EARLY DRAFT 1

Sharing and Stealing 2

Jessica Litman* 3

The purpose of copyright is to encourage the creation and mass dissemination of a wide variety of works. Until recently, most means of mass dissemination required a significant capital investment. Disseminators needed printing presses, trains or trucks, warehouses, broadcast towers, or communications satellites. It made economic sense to channel the lion’s share of the proceeds of copyrights to the publishers and distributors, and the law was designed to facilitate that.[1] Digital distribution raises the possibility of mass dissemination without the assistance of professional distributors, via direct author-to-consumer and consumer-to-consumer dissemination. Digital distribution, thus, invites us to reconsider the assumptions underlying the conventional copyright model.  4

We are still in the early history of the networked digital environment, but already we’ve seen experiments with both direct and consumer-to-consumer distribution of works of authorship. Direct author distribution – by itself – has not yet garnered a lot of attention because the most publicized efforts have been less than wholly successful.[2] When direct author distribution is augmented by consumer-to-consumer distribution, though, the combination has the potential to revolutionize the distribution chain. That potential has not escaped the attention of professional distributors. Consumer-to-consumer dissemination, especially in the form of peer-to-peer file sharing, has been met with hostility and panic. Legislation pending in Congress seeks to deter consumers from engaging in peer-to-peer file sharing.[3] Meanwhile, representatives of the music, recording and file industry have sued the purveyors of peer-to-peer file sharing software,[4] the Internet service providers who enable consumers to trade files,[5] and 341 individual consumers accused of making recorded music available to other consumers over the Internet.[6] 5

In this paper, I propose that we look for some of the answers to the vexing problem of unauthorized exchange of music files on the Internet in the wisdom intellectual property law has accumulated about the protection and distribution of factual information. In particular, I analyze the digital information resource that has developed on the Internet, and suggest that what we should be trying to achieve is an online musical smorgasbord of comparable breadth and variety.  6

Ten years ago, an influential government task force proposed enhancing the scope of intellectual property rights in the digital environment as a device to encourage investment in the infrastructure underlying a national digital network. [7] As the task force explained, the cost of constructing such a network was beyond the federal government’s ability to fund, and the construction would need to be undertaken by the private sector. The private sector, however, would be reluctant to invest its resources unless it saw profits to be made. The network would be commercial only if large numbers of people could be persuaded to subscribe to digital network services, which would require a killer application to draw people online. In the view of the task force, that application was the possibility consumers could enjoy movies, music and other content on demand. Enhanced copyright protection would be needed to persuade the producers of movies, music and other content to make the investment in making their material available over the national digital network. In order to create a viable online information and entertainment resource, the task force concluded, the United States needed to promise the distributors of copyrighted works a larger share of the copyright pie – only then would they invest the resources needed to develop digital content that would be sufficiently compelling that ordinary consumers would want it enough to pay for Internet access.[8] With the benefit of hindsight, it’s become clear that most of the assumptions underlying that argument were wrong.  7

Greatly expanded copyright has not yet encouraged movies or music online – there is an enormous variety of music and movies available over the Internet, but the overwhelming majority of what’s there is there over the vehement objections of the content owners. Nonetheless, the network has grown at an unbelievable rate. The killer application that fueled the growth of the Internet wasn’t digital movies, after all. Instead, it was communication – email, chat, online forums and personal web pages. It turns out that people want to communicate with one another, and that they love to share. The information space that has grown up on the world wide web is largely the result of anarchic volunteerism – not to build the pipes, which have been constructed by telephone and cable companies to meet consumer demand for broadband Internet access,[9] but to supply the information that runs through them. Anecdotal evidence indicates that at least for some material, untamed digital sharing turns out to be a more efficient method of distribution than either paid subscription or the sale of conventional copies. If untamed anarchic digital sharing is a superior distribution mechanism, or even a useful adjunct to conventional distribution, we ought to encourage it rather than making it more difficult. 8

Part I of this essay explores the burgeoning digital information space that has grown up on the Internet in the last two decades. In part II, I review the legal obstacles preventing us from simply treating digital music the way we treat digital information. Amendments to the copyright law enacted over the past 30 years have erected legal barriers to consumer-to-consumer distribution that make lawful exchange of copyrighted material extremely difficult. Part III tells a true story about my son’s third grade classroom, and spends a brief moment looking at his teacher’s use of the resources she finds on the Internet. Part IV suggests that we look to the digital information space described in Part I as a model for crafting a solution to the controversy over peer-to-peer file sharing, and reviews some of the proposals made in recent copyright scholarship. Finally, Part V briefly outlines a particular solution that is in some respects different from those discussed earlier. 9

I. Someone knows what I want to know 10

Someone knows what I want to know. Someone has the information I want. If I can find her, I can learn it from her. She will share it with me. 11

Which came first, the computer or network television?[10] 12

I could try to find the answer in a reference book instead. On my bookshelf, I have two editions of the Encyclopedia Britannica, one published in 1989 and one assembled at a public library used-book sale from individual volumes published in 1964 and 1966. I no longer consult either of them with any frequency. In a jewel case somewhere near my desk, I have a multimedia CD ROM version of the Britannica that I received as a gift in 1998. I never look at it at all, and haven’t since the month that I received it. I used to buy an Almanac each year to look up quick facts (what’s the population of New Zealand?[11] How old is Senator Barbara Mikulski?[12]), but between 1994 and beginning work on this project, I didn’t bother. I stopped relying on these books as it became possible to find specific answers to specific questions online, because the person or persons who knew what I wanted to know had been generous enough to post the answer where it was easy for me to find it. The search was quicker, and commonly yielded more accurate information, than consulting whatever reference books were handy. 13

Although the Internet’s usefulness as an engine of commerce has flowed and ebbed, its value as a repository of information has continued to grow exponentially. What we used to class as trivia (and therefore useless information) becomes a matchless resource when it is combined with other trivia in searchable form.[13] Volunteers, most of them amateurs, have collected an unimaginable variety of information and are eager to share it with the world. What I want to know may not be in any book on my shelf or in my university’s libraries. I can probably find it on the Internet in less than an hour. 14

