114 Harv. L. Rev. 606, *

 

Copyright (c) 2000 The Harvard Law Review Association
Harvard Law Review

 

December, 2000

 

114 Harv. L. Rev. 606


LENGTH: 15906 words

BOOK REVIEW: PERFECT FREEDOM OR PERFECT CONTROL?

Reviewed by Charles Fried*
 
*Beneficial Professor of Law, Harvard Law School. Thanks to Christine Jolls, Charles Nesson, Mark Ramseyer, Hal Scott, Eugene Volokh, and Jonathan Zittrain, and to Rocky Tsai, Harvard Law School '02.

SUMMARY:
... 1 All these rich influences have left their deposit and show up in this book. ... Other threats to First Amendment "values" 68 are the pervasively "propertized" regime of intellectual property I have already discussed 69 and the more obvious threat that major intermediaries - for instance, those who operate the spine of the Internet, or the service providers - might take it upon themselves to engage in more active and fine-grained filtering than they do now. ... B. The State Action Doctrine and Free Speech on the Internet ... 90 It is not clear, therefore, that the state action doctrine presents quite the obstacle Lessig suggests to the maintenance of the intellectual commons. ...  

Code, and Other Laws of Cyberspace. By Lawrence Lessig. New York: Basic Books. 1999. Pp. xii, 297. $ 30.00.

TEXT:
 [*606] 

This is a book that lawyers, law students, and legal scholars should read. Lawrence Lessig is a leading constitutional law scholar and generalist of the generation that studied at Yale Law School long after the death of Alex Bickel, when not only Bruce Ackerman and Owen Fiss had attained the status of the intellectual barony, but also a younger cohort - of which Akhil Amar is perhaps the outstanding member - had already marked its place there. As a law clerk Lessig served under two extraordinary, highly flavored judges, Richard Posner and Antonin Scalia, and he spent his first six years of teaching in the distinctive atmosphere of the University of Chicago Law School, with its long tradition of hard-headed law and economics softened by the current ascendancy of the "New Chicago School" led by Cass Sunstein. 1 All these rich influences have left their deposit and show up in this book. And its subject is the signal phenomenon of the day: that technologically complex, conceptually elusive, and legally amorphous realm known as cyberspace, or the Internet. Lessig's subject is the intersection of that phenomenon with the traditional pillars of constitutional law - privacy, freedom of expression, property rights, economic regulation - and with legal and political theory generally. In particular, his organizing concern is the relation between liberty and regulation as exemplified in a novel context: the practices, legal structures, and regulatory struggles concerning the Internet.

The book should be read for several reasons: first, because the Internet is so important, and this is one of the rare lucid guides to it written by someone with a profound legal understanding, as well as a broad cultural sensibility; second, because the relation between liberty and regulation is significant and fundamental well beyond the context of the Internet, and Lessig has many interesting things to say about this relation in its own right; third, because this is a first book in  [*607]  which a young scholar of power and ambition announces his intellectual agenda. Finally, and perhaps most importantly, this book should be read because it is fun. It left me full of questions, objections, and proposals for further work. I kept wishing the author were in the room with me so I could ask, "But Larry, what do you mean by this? And how would you answer that? And how about this other thing?" In short, Code invites - by both its style and its content - a conversation, and that, if I have caught the wave, is exactly the spirit of cyberspace. So, here goes.

I. Liberty and the Internet
 
Code announces its theme in opposition. Lessig begins by reminding us of the wave of triumphalist enthusiasm that swept up liberals (in the original, nineteenth-century sense of that term, not the degraded, contemporary, journalistic sense) after the fall of the "Evil Empire" in late 1989, in the belief that this turn would usher in a utopia of political liberty and capitalist prosperity (pp. 3-4). In much the same way, enthusiasts have spilled barrels of ink (oops! spewed billions of bytes) announcing that cyberspace is to be the realm of perfect freedom - anarchy without chaos - where government need not and, indeed, cannot (effectively) intrude (pp. 4, 24). And just as the hopes for Eastern Europe and Russia were soon to be disappointed, so - it is Lessig's theme - liberal triumphalism is out of place in cyberspace (pp. 5-6). What is quite systematically ambiguous throughout the book is whether this is a matter for Schadenfreude or regret. Although I do not read Lessig as hostile to the party of liberty, he regularly invokes the virtues of democratic (that is, political) control over human interaction (p. 241 n.4) - as do his teacher and his erstwhile colleague, Fiss and Sunstein. The theory underlying this dismal and debunking theme is one made familiar not only by ambivalent moderates like Sunstein, but also by the Critical Legal Studies movement and before it by Legal Realists such as Robert Hale and Morris Cohen: There is no such thing as natural liberty. All choices, including the choice that government not regulate, are political choices establishing different political regimes. Correspondingly, the distinction between the public and the private, on which liberal theory depends, is an illusion. 2

 [*608]  The libertarians - this is Lessig's designation (p. 4) - of cyberspace imagine that persons interacting on the Internet can form and dissolve communities of communication freely and virtually costlessly, and that these communities constitute ways of life, "spaces" in which people can expend their emotional and intellectual energies - in other words, in which they can live (p. 63). (This is a perfect instantiation of what Robert Nozick described in the third part of his Anarchy, State, and Utopia as a purely imaginary, science fiction utopia.) 3 Since these interactions are entirely voluntary and costless, there is no need for government to regulate them. 4 And because of the anonymity and wide dispersion of the Internet, government could not regulate it even if it tried. Lessig demonstrates this vision to be at once a fantasy and a fallacy.

A. What the Internet Is: Code
 
Communication on the Internet depends on the decomposition of the things communicated - whether text, images, or sounds - into packets of electronic data (ultimately, the 0s and 1s of binary code), which are then routed across a network of interconnected computers to their destination "address," where they are reassembled back into the form intended by the sender. A congeries of complex protocols governs the process of information translation at both ends of the communication. The original message must first be translated into something resembling natural language and then must be decomposed again until it ultimately takes the basic digital form that a computer understands. A child's drawing of a house may seem the very picture of simplicity,  [*609]  but breaking it down to its constituent bits on one end and then reassembling these data on the other require a large number of instructions in order to corral the perhaps millions of bits of digital information involved. At both ends, this remarkable juggling act is governed by the programmed routines of the machines employed. Those routines - different ones for text, images, and sounds - are automatically carried out by the machine because it is instructed, in a very specific way, how to translate, say, text to bits and back again. These instructions, which govern a computer's actions (assuming no bugs) with the inexorable force of law, are what Lessig refers to as "code," or "architecture." Code can itself be - and at some levels of the translation process almost always is - a set of programmed instructions to the wires and switches of the machine. This is the software. But ultimately some of the code must be built into the physical configuration of the machine - and networks of machines - themselves (the hardware). The term "code" is most intuitively applied to software, the instructions to the machine; "architecture," to hardware, the physical design of the machine itself. But Lessig uses the terms interchangeably - rightly so, it seems to me, for hardware can be conceived of as essentially a set of instructions at the level of the machine itself, its architectural backbone, while the layer of instructions that activates that hardware (for example, an operating system like Windows or OS9) is itself fixed relative to the layer above it (for example, an application program like WordPerfect), whose "text" it translates into a form usable by the layer below. It is all code and all architecture.

Lessig's mantra is "code is law." Once again he is certainly correct, in the sense that there is nothing preconventional about how communication takes place between a user and her machine, and therefore also between that user and the millions of machines to which the Internet may connect her. Every step is constructed, designed by engineers for some purpose and according to whatever constraints the designer must observe (say, the technical limits of the machine, or the fact that electromagnetic signals cannot travel faster than 186,000 miles per second) or chooses to observe (for instance, considerations of cost).

B. The Evolution of the Internet
 
The early enthusiasts who proclaimed cyberspace the realm of perfect freedom were thinking of the original arrangements, born (one must remember) out of a project designed, and paid for, by the Department of Defense. 5 In that Edenic time, anyone with a modem and  [*610]  a phone line could access the global Internet backbone for the price (often free) of a local phone call and communicate with anyone else similarly connected. Initially, technical limitations restricted communication to typed text. Transmitting images and sounds was impracticable because it would have taken literally hours to disassemble them into digital form and reassemble them. But still there was a sense of freedom about this capacity to be instantly in touch with people - singly or in groups - all over the globe, who could respond more or less instantly, giving a semblance of conversation, albeit through the medium of typed text. There were no stamps to buy, no postal service to employ, no phone operators to set up a conference call (which at any rate could involve at best many fewer people), and no billing (except for the local phone call). These enthusiasts tended to forget that the network had been designed and built and was maintained by someone - originally by the Department of Defense and later by giant communications companies like GTE. More important, the thing only worked because messages could be disassembled and sent in packets over a variety of alternative routes, only to be reassembled at or close to the designated address. The technology for accomplishing these tasks is complex and depends on instructions (code) to computers located along the Internet backbone, which in turn depend on a certain standardization of addresses (more code).

Lessig points out that whatever freedom the early Net allowed (and it was less than the enthusiasts proclaimed), it was also, to a precisely equivalent degree, a technology highly amenable to control (p. 25). The early situation was Edenic because nobody had much incentive to assert control, but very soon all kinds of people woke up to the possibilities - both hopeful and disturbing - of a medium permitting virtually instantaneous worldwide communication with huge numbers of people. And with that awakening, the controls began to appear.

C. Commerce
 
The worm in paradise was commerce (p. 30). 6 Entrepreneurs saw a variety of opportunities. Consider just these few: A retail merchant could offer its wares to a vast potential clientele without the expense of opening and operating shops wherever there might be buyers, or of printing, mailing, and updating catalogues by the millions. Information stored in books and other searchable databases could be made instantly available and precisely tailored to the searcher's criteria. The very freedom and anonymity that the early Web enthusiasts celebrated created a common set of problems for those who would exploit the  [*611]  commercial possibilities of the Internet: how to charge for the goods or services provided, how to get paid, and how to let potential customers know that they had something to offer that these customers might want (that is, how to target advertising).

