By Lawrence Lessig
Public Domain Books
It used to be that the failure to include a copyright notice on a creative work meant that the copyright was forfeited. That was a harsh punishment for failing to comply with a regulatory rule—akin to imposing the death penalty for a parking ticket in the world of creative rights. Here again, there is no reason that a marking requirement needs to be enforced in this way. And more importantly, there is no reason a marking requirement needs to be enforced uniformly across all media.
The aim of marking is to signal to the public that this work is copyrighted and that the author wants to enforce his rights. The mark also makes it easy to locate a copyright owner to secure permission to use the work.
One of the problems the copyright system confronted early on was that different copyrighted works had to be differently marked. It wasn’t clear how or where a statue was to be marked, or a record, or a film. A new marking requirement could solve these problems by recognizing the differences in media, and by allowing the system of marking to evolve as technologies enable it to. The system could enable a special signal from the failure to mark—not the loss of the copyright, but the loss of the right to punish someone for failing to get permission first.
Let’s start with the last point. If a copyright owner allows his work to be published without a copyright notice, the consequence of that failure need not be that the copyright is lost. The consequence could instead be that anyone has the right to use this work, until the copyright owner complains and demonstrates that it is his work and he doesn’t give permission.  The meaning of an unmarked work would therefore be “use unless someone complains.” If someone does complain, then the obligation would be to stop using the work in any new work from then on though no penalty would attach for existing uses. This would create a strong incentive for copyright owners to mark their work.
That in turn raises the question about how work should best be marked. Here again, the system needs to adjust as the technologies evolve. The best way to ensure that the system evolves is to limit the Copyright Office’s role to that of approving standards for marking content that have been crafted elsewhere.
For example, if a recording industry association devises a method for marking CDs, it would propose that to the Copyright Office. The Copyright Office would hold a hearing, at which other proposals could be made. The Copyright Office would then select the proposal that it judged preferable, and it would base that choice solely upon the consideration of which method could best be integrated into the registration and renewal system. We would not count on the government to innovate; but we would count on the government to keep the product of innovation in line with its other important functions.
Finally, marking content clearly would simplify registration requirements. If photographs were marked by author and year, there would be little reason not to allow a photographer to reregister, for example, all photographs taken in a particular year in one quick step. The aim of the formality is not to burden the creator; the system itself should be kept as simple as possible.
The objective of formalities is to make things clear. The existing system does nothing to make things clear. Indeed, it seems designed to make things unclear.
If formalities such as registration were reinstated, one of the most difficult aspects of relying upon the public domain would be removed. It would be simple to identify what content is presumptively free; it would be simple to identify who controls the rights for a particular kind of content; it would be simple to assert those rights, and to renew that assertion at the appropriate time.
2. Shorter Terms
The term of copyright has gone from fourteen years to ninety-five years for corporate authors, and life of the author plus seventy years for natural authors.
In The Future of Ideas, I proposed a seventy-five-year term, granted in five- year increments with a requirement of renewal every five years. That seemed radical enough at the time. But after we lost Eldred v. Ashcroft, the proposals became even more radical. The Economist endorsed a proposal for a fourteen-year copyright term.  Others have proposed tying the term to the term for patents.
I agree with those who believe that we need a radical change in copyright’s term. But whether fourteen years or seventy-five, there are four principles that are important to keep in mind about copyright terms.
(1) Keep it short: The term should be as long as necessary to give incentives to create, but no longer. If it were tied to very strong protections for authors (so authors were able to reclaim rights from publishers), rights to the same work (not derivative works) might be extended further. The key is not to tie the work up with legal regulations when it no longer benefits an author.
(2) Keep it simple: The line between the public domain and protected content must be kept clear. Lawyers like the fuzziness of “fair use,” and the distinction between “ideas” and “expression.” That kind of law gives them lots of work. But our framers had a simpler idea in mind: protected versus unprotected. The value of short terms is that there is little need to build exceptions into copyright when the term itself is kept short. A clear and active “lawyer-free zone” makes the complexities of “fair use” and “idea/expression” less necessary to navigate.