What are sesame seeds?[14] 15

Networked digital technology has transformed information and the way that we interact with it. Digital information is extraordinarily accessible. If I have a question, I don’t need to make up an answer that seems plausible, or reason out what it’s likely to be. I don’t need to go to the library and ask the reference librarian if I can see the library’s only copy of a reference book that ought to have the answer. I can just turn to my computer and look it up on the web.  16

Digital information, moreover, is shared. Ten years, even five years ago, it was conventional to talk about the Internet as a tool for disintermediation. Authors and musicians would be able to use digital networks to sending stuff directly to their readers and listeners. (Remember Stephen King’s The Plant?[15]) There’s some of that. People post content on their websites for the rest of the world to view. Academics exchange drafts of scholarly papers that way,[16] and independent musicians and composers make recordings of their work available for sampling and download.[17] But, while we’ve seen a small but appreciable amount of direct distribution, there’s even more consumer-to-consumer distribution. The “blog” (or weblog) is an increasingly popular art form in which people post an ongoing, public, hyperlinked diaries of things they find interesting and want to share.[18] Readers of the blogs write in to contribute their own comments.[19]  17

Someone has the recording I want. If I can find her, she’ll share it with me. I can copy it and pass it on. Someone knows the answer to my question. If I can locate her, she’ll tell me. I can learn it and pass it on. Someone has seen the source I want to consult. She can tell me where to find it. 18

Where is a copy of the complaint in Hamilton v. Microsoft?[20] 19

The most powerful engine driving this information space turns out not to be money – at least if we’re focusing on generating and disseminating the content rather than constructing the pipes that it moves through. What seems to be driving the explosive growth in this information space is that people like to look things up, and they want to share. This information economy is largely a gift economy. The overwhelming majority of the information I’m talking about is initially posted by volunteers. Many of them are amateurs, motivated by enthusiasm for their topics, a desire to share, and, perhaps, an interest in attention and the benefits it may bring. When one is a volunteer, the time and effort one is willing to put into contributing to the information space can seem limitless. Volunteers move on, of course: they get bored, or broke, or caught up in other things, but there seems to be an inexhaustible supply of new volunteers to take their places, and, luckily, the new volunteers are able to build on earlier volunteers’ foundations.[21] I potentially know all of the information the other participants know. Their knowledge can be my knowledge with a few clicks of a mouse. In return, I make my knowledge available to anyone who happens by. Each of us can draw on the information stores of the others.[22]  20

The rate at which people have adopted the Internet as their research tool of choice is astonishing. People find the easy availability of all that information empowering. People want to know how old Steven Spielberg is. They want to know the history of early radio. They want to know what traveling musicians wore in 15^th Century Europe. They want to know how to make Shaker Lemon Pie. They want to know what the Constitution actually says. If it’s quick and easy to do so, they’ll look it up. They enjoy discovering new stuff. The system has been evolving as we watch: consumer-to-consumer interaction is leading to more information, better information, and more accessible information; more complete and deeper archives; wider ranges of divergent sources. [23] People appreciate the instant gratification of learning answers in a moment. Probably more important than the speed of the system, however, is its breadth and depth. Because of the disparate contributions of a host of volunteers, one can find information that would not appear in conventional reference sources.[24] 21

Ten years ago, not only Washington but the entire journalism business believed that the burgeoning digital network (which went by the name, back then, of the “National Information Infrastructure”) would develop into a 500 channel interactive television system, with “interactive” meaning that it would incorporate a method for ordering and charging purchases and receiving targeted advertising.[25] There are a number of businesses out there that are continuing to try to shove the Internet in that direction,[26] but the fact that it isn’t yet anything like a 500 channel TV is largely because of the way that lots and lots of people have come to interact with information.  22

What is “the fuct of Pepsiman”?[27] 23

Let’s pause for a word from our friendly reference book publishers. Speed and convenience are all very well, but doesn’t selecting the speediest research tool ignore the quality and reliability of the information I retrieve? The Internet, after all, is an infamous source of falsehood and untruth.[28] Books and periodicals have editors and fact checkers to screen out misinformation; websites need not.[29]  24

The story, as stories often do, turns out to be more complicated. The efforts of editors and fact checkers have apparently not, for example, prevented periodicals from reprinting Internet untruths as if they were fact.[30] That should not be surprising. Many editors and fact checkers are neither well-paid nor well-qualified to assure the accuracy of the information their employers print. Not all publications use them. Often, fact checkers must rely on authors to direct them to corroborating sources. The customs of different disciplines may control how carefully content is checked.[31] Correcting errors in print publications is difficult and expensive. Except in cases egregious enough to merit a recall,[32] the corrections must be put off until the publication of a later edition. 25

On the World Wide Web, in contrast, correcting errors and revising documents is simple and inexpensive. One can rewrite a file several times each day and spend nothing more than the time that it takes to enter the revisions and transmit the file to the server. If one makes a mistake, there are dozens of eager volunteers likely to send one an email offering corrections.[33] Indeed, the feedback of knowledgeable readers is a powerful force promoting accuracy on the web. With the world looking in, errors are much more likely to be identified, and correcting them is easy.  26

Stepping back to look at the whole dynamic information space, it becomes clear that the remote participation of readers doesn’t stop at writing comments in other people’s blogs, or even at writing in to correct errors or misstatements. Fellow enthusiasts are likely to reuse the information they find in one web page – or a dozen -- in their own web pages. A reader may simply post a hyperlink to someone else’s page, or she may bodily appropriate some prose, combine it with her own prose and additional prose lifted from some other sites, and post the amalgam as her own (with or without attribution). Thus does information spread. What makes this economy so astonishingly useful is information sharing – we’re not each of us downloading facts from some giant Encyclopedia Britannica in the sky, we’re both finding what we need and also making available material that we’ve generated or assembled.  27

