Getting Copyright Right
Mandatory copyright licensing legitimized the early radio and cable TV industries. Can it do the same for the Internet?
In 1842, Charles Dickens implored his U.S. lecture tour audiences to support a bilateral copyright treaty. Works like his Pickwick Papers were wildly successful on both sides of the Atlantic, but no royalties came his way from sales on the western side. U.S. copyright law concerned only U.S. works. Without a treaty, no native law was broken when a U.S. publisher pirated his stories.
In the end, Dickens gave up on legal remedies. Instead, he sent one or another U.S. publisher an early copy of a work--for example, Harper Brothers paid him UK £1250 (at least UK £150 000 today) for an advance proof of Great Expectations. If someone was going to take a single copy and commercially reproduce it, Dickens wanted to make as much from that one copy as he could.
Copyright used to be a law governing the rights and duties of commercial publishers, because only they had the means to produce and distribute books and other works. Now, however, instead of a handful of companies with printing presses, there are over 500 million other "publishers"--people with access to the Internet--who can take a single copy of a digital book, a song, or even a film, and distribute any number of perfect copies. How, then, are that first copy's creators--authors, songwriters, filmmakers, and publishers--to be rewarded?
Digital music and e-book publishers have chosen to use complex digital rights management technologies to restrict when, how long, and on what devices consumers can listen and read. Such restrictions, by and large, do not exist for CDs and bound books [see "Making Music Pay," IEEE Spectrum, October 2001, pp. 41-46]. Moreover, copyright law has grown too complex for even the lawyers to fully understand, while the big record companies themselves are struggling to assemble all the licenses needed to form their own Napster-like music services.
In the past, the U.S. Congress, the Canadian Parliament, and other governments have found an effective solution to media licensing problems: mandatory licensing schemes that remunerate the publisher without individually licensing each distributor. Under such a regime, record companies and other rights holders could not withhold licenses for the music files created by the new Internet-based peer-to-peer exchanges, which, however, would charge participants a modest monthly fee. Composers, musicians, and other music artists would be paid mandatory royalties doled out by authorized agencies. Monitoring and metering software would record how often a music file was downloaded and the artists would be paid accordingly.
A classical paradox
Copyright balances the competing interests of authors and their publishers, who want their work rewarded, and the public, which wants artistic works to be available. As expressed in the U.S. Constitution, for example, the goal of copyright is "to promote the progress of science and useful arts." The means has been, paradoxically, to induce an artificial scarcity by outlawing all but one commercial source of a copyrighted work. In this way alone will the market reward an author enough to motivate him or her to compose new works. Never has the scarcity seemed more artificial than today, when it only takes seconds to download a song from a stranger's hard disk.
All the same, copyright has always been limited in scope and duration. According to Stanford University law professor Lawrