perhaps not under Singaporean law. The implications of this line of reasoning are clear: the US Congress cannot assert a legitimate claim to control that behavior or the distribution of that material on the net any more than the Singaporeans can do so with respect to the contents of my speech.
"Fine," you say -- Singapore will not be able to apply its law to Web servers in Ithaca, and Ithaca will not be able to apply its law to Web servers in Singapore. Perhaps the answer is simple: the law of the United States applies to Web servers in Ithaca, and the law of Singapore to web servers in Singapore. Surely that gives us an orderly world, a world in which people understand and will largely comply, or can be made to comply, with their legal obligations; indeed, that's more or less the world we live in now, where the editors of the Ithaca Journal look to US law, and the editors of the Singapore Times to Singapore law, to determine what is or is not wrongful. So what's the big problem with that?
A good question, to which I have two answers. The first is a practical one. In a world such as cyberspace where location does not matter, it is far too easy to avoid compliance with any obligations one finds objectionable simply by moving around to make this scheme workable in the long run. This has a fancy descriptor in the academic literature -- "regulatory arbitrage" -- but is a simple concept: if it makes no difference at all whether the Ithaca Journal Web site is physically located in downtown Ithaca or in Brazil or Belize, the proprietors of that web site can simply find the jurisdiction most congenial to their view of what the law should be. If U.S. law prohibits them from doing something they want to do, why, then, they can simply move to a more welcoming jurisdiction.
But my second response is perhaps more important: while application of the "law of the place" produces order, it does not -- at least as applied to cyberspace -- produce legitimate order. The people of Singapore should have a say in the rules governing my conduct if and to the extent they are indeed affected by it. We should be governed by rules that have been created by a process that takes the interests of all affected parties somehow into account; that is, in a sense, nothing more -- but nothing less -- than a re-statement of the principle of self-determination itself, a principle that goes to the heart of what we consider legitimate in social and political life. It doesn't mean, of course, that the army of Singapore is entitled to come in and force me to comply with its law, because the process by which that law was created didn't take my interests into account -- or the interests of the others across the globe who are indeed affected by my actions. Where is the law-making process that can do that?
It is all enough to make one's head swim. Existing legal and political doctrine and institutions appear to be inadequate for the task of crafting effective and legitimate rules for this new environment. The advent of Cyberspace heralds what Walter Wriston called, some years back, the "twilight of sovereignty," quite possibly the final days of a governance system relying on individual sovereign states as primary law-making authority, a system that has served us, often for better and sometimes for worse, for the last half millennium. But the basic problem of social life remains: how can people order their collective affairs to achieve results that they cannot accomplish on their own? Anarchy, after all, has its costs; random results -- the systematic misrouting of messages, say -- don't encourage trade or continuing interactions among people or entities. Large numbers of users will not visit online spaces if they encounter systematic fraud or vandalism or other activities they view as harmful or antisocial. If one model of governance is inadequate for the task, others need to be, and will be, created and tried. And I want to suggest that what is perhaps most exciting about the rise of cyberspace is precisely the glimpse it gives us of another possibility, of a kind of "decentralized, emergent law," that cannot impose order on the electronic world but through which order can emerge.
To illustrate, and to bring this discussion back down to earth, let us return to the federal courtroom in Philadelphia -- not to the Communications Decency Act case but to another case now, coincidentally, being heard in that same courtroom. This other case, CyberPromotions, Inc. v. America Online, is a simple and rather mundane commercial dispute, with nothing, on the surface at least, as grand as the issues raised in ACLU v. Reno. Yet I think it has, ultimately, far greater implications for the question of law and order on this new electronic frontier.
The defendant, America Online, is familiar, I assume, to you all as a large commercial online service provider. Like most OLSPs, America Online offers a service to its subscribers -- the ability to receive, and to send, electronic mail, not merely to other AOL subscribers but over the Internet. The engineers describe this as providing a "gateway" between the closed America Online system and the open, worldwide Internet. CyberPromotions, Inc., the plaintiff, is an entrepreneurial operation that takes advantage of the very speech-enhancing features of the net that the ACLU v. Reno court so ringingly endorsed; it is a bulk e-mail operation, offering businesses the opportunity to send electronic mail to millions of recipients at a minuscule fraction of the cost of such promotional activities in the real world.
America Online claims that many of its subscribers have complained about receiving this "virtual junk mail" -- or "spam," as it is sometimes called in online parlance. In response, it has configured its Internet gateway to reject any and all electronic mail that arrives bearing a CyberPromotions return address. CyberPromotions has cried foul (and filed suit), claiming that America Online's actions improperly interfere with its "freedom of speech," its ability to engage in commercial activity on the net.