Who invented the phonograph?[34] 28

This information system is vital and dynamic because information sharing is almost frictionless. Material is passed along at low cost with few practical or legal barriers. Jeff Dalehite, webmaster of <scratchdj.com>, is free to post the details of the early history of the phonograph without seeking the consent of his sources. Dalehite’s site tells us that Thomas Edison invented the cylinder phonograph in the 1870s and patented it in 1878. Dalehite recounts the details of the commercial standards competition between Edison’s phonograph and the disk gramophone introduced to the U.S. market in 1901 by the Victor Talking Machine Company.[35] He attributes none of his sources; he need not even know whether the information he has abstracted was original to the references he used or derived by them from some other source. Technical writer Samuel Berliner III has posted a site honoring famous people throughout history named Berliner. His site reports that the disk gramophone was invented by Emile Berliner in 1887. Berliner needs no permission from Frederick W. Nile, the author of a 1926 biography of Emile Berliner,[36] nor the National Inventors Hall of Fame, who have posted a short profile of Berliner,[37] from whom he initially learned that information.[38] Neither Dalehite nor Berliner has secured a license from Tommy Cichanowski for any facts they might have learned by studying Tommy’s History of Western Technology,[39] nor have they sought the blessing of the periodical Electronic Design, whose February 1976 issue commemorating the U.S, bicentennial[40] furnished many of the dates that Cichanowski reports. If one were unable to post facts without determining who controlled them and obtaining a license to pass those facts on, this online information space would not exist. 29

II. Formalities and Default Rules 30

Who wrote “When I was One-and-Twenty”?[41] 31

The purpose of copyright is to promote the progress of science, by encouraging the production and dissemination of works of authorship.[42] The contributions of this networked digital information space to the “Progress of Science” are difficult to overestimate. Already, a network of people sharing what they know has made many of the most popular reference sources obsolete. Thus, one might reasonably hope and expect that a law designed to promote the Progress of Science would encourage the robust growth and prodigious use of this network to exchange the full spectrum of interesting material. 32

Under current law, though, the information space I’m talking about is lawful only because so much of its content – the facts, information, ideas – is in the public domain. To the extent that the material in this information space is in the public domain, we can all share it, use it, and reuse it. To the extent it’s protected by copyright, on the other hand, we would need permission to do all of that, and, as a practical matter, it isn’t possible to secure that permission. One of the most salient lessons from the copyright wars of the last few years is that if express permission is required before one can post a collection of anything on the Internet, one will be unable to do it.[43]  33

To appreciate the extent of the problem, it’s helpful to review key changes in the copyright law and the information space over the past thirty years. Today, facts are some of the only material solidly part of the public domain, but 30 years ago, that wasn’t true. In 1973, copyright protection was not automatic. To get it, you needed to distribute copies of your work to the public, and the copies needed to be marked with a copyright notice.[44] Notice of copyright – the familiar C-in-a-circle, along with the name of the copyright owner and the date the work was first published – secured copyright. Distributing copies without notice caused the work to enter the public domain.[45] Indeed, the point of the copyright system was to offer authors protection for a limited time as an incentive to encourage them to distribute their works to the public, and simultaneously to ensure that as many works entered the public domain as soon as possible, so that the public could make unfettered use of them.[46] Copyright law was designed to separate works whose authors wanted copyright protection enough to follow a few simple rules for preserving it, from works that would have been created and distributed anyway. 34

Thirty years ago, when you saw something you wanted to use or share, the default rule was that you were entitled to do so. Unless the object was marked “do not copy” you were, with some modest exceptions, entitled to assume it was in the public domain, because the absence of a copyright notice ensured that it was in the public domain (even if it hadn’t been before.)[47] Not only that, but the notice had to be accurate, had to tell you when the copyright was scheduled to expire, and had to tell you to whom you needed to address any request for permission.[48] The overwhelming majority of potentially copyrightable works didn’t have this notice and entered the public domain the minute copies were publicly distributed. Of the ones that bore the prescribed copyright notice, only a fraction were registered, and of the fraction that were registered, only 15% were renewed, so for most of the copyright-protected works that had the requisite notice, copyright protection lasted only 28 years.[49] 35

When was the U.S. army first officially racially segregated?  36

When was it officially integrated?[50] 37

The formalities get a bad rap these days. We’ve left that sort of thinking behind us; we’re more enlightened now. We know better than to condition copyright protection on a bunch of technical requirements.[51] (We feel more comfortable conditioning use of copyrighted works on a bunch of even more technical requirements.[52]) The formalities have been so thoroughly discredited that some of us have even stopped teaching them.[53]  38

What we miss when we dismiss the formalities as characteristic of a provincial and outmoded attitude is that the formalities were the principal method embodied in US copyright law for preserving the public domain and encouraging the public to use, reuse and share potentially copyrightable material. If you read older copyright cases, textbooks and law review articles, you find a broad consensus that copyright law was designed to encourage the growth of the public domain.[54] The theory underlying the system was that a rich public domain was essential to the progress of knowledge. By offering copyright for a limited time to authors who distributed their works to the public, copyright bribed them to generate material for the public domain.  39

The old rules worked to preserve copyright for works whose owners wanted it enough to take the affirmative steps required to assert it. The law made copyright subject to exceedingly modest requirements to claim protection and put the public on notice. It was designed to force everything else into the public domain, so that everyone else could make whatever use of it they wanted. Copyright wasn’t automatic, but it was easy to secure. Putting a notice on publicly distributed copies might not be trivial, but it was far easier than the effort involved in applying for a patent or registering a trademark. Retaining copyright after the initial 28 year term was a little harder, but again, not very hard. Nothing one would need a lawyer for. Meanwhile, the rules were designed to make it easier for people who wanted to negotiate a license to use a work protected by copyright to know whether and whom they needed to ask. Again, for most licenses, a lawyer would be strictly optional. 40