(3) Keep it alive: Copyright should have to be renewed. Especially if the maximum term is long, the copyright owner should be required to signal periodically that he wants the protection continued. This need not be an onerous burden, but there is no reason this monopoly protection has to be granted for free. On average, it takes ninety minutes for a veteran to apply for a pension.  If we make veterans suffer that burden, I don’t see why we couldn’t require authors to spend ten minutes every fifty years to file a single form.
(4) Keep it prospective: Whatever the term of copyright should be, the clearest lesson that economists teach is that a term once given should not be extended. It might have been a mistake in 1923 for the law to offer authors only a fifty-six-year term. I don’t think so, but it’s possible. If it was a mistake, then the consequence was that we got fewer authors to create in 1923 than we otherwise would have. But we can’t correct that mistake today by increasing the term. No matter what we do today, we will not increase the number of authors who wrote in 1923. Of course, we can increase the reward that those who write now get (or alternatively, increase the copyright burden that smothers many works that are today invisible). But increasing their reward will not increase their creativity in 1923. What’s not done is not done, and there’s nothing we can do about that now.
These changes together should produce an average copyright term that is much shorter than the current term. Until 1976, the average term was just 32.2 years. We should be aiming for the same.
No doubt the extremists will call these ideas “radical.” (After all, I call them “extremists.”) But again, the term I recommended was longer than the term under Richard Nixon. How “radical” can it be to ask for a more generous copyright law than Richard Nixon presided over?
3. Free Use Vs. Fair Use
As I observed at the beginning of this book, property law originally granted property owners the right to control their property from the ground to the heavens. The airplane came along. The scope of property rights quickly changed. There was no fuss, no constitutional challenge. It made no sense anymore to grant that much control, given the emergence of that new technology.
Our Constitution gives Congress the power to give authors “exclusive right” to “their writings.” Congress has given authors an exclusive right to “their writings” plus any derivative writings (made by others) that are sufficiently close to the author’s original work. Thus, if I write a book, and you base a movie on that book, I have the power to deny you the right to release that movie, even though that movie is not “my writing.”
Congress granted the beginnings of this right in 1870, when it expanded the exclusive right of copyright to include a right to control translations and dramatizations of a work.  The courts have expanded it slowly through judicial interpretation ever since. This expansion has been commented upon by one of the law’s greatest judges, Judge Benjamin Kaplan.
“So inured have we become to the extension of the monopoly to a large range of so-called derivative works, that we no longer sense the oddity of accepting such an enlargement of copyright while yet intoning the abracadabra of idea and expression.” 
I think it’s time to recognize that there are airplanes in this field and the expansiveness of these rights of derivative use no longer make sense. More precisely, they don’t make sense for the period of time that a copyright runs. And they don’t make sense as an amorphous grant. Consider each limitation in turn.
Term: If Congress wants to grant a derivative right, then that right should be for a much shorter term. It makes sense to protect John Grisham’s right to sell the movie rights to his latest novel (or at least I’m willing to assume it does); but it does not make sense for that right to run for the same term as the underlying copyright. The derivative right could be important in inducing creativity; it is not important long after the creative work is done.
Scope: Likewise should the scope of derivative rights be narrowed. Again, there are some cases in which derivative rights are important. Those should be specified. But the law should draw clear lines around regulated and unregulated uses of copyrighted material. When all “reuse” of creative material was within the control of businesses, perhaps it made sense to require lawyers to negotiate the lines. It no longer makes sense for lawyers to negotiate the lines. Think about all the creative possibilities that digital technologies enable; now imagine pouring molasses into the machines. That’s what this general requirement of permission does to the creative process. Smothers it.
This was the point that Alben made when describing the making of the Clint Eastwood CD. While it makes sense to require negotiation for foreseeable derivative rights—turning a book into a movie, or a poem into a musical score—it doesn’t make sense to require negotiation for the unforeseeable. Here, a statutory right would make much more sense.