In a stunningly wrongheaded move, the court just this past week issued a preliminary ruling siding with CyberPromotions, ordering America Online to open up its mail gateway to incoming mail from the CyberPromotions site, at least pending a full trial. Wrongheaded because what America Online has done here is precisely what Internet governance requires. They have promulgated a rule -- dare I say a 'law'? -- about junk e-mail. Is it a "good" rule to deal with this rather novel social problem? Is it the "right" rule? I have no idea -- all I know is that those who believe it is a good rule can remain as members of the America Online community, and those who believe it is otherwise (including those who believe it is infringing some right of entities like CyberPromotion) can move elsewhere, to any one of the thousands of other providers of Internet mail service. It is a rule, in other words, that individual America Online subscribers have voluntarily chosen -- if you will, it is an expression of their collective desires to order their electronic world in a certain way. And I am confident that it adequately expresses their collective will, because America Online's ability to "impose" rules contrary to that collective will is severely constrained by the ability of its subscribers to move somewhere else -- to "vote with their electrons." It is true, of course, that if many Internet service providers implement a similar rule, and if those service providers flourish and attract many subscribers, the ability of entities like CyberPromotions will be drastically curtailed. But if many Internet service providers implement a similar rule, and if those service providers flourish and attract many subscribers, who would venture to say -- who has the right to say -- that that isn't the best rule to deal with this problem? What does your personal First Amendment say about this form of communication? Rather than an abstract argument about whether there exists some 'right' to send unsolicited bulk e-mail, let's see what kind of communities people build and what the rules in those communities are.
What I envision, then, is a radical decentralization of law-making, a kind of electronic federalism in which individual network access providers -- the America Onlines and the thousands of other Internet service providers -- rather than territorially-based states, become the essential units of governance. Users like you and me delegate the task of rule-making to them -- we confer sovereignty on them -- and we choose among them according to our own individual views of the constituent elements of an ordered society. Mobility -- our ability to move unhindered into and out of these networks with their distinct 'law' -- is a powerful guarantee that the resulting distribution of rules is a just one; indeed, our very conception of what constitutes justice may change as we observe the kind of law that emerges from uncoerced individual choice. The "law of the Internet" thus emerges, not from the decision of some higher authority, but as the aggregate of the choices made by individual system operators about what rules to impose, and by individual users about which online communities to join. Or perhaps we should think of this as the law of the Internets, for one possible (or even likely) consequence of this evolutionary development is the emergence of multiple network confederations, each with their own "constitutional" principles -- some permitting and some prohibiting, say, bulk electronic mail, or anonymous communications, some imposing strict rules regarding redistribution of information and others allowing freer movement, etc. -- enforced by means of electronic fences prohibiting the movement of information across confederation boundaries.
We know that decentralized systems are capable of generating remarkable degrees of coordination -- think, most obviously, of biological systems, which evolve without any centralized decision-making at all. And if you think that decentralized governance systems cannot likewise produce a stunning degree of order and harmony, consider how the Internet arose in the first place. The Internet is itself a remarkable triumph of international coordination and cooperation, a complex adaptive system that has achieved a previously unimaginable degree of inter-connection without reliance on top down, hierarchical control. The Internet itself exists only because a very large number of individual computer networks voluntarily adopted a new language -- the "Internet Protocols," the technical rules governing the way that messages are to be formatted and routed from one machine to another -- that allows those networks to communicate with one another. This language is itself a kind of law, a set of rules governing communication -- but no sovereign authority with the power to compel obedience among its subjects promulgated that law, no treaty decreed that a specific set of such standards must be used in order to link each of the diverse individual networks together into a single global web. Instead, under the banner of "rough consensus and working code," groups like the Internet Engineering Task Force and the World Wide Web consortium -- unofficial, unsanctioned, collections of interested volunteers -- published proposed communication standards that became the "law of the net" as large numbers of individual network administrators voluntarily adopted the proposed rules. Each individual network remains free to impose its own technical standards on its users -- the Microsoft Network uses different communications principles for its intra-network communication than does America Online or the Cornell University Local Area Network -- subject to the overriding mandate that if it wants to enable communication with other similarly-situated networks, it must adopt the basic communication protocols that those other networks have adopted
* * * * * * * * * *
1. Email: DPost@vm.temple.edu. Thanks to David R. Johnson for his help formulating the argument developed here.