Congress abandoned many of the formalities when it enacted the 1976 Copyright Revision Act,[55] and ditched the rest of them in 1989 when we acceded to the Berne Convention.[56] In 1976, we essentially abolished the rule that publication without notice or with inaccurate notice sent the work into the public domain[57] and in 1989, we abolished the notice requirement entirely.[58] We also made other changes to the law that, cumulatively, reversed the default rule. Today, all potentially copyrightable works are protected by copyright, whether their authors want copyright protection or not.[59]  41

How do you make Shaker lemon pie?[60] 42

A second, less obvious but still crucial, change transformed the US copyright system from one designed to ensure the enhancement of the public domain to one designed to support the indefinite proprietary treatment of articulated thought. In 1976, Congress adopted divisibility of copyright.[61] So far as I can tell, the change was completely uncontroversial. Divisibility is all sorts of useful. It’s the biggest reason that authors don’t need to sign over their copyrights when they publish things. It allows the author to keep control over different sorts of exploitation of her work by different entities. The problem with divisibility is that it potentially requires multiple licenses for any single use of a copyrighted work, while simultaneously making it very difficult to tell who owns the rights one needs to license.[62] Professor Lydia Loren recently summed up the problems the music industry faces in putting its works online: 43

[T]here are too many vested industry players for downstream users to be able to efficiently obtain the authorizations needed for downstream use of recorded music. Second, the divisible yet overlapping rights granted to copyright owners leads to industry gridlock and problems with holdout behavior. Finally, the demands for payment from the downstream user by too many vested industry players, combined with industry consolidation, result in the price being too high to achieve the goal of copyright.[63] 44

There once was an interesting Internet start-up named MP3.com, which specialized in making both major-label and unsigned music available in the MP3 format. MP3.com intended to stream copyrighted music to its subscribers, and bought ASCAP and BMI public performance licenses to allow it to do so. That seems right. If you look at the statutory definition of public performance, it appears to cover Internet streaming quite nicely.[64] MP3.com got sued for willful infringement (and lost) because it didn’t also license the reproduction rights to those songs, which are controlled by a different entity.[65] 45

This is much worse in the Internet context because copyright owners have asserted, so far successfully, that every time a work is made available over the Internet, someone has reproduced the work, distributed the work, and publicly performed or displayed the work.[66] Anyone who wants to post a work on the web, thus, needs a license from the owners of each of these rights, plus a license from the owners of each of these rights in any underlying works that are incorporated within the work.[67] Under the current leading analysis of how copyright law interacts with the Internet, making any material available over the Internet (whether via posting it on a website, sending it through email, posting it to Usenet news, typing it on Internet relay chat or making it available in a share directory associated with a peer-to-peer file trading application) constitutes a reproduction of the material, a distribution of the material to the public, and a public display or performance of the material. It is therefore illegal unless done with the authorization of the copyright owners of the reproduction right, the public distribution right, and the public display or public performance right, as well as the copyright owners of those rights in any underlying material.[68] It counts as an actionable copy notwithstanding the fact that the reproduction may be ephemeral (what the law used to deem unfixed).[69] It counts as a distribution to the public notwithstanding the fact that no tangible copy of the material is transferred (what the law used to deem a display or performance rather than a distribution).[70] It constitutes a public display or performance notwithstanding the fact that any display or performance may occur only between two individual computers (what the law used to deem private).[71] 46

Indeed, there’s some indication in the case law that making a hyperlink to material available over the Internet may be deemed to be a reproduction, public distribution, and public performance or public display, requiring the permission of the owners of the reproduction, distribution and public performance and display rights in the material on the other end of the link.[72] Moreover, the theory underlying the recording industry’s recent service of more than a thousand subpoenas[73] on Internet service providers and universities appears to be that merely possessing an unauthorized digital copy of a protected work can itself be infringing distribution, because a member of the public could download a copy of the work from the possessor’s hard disk. A bill recently introduced in Congress extends that argument further. Under Congressman Conyers’ Author, Consumer, and Computer Owner Protection and Security Act, possessing an unauthorized digital copy could constitute felony distribution.[74]  47

And (as if that weren’t troubling enough) largely because of the adoption of divisibility of copyright, in many if not most cases, it can be difficult and sometimes impossible to discover who the copyright owners of all of those rights are.[75] One of the more disturbing revelations of the Napster litigation was that record companies insisted that they were unable to generate a list of the copyrighted works they claimed to own.[76] (This is particularly disquieting because one would assume they kept records in order to send out those royalty checks they’re supposed to be sending out, but apparently not.) Some of the problem, apparently, is record keeping, but not most of it. In addition to difficulties caused by lost or misfiled records, there is significant legal uncertainty about the ownership of rights to control digital exploitation of works that are subject to contracts contemplating conventional exploitation.[77] Record companies, for example, have claimed to own all copyright rights in the recorded music they distribute under the work made for hire doctrine, but most experts agree that those claims are unpersuasive.[78] A successful effort to amend the copyright law to strengthen the record labels’ work-made-for-hire arguments excited so much outrage among musicians that the recording industry persuaded Congress to repeal the amendment the following year.[79] Without the benefit of a work-made-for-hire claim, though, the record labels’ claims to own the digital rights to the recordings they produce requires a work-by-work, contract-by-contract analysis. New York Times v. Tasini[80] and Random House v. Rosetta Books[81] teach us that contractual assignments of copyright may not necessarily include the electronic rights. We’d have to examine the contracts to be sure. We might need to know whether the case would be coming up on the east coast or the west coast.[82] We’d also need to see the contract between the composer and the music publisher for each song on the recording, and the contracts between each of the music publishers and the record company that recorded each song. Those contracts aren’t publicly available. One suspects that a large number of them are no longer in anyone’s file cabinets either. Bottom line: we don’t know with any certainty who owns the digital rights in any number of recorded musical performances. That’s why record companies have scrambled to settle cases when their ownership of sound recordings is actually put in issue.[83] If I want to share my music collection with my newfound friend who was able to tell me that the “Fuct of Pepsiman” is a promotional toy released in Japan by the Pepsi Cola company, there isn’t any way for me to figure out whose permission I need to ask.  48

Today, in short, everything is protected by copyright and it is almost impossible to figure out whom to ask for permission. Just as we built a communications network that would permit us, if we chose to, to dispense with a complicated and expensive distribution infrastructure, we ditched the legal rules that would have permitted us to do so without high legal barriers.[84]  49

III. The Music of Room A-9 50

What are the lyrics to “The Syncopated Clock?”[85] 51

Last year, my son was in third grade, and one of his assignments required him to conduct research on the flora, fauna, and climate of the alpine tundra. His teacher didn’t send him to look it up in books – indeed, the school library didn’t have a lot of information to offer on the alpine tundra. My son’s teacher sent him to look it up on the Web. She gave him a list of URLs for some websites that were likely to lead him to the information he needed, and sat him down in front of a computer to do his research. At the end of the school year, this teacher said goodbye to the class and presented all of the students with a souvenir: A home-burned CD full of Room A-9’s favorite songs. Where did the songs come from? My son’s elementary school teacher had downloaded them from the Internet herself so the class could enjoy them. Room A-9 apparently especially liked the Sugar Beats’ rendition of “Put A Little Love in Your Heart.”[86] 52

When an elementary school teacher helps her class to download information about the animals that inhabit the tundra, we all agree that that’s admirable. When she teaches the class to download “Put a Little Love in Your Heart,” at least some of us would argue that that’s reprehensible. Collecting information on the Internet is “learning.” Posting information on the net is “sharing.” Try exactly the same thing with recorded music and it’s “stealing.” When my son’s teacher downloads information from the Internet and shares it with her students, that’s the sort of thing the law is supposed to encourage; when she downloads music from the Internet and shares it with her students, that’s the sort of thing the law is supposed to prevent. The law treats the two acts differently because facts are in the public domain, while music is someone’s property. Information cannot be owned, we’re told, because, unlike music, facts aren’t original.[87] From my son’s teacher’s point of view, though, what she’s doing is the same: she’s sharing.[88] From her point of view, there’s no reason to think that it would make intuitive sense that downloading information to share with her students would be good, while downloading music to share with her students would be bad. Those of us who teach copyright know that the distinction between unprotected fact and protected expression is as elusive and counterintuitive as anything in the copyright course. There’s a wealth of literature challenging the rule that information is unlike music in any way that’s important to whether we should give it intellectual property protection.[89] Any originality-based distinction between facts and notes is untenable, we’re told, since unearthing and assembling facts takes at least as much creativity and often lots more money than writing a song.[90] Scholar after scholar has deconstructed the supposed rationales for giving factual information different treatment from fiction, and concluded that the asserted differences can’t be defended. The inescapable conclusion, they’ve told us, is that we need to give comparable intellectual property protection to information.[91] There’s a perennial bill pending in the U.S. Congress that threatens to do just that;[92] it’s even passed the House of Representatives once or twice.[93] 53

Copyright scholars never seem to reverse the syllogism. You never run into an argument that says: if facts and music are equivalent in the respects that matter, and we have an ample, readily accessible and diverse supply of facts when the law gives them no protection, shouldn’t we at least investigate what sort of musical smorgasbord we might develop if we treated music comparably? 54

IV. Resetting the Default Rules 55

Who are the Sugar Beats?[94] 56

We have a mature information market on the Internet that allows almost anyone with a net connection to find the answer to almost any question by consulting what would a generation ago have been an unimaginable wealth of information resources. This information space has sprung up not despite but because of the absence of any copyright protection for facts. (If you doubt me, stop a moment for a thought experiment, and imagine what this information space would look like if we adopted and enforced a legal rule that no fact could be posted without the permission of the originator of that fact or his employer or assignee.) At worst, this information space is an invaluable adjunct to the library of reference books, and at best it’s a superior alternative for retrieving and disseminating information. If consumer-to-consumer dissemination creates a superior information marketplace, shouldn’t we give serious consideration to the idea that it would create a superior music marketplace? The digital information space is compelling at least as much because of the variety and ecology of shared information, as because of the convenience and speed that might be supplied by an online “encyclopedia world,” containing the digitized text of the Encyclopedia Britannica and a dozen of its competitors. Consumer-to-consumer dissemination of music might enable the evolution of a music space with comparable variety. That potential is more exciting than the advantages of instant gratification that accompany the ability to download whatever music the record labels are currently selling. Just as we wouldn’t want to get all of our facts from some giant Encyclopedia Britannica in the sky, there’s no need to cabin our musical tastes to reflect what’s currently selling in online or offline stores. 57

I’m not seriously suggesting anything as radical as that we treat music exactly the way we treat facts, or that we dump all recorded music into the public domain. Copyright has a number of virtues along with its vices. Rather, I’m suggesting that we apply some of the insight we’ve gained from watching the expanding exchange of information over the Internet. Creation and dissemination may flourish without the incentives supposedly supplied by producer control.[95] One of the lessons we can take from the vibrant commerce in facts that goes on over the Internet is that allowing, indeed encouraging individuals to share music, trade music -- engage in non-commercial “stealing” of music if you prefer – without legal liability is not necessarily going to bring the progress of science and the useful arts to a crashing halt, and it has lots of advantages over the distribution system that preceded it. 58

One non-trivial advantage is that consumer-to-consumer distribution is a lot less costly, and may allow us to free up resources now spent on CD burning, shipping, storage, shelf space and radio payola, not to mention the huge cost of legal efforts to stamp out what is commonly called “piracy.” That money could be used to pay the people who create the music – something the record companies insist they can’t really afford to do very well under the current system.[96]