In each of these cases, the law should mark the uses that are protected, and the presumption should be that other uses are not protected. This is the reverse of the recommendation of my colleague Paul Goldstein.  His view is that the law should be written so that expanded protections follow expanded uses.
Goldstein’s analysis would make perfect sense if the cost of the legal system were small. But as we are currently seeing in the context of the Internet, the uncertainty about the scope of protection, and the incentives to protect existing architectures of revenue, combined with a strong copyright, weaken the process of innovation.
The law could remedy this problem either by removing protection beyond the part explicitly drawn or by granting reuse rights upon certain statutory conditions. Either way, the effect would be to free a great deal of culture to others to cultivate. And under a statutory rights regime, that reuse would earn artists more income.
4. Liberate the Music—Again
The battle that got this whole war going was about music, so it wouldn’t be fair to end this book without addressing the issue that is, to most people, most pressing—music. There is no other policy issue that better teaches the lessons of this book than the battles around the sharing of music.
The appeal of file-sharing music was the crack cocaine of the Inter-net’s growth. It drove demand for access to the Internet more powerfully than any other single application. It was the Internet’s killer app—possibly in two senses of that word. It no doubt was the application that drove demand for bandwidth. It may well be the application that drives demand for regulations that in the end kill innovation on the network.
The aim of copyright, with respect to content in general and music in particular, is to create the incentives for music to be composed, performed, and, most importantly, spread. The law does this by giving an exclusive right to a composer to control public performances of his work, and to a performing artist to control copies of her performance.
File-sharing networks complicate this model by enabling the spread of content for which the performer has not been paid. But of course, that’s not all the file-sharing networks do. As I described in chapter 5, they enable four different kinds of sharing:
A. There are some who are using sharing networks as substitutes for purchasing
B. There are also some who are using sharing networks to sample, on the way to purchasing CDs.
C. There are many who are using file-sharing networks to get access to content that is no longer sold but is still under copyright or that would have been too cumbersome to buy off the Net.
D. There are many who are using file-sharing networks to get access to content that is not copyrighted or to get access that the copyright owner plainly endorses.
Any reform of the law needs to keep these different uses in focus. It must avoid burdening type D even if it aims to eliminate type A. The eagerness with which the law aims to eliminate type A, moreover, should depend upon the magnitude of type B. As with VCRs, if the net effect of sharing is actually not very harmful, the need for regulation is significantly weakened.
As I said in chapter 5, the actual harm caused by sharing is controversial. For the purposes of this chapter, however, I assume the harm is real. I assume, in other words, that type A sharing is significantly greater than type B, and is the dominant use of sharing networks.
Nonetheless, there is a crucial fact about the current technological context that we must keep in mind if we are to understand how the law should respond.
Today, file sharing is addictive. In ten years, it won’t be. It is addictive today because it is the easiest way to gain access to a broad range of content. It won’t be the easiest way to get access to a broad range of content in ten years. Today, access to the Internet is cumbersome and slow—we in the United States are lucky to have broadband service at 1.5 MBs, and very rarely do we get service at that speed both up and down. Although wireless access is growing, most of us still get access across wires. Most only gain access through a machine with a keyboard. The idea of the always on, always connected Internet is mainly just an idea.
But it will become a reality, and that means the way we get access to the Internet today is a technology in transition. Policy makers should not make policy on the basis of technology in transition. They should make policy on the basis of where the technology is going. The question should not be, how should the law regulate sharing in this world? The question should be, what law will we require when the network becomes the network it is clearly becoming? That network is one in which every machine with electricity is essentially on the Net; where everywhere you are—except maybe the desert or the Rockies—you can instantaneously be connected to the Internet. Imagine the Internet as ubiquitous as the best cell-phone service, where with the flip of a device, you are connected.