In identifying commerce as the cause of the Internet's fall from its original state of grace, Lessig may have intended to be no more than descriptive, but one detects a certain tone of sympathy with those who deplored abandonment of the original principle that the Web was to be a commerce-free zone. Whether or not Lessig sympathizes, that principle was unreasonable. Commerce is just buying and selling, and buying and selling are just the activities of free people offering to others the products of their labor and those others giving the products of their labor in return. Only in a world where producers have no needs or wants, or where others supply those needs and wants on a wonderfully voluntary basis, is there no need for buying and selling. So commerce sums up a quintessentially human activity, and to exclude it from the Net would have required a measure of control more draconian and stultifying than anything that cyber-libertarians have subsequently deplored. 7

But commerce requires pricing and payment systems and advertising. In satisfying those requirements, commerce required changes in the code or the architecture of the Web. Take payment, for example. Consider how a retailer is to get paid for merchandise he offers on the Web (pp. 39-40). He could sell on credit, sending out a bill with the goods, but the whole point of Web retailing is that the retailer can reach hundreds of millions of potential customers in many countries, whose creditworthiness he cannot, and does not want to have to, check. He could send out goods C.O.D., but many carriers will not perform that service, and the Postal Service, which will, charges for it. Also, such a system slows down delivery, as it requires the customer to be at home at the time the goods are delivered. That leaves the credit or debit card, the unsung hero of retail e-commerce. The card issuer bears the risk, not the merchant, who is assured of rapid and regular payment. But the use of credit cards requires security measures: the card company must be willing to make the charge without the cardholder's signature, and that means it must have some assurance that  [*612]  the cardholder has authorized the charge. This assurance in turn requires systems of identification and authentication quite foreign to the original, free-wheeling, anonymous (or pseudonymous) world of the Web (p. 40). Systems of authentication and identification are effected by changes in code - not necessarily in the systems that route messages around the world, but in the systems at either end. The changes need not be universal; they can apply only to those transactions that require security and authenticity.

Now consider the cases in which the merchandise is not just ordered on the Web and delivered by UPS, but is actually delivered by the Web itself. Certainly the data could be made available for free, as by a public library, a government agency, or a nonprofit institution. But publishers, authors, and compilers of information have always provided this service as a business, and there is no obvious reason why this new mode of dissemination should change that. Unlike sellers of merchandise, who are eager to have as many visitors as possible on the Web, many of these "e-publishing" businesses choose to restrict access to those who pay either a periodic subscription fee or a fee measured by features of the search. The latter is the metered system that databases like Westlaw and Lexis-Nexis used long before the Web came into common use. Subscribers accessed them by modem over ordinary phone lines. These payment methods require some sort of electronic turnstile that allows access only to registered users (p. 34). Such an electronic turnstile, too, is constructed of code.

There is another way that sellers of information on the Web can get paid for their efforts: by relying on advertising, like traditional radio and television stations (as opposed to premium cable and satellite channels). The most natural advertisers are Internet retailers, whose advertisements are so devised (again by code) that by a single click you can move from the advertisement to the merchandising website advertised. But the very profusion of potential advertisers, combined with the unprecedented size and diversity of the audience, suggests that advertisers would do well to be as selective as possible. If there were some way to identify which visitors to an information site were statistically more likely to be interested in a particular line of goods, then advertisements on that site would be more productive, and the site could charge more to display them. To accomplish this targeting, a mechanism has been created that works in quite a different way from the turnstiles, authenticators, and identifiers. This mechanism collects information on visitors to the information site from various traces those visitors leave as they cruise the Internet (pp. 34-35). Those traces may be left on the information site, so that search engines like Yahoo! may keep a record of who asked questions about what. But they may also be left on the user's own machine (in "cookie" files). That information may be compiled and analyzed so that when the user visits the information site, he will be shown advertisements that analysis  [*613]  suggests may be of particular interest to him. After all, food and wine magazines carry advertisements different from those found in camping magazines or magazines catering to Civil War buffs, and that targeting benefits advertisers and magazine readers alike. But on the Internet, for such a system to work best, the prospective compiler must be able to "worm" his way into the user's computer to glean the information revealing his preferences and interests. All of this is effected by code.

D. Choices About Code
 
Lessig argues that these varieties of code - spawned largely to meet the needs of commerce - can change the Internet environment from one of perfect freedom to one of perfect control: everything you send or receive can be monitored, and your access to sites can be blocked or conditioned depending on who you are and what you are willing to pay (pp. 30-42). It is even possible to devise code that filters certain types of material, thus controlling the content of what you can send or receive (p. 177). Most of these varieties of code are already with us, and others (like highly sophisticated filters) are right around the corner. Although commercial needs sparked much of this transformation of the Internet environment, Lessig shows that the government may use these varieties of code for the whole gamut of its objectives: for example, censorship (p. 178), restriction of access to sensitive information (p. 56), and monitoring for law enforcement purposes (p. 48). Because the Internet is organized by code, and code can either facilitate or disable all of these varieties of control - whether by government or private institutions - Lessig argues that there is no such thing as a default position of freedom, but that choices must be made at every point (p. 59). The rhetoric of perfect freedom in the primeval Internet has obscured this necessity for choice, with the result that we have already partially foreclosed our options without adequate public deliberation and risk doing so further in even more preclusive and perhaps undesirable (or at least undesired) ways. The analogy that comes to mind is the way that the post-World War II emphasis on road-building and the automobile shaped our national environment without any broad public awareness, much less choice in the matter. Above all else, this book is a call to public awareness, public deliberation, and public choice. In this mission it succeeds, and for that reason alone it is an important book.

II. Choices: Public or Private?
 
I do not disagree with Lessig's prescriptions - for example, his desire to preserve an "intellectual commons" on the Net (p. 141), his preference for zoning over filtering (p. 181), and his championing of "open code" (pp. 107-08). I quarrel with some of his analyses of particular  [*614]  problems. But more generally I quarrel with Lessig's barely explicit but detectable bias toward public decisionmaking - by which I mean political decisionmaking, as opposed to the disaggregated private decisionmaking of the market - about the design of code for the Internet (pp. 58-59). This bias is just the natural outgrowth of Lessig's embrace of the distinctive stance of scholars like Fiss and Sunstein (pp. 180, 275 n.1, 279 n.50). These scholars maintain that there is no distinction between the public and the private (and in particular, with respect to freedom of expression, that a "New Deal" is needed to place private power over expression under the same constitutional strictures applied to government) and that the constitutional doctrine of state action is an impediment to sound analysis and desirable outcomes. 8 But Lessig does not entirely disavow the disaggregated liberty celebrated by the first wave of Internet enthusiasts. He embraces "open code," precisely because it would draw the teeth of the threat that both public and private power might represent (pp. 107-08).

As I have said, this book announces a talented thinker's intellectual program and opens up a realm of puzzles and choices to its readers. One thing, however, is clear: it is a premise of Lessig's book, as of so much else that is written on this subject, that the Internet promises (or threatens) rapid, order-of-magnitude changes in the way we live our lives and relate to the world and each other.

Lessig's premise leads him to argue that the phenomenon of the Internet upsets familiar conceptual structures, including those in the law: at a general level, the distinctions the law makes between public and private power, and more specifically, the way we think about particular areas, such as sovereignty, intellectual property, free speech, and privacy. In the next section, I consider the premise itself, its bearing on the public/private distinction, and its application to one of those specific areas, sovereignty. In the following sections, I consider Lessig's discussions of intellectual property and free speech. (I do not take up his excellent treatment of privacy.) But before embarking on those particular discussions, I put before you Lessig's important and distinctive methodological treatment, contained in his chapter called "Translation" (pp. 111-21), of how the law does and should go about adjusting familiar concepts to radically altered circumstances. 9

 [*615] 

A. Translation
 
Lessig argues that throughout our history, when we have needed to apply the Constitution in contexts radically different from those imagined by its framers, we have found it necessary to "translate" the text into its new context (p. 114). Lessig's position reminds me of Gadamer's argument about the translation of biblical texts into modern times. 10 The more radically different the context, the further one must move from the actual rules stated in the text, or that have grown up around it, to implement its underlying values (pp. 115-16). Lessig's best example is the application of the Fourth Amendment protection against unreasonable searches and seizures to wire-tapping in Katz v. United States (pp. 116-18). 11 If the circumstances are too different from those that obtained at the Framing, then translation becomes choice (p. 118).

For each of the topics that Lessig addresses, the questions are as follows: Just how far afield are we from the framing context? Are we at least close enough that we may discern the values implicit in the texts so as to apply them, or are we so far off that we are more in the domain of original choice? The textualist's answer, that we must rigorously apply the text and leave it to constitutional amendment to adapt that text to modern times, will not do, because in such cases, strict application of the text leads to random or perverse results.

Here we join up with Sunstein's and the Critical Legal Studies movement's analysis of the assumed background of private law institutions. Under that analysis, if the private law institutions have themselves changed radically or are inapplicable to the new context, then their nature as engines of social control stands revealed, and they invite deliberate evaluation and revision in terms of public law values. Given the tone of this book, Lessig must be taken to think that our situation is, or is heading, pretty far out, so that we are in or near the realm of pure choice. His very last sentences affirm that proximity: "We are entering a time when our power to muck about with the structures that regulate is at an all-time high. It is imperative, then, that we understand what to do with this power. And, more important, what not to do" (p. 239). But as I read his excellent and lucid discussions of intellectual property, privacy, and free speech, I felt far nearer Earth than he would have me be - near enough that the text of the Constitution remained in sight.