 59

What’s the name of that song that keeps going through my head?[97] 60

There are vast differences between music and information, but outside of the fact that the owners of music and sound recording copyrights have a lot more brute political clout than, say, Reed Elsevier, I’m not sure that any of those differences undercut the basic insight: If music in a digital world shares many of the attributes of information, it may be useful to apply some of the wisdom IP law has developed over the protection and distribution of information. In particular, we should remember that widespread sharing is the goal; facilitating the sale of copies is only the means the law has adopted to further that goal.[98] If sharing is a more effective method of dissemination than selling copies, then prohibiting sharing to protect the market for copy sales is exactly backward.[99]  61

If we can agree on that, I think it’s relatively easy to work out the details of a compromise we can live with.[100] A surprising consensus has emerged that P2P is exciting technology with one serious flaw – creators aren’t getting paid. [101] (That flaw characterizes much conventional distribution as well.[102]) The current conventional system of music distribution has been successful in disseminating a broad range of music to consumers and less successful in compensating the individuals who create that music. Peer-to-peer file trading has so far proved to be a far more effective distribution mechanism for a broader range of music, but is even worse than the conventional system at compensating creators.[103] Tweaking peer-to-peer file trading to incorporate a mechanism for compensating creators is relatively straightforward – there are a host of recent thoughtful suggestions outlining ways to do that.[104] If the only reason we care about compensation for composers and musicians is to induce them to make music, the most efficient option is probably to legalize peer-to-peer file trading, prohibit well-poisoning and leave creator compensation untouched – the recording industry has demonstrated that artists make music even when money is not forthcoming. If our sense of fairness impels us to compensate creators because they deserve to be paid, then extracting creator compensation from peer-to-peer file trading would probably be an easier route than reforming the recording and broadcast industries.[105] 62

A number of scholars have floated thoughtful proposals urging systems that would permit peer-to-peer file sharing, charge money to the people who enjoy it (or the businesses that profit from it), and use those funds to compensate creators and copyright owners.[106] Professor Neil Netanel suggests allowing consumers to engage in unrestricted noncommercial use, adaptation, and peer-to-peer exchange of all types of communicative expression, and imposing a noncommercial use levy to compensate copyright owners. Netanel would impose the levy on the sale of products and services whose value is enhanced by peer-to-peer file sharing; the Copyright Office would divide the levy proceeds among copyright owners using both sampling and digital tracking technologies.[107] Professor Terry Fisher proposes a slightly different solution. Fisher would encourage copyright owners of music recordings and films to register their works with the Copyright Office, which would assign every registered music recording or film a unique registration number. Copyright owners would incorporate the registration number into the names of the digital file containing the registered work. The Copyright Office would be responsible for administering a tax on digital recording devices, digital storage media and Internet access services, and would divide the proceeds of the tax among owners of the copyright in registered works by tracking downloads of files by registration number and using sampling to estimate offline consumption. Anyone would be permitted to reproduce, distribute or perform audio and video recordings over the Internet. Professor Fisher suggests that the initial deployment of his proposal be completely voluntary, but he envisions that it would ultimately replace the current copyright law completely.[108] Professor Raymond Ku argues that the current copyright law makes no sense in the context of digital distribution. Ku would retain the current law for analog distribution, but would replace copyright in the Internet context with a statute imposing levies on sales of Internet service and on computer, audio, and video equipment.[109] Professor Glynn Lunney argues that private digital copying probably does more good than harm, but suggests that if that harm must be redressed, a levy imposed on devices and blank storage media is the best available solution.[110] 63

The differences between these proposals are not, in fact, that large. Netanel would permit unrestricted noncommercial use of a majority of copyrighted material,[111] while Fisher limits his proposal to audio and video recordings, but would allow commercial as well as noncommercial uses.[112] Netanel and Fisher would direct their alternative compensation to copyright owners, while Ku would reserve it for musicians and songwriters. Netanel, Fisher and Ku all, however, rely on a government-imposed, Copyright Office-administered fee on the sale of digital goods and services to provide compensation for missed sales and royalties. Netanel would leave the current copyright law untouched except for his noncommercial user privilege and levy. Ku would retain the current law for analog distribution so long as works were distributed in the analog as well as the digital channel.[113] Fisher envisions his system’s ultimately superseding the current statutory copyright and calls for the copyright statute’s eventual repeal.[114] Despite these differences, however, the core of all four proposals is to permit, indeed encourage consumers to engage in consumer-to-consumer distribution while compensating creators from a fund financed by the sales of related equipment and services. Considered in the context of music, the proposed change is a modest one. Consumers already have a privilege to make non-commercial digital copies of musical recordings,[115] and the right to distribute those copies to members of the public.[116] Consumers, moreover, already pay a levy intended to compensate composers, musicians and record companies for the sales lost through private consumer copying.[117] Netanel, Fisher, Lunney and Ku would extend both the privilege and the levy to copying and dissemination over digital networks. 64

From the viewpoint of the individuals who make the music, moreover, these proposals are remarkably similar to the devices we rely on today to pay money to composers and musicians. In Canada, Europe and Japan, musicians and composers rely heavily on collecting societies. In the United States, a patchwork combination of compulsory licenses, blanket licenses, standard trade practices and rate courts add up to much the same thing. The proposals to enact a new license to permit peer-to-peer file sharing and compensate creators through a levy, tax, or uniform royalty have inspired heated philosophical and economic debates over the flaws in any compulsory or collective licensing system. The objections tend to ignore the fact that composers and performers of music currently receive most of their income through a combination of standardized, compulsory and collective licenses administered by intermediaries (music publishers, record companies, performing rights societies) in return for payment. From the vantage point of music creators, replacing the theoretical control they enjoy under the copyright law with an enforceable promise of payment makes them no worse off, and makes most of them better off.  65