In that world, it will be extremely easy to connect to services that give you access to content on the fly—such as Internet radio, content that is streamed to the user when the user demands. Here, then, is the critical point: When it is /extremely easy to connect to services that give access to content, it will be /easier to connect to services that give you access to content than it will be to download and store content on the many devices you will have for playing content. It will be easier, in other words, to subscribe than it will be to be a database manager, as everyone in the download-sharing world of Napster-like technologies essentially is. Content services will compete with content sharing, even if the services charge money for the content they give access to. Already cell-phone services in Japan offer music (for a fee) streamed over cell phones (enhanced with plugs for headphones). The Japanese are paying for this content even though “free” content is available in the form of MP3s across the Web. 
This point about the future is meant to suggest a perspective on the present: It is emphatically temporary. The “problem” with file sharing—to the extent there is a real problem—is a problem that will increasingly disappear as it becomes easier to connect to the Internet. And thus it is an extraordinary mistake for policy makers today to be “solving” this problem in light of a technology that will be gone tomorrow. The question should not be how to regulate the Internet to eliminate file sharing (the Net will evolve that problem away). The question instead should be how to assure that artists get paid, during this transition between twentieth-century models for doing business and twenty-first-century technologies.
The answer begins with recognizing that there are different “problems” here to solve. Let’s start with type D content—uncopyrighted content or copyrighted content that the artist wants shared. The “problem” with this content is to make sure that the technology that would enable this kind of sharing is not rendered illegal. You can think of it this way: Pay phones are used to deliver ransom demands, no doubt. But there are many who need to use pay phones who have nothing to do with ransoms. It would be wrong to ban pay phones in order to eliminate kidnapping.
Type C content raises a different “problem.” This is content that was, at one time, published and is no longer available. It may be unavailable because the artist is no longer valuable enough for the record label he signed with to carry his work. Or it may be unavailable because the work is forgotten. Either way, the aim of the law should be to facilitate the access to this content, ideally in a way that returns something to the artist.
Again, the model here is the used book store. Once a book goes out of print, it may still be available in libraries and used book stores. But libraries and used book stores don’t pay the copyright owner when someone reads or buys an out-of- print book. That makes total sense, of course, since any other system would be so burdensome as to eliminate the possibility of used book stores’ existing. But from the author’s perspective, this “sharing” of his content without his being compensated is less than ideal.
The model of used book stores suggests that the law could simply deem out-of- print music fair game. If the publisher does not make copies of the music available for sale, then commercial and noncommercial providers would be free, under this rule, to “share” that content, even though the sharing involved making a copy. The copy here would be incidental to the trade; in a context where commercial publishing has ended, trading music should be as free as trading books.
Alternatively, the law could create a statutory license that would ensure that artists get something from the trade of their work. For example, if the law set a low statutory rate for the commercial sharing of content that was not offered for sale by a commercial publisher, and if that rate were automatically transferred to a trust for the benefit of the artist, then businesses could develop around the idea of trading this content, and artists would benefit from this trade.
This system would also create an incentive for publishers to keep works available commercially. Works that are available commercially would not be subject to this license. Thus, publishers could protect the right to charge whatever they want for content if they kept the work commercially available. But if they don’t keep it available, and instead, the computer hard disks of fans around the world keep it alive, then any royalty owed for such copying should be much less than the amount owed a commercial publisher.
The hard case is content of types A and B, and again, this case is hard only because the extent of the problem will change over time, as the technologies for gaining access to content change. The law’s solution should be as flexible as the problem is, understanding that we are in the middle of a radical transformation in the technology for delivering and accessing content.
So here’s a solution that will at first seem very strange to both sides in this war, but which upon reflection, I suggest, should make some sense.
Stripped of the rhetoric about the sanctity of property, the basic claim of the content industry is this: A new technology (the Internet) has harmed a set of rights that secure copyright. If those rights are to be protected, then the content industry should be compensated for that harm. Just as the technology of tobacco harmed the health of millions of Americans, or the technology of asbestos caused grave illness to thousands of miners, so, too, has the technology of digital networks harmed the interests of the content industry.