 [*616] 

B. The Public/Private Distinction
 
The standard argument against the public/private distinction goes like this: The realm of the "private" is in fact constituted by legal rules - of contract, tort, and property - that represent decisions about how entitlements should be defined, allocated, and transferred. 12 But these decisions are as much decisions about the structure of our social world as are the decisions of public law, such as the level and forms of taxation, the provision of social welfare, and the regulation of the professions, the environment, product safety, and the sale of securities. Applying this trope to freedom of expression, Sunstein argues that it is anomalous that under First Amendment jurisprudence a speaker may not be barred from a street corner on the basis of the content of his speech but may be excluded from the pages of the only newspaper in town. 13 The street corner is a public space, and the bar would be public law - state action - subject to the First Amendment. The newspaper is private property, and the exclusion would be a private decision. This distinction is argued to be anomalous because the system of rules that confers the ownership rights over the newspaper is as much the product of social choices as are the rules that regulate access to the street corner or the newspaper's waste disposal practices. 14 Sunstein argues that the First Amendment must therefore be reconceived - this is his "New Deal" - in terms of the values it was meant to protect: the democratic values of open public discourse on matters of public concern. 15 The First Amendment should be understood to protect that free and equal discourse against exercises of power, whether designated public or private. At the very least, the First Amendment should not stand as an obstacle to public rules designed to discipline (what is called) private power in the interest of free and equal democratic deliberation.

The public/private distinction may be defended against the attacks of Sunstein and Critical Theorists in this way: granted, the rules of private law - contract, tort, and property - are themselves rules of law and therefore public rules, but they represent a relatively stable (or at least slow-moving), natural-seeming, and therefore intuitively  [*617]  graspable (at least in their broad outlines) foundation on which individuals may securely plan their economic and personal lives. 16

C. Is the Internet a Revolution?
 
This is where Lessig's premise about the radical change in virtually everything touched by the Internet comes in. If the Internet represents the manmade equivalent of something like a reversal of the laws of gravity, then it radically and definitively disrupts the stable, intuitive baseline upon which the public/private distinction rests - and the plans we make to organize our future starting from that baseline. For Lessig and other Internet theorists, we are entering something like those science fiction worlds in which the laws of physics and biology and psychology, our most fundamental assumptions about the forms and limits of our universe, no longer hold. That is why he and many others use the terms Internet and cyberspace almost interchangeably: the Internet has worked such an alteration in our environment that it has propelled us into a new domain, which is what the term cyberspace denotes. In such a changed world, it makes no sense to seek to maintain privacy, freedom of expression (or liberty generally), or property rights (especially intellectual property rights) by applying the same old rules. If we insist on using the pre-Internet world's conceptual and doctrinal framework, we will in this new world get results that are at best randomly related to the values we seek to uphold, and we will more probably end up systematically undermining those values. This is Lessig's premise.

Certainly our situation today is very different from what it was at the end of World War II. But has the Internet introduced a radical discontinuity into the process of change? It is worth pausing for a moment to ask just what it is about this new technology that portends such a revolution in the way we live. It is not just the ease of communication, because we had cheap and ever cheaper communications networks ten and twenty years ago. The fax machines of that time allowed the instant transmittal of documents that previously would have required days to reach their destination. Radio for almost a century, and television for a half century, have allowed the transmission of sounds and then images anywhere, instantly, though they were only one-way lines of communication: the audience could not talk back. And air travel - which has not gotten quicker, though it is cheaper than a quarter century ago 17 - knitted together peoples and places  [*618]  previously made strange and unreachable by long distances. Does the revolution lie in the ability to access and manipulate vast amounts of information? That is the work not of the Internet, but of the computer, which entered the scene in a general way in the late 1960s and 1970s. I recall using Lexis and Nexis for research purposes in the early 1980s. At that same time, airline reservation systems and FBI computers made possible the nearly instantaneous manipulation of so much data that even then people were speaking in the same millenarian terms about the computer - its threats and its promise - as they are now about the Internet. 18

It seems to me that what is special about the Internet comes from the confluence of the computer, which affords near-instant access to and easy manipulation of vast amounts of information, and today's improved, almost costless channels of communication. It is this combination that makes it possible for large numbers of people to reach each other instantly and interactively, in whatever groupings they wish. Largely because of their commercial potential, the lines of communication have been progressively improved so that they can now quickly accommodate very complex "messages" consisting of images - including moving images - and sound. In the past, hundreds of millions of people could talk to each other on the phone or receive sounds and images broadcast from anywhere on the globe; now these same people can broadcast as well as receive images and sounds, reaching as wide an audience as wishes to watch or listen. Of course this is a vast change for commerce, for the exchange of ideas and information, and perhaps even for the possibilities of human social connection. Is it such a change that we should treat it as a metaphorical new world and give it the name cyberspace, a new place where people live? If it is, then Lessig's premise does indeed entail his claim that we must rethink the baselines.

1. Is Cyberspace a Place? - First, let me vent some irritation at what seems to me a hyperbolic, if not somewhat fatuous use of the cyberspace metaphor. Lessig tells several stories, some real, about persons who commit harm in cyberspace. For example, there is the famous case of the cyberspace rapist (pp. 74-78). 19 One of the things people have done with the virtually instant, interactive, multi-person  [*619]  communications allowed by the Internet is to tell stories to each other. The participants create characters - one for each of them - who interact with each other, in the way that children sitting around a campfire might devise a round-robin or chain story. It seems that one particularly obnoxious participant in such an exercise had his fictitious character rape the characters of some of the other participants. Now, this may be - it almost certainly is not - punishable as obscenity; even more remotely, but still possibly, it may count as the communication of some sort of threat. 20 But one thing it is not is rape. Or, rather, it is rape only in the same misleading and metaphorical sense that, as Catharine MacKinnon has argued, pornography violates the civil rights of women. 21 It is not rape for the simple reason that no actual human beings came into physical contact with each other. It is offensive, in the same way that sending this kind of story to someone in the mail would be, and that is why the threat charge comes closer to the mark. There is a further offense here, too. The participants in this group had obviously invested time and ingenuity in this joint effort. They trusted each other as the participants in any joint activity - say, a touch football game - trust each other. This jerk violated their trust, spoiled the fun, and made the others feel bad. All that is obnoxious, but it is not rape. I am afraid people disappoint and offend each other and spoil games, reading groups, and friendships casual and longstanding all the time. These are the sour notes in the symphony of life. Except when such things go over the line and become threats of actual physical harm, it would be intolerable if the law had anything to say about them. That it happens on the Internet changes nothing, yet Lessig is too willing to indulge, even for consciousness-raising purposes, in this kind of oratorical hypertrophy. 22 It does not help.

Lessig says things like "Cyberspace is a place. People live there" (p. 190). (He also says "Cyberspace is not, of course, a place; you don't go anywhere when you are there" (p. 17).) People live in Cambridge, Massachusetts, or Palo Alto, California, or wherever, and from their computers there they communicate over the Internet. They tell stories, write love letters, buy and sell goods, search for data on rates of tuberculosis infection in Angola - just as they might by telephone or letter. But we do not speak of people living in telephone space or postal space. True, there are people who spend hours before their computers  [*620]  - to the detriment of their relations with the human beings with whom they might otherwise come into physical contact. But there were people in the past whose whole affective lives took place via correspondence - just look at the genre of the epistolary novel. No big deal here.

2. Sovereignty. - Lessig's overstatement of the implications of the Internet for the broad public/private distinction is mirrored in the magnitude of the change that he foresees in the traditional institutions that take that distinction for granted. Lessig's entire chapter on sovereignty (pp. 188-209) is not hard-headed enough about the Internet hype. In it he argues that the Internet has radically undermined the concept of national sovereignty. Now we are coming to one of those phenomena that Lessig claims have changed the environment to such an extent that the old categories do not retain their meaning. Lessig considers Internet gambling (pp. 54-55, 198), Internet pornography (p. 191), Internet commerce (pp. 196-97), and what some regimes might deem Internet sedition (p. 205). His point is that, to the extent that the Internet makes it harder for the state to regulate these activities, it undermines the state's sovereignty, defined as the ability to lay down the rules according to which activities within a jurisdiction go forward. And he argues that the Internet has indeed made such regulation harder (p. 192). I am not convinced. Lessig distinguishes between sophisticated and unsophisticated actors (p. 197), the latter needing more protection from the state. The sophisticated can take care of themselves. I would distinguish first between transactions/communications between willing (even eager) parties and those between parties who are not or do not remain willing participants.

Take Lessig's example of gambling (pp. 14, 54-55). A state may wish to prevent its citizens from gambling, and at the level of the local numbers game or betting parlor, it has some ability to enforce its judgment. But on the Internet, Lessig tells us, the betting parlor can put itself in another jurisdiction, beyond the reach of the state, and the state's policy judgment is thus frustrated (pp. 54-55). But how is this different from citizens traveling to Las Vegas and dissipating the family nest egg there? Or buying lottery tickets, or betting on the Super Bowl by mail or by telephone? Well, it is very much easier, and being easier it is more likely to cater to the impulsive gambler, who may be just the person the state wishes to protect. Yes, it is easier - for the gambler, not for the gambling "house," which must be paid. Enter the unsung hero (unsung by Lessig) of Internet commerce - whether of retail merchandising or of the provision of pornography and gambling: the customer's credit or debit card. The satisfied customer presents the greater problem for the state regulatory policy (and so he should),  [*621]  because the credit card issuer is also likely to be beyond the state's jurisdiction and the customer will pay his monthly bill to the issuer by a check mailed out of the state. 23 But if the customer later balks, he may be able to countermand the charge or refuse to pay it. Then the merchant or card issuer will have to go after the customer, at which point the state can implement its sovereign authority over the transaction very much as if it had taken place on Main Street between two local residents.

The online gambling establishment could defeat local regulation by requiring the consumer to deposit funds in an account in the same offshore bank used by the provider. The consumer would then authorize an immediate debit to his account with each transaction. This device is cumbersome and requires the consumer to pay in advance and accept the risk of improper debits. How he is to repatriate his winnings, if any, is yet another problem. Lessig says that to the extent that we have had tools to deal with these issues, they have addressed interstate conflicts between institutions or large repeat players, not conflicts at the level of small-time, individual consumers (p. 193). This is surely wrong. There have been interstate consumer transactions as long as there have been interstate travel and freight shipment, and the books are full of cases involving conflicts of law rules and interstate judgments at just this level. Moreover, to the extent that credit or debit card issuers become involved, we are indeed at the level Lessig says the system is used to.