The intermediaries who hold control over musical works and recordings are also in it for the money, and one might expect them to be delighted to hand over their control in return for more cash. Not a bit of it. The current dominant forces in the music and recording business may no longer need record pressing plants, CD burning plants, warehouses and trucks to distribute music, but they have a huge stake in ensuring that digital distributors be limited to those who used to rely on record pressing plants, CD burning plants, warehouses and trucks. They rest of us, however, don’t share that stake. Indeed, new distributors who never assumed those expenses may be in a position to experiment with new variations on digital distribution and still pay a larger percentage of proceeds to the creators of the material. 66

The Fisher, Ku, Lunney and Netanel proposals would improve the law by allowing frictionless, consumer-to-consumer dissemination and collecting royalties to compensate creators from those who in a broad sense may be described as commercially exploiting copyrighted works.[118] Moreover, when their schemes are limited to music, currently the most vexing case of consumer-to-consumer dissemination, the proposals are modest extensions of devices contained in current law and business practice. In drawing on their analyses, I end up suggesting a variant solution pegged at least initially only to music sharing, but my pursuit of some different choices shouldn’t obscure the importance of their work. The politics of copyright legislation will likely prevent us from adopting any of the four proposals they advance, but our copyright law would be much improved if we did.  67

V. Sharing and Hoarding

How violent is next week’s episode of Dragonball Z?[119] 68

If I’m persuaded that politics would prevent the adoption of a Netanel/Fisher/Ku/Lunney solution, why am I bothering to articulate my own variation? As consensus builds around the idea of paid peer-to-peer, it seems increasingly plausible that some legislation will emerge with enough support from the music, recording, computer, and consumer electronic industries to have a fair chance of enactment. I expect that that legislation will include both consumer downloads of music and collective licenses to pay for them. Such a bill is less likely to resemble the proposals advanced by Netanel, Fisher, Lunney and Ku, however, than it is to be designed to maintain the current recording and music industry distributors in their market dominant position. Most importantly, it is less likely to incorporate a privilege for consumer-to-consumer dissemination than it is to include measures designed to prevent it. If we are willing to give up consumer-to-consumer dissemination in return for the instant gratification of licensed direct downloads, the recording industry is probably willing to sell us copy-protected files replicating at least a portion of the music it makes available in stores.  69

The prospect of downloading copy-protected versions of music otherwise available in stores is not particularly enticing. This is the music version of the online encyclopedia world, and we can do better. The promise of being able to find music that is not available in stores, and to share it with other consumers, in contrast, is compelling. Lots of music is not available in any store, because it's old, it's obscure, it has little commercial potential, or the rights can't be cleared without a statutory license or privilege because it's just too difficult to figure out who owns them. Consumer-to-consumer music dissemination makes it possible to find and share that music.  70

The fact that more than sixty million consumers are currently exchanging music over peer-to-peer networks in the U.S. gives them a stake in the building consensus and both a moral and a political claim to a seat at the copyright bargaining table. The details of any proposal for an online music system will determine the extent to which it promotes unfettered consumer-to-consumer exchange, allows untethered consumer use, encourages the broad dissemination of a wide variety of music of disparate types, takes advantage of the economies made possible by digital distribution, and pays composers and musicians. The details of such a system will also determine whether and to what extent it requires copyright police to enforce its rules. The more conversations that people who are not copyright lobbyists can have about the details of a revised copyright bargain, the better positioned they will be to shape the law Congress may enact.[120] The devil will be in the details, and focusing on the details allows us to figure out which ones are most important. 71

I suggest that we should try to build a music space that resembles the current digital information space in the ubiquity of music it contains and the ease with which music may be shared, and that we should devise a combination of blanket fees or levies designed to compensate the creators of the music we exchange. In order to achieve the breadth and diversity of music, and the community of consumers who enjoy it, that has evolved in the Internet information space, we will need to rely on consumer-to-consumer dissemination as well as licensed downloads or streams. If we as consumers want to pay for the music we exchange, we need some form of blanket fee or levy to enable us to do so. Because some creators and copyright owners find the idea of consumer-to-consumer dissemination unacceptable, I suggest that we devise a way to allow them to withhold their music from the system. To discourage them from electing that option, I believe we should optimize the legal infrastructure for sharing. I’ve drawn the details of that infrastructure with an eye toward recapturing some of the lost advantages of notice and indivisibility. 72

So, should I have a drink with that guy or not?[121] 73

We should adopt a legal architecture that encourages but does not compel copyright owners to make their works available for widespread sharing over digital networks. We should incorporate into that architecture a payment mechanism designed to compensate creators and to bypass unnecessary intermediaries.[122] That mechanism should have sufficient flexibility to allow current and new upstart intermediaries to devise useful value-added flavors of intermediation and collect dollars accordingly. Fisher, Ku and Netanel have done a pretty good job of identifying the goods and services that will need to be subject to a copyright “tax,” “fee,” “levy,” or “royalty. ”[123] The pot of money generated by the fees should be disbursed by whatever agency Congress decides should replace the current unworkable Copyright Arbitration Royalty Panels.[124] In the first instance, money should be disbursed to the musicians and composers who author music and recordings.[125] 74

If the legal architecture encourages sharing but permits what we might want to call “hoarding,” then consumer-to-consumer exchange can develop without difficult legal or technological barriers.[126] Thus, I’d be willing to incorporate a limited, carefully structured, notice-based opt out for copyright owners who prefer control to payment. If hoarded music is indeed superior, it will be able to compete with the “free” stuff. (If it can’t compete with the “free” stuff, then overall welfare is probably enhanced if we refuse to subsidize it with expensive legal barriers and copyright police.) To achieve a legal regime that encourages sharing but permits hoarding, we should impose a requirement that copyright owners who decide to hoard must forgo any payment for hoarded works from the common payment system, and must take affirmative but relatively modest steps to exclude their works from the network and enable consumers to quickly and painlessly verify that those works may not lawfully be shared.  75