I love the Internet, and so I don’t like likening it to tobacco or asbestos. But the analogy is a fair one from the perspective of the law. And it suggests a fair response: Rather than seeking to destroy the Internet, or the p2p technologies that are currently harming content providers on the Internet, we should find a relatively simple way to compensate those who are harmed.
The idea would be a modification of a proposal that has been floated by Harvard law professor William Fisher.  Fisher suggests a very clever way around the current impasse of the Internet. Under his plan, all content capable of digital transmission would (1) be marked with a digital watermark (don’t worry about how easy it is to evade these marks; as you’ll see, there’s no incentive to evade them). Once the content is marked, then entrepreneurs would develop (2) systems to monitor how many items of each content were distributed. On the basis of those numbers, then (3) artists would be compensated. The compensation would be paid for by (4) an appropriate tax.
Fisher’s proposal is careful and comprehensive. It raises a million questions, most of which he answers well in his upcoming book, Promises to Keep. The modification that I would make is relatively simple: Fisher imagines his proposal replacing the existing copyright system. I imagine it complementing the existing system. The aim of the proposal would be to facilitate compensation to the extent that harm could be shown. This compensation would be temporary, aimed at facilitating a transition between regimes. And it would require renewal after a period of years. If it continues to make sense to facilitate free exchange of content, supported through a taxation system, then it can be continued. If this form of protection is no longer necessary, then the system could lapse into the old system of controlling access.
Fisher would balk at the idea of allowing the system to lapse. His aim is not just to ensure that artists are paid, but also to ensure that the system supports the widest range of “semiotic democracy” possible. But the aims of semiotic democracy would be satisfied if the other changes I described were accomplished—in particular, the limits on derivative uses. A system that simply charges for access would not greatly burden semiotic democracy if there were few limitations on what one was allowed to do with the content itself.
No doubt it would be difficult to calculate the proper measure of “harm” to an industry. But the difficulty of making that calculation would be outweighed by the benefit of facilitating innovation. This background system to compensate would also not need to interfere with innovative proposals such as Apple’s MusicStore. As experts predicted when Apple launched the MusicStore, it could beat “free” by being easier than free is. This has proven correct: Apple has sold millions of songs at even the very high price of 99 cents a song. (At 99 cents, the cost is the equivalent of a per-song CD price, though the labels have none of the costs of a CD to pay.) Apple’s move was countered by Real Networks, offering music at just 79 cents a song. And no doubt there will be a great deal of competition to offer and sell music on-line.
This competition has already occurred against the background of “free” music from p2p systems. As the sellers of cable television have known for thirty years, and the sellers of bottled water for much more than that, there is nothing impossible at all about “competing with free.” Indeed, if anything, the competition spurs the competitors to offer new and better products. This is precisely what the competitive market was to be about. Thus in Singapore, though piracy is rampant, movie theaters are often luxurious—with “first class” seats, and meals served while you watch a movie—as they struggle and succeed in finding ways to compete with “free.”
This regime of competition, with a backstop to assure that artists don’t lose, would facilitate a great deal of innovation in the delivery of content. That competition would continue to shrink type A sharing. It would inspire an extraordinary range of new innovators—ones who would have a right to the content, and would no longer fear the uncertain and barbarically severe punishments of the law.
In summary, then, my proposal is this:
The Internet is in transition. We should not be regulating a technology in transition. We should instead be regulating to minimize the harm to interests affected by this technological change, while enabling, and encouraging, the most efficient technology we can create.
We can minimize that harm while maximizing the benefit to innovation by
1. guaranteeing the right to engage in type D sharing;
2. permitting noncommercial type C sharing without liability, and commercial type C sharing at a low and fixed rate set by statute;
3. while in this transition, taxing and compensating for type A sharing, to the extent actual harm is demonstrated.
But what if “piracy” doesn’t disappear? What if there is a competitive market providing content at a low cost, but a significant number of consumers continue to “take” content for nothing? Should the law do something then?