These are issues to be gone into more deeply. I do not fault Lessig for only whetting our appetites, but I do fault him for the tone of his discussion. He does not mention these aspects of the question, but writes as if, were we to go into it, we would find that the Internet does indeed seriously undermine the state's ability to govern its citizens. 24 He gives us no way to judge how true this is. And I certainly fault him for the consequently unjustified leap into the realm of metaphor when he asks us to view cyberspace as a sovereignty overlapping traditional sovereignties, to be governed by a legal regime appropriate to that level (pp. 198-203). True, the cyber-vandal who invades your hard drive from his sanctuary in the Philippines may be hard to get at, but no more so than the terrorist who puts a bomb aboard a plane in Malta, timed to explode over Lockerbie, Scotland. Crimes and torts in which the harm is started in one jurisdiction and produced in another have long been a staple of the conflict of laws.

 [*622]  Although there is nothing qualitatively new in these interstate transactions, Lessig is correct in one respect: the growing ease of interstate crimes and torts means that their quantity will increase (perhaps by orders of magnitude). That itself may pose novel regulatory problems. Similarly, wire, fax, and telephone transfers of funds and the use of coded messages authenticated by secret passwords are nothing new, but their greater ease has made international money laundering - as well as transfers not associated with criminal activity - orders of magnitude harder to track and control. Harder, but not impossible: witness the unraveling of at least some part of the Bank of New York scandal and the resulting successful criminal prosecutions. 25

In the meantime, as Lessig tells the tale, code writers have themselves provided some of the tools that have allowed government surveillance to catch up with Internet communication (p. 177). Sophisticated filters are now able to search through millions of messages for telltale bits of code sufficient to attract attention and to warrant further inquiry. 26 And the code writers on the other side are busy finding ways to defeat these filters. 27 But cops and robbers is an old game, too. Once again Lessig whets our appetites, but the real meal is in the details, which he does not supply.

III. Intellectual Property

A. The Old World and the Brave New World
 
Copyright presents a signal test for Lessig's thesis that the regulation that code enables threatens to alter radically the current balance between liberty and regulation. Although he devotes a compelling and extensive discussion to it (pp. 122-41), I would peel the subject all the way back to reveal what it would look like under the baseline intuitive concepts of property and contract. You write a novel or paint a picture. The baseline concepts are very material-minded: the book or painting is a material object that belongs to you, like the fruit you grow in your orchard or the false teeth you buy from the dentist or your spare kidney. You can show these objects to whomever you wish. You may let them out of your possession by lending them to a friend (bailment - gratuitous or not) or transferring them by sale or gift.  [*623]  Each of these transactions may be accompanied by contractual terms: for example, that the transferee not show them to anyone else, not copy them, or pay a fee each time he looks at them or lets someone else look at them. But if there are no such terms, the book or painting is now in the possession of someone else, who violates no property right of yours - viewed in these materialist terms - if without paying you a penny he makes and sells a million copies of your book, charges admission to view your painting, or uses the theme, title, and characters of your novel to make a movie. If you did not want any of these things to happen - at least without payment to you - you should have kept your book or painting to yourself or made a contract restricting what might be done with it. If someone had broken into your home and read your novel, copied it, or taken it away, you could sue him for trespass, and his earnings might be part of your consequential damages. If you had made a contract when you transferred your book or painting, you could sue to enforce that contract. But if you did not, you have nothing to complain about. And that is about how things stood before the introduction of copyright law in the sixteenth century. 28

Today, the availability of some kind of protection for my intellectual property seems almost as natural as protection for my false teeth, but that is just the point of the analyses of the Critical Theorists and Sunstein, 29 which Lessig adopts: that property rights are socially constructed, and deliberately so, for a particular purpose (p. 131). Article I, Section 8 of the Constitution includes among the powers of Congress the power "to promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries." 30 The socially constructed, instrumental point of this grant of power could hardly be more explicit. It stands in contrast, I would point out, to Section 10 of the same article, which forbids states from "impairing the Obligation  [*624]  of Contracts," 31 and to the Fifth Amendment, which prohibits the national government from taking "private property ... for public use, without just compensation." 32

Lessig follows Stephen Breyer's canonical account in explaining the "uneasy" instrumental case for copyright protection (p. 133). 33 The heart of that case, obviously, is the socially enriching effect of providing an incentive to authors and inventors, just as the Constitution says. But there is a large social advantage to free access to intellectual goods as well, and Lessig pauses lovingly over that side of the equation (pp. 134-35, 141). In the end, the law strikes a balance, again just as the Constitution invites, by providing that copyright protection should last for only a "limited time" (p. 133). And as Lessig also points out, there are other elements in the balance too. He particularly emphasizes the copyright law's doctrines of "fair use," which allows a limited amount of copying and the use of short quotations in one's own writings, and "first sale," which gives purchasers of printed works the right to give away or sell their copies (p. 134). I would add the bedrock but fuzzy doctrine that only expressions of ideas, not the ideas themselves, may be copyrighted. 34 These features of the law strike a balance between incentives for authors and ready access for the public that the law has determined best benefits society. None of the items on this balance is natural, although some may see something natural about an author's right to profit from his work (even after he has let it out of his hands?), and others may see something natural about unrestricted access to and dissemination of expression (even if it resides in a single copy in my desk drawer?).

B. Intellectual Property on the Internet
 
Lessig's chapter on intellectual property (pp. 122-41) shows how the Internet threatens to upset this balance. Doomsayers initially feared that it would do so by undermining authors' rights. Their frightful picture was of copyrighted material posted on the Net and available for free to millions. The opportunities for dissemination were enormous, and in this sense represented a boon to authors, at least those who did not care about getting paid. (Some authors do not care. Others - like academics, sometimes - are paid to write and not paid for their writings.) Lessig points out that this free availability was an artifact of code, of the Internet's architecture (pp. 139-40). But  [*625]  what if code could be written to reverse the situation completely, so that Internet publishing could be the scene not of perfect freedom, but of perfect control? Code could be written to enable publishers to preclude or charge for downloading, or charge by the second (as Lexis and Westlaw now do) for each viewing. And indeed, code has already been written that allows a publisher to condition access to ("clickwrap") his material on the user's agreement to the publisher's terms regarding the material's use and further dissemination. Lessig worries that the possible totally "propertized" Internet would upset the balance our present copyright law strikes and destroy the extensive "intellectual commons" we presently enjoy, with unknown but probably terrible effects on the circulation of ideas and freedom of expression (pp. 140-41). As Lessig says, "the law was never designed to simply do the author's bidding" (p. 129). 35

Let us step back and ask whether it might be said that a Web publisher who successfully uses such a technology of perfect control is not making an end run around the balance struck by the copyright law because he is not even relying on their protection. He is going all the way back to the minimal protection that the background laws of property and contract offer. Imagine that the publisher kept his material in a locked drawer in his office and invited you to read it there only if you agreed to strip to your underwear and come in without paper or pencil. (Lawyers and government agencies sometimes follow a version of this arrangement in allowing the inspection, but not the duplication, of documents they would rather not let you see, but which they doubt they could keep from you in an all-out legal fight.) If a lot of important information were available only on this basis, this would certainly also upset the balance struck by the copyright law's time limit, first sale, and fair use provisions. Or imagine that the publisher conditions your inspection of his material on your signing an agreement about its dissemination. Is this not just what the Web publisher is doing? And if the underwear ploy is compatible with the existing legal balance, why is code-or clickwrap-restricted access to a website not similarly consistent?

Lessig would be right if publishers could substitute the protections of code for those of copyright and thus, in effect, substitute the locked drawer of architecture for the nuanced, publicly enacted restraints of law. The balance between control and free access to information would be drastically upset. It is much the same point as I made about e-commerce: the Net allows vast, instant, and almost costless (certainly  [*626]  to the marginal recipient) dissemination. The reason publishers embraced copyright was that controlled (underwear) access and access conditioned on individual contracts were incompatible with any but the most limited dissemination. Publishers gave up the absolute legal control of property and contract law for the limited control of copyright law because it was an exceedingly advantageous trade all around. So Lessig is right, because the context of the Web has made a difference of degree so vast that it surely counts as a difference of kind. The attractions of Web publishing are so great, indeed, that it may lead publishers to decline the copyright compromise; it is not fanciful to see the Web replacing an important segment of paper publishing. And if that happens, then a fully propertized regime might well replace our present partially and imperfectly propertized regime. Lessig does not demonstrate that such a change necessarily would be a bad thing, but he insists that it should be made with no less public deliberation and control than have accompanied the development of the copyright law thus far (p. 140).

C. How New Is the Brave New World?
 
It is a theme of this book that there are public choices to be made and that some of these choices have constitutional dimensions. Well, how exactly would we go about exercising this public control? As for clickwrapping, it is not hard to imagine a doctrinal development that would refuse enforcement of some of these contracts as against public policy (recall my discussion of online gambling n36) or as preempted by the copyright law. Oe distinguished judge who has considered this issue has decided that such contracts are like any other contract and should not be subject to any special legal regime. 37 (This is the same judge who has written that, more generally, there is no need to develop any special legal regime for the Internet.) 38

Such legal responses are not particularly novel. After all, the publisher who tries to enforce a clickwrap contract invokes the state's help, and he must expect that the state will impose its conditions and policies in granting it. But what of restrictions that the publisher effects by resort to code alone - the equivalent of the locked drawer - so that access depends on payment and usage is metered? Once again, as in Internet gambling, the restrictions' success depends on enforcing payment of the metered charge, usually by credit or debit card, and the state may declare some of these arrangements void as against public  [*627]  policy and deny enforcement of the charges. The conceptual crunch comes only if a publisher denies access to his site altogether, or allows access only to certain (code-designated) users, or allows access but codes his offering in such a way that the user cannot download or retransmit it. Legal intervention at this point would be a more radical departure from traditional property concepts. It would be the equivalent of forcing me to allow the public to enter my private library or to open my desk drawer and read my documents (not in their underwear). A state requirement that publishers grant access to their websites might be said to violate the Fifth Amendment's prohibition against an uncompensated taking of property, just as did an order by the California Coastal Commission requiring that landowners allow strangers to pass back and forth across their beach. 39 This argument, then, would be taking Lessig's (sometime) claim that cyberspace is a place, one that can be zoned and subjected to the same legal regimes we apply to other places, quite literally and to a different conclusion (one that argues against state intervention) than the one for which Lessig generally argues in this book.