My reasons for preferring a system that copyright owners can choose not to participate in are at least in part pragmatic. First, so long as the legal and technological architecture are optimized for sharing, allowing copyright owners to withhold their works does little harm. Paid subscription information and news sites on the Internet coexist comfortably with sites that are open to the public and free of charge; if we can duplicate that peaceful coexistence for digital music, it seems sensible to try to do so. Second, if we design an alternative compensation system to collect enough money to compensate the proprietors of mega-hits for all of their forgone income, we can expect that the expense of such a system will be unreasonably high, and that the compensation paid to the creators of more modestly successful music will be unreasonably low. Third, if such a system allows copyright owners to decline to participate, it seems more likely that it will be deemed at least arguably compliant with our treaty obligations under the Berne Convention and the WIPO Copyright Treaty. Finally, my proposal is motivated in part by my conviction that composers and musicians have been ill-served by the current system. If they nonetheless prefer the dysfunction they know to a new and unproved system, and we can make the system work without including them, I see no important policies that will be served by forcing them to continue to participate.[127]  76

To enable an opt-out mechanism that won’t deform the legal and technical architecture encouraging sharing, I suggest that we try to reproduce the functions that notice and indivisibility provided before we abandoned them.[128] The key to the opt-out mechanism I propose is the selection of a single digital file format or family of formats capable of conveying copyright management information as defined in section 1202 of the copyright act.[129] The format will probably incorporate digital rights management capability because the people who will be using it will desire that feature, but there’s no need for any copy-protection to be hack-proof, or even exceptionally durable. It should also be compatible with the current generation of digital playback devices, including CD players.[130] I’ll call the format “.drm” for short.[131] Any musical work or sound recording that is made available to the public, under the copyright owner’s authority, only in *.drm format will be ineligible for sharing or compensation. To protect copyright owners from having their works kidnapped into sharable file formats, we should require them to enter the copyright management information for works made available to the public only in *.drm format into a single, universal, searchable index of works that have opted out of digital sharing. The authoritative copy of the database should probably be maintained by the Copyright Office, although other computers should be encouraged to mirror it. At such time as the creators or copyright owners of a work desire to participate in the revenue earned from digital sharing, they may remove their listing from the database, publish the work in another format and become eligible to collect compensation. 77

Whatever happened to Herman’s Hermits?[132] 78

With the exception of works released only in *.drm format, consumer-to-consumer dissemination and any reproduction, distribution or public performance or display that it entailed, would be completely legal. Any music that’s already been released in other formats could not be recaptured by releasing it in a *.drm file, so the music currently being shared over peer-to-peer networks could not be locked back up. It would, however, be eligible for compensation. Creators of new releases could choose to make them available for sharing or they could hoard them and forgo both the free distribution and the additional income that sharing would generate. What happens when a noncompliant consumer trades a *.drm file over a peer-to-peer network? Copyright owners would have the options they have now for pursuing consumers who exchanged material illegally. Notwithstanding that record labels are currently pressing copyright infringement suits against individual peer-to-peer file traders, why would they settle for a system that gives them no more effective an enforcement mechanism than the one they have now? My answer is that if hoarding were reserved for new releases with significant commercial potential, I believe that consumers would be far more likely to respect the choice to hoard, and I believe that a law with broad consumer support is easier to enforce than one that lacks it.[133] 79

The use of a single file format will enable consumers easily to identify files they may not share and facilitate efforts of software designers to create file sharing software that blocks transfers of proprietary files. The maintenance of a master searchable index will enable consumers to ascertain whether particular files are unauthorized counterfeits. The use of a single, identified file format augmented by a searchable index captures some of the public benefits of copyright notice and registration. By allowing copyright owners to opt out of file sharing so long and only so long as their work remains published in the single *.drm format, the proposal mitigates the effects of divisibility by requiring the copyright owners to coordinate with one another in exploiting their works. 80

To preserve the integrity of the system, it will be necessary to impose a penalty for entering or maintaining inaccurate copyright management information in the database. This is especially important given that the system contemplates that copyright owners will elect to stop hoarding and start sharing, and will need an incentive to ensure that their database entries reflect the change in their works’ status. The penalty should be significant, but it need not nor should not be as harsh as the classic penalty for inaccurate copyright notice. Rather than forfeiting copyright entirely, inaccuracies in the database should be punished by allowing, for some appropriate period of time, the assertion of a defense in the nature of copyright misuse to actions to enforce the copyright in the work.[134] Since copyright owners will either be collecting compensation derived from consumer-to-consumer distribution, or keeping their works ineligible for peer-to-peer sharing, there will be no legal excuse for the sort of self-help shenanigans we’ve recently seen deployed as a high-tech version of copyright-owner vengeance against peer-to-peer file traders. Since we’re relying on the consumer-to-consumer distribution network, moreover, any poisoning of the well is likely to spoil legitimate file trading for consumers and other creators. Any spoofing, interdiction, denial of service attacks, decoys and similar self help, if undertaken under the authority of the copyright owner, should be grounds for asserting a similar copyright misuse defense. 81

My specific proposal is inspired by an impulse to see whether an architecture like the one that has permitted the Internet to flourish as an information space can define a thriving music space. The U.S. recording industry’s recent enforcement campaign seems to seek to move us in a very different direction. It promises us something we would all agree is desirable if we only renounce what to many of us, it has become clear, is crucial. The recording industry appears poised to accept a world in which we agree to allow consumer downloading (either for free or for a price) but not what the recording industry is calling “uploading” – which is the state of having on your hard disk a music file that someone else can search for and copy from you. Just as the idiosyncratic interests of large numbers of individuals who want to share is directly responsible for the wealth and incredible variety of information we can find when we go looking for it, I think that consumer-to-consumer file trading has the potential to make it economically feasible to distribute a much broader variety of music to a much larger audience. I’d hate to lose that potential just because it’s strange, new, unproven, and not yet well represented by lobbyists. 82