Yes, it should. But, again, what it should do depends upon how the facts develop. These changes may not eliminate type A sharing. But the real issue is not whether it eliminates sharing in the abstract. The real issue is its effect on the market. Is it better (a) to have a technology that is 95 percent secure and produces a market of size x, or (b) to have a technology that is 50 percent secure but produces a market of five times x? Less secure might produce more unauthorized sharing, but it is likely to also produce a much bigger market in authorized sharing. The most important thing is to assure artists’ compensation without breaking the Internet. Once that’s assured, then it may well be appropriate to find ways to track down the petty pirates.
But we’re a long way away from whittling the problem down to this subset of type A sharers. And our focus until we’re there should not be on finding ways to break the Internet. Our focus until we’re there should be on how to make sure the artists are paid, while protecting the space for innovation and creativity that the Internet is.
5. Fire Lots of Lawyers
I’m a lawyer. I make lawyers for a living. I believe in the law. I believe in the law of copyright. Indeed, I have devoted my life to working in law, not because there are big bucks at the end but because there are ideals at the end that I would love to live.
Yet much of this book has been a criticism of lawyers, or the role lawyers have played in this debate. The law speaks to ideals, but it is my view that our profession has become too attuned to the client. And in a world where the rich clients have one strong view, the unwillingness of the profession to question or counter that one strong view queers the law.
The evidence of this bending is compelling. I’m attacked as a “radical” by many within the profession, yet the positions that I am advocating are precisely the positions of some of the most moderate and significant figures in the history of this branch of the law. Many, for example, thought crazy the challenge that we brought to the Copyright Term Extension Act. Yet just thirty years ago, the dominant scholar and practitioner in the field of copyright, Melville Nimmer, thought it obvious. 
However, my criticism of the role that lawyers have played in this debate is not just about a professional bias. It is more importantly about our failure to actually reckon the costs of the law.
Economists are supposed to be good at reckoning costs and benefits. But more often than not, economists, with no clue about how the legal system actually functions, simply assume that the transaction costs of the legal system are slight.  They see a system that has been around for hundreds of years, and they assume it works the way their elementary school civics class taught them it works.
But the legal system doesn’t work. Or more accurately, it doesn’t work for anyone except those with the most resources. Not because the system is corrupt. I don’t think our legal system (at the federal level, at least) is at all corrupt. I mean simply because the costs of our legal system are so astonishingly high that justice can practically never be done.
These costs distort free culture in many ways. A lawyer’s time is billed at the largest firms at more than $400 per hour. How much time should such a lawyer spend reading cases carefully, or researching obscure strands of authority? The answer is the increasing reality: very little. The law depended upon the careful articulation and development of doctrine, but the careful articulation and development of legal doctrine depends upon careful work. Yet that careful work costs too much, except in the most high-profile and costly cases.
The costliness and clumsiness and randomness of this system mock our tradition. And lawyers, as well as academics, should consider it their duty to change the way the law works—or better, to change the law so that it works. It is wrong that the system works well only for the top 1 percent of the clients. It could be made radically more efficient, and inexpensive, and hence radically more just.
But until that reform is complete, we as a society should keep the law away from areas that we know it will only harm. And that is precisely what the law will too often do if too much of our culture is left to its review.
Think about the amazing things your kid could do or make with digital technology—the film, the music, the Web page, the blog. Or think about the amazing things your community could facilitate with digital technology—a wiki, a barn raising, activism to change something. Think about all those creative things, and then imagine cold molasses poured onto the machines. This is what any regime that requires permission produces. Again, this is the reality of Brezhnev’s Russia.
The law should regulate in certain areas of culture—but it should regulate culture only where that regulation does good. Yet lawyers rarely test their power, or the power they promote, against this simple pragmatic question: “Will it do good?” When challenged about the expanding reach of the law, the lawyer answers, “Why not?”
We should ask, “Why?” Show me why your regulation of culture is needed. Show me how it does good. And until you can show me both, keep your lawyers away.