But in fact, events have very recently taken a turn in quite a different direction. Universal City Studios, Inc. v. Reimerdes, 40 for example, shows that publishers will indeed try to exert perfect control over their product, just as Lessig fears. But it also shows that code (or architecture) will likely be unable to do the job for them and that the publishers will have to turn to the law after all, just as they did centuries ago, when they turned to the law to enact the regime of copyright. The point is implicit in an argument that Lessig himself makes about encryption (pp. 35-40): the more secure an encryption system is against hackers who would defeat it, the more incompatible with mass distribution it becomes, just as the locked drawer is very secure, but quite incompatible with mass dissemination - and mass dissemination is how you make money in this business. Reimerdes concerned the motion picture industry's marketing of digital versatile disks (DVDs), on which full-length motion pictures are stored and played back with far greater fidelity and convenience than videotapes permit. The same digital technology, however, that makes this convenience and fidelity possible also makes possible the almost perfect reproduction of the disks. (This contrasts with videotapes, whose quality degrades with each successive copy.) To protect their product from such ready copying while maintaining mass distribution, the publishers encrypted the  [*628]  information on the disks, using a code called CSS, and released the key only to those manufacturers who agreed to design DVD players so that the message on the disks could not be copied. 41 In effect, this was an attempt to allow only a kind of limited "underwear" access to purchasers of DVDs. But it was not long before ingenious hackers (the Reimerdes case involved a fifteen-year-old hacker in Norway) were able to write code, DeCSS, that defeated the encryption and copy-protection system and thus allowed widespread unauthorized distribution of the publishers' product via homemade discs or over the Internet. 42 The defendants in the case had posted on their website this decryption code, as well as links to other websites with decryption codes. 43 The availability of this decryption code created an intolerable situation for the publishers (or rather motion picture distribution companies), who faced losing control of their product and of the revenue they hoped to earn from it.

The story of CSS and DeCSS illustrates why, contrary to Lessig's assertion, code need not be a technology of perfect control after all: as in physics, so in cyberspace - for every particle there seems to be an antiparticle. And in fact, the DVD publishers did not rely only on architecture (very high cyber-walls, or tightly locked cyber-drawers) to maintain control over their product. They had recourse to the law, just as publishers in an earlier day had recourse to the law to enact and enforce copyright protection. In this case, foreseeing the eventuality of something like DeCSS, the publishing industry, pursuant to the World Intellectual Property Organization Treaty, 44 had obtained from Congress something called the Digital Millenium Copyright Act (DMCA), 45 which made it unlawful to "circumvent a technological measure that effectively controls access to a work protected under this title." 46 Analogizing the use of DeCSS to picking a lock, 47 a federal district court permanently enjoined the defendants from making DeCSS available, via the Net or otherwise, and from publishing links to other websites where DeCSS might be available. 48

No doubt if someone picked the lock on a publisher's desk drawer, he would indeed be a thief under traditional legal categories. Similarly, if he supplied the lockpick, he would be an accomplice, an accessory before the fact of burglary, and criminally liable for the possession  [*629]  of burglary tools. But the use of the analogy exactly begs the question: breaking into a locked drawer is an unambiguous violation of traditional property rights. Those rights are part of a web of laws that strike a familiar balance among a large variety of claims: you may not break into another's desk drawer, but if he leaves his information lying about in a common area, or lends it to you, or tells you about it and you have a perfect memory, he has no complaint. Copyright laws, as Lessig shows, extend that protection but maintain a balance. The publisher can now let his product out of his hands, secure in the knowledge that the law will protect him - up to a point: the protection covers only certain aspects of the work; it lasts only for a limited time; and (most relevant here) the protection does not prevent the reader from making fair use, including limited dissemination for classroom use, of short quotations in reviews and commentary. 49 The analogy to picking locks cuts off consideration of the fair use limits on copyright protection. Fair use does not protect a burglar but does sometimes protect a copyist. In the DVD context, since the copyist has bought and paid for the DVD, the court's analogy assumes that the object stolen is the information on the DVD and not the DVD itself, an assumption that cuts off consideration of the copyright law's first sale doctrine, which allows a purchaser of copyrighted material to pass it on to others. Thus the prior balance is threatened and a new balance, impinging upon what Lessig calls the intellectual commons, is struck. 50 But it is not code that does it; it is plain, old-fashioned law. In Reimerdes, the district court did not reach the defendants' fair use claims, finding that there was no evidence that anyone's fair use had in fact been inhibited or that the material was not available in other formats  [*630]  or not available for an additional fee. 51 The latter two findings either are beside the point or beg the question.

It will be interesting to read how Lessig develops his general notions to address the many specific questions this area of intellectual property raises. It is, I understand, the subject of his next book.

IV. Free Speech

A. Architectures of Control
 
Lessig's discussion of free speech is at once the most suggestive and the most inconclusive part of Code. He offers a veritable paean to free speech on the Internet:
 
But on top of this list of protectors of speech in cyberspace is architecture. Relative anonymity, decentralized distribution, multiple points of access, no necessary tie to geography, no simple system to identify content, tools of encryption - all these features and consequences of the Internet protocol make it difficult to control speech in cyberspace. The architecture of cyberspace is the real protector of speech there; it is the real "First Amendment in cyberspace," and this First Amendment is no local ordinance.

Just think about what this means. For over fifty years the United States has been the exporter of a certain political ideology, at its core a conception of free speech. Many have criticized this conception: some found it too extreme, others not extreme enough. Repressive regimes - China, North Korea - rejected it directly; tolerant regimes - France, Hungary - complained of cultural decay; egalitarian regimes - the Scandinavian countries - puzzled over how we could think of ourselves as free when only the rich can speak and pornography is repressed.

This debate has gone on at the political level for a long time. And yet, as if under cover of night, we have now wired these nations with an architecture of communication that builds within their borders a far stronger First Amendment than our ideology ever advanced. Nations wake up to find that their telephone lines are tools of free expression, that e-mail carries news of their repression far beyond their borders, that images are no longer the monopoly of state-run television stations but can be transmitted from a simple modem. We have exported to the world, through the architecture of the Internet, a First Amendment in code more extreme than our own First Amendment in law. (pp. 166-67) 52
 
But as throughout this book, the picture soon darkens: "The right to free speech is not the right to speak for free.... In the United States ... constitutional protection [of free speech] is a protection against the government. Nevertheless, a constitutional account of free  [*631]  speech that thought only of government would be radically incomplete" (p. 164). Here Lessig invokes Fiss and Sunstein (p. 275 n.1). Those authors have regularly insisted on several points. Principally, they argue that private power can be as effective a censor as the government. Far more radical is what I call their "drown out" argument: 53 that private power, even when it does not actually block anyone's speech, drowns it out by flooding the channels of communication with more readily accessible or more entertaining, less substantial subject matter. MTV beats out C-SPAN every time, even though both are equally available at the click of a remote. 54 Both arguments show up in Lessig's account (pp. 164, 185). And, of course, the villain of this analytical piece is the doctrine of state action. The First Amendment says, "Congress [that is, government] shall make no law ...abridging the freedom of speech, or of the press." 55 But if private power cannot be disciplined, then we cannot realize what Sunstein and others, 56 possibly Lessig himself - though he is a bit noncommittal on this - call the Madisonian conception of the First Amendment (p. 180), which sees the protection of speech as directed at ensuring democratic deliberation through open and general political discourse. 57 This means that we must get past the state action doctrine and discern constitutional protections against private as well as public power and, more important, that we must be willing to accept government limitations on private speech in order to guarantee an equal hearing to all points of view. Lessig - once again ambiguously and haltingly - appears to endorse this argument (p. 181).

As throughout this book, Lessig is torn between the dark and the bright sides of the Internet, both visions enabled by code. Using the  [*632]  example of the perceived need to control children's access to indecent material on the Internet, Lessig lays out alternative modes of control: zoning and filtering (p. 173-74). What Lessig calls zoning (p. 175) 58 is a system of code that restricts access to defined categories of people. Government might, as it did in the Communications Decency Act, 59 make punishable the transmission of indecent material. 60 Broadcasters would then have to institute a system allowing access to such material only to persons holding an appropriate credential - for example, a credit card or password - that is unlikely to fall into the hands of children. 61 True, there would be a great deal of leakage, but the same is true of the purveying of such material by traditional means. The Supreme Court has invalidated regulation of this kind because it imposes too heavy a burden on potential receivers, who must identify themselves, and on broadcasters, who must both monitor the content of material and maintain costly and imperfect systems of authentication to enforce the zoning restrictions. 62 The alternative method of control is a filtering mechanism, which can be placed anywhere along the line to detect and block further transmission of material having whatever characteristics the filter's authors code into it (pp. 177-78). 63 Either the government or private parties may institute zoning or filters. China and North Korea may - and probably do - filter out politically disturbing material. 64 Internet service providers might filter out certain transmissions to save their subscribers from annoying commercial e-mail (spam). End-users might place filters for a variety of purposes: parents, to filter out indecent, or violent, or anti-religious, or proselytizing material from their children's machines; businesses, to filter out messages sent by particular individuals or corporate competitors from their employees' machines.

 [*633]  Lessig emphasizes that both technologies could be used by government and by private entities (p. 181). He expresses a preference for zoning because it is transparent relative to filtering (p. 181); that is, you may not get the material either way, but zoning at least lets you know that you have been kept away from it, and such knowledge is the precondition of opposition and resistance. As to government, Lessig certainly has a point - it is his general point about cyberspace potentially being a place not of perfect freedom but of perfect control. But government cannot insert filters into the Internet in secrecy. There is little government can do in total secrecy; it is simply too big. It might openly mandate such filters, and then it would be susceptible to First Amendment challenges. Or it might order service providers, backbone operators, or even machine manufacturers to install such filters (in a similar manner as legislation that requires television manufacturers to install the so-called "V-chip"). This, too, would become apparent and provoke resistance. I doubt, however, that the government could sneak filters into the system via some kind of virus-like device - at least not without being detected and inviting countermeasures by ingenious code-writers.

The real specter of total control comes not from government, then, but from private entities entrenched along the pathways of the Internet. Here is where Lessig's theses about the Internet as a whole new realm of existence and about the need to "translate" constitutional doctrines into this new realm come together. I turn to Lessig's fascinating discussion about the role of editors in the dissemination of speech (pp. 171-72). He points out that at the Framing there were - to use antitrust lingo - very few barriers to entry into the marketplace of ideas (p. 183). All it took was a printing press, a pulpit, or a platform. People read a large number of newspapers and pamphlets and listened to sermons and public orators. More recently, until the advent of the Internet, things had changed because of the vast proliferation of speakers and the rapid means of reaching distant audiences. Audiences felt they needed filters to help them sort out the information that really interested them and to certify the reliability of the information they received. These are functions editors supply: for some of us, The New York Times; for others, People or network news. These media channels - the large dailies, the popular magazines, the TV networks - not only performed these functions, but also overwhelmed other, smaller, more idiosyncratic speakers (pp. 171-73), as I have described, "drowning them out."

 [*634]  The upshot for Fiss, Sunstein, and others was a challenge to what they called First Amendment "values." 65 The public heard only certain voices, and private parties had inordinate power to choose which voices and thus to limit and shape the processes of democratic deliberation. The Internet has changed all that. Barriers to entry are once again very low; practically anyone can become a publisher and reach an audience of millions. But not only is the audience orders of magnitude larger than at the time of the founding, but so is the universe of speakers. Who will perform the editing and certifying functions in cyberspace? Lessig points out that the same filtering codes that can keep dirty pictures and violent stories from our kids could not only filter out for us boring screeds by cyberspace crackpots, but also shelter us from Republican or Democratic opinions or, indeed, filter out everything but communications about golf and our grandchildren (p. 180). 66 Designing and "customizing" cyberspace filters could become an important and lucrative service. 67 But it has the entailment for First Amendment "values" that I can insulate myself from unwanted messages - not only I, but millions like me - and thus that we are back in the pre-Internet, post-Framing condition, in which only a few voices get through and "editors" plug the channels of equal and open democratic deliberation.

Other threats to First Amendment "values" 68 are the pervasively "propertized" regime of intellectual property I have already discussed 69  [*635]  and the more obvious threat that major intermediaries - for instance, those who operate the spine of the Internet, or the service providers - might take it upon themselves to engage in more active and fine-grained filtering than they do now. Is Lessig right that both developments represent so radical a change in the communications environment that we must once again translate the Framers' work to make the First Amendment relevant to the new context?

B. The State Action Doctrine and Free Speech on the Internet
 
The state action doctrine is the key to Lessig's argument (p. 217). Lessig believes that some revision of the doctrine is essential, and he invokes the German example to show that it can be done (p. 218). 70 I am not so sure. Note first of all that the state action doctrine does not only limit the power of courts to protect persons from private power that interferes with public freedoms. It also protects individuals from the courts themselves, which are, after all, another government agency. By limiting the First Amendment to protecting citizens from government (and not from each other), the state action doctrine enlarges the sphere of unregulated discretion that individuals may exercise in what they think and say. In the name of First Amendment "values," courts could perhaps inquire whether I must grant access to my newspaper to opinions I abhor, 71 must allow persons whose moral standards I deplore to join my expressive association, 72 or must remain silent so that someone else gets a chance to reach my audience with a less appealing but unfamiliar message. Such inquiries, however, would place courts in the business of deciding which opinions I would have to publish in my newspaper and which would so distort my message that putting those words in my mouth would violate my freedom of speech; what an organization's associational message really is and whether forcing the organization to accept a dissenting member would distort that message; 73 and which opinions, though unable to attract an audience on their own, are so worthy that they must not be drowned out by more popular messages. I am not convinced that whatever changes the Internet has wrought in our environment require the courts to mount this particular tiger.

 [*636]  As to the sheer volume of messages in cyberspace, the potential number and variety of filtering and sorting codes from which one may choose constitute an adequate guarantee that listeners are free to listen to what they want. This system is different from, and freer than, the dubious Red Lion regime, under which the government allocated the broadcast spectrum according to its conception of "public "convenience, interest, or necessity.'" 74 In cyberspace, each user devises or chooses his own criteria for preventing channel overload - much as we decide which magazine to buy when we view the array at a well-stocked urban newsstand. It may be reasonable to worry that such a regime will filter out what some wise man believes the consumer should hear, but it is an Orwellian use of language to seek to regulate such filtering in the name of "free speech." This is the argument from "false consciousness," which I had thought went down with the Berlin Wall. It seems to me that in this respect Lessig has sketched a technology of near-perfect freedom.

Similar to the threat posed by a totally "propertized" regime of ideas is the problem presented by the Internet service provider, which may choose to filter out certain messages. Lessig is right about the danger that free speech would face if, in a Microsoft-or AOL-dominated world, the dominant entity could install filters that controlled what all of us could see and hear. This is where another of Lessig's several fields of expertise, antitrust, comes in. 75 The antitrust laws should ensure enough competition in cyberspace that if we do not like AOL's filtering policies, we can switch to MediaOne. But even this kind of consumer choice may not be enough in a transformed world in which everyone depends on the Internet to send and receive messages. After all, competitive regimes serve consumer preferences only in the aggregate. If some consumers want channels that filter out violence, others indecency, and still others extremist hate messages, there may be options that cater to each of them, but what of the minority that wants to filter out none of these things? (Competition in the automobile industry gives me a pretty wide range of choice, yet it does not guarantee me just the shape, color, and kind of car I might like.) In First Amendment law, the doctrine of the public forum guarantees a space where the most marginal and idiosyncratic of speakers may try to attract an audience. 76 But the public forum is traditionally  [*637]  a government-owned space, and the parks and streets of the Internet are privately owned. The law might respond to this situation in the name of First Amendment "values," and without abandoning the state action doctrine, by designating AOL and the like as common carriers and obliging them to carry any message that will pay its way. 77 The law might, but then again it might not. We would then have to decide whether a private citizen might sue AOL by invoking the constitutional provision that "Congress shall make no law ... abridging the freedom of speech," 78 even though Congress had made no law (code) - only AOL had. A possible argument would be that AOL was the cyberspace equivalent of the company town, which the Supreme Court held in Marsh v. Alabama 79 was subject to the First Amendment like any other town. 80 That would be a stretch, because the company town in Marsh had assumed all the functions of a usual municipal government. 81 A later case, Jackson v. Metropolitan Edison Co., 82 held that even a regulated monopoly that performs services that government may, but need not necessarily, perform itself, is not for that reason a state actor subject to constitutional strictures. 83

Here is one scenario in which free expression could triumph over AOL without our abandoning the state action doctrine: An ingenious hacker writes code that can both detect filters and defeat them. 84 He sells it or - in the spirit of cyber-anarchy - gives it away. AOL sues in contract or tort to block the use and dissemination of the code, and a state court agrees. Might the hacker and AOL subscribers not argue in the Supreme Court that such a court judgment violates the First Amendment, citing New York Times Co. v. Sullivan 85 and Shelley v.  [*638]  Kraemer? 86 Could AOL fight back, citing Jackson? This is a version of the scenario that is playing out with respect to DeCSS, as I have discussed above. 87 The DVD publishers there turned to the law to enforce what they considered their rights. Shelley makes clear that an injunction is state action, 88 and so it would be in the DVD case too. Of course, the publishers who obtained the injunction in Reimerdes relied on a statute, the DMCA, which the defendants argued violates the First Amendment and the limits implicit in the Constitution's Copyright Clause; 89 an injunction obtained to enforce the prohibitions of that statute surely counts as state action. 90 It is not clear, therefore, that the state action doctrine presents quite the obstacle Lessig suggests to the maintenance of the intellectual commons.

* * *

Lessig raises many questions. For answers to these questions and new questions arising from those answers, read this book, and then read what Lessig writes next.

FOOTNOTES:
n1. See generally Lawrence Lessig, The New Chicago School, 27 J. Legal Stud. 661 (1998) (introducing the "New Chicago School" approach to the question of regulation).

n2. See generally Cass R. Sunstein, Democracy and the Problem of Free Speech (1993) [hereinafter Sunstein, Democracy] (discussing the relationship between political sovereignty and the free speech principle); Cass R. Sunstein, The Partial Constitution (1993) [hereinafter Sunstein, Partial Constitution]; Morris R. Cohen, The Basis of Contract, 46 Harv. L. Rev. 553 (1933) (discussing the overlap of public and private law as it relates to a comprehensive theory of contract); Morris R. Cohen, Property and Sovereignty, 13 Cornell L.Q. 8 (1927-1928) (discussing the changing meaning of "property" and arguing that a certain domain of sovereignty has passed from the state to private individuals); Robert L. Hale, Bargaining, Duress, and Economic Liberty, 43 Colum. L. Rev. 603 (1943) (arguing that government and law play a more significant role in a "free economy" than is generally realized); Robert L. Hale, Coercion and Distribution in a Supposedly Non-Coercive State, 38 Pol. Sci. Q. 470 (1923) (arguing that coercive restrictions on individual freedom permeate laissez-faire systems); Robert L. Hale, Force and the State: A Comparison of "Political" and "Economic" Compulsion, 35 Colum. L. Rev. 149 (1935) (arguing that the political power of the state is the same kind of power that some private individuals can exercise against other private individuals); Robert L. Hale, Value and Vested Rights, 27 Colum. L. Rev. 523 (1927) (discussing the effects of government action on property rights); Morton J. Horwitz, The History of the Public/Private Distinction, 130 U. Pa. L. Rev. 1423 (1982) (arguing that the erosion of the public/private distinction stemmed from the perception that private institutions were acquiring coercive power that had been reserved for governmental entities); Duncan Kennedy, The Stages of the Decline of the Public/Private Distinction, 130 U. Pa. L. Rev. 1349 (1982) (describing the later stages of the distinctions, particularly the public/private distinction, that constitute the "liberal" mode of thought).

n3. Robert Nozick, Anarchy, State, and Utopia 307-09 (1974).

n4. For elaborations of the cyber-libertarian argument, see Llewellyn J. Gibbons, No Regulation, Government Regulation, or Self-Regulation: Social Enforcement or Social Contracting for Governance in Cyberspace, 6 Cornell J.L. & Pub. Pol'y 475, 509-32 (1997); I. Trotter Hardy, The Proper Legal Regime for "Cyberspace", 55 U. Pitt. L. Rev. 993, 1026-36 (1994); David R. Johnson & David Post, Law and Borders - The Rise of Law in Cyberspace, 48 Stan. L. Rev. 1367, 1378-87 (1996); and David G. Post, Governing Cyberspace, 43 Wayne L. Rev. 155, 165-71 (1996).

n5. See Steve Bickerstaff, Shackles on the Giant: How the Federal Government Created Microsoft, Personal Computers, and the Internet, 78 Tex. L. Rev. 1, 38 (1999).

n6. Lessig also says that he does not "pick out commerce to pick on commerce .... [It] will flourish in the future of the Internet, and no doubt that is a good thing" (p. 30).

n7. Although I am using the terms Internet (or Net) and Web interchangeably, they are in fact both technically and historically distinct. The Internet consists, simply speaking, of the global sum of all computers linked together in a specific way (namely, the so-called TCP/IP communications protocol) to enable the exchange of information. The Internet originated in the 1960s with a Department of Defense contract to explore the possibilities of distributed data communication among computers remote from each other. The World Wide Web, invented in 1990, is the software that enables a user to point to any given document, image, or multimedia file on the Internet and to link directly to it. Before the Web, navigating the Internet in search of distant files was a high art; with the Web, even law professors can find and retrieve information on the Net with ease.

n8. See generally Owen M. Fiss, The Irony of Free Speech (1996) [hereinafter Fiss, Irony] (arguing that government interference with free speech rights may be necessary to ensure equal access to the public discourse); Owen M. Fiss, Liberalism Divided: Freedom of Speech and the Many Uses of State Power (1996) [hereinafter Fiss, Liberalism Divided] (examining the conflicting individual and societal interests inherent in the Free Speech Clause); Sunstein, Democracy, supra note 2, at 17-52 (calling for a "New Deal" with respect to speech, under which government control might be tolerated to the extent necessary to ensure diversity in public debate).

n9. This chapter builds on an equally splendid article by Lessig, Fidelity in Translation, 71 Tex. L. Rev. 1165 (1993).

n10. Cf. Hans-Georg Gadamer, Truth and Method (W. Glen-Doepel trans., Joel Weinsheimer & Donald G. Marshall revs., Crossroad 2d rev. ed. 1989) (1960).

n11. 389 U.S. 347 (1967); see id. at 352-53.

n12. See Horwitz, supra note 2, at 1426; Kennedy, supra note 2, at 1351-52.

n13. See Sunstein, Democracy, supra note 2, at 101-03, 107-08; Sunstein, Partial Constitution, supra note 2, at 225-26; see also Fiss, Irony, supra note 8, at 50-53; Fiss, Liberalism Divided, supra note 8, at 49-51.

n14. See Sunstein, Democracy, supra note 2, at 30-32; Sunstein, Partial Constitution, supra note 2, at 51-53.

n15. See Sunstein, Democracy, supra note 2, at xix, 17-51.

n16. For an elaboration of this point, see Charles Fried, Is Liberty Possible?, in 3 Tanner Lectures on Hum. Values 91, 110-35 (1982).

n17. See, e.g., Steven A. Morrison & Clifford Winston, The Evolution of the Airline Industry 11-12 (1995) "The standard measure of fares, yield, is the average fare per mile for trips by paying customers.... After more than a decade and a half of deregulation, real yield in 1993 was two-thirds of its value in 1976.").

n18. See, e.g., Robert W. Lucky, The Social Impact of the Computer, in Computer Culture: The Scientific, Intellectual, and Social Impact of the Computer 1, 1 (Heinz R. Pagels ed., 1984) "I believe that computers will provide life enrichment for mankind.... At best, computers can pull us up the evolutionary ladder .... At worst they will steal our jobs, and offer us in return a vicarious existence in the form of an electronic nirvana.").

n19. See Julian Dibbell, A Rape in Cyberspace, at http://www.humanities.uci.edu/mposter/syllabi/readings/rape.html (1993).

n20. Cf. United States v. Baker, 890 F. Supp. 1375, 1390-91 (E.D. Mich. 1995) (holding that the "e-mail rapist's" stories were not threats), aff'd on other grounds sub nom. United States v. Alkhabaz, 104 F.3d 1492 (6th Cir. 1997).

n21. See Catharine A. MacKinnon, Feminism Unmodified: Discourses on Life and Law 146 (1987); Catharine A. MacKinnon, Pornography, Civil Rights, and Speech, 20 Harv. C.R.-C.L. L. Rev. 1, 18<NDASH>19 (1985).

n22. See also Lawrence Lessig, Foreword, 52 Stan. L. Rev. 987, 988-89 (2000).

n23. If the issuer is a local bank, the state's regulatory grip may be surer, as it may forbid local issuers from honoring charges to certain merchants - if it can keep up with their phony names - or require local issuers to do this policing for them. Banks and credit card networks are easy marks for local regulation.

n24. It does seem that former Citibank chairman Walter Wriston agrees with Lessig (p. 206).

n25. See Noelle Knox, Swiss Judge Traces Trail of Laundered Russian Funds, USA Today, Aug. 24, 2000, at 1B; Greg B. Smith, 2<fr1/2> Weeks for Money Laundering, N.Y. Daily News, July 12, 2000, at 25.

n26. See, e.g., Stephen Labaton with Matt Richtel, Proposal Offers Surveillance Rules for the Internet, N.Y. Times, July 18, 2000, at A1.

n27. Cf. Microsystems Software, Inc. v. Scandinavia Online AB, 226 F.3d 35, 38 (1st Cir. 2000) (summarizing litigation against two programmers who reverse-engineered a piece of blocking software and created a bypass to it).

n28. See generally Benjamin Kaplan, An Unhurried View of Copyright (1967) (discussing the historical origins and development of copyright law); Howard B. Abrams, The Historic Foundation of American Copyright Law: Exploding the Myth of Common Law Copyright, 29 Wayne L. Rev. 1119 (1983) (describing the historical basis for the American copyright system); W.S. Holdsworth, Press Control and Copyright in the 16th and 17th Centuries, 29 Yale L.J. 841 (1920) (exploring the methods that the Tudors and early Stuarts used to control the press as a means for tracing the origins of copyright law); Brander Matthews, The Evolution of Copyright, 5 Pol. Sci. Q. 583 (1890) (tracing the development of copyright law); W.F. Wyndham Brown, The Origin and Growth of Copyright, 34 Law Mag. & Rev. 54 (1909) (discussing early English copyright laws and the development of common law property rights in published material).

n29. See, e.g., Sunstein, Democracy, supra note 2, at 36-38; Sunstein, Partial Constitution, supra note 2; Duncan Kennedy, Form and Substance in Private Law Adjudication, 89 Harv. L. Rev. 1685, 1748-49 (1976).

n30. U.S. Const. art. I, 8, cl. 8.

n31. Id. art. I, 10, cl. 1.

n32. Id. amend. V.

n33. Stephen Breyer, The Uneasy Case for Copyright: A Study of Copyright in Books, Photocopies, and Computer Programs, 84 Harv. L. Rev. 281, 281 (1970).

n34. 17 U.S.C. 102(b) (1994) "In no case does copyright protection ...extend to any idea ...."); see William F. Patry, Latman's The Copyright Law 30-35 (6th ed. 1986).

n35. Congress, in a not uncommon spasm of stupidity and special interest servility, has recently taken a large step toward distorting that balance in passing the so-called Sonny Bono Copyright Term Extension Act, greatly extending the length of copyright protection. See Pub. L. No. 105-298, 201, 112 Stat. 2827, 2827-28 (1998) (codified at 17 U.S.C. 301-304 (Supp. IV 1998)).

n36. See supra pp. 620-21.

n37. ProCD, Inc. v. Zeidenberg, 86 F.3d 1447, 1450 (7th Cir. 1996) (Easterbrook, J.).

n38. See Frank H. Easterbrook, Cyberspace and the Law of the Horse, 1996 U. Chi. Legal F. 207. But see Lawrence Lessig, The Law of the Horse: What Cyberlaw Might Teach, 113 Harv. L. Rev. 501 (1999).

n39. Nollan v. Cal. Coastal Comm'n, 483 U.S. 825, 831 (1987); see also Loretto v. Teleprompter Manhattan CATV Corp., 458 U.S. 419, 434-41 (1982) (holding that New York's statute requiring that landlords permit a cable company to install cables on their property was a compensable taking of private property).

n40. 111 F. Supp. 2d 294 (S.D.N.Y. 2000).

n41. Id. at 310.

n42. Id. at 311.

n43. Id. at 311-12.

n44. World Intellectual Property Organization Treaty, Apr. 12, 1997, art. 11, S. Treaty Doc. No. 105-17 (1997).

n45. Pub. L. No. 105-304, 112 Stat. 2860 (1998) (codified in scattered sections of 17 U.S.C.).

n46. Id. 103, 112 Stat. at 2863-64 (codified at 17 U.S.C. 1201(a)(1)(A) (Supp. IV 1998)).

n47. Reimerdes, 111 F. Supp. 2d at 315.

n48. Id. at 343-45.

n49. See generally David Nimmer, A Riff on Fair Use in the Digital Millennium, 148 U. Pa. L. Rev. 673, 702-39 (2000) (discussing user rights under the DMCA, the WIPO Treaties Act, and the Copyright Act of 1976).

n50. As applied, the DMCA may prevent access to some works that are not and could not be copyrighted, see Yochai Benkler, Free as the Air to Common Use: First Amendment Constraints on Enclosure of the Public Domain, 74 N.Y.U. L. Rev. 354, 414-29 (1999), even if it states that it applies only to works "subject to this Title" - that is, copyrighted works - because, as in Reimerdes, it may bar the dissemination of decoding technologies that may be used against both copyrighted and uncopyrighted works. In the latter case, there surely is a First Amendment issue to which the court did not adequately advert. The court analogized the use of DeCSS to breaking into a locked room. Reimerdes, 111 F. Supp. 2d at 316. But a metaphor is not an explanation; at best, it points to or summarizes an explanation. See generally Scott Brewer, Exemplary Reasoning: Semantics, Pragmatics, and the Rational Force of Legal Argument by Analogy, 109 Harv. L. Rev. 923 (1996). Here, the metaphor begs the question entirely. CSS is no more a locked room than a "Private Property" sign is a fence. The sign does not physically exclude one who contests the sign-poster's rights, and it certainly does not of its own force make the property the poster's. The law does that, just as the law of the DMCA makes what the movie and record industry "lock" behind CSS their property. And the question remains whether, in light of First Amendment protections, this is a valid law. If it is not, then of course the locked-room metaphor fails.

n51. Reimerdes, 111 F. Supp. 2d at 322-24.

n52. Footnotes have been omitted.

n53. Charles Fried, The New First Amendment Jurisprudence: A Threat to Liberty, 59 U. Chi. L. Rev. 225, 250-53 (1992) [hereinafter Fried, New First Amendment Jurisprudence]; Charles Fried, Perfect Freedom, Perfect Justice, 78 B.U. L. Rev. 717, 736-38 (1998); see also Ronald M. Dworkin, Liberty and Pornography, N.Y. Rev. Books, Aug. 15, 1991, at 12, 14 (discussing Catharine MacKinnon's argument that pornography leads to the political subjugation of women by producing a political climate that portrays women as submissive).

n54. Indeed, C-SPAN is now available at the click of a mouse as well. See Watch C-SPAN, at http://www.c-span.org/watch/cspan.asp (last visited Nov. 5, 2000).

n55. U.S. Const. amend. I.

n56. Sunstein, Democracy, supra note 2, at 122-23, 132-37; see also, e.g., Alexander Meiklejohn, Political Freedom: The Constitutional Powers of the People (1960) (distinguishing between freedom of speech in private arenas and freedom of speech on public issues and discussing the meaning of intellectual freedom); Robert H. Bork, Neutral Principles and Some First Amendment Problems, 47 Ind. L.J. 1, 2<NDASH>3 (1971); Frank I. Michelman, Conceptions of Democracy in American Constitutional Argument: The Case of Pornography Regulation, 56 Tenn. L. Rev. 291 (1989); Robert Post, Racist Speech, Democracy, and the First Amendment, 32 Wm. & Mary L. Rev. 267 (1991) (illustrating the harm caused by racist speech and the possible First Amendment issues raised by regulation and lack of regulation).

n57. Dworkin has recently joined these ranks. See Ronald M. Dworkin, Sovereign Virtue: The Theory and Practice of Equality 351-85 (2000).

n58. Justice O'Connor, in Reno v. ACLU, 521 U.S. 844 (1997), picked up this terminology and acknowledged its origin in Lessig's writing. Id. at 889-90 (O'Connor, J., concurring in the judgment in part and dissenting in part).

n59. Communications Decency Act of 1996 502, 47 U.S.C. 223 (Supp. II 1996). The Court ruled subsections 223(a) and 223(d) unconstitutional. Reno, 521 U.S. at 885.

n60. Government may constitutionally limit the dissemination of indecent material to children but not adults. Ginsberg v. New York, 390 U.S. 629, 640 (1968); cf. FCC v. Pacifica Found., 438 U.S. 726, 736-38 (1978) (finding no First Amendment violation in the FCC's prohibition on obscene, indecent, or profane broadcasting).

n61. Institutions might use similar methods to safeguard sensitive information. Only persons with authorized passwords would be able to gain access to, say, hospital patient records.

n62. See Reno, 521 U.S. at 876-77.

n63. The government may also place filters not to block further transmission of certain messages, but to route copies to, say, the FBI or NSA for surveillance purposes (p. 144).

n64. See, e.g., John Gittings, Shanghai Noon, Guardian (London), Aug. 24, 2000, Guardian Online Pages, at 2, LEXIS, News Library, GUARDN File; Deb Price, Internet Vexes Countries with Repressive History, Detroit News, May 11, 2000, at 4, available at 2000 WL 3477505.

n65. See Fiss, Irony, supra note 8, at 16-18; Fiss, Liberalism Divided, supra note 8, at 26-30; Sunstein, Democracy, supra note 2, at 53-92; Sunstein, Partial Constitution, supra note 2, at 198-223.

n66. I am reminded of the furor surrounding Caller ID. Some claimed that requiring Caller ID would violate the First Amendment right to anonymous speech. See, e.g., Barasch v. Pa. Pub. Util. Comm'n, 576 A.2d 79, 89 (Pa. Commw. Ct. 1990), aff'd on other grounds, 605 A.2d 1198 (Pa. 1992). But I have learned from Lessig that such restrictions are easily circumvented. I can program my phone so that any caller who refuses to reveal his number to my Caller ID system encounters a message, possibly a very rude one, to the effect that, "If you don't want to tell me who you are, I don't want to take your call. And don't bother to call back." Since it is I and not the government who installs this bit of code, the caller has no First Amendment complaint. Indeed, I might have one if the government tried to preclude me from installing the device.

n67. Search engines like Yahoo!, Google, and Ask Jeeves perform an analogous function. Search engines are indispensable tools for picking through the millions of available sites in cyberspace. A search engine that indexed only paying sites with or without notifying searchers could seriously distort access on the Net.

n68. I place this term in shudder quotes because I view its use as an insidious threat to the First Amendment itself. See generally Fried, New First Amendment Jurisprudence, supra note 53. By encouraging courts to consider what they call First Amendment values, authors such as Fiss, Michelman, and Sunstein further their favored values of equality and what they call deliberative democracy, and they divert attention from the First Amendment itself, which plainly speaks of government restriction on freedom of speech. See, e.g., Nixon v. Shrink Mo. Gov't PAC, 120 S. Ct. 897, 910-11 (2000) (Breyer, J., concurring); Turner Broad. Sys. v. FCC, 520 U.S. 180, 226-27 (1997) (Breyer, J., concurring).

n69. See supra section III.B, pp. 624-26.

n70. The German example is a good bit more halting and ambivalent than Lessig's message would suggest. The principal case in the German Constitutional Court, the L<um u>th case, is quite similar to New York Times Co. v. Sullivan, 376 U.S. 254 (1964). See David P. Currie, The Constitution of the Federal Republic of Germany 184 (1994).

n71. See, e.g., Miami Herald Publ'g Co. v. Tornillo, 418 U.S. 241, 243-44 (1974).

n72. See, e.g., Boy Scouts of Am. v. Dale, 120 S. Ct. 2446, 2449-50 (2000).

n73. This sort of inquiry is what Justice Stevens's dissenting opinion in Dale proposed. See id. at 2470 (Stevens, J., dissenting).

n74. Red Lion Broad. Co. v. FCC, 395 U.S. 367, 377 (1969) (quoting Radio Act of 1927, ch. 169, 4, 44 Stat. 1162, 1163).

n75. At the request of the court, Lessig submitted a brief in the antitrust case against Microsoft. See Brief of Amicus Curiae Lawrence Lessig, United States v. Microsoft Corp., 147 F.3d 935 (D.C. Cir. 1998) (Nos. 97-5343, 98-5012), available at http://cyber.law.harvard.edu/works/lessig/ab.pdf.

n76. See, e.g., Int'l Soc'y for Krishna Consciousness, Inc. v. Lee, 505 U.S. 672, 702 (1992) (Kennedy, J., concurring) "The danger of allowing the government to suppress speech is shown in the cases now before us. A grant of plenary power allows the government to tilt the dialog heard by the public, to exclude many, more marginal, voices."); see also Benefit v. City of Cambridge, 679 N.E.2d 184, 185-86 (Mass. 1997) (holding unconstitutional a statute prohibiting begging without a license in public spaces).

n77. The carriers may respond, as did the cable companies, that such a regime would violate their First Amendment rights as publishers to choose what content they will and will not carry. The Supreme Court narrowly and ambiguously rejected this argument in the two Turner Broadcasting cases. See Turner Broad. Sys. v. FCC, 520 U.S. 180, 213-16 (1997); Turner Broad. Sys. v. FCC, 512 U.S. 622, 643-46 (1994). Although AOL does engage in activities that some might call publishing, most Internet service providers and the companies that operate the system backbone seem less likely than a cable company to find shelter in the First Amendment. Such companies are more like telephone companies, which have long been subject to the common carriage regime.

n78. U.S. Const. amend. I.

n79. 326 U.S. 501 (1946).

n80. Id. at 508-09.

n81. See id. at 503.

n82. 419 U.S. 345 (1974).

n83. Id. at 358-59.

n84. Lessig's championing of open code would facilitate this hacking.

n85. 376 U.S. 254, 279-80 (1964) (holding that absent malicious intent, the First Amendment precluded tort liability for a newspaper advertisement libeling a public figure).

n86. 334 U.S. 1, 18-20 (1948) (holding that an injunction enforcing a racially restrictive covenant violated the Equal Protection Clause).

n87. See supra pp. 627-30.

n88. Shelley, 334 U.S. at 19-20.

n89. See Brief of Amicus Curiae Openlaw Participants at 12, Universal Studios, Inc. v. Corley, 111 F. Supp. 2d 346 (S.D.N.Y. 2000) (No. 00 CIV 0277 (LAK)), available at http://cyber.law.harvard.edu/openlaw/DVD/filings/NY/openlaw-amicus/0530-openlaw-amicus.pdf; supra pp. 627-28.

n90. This argument stands in contrast to Flagg Bros. v. Brooks, 436 U.S. 149 (1978), which found that a warehouseman who sold the plaintiff's property to enforce a lien, as the Uniform Commercial Code authorized him to do, was not a state actor but a private actor engaging in nonactionable self-help. Id. at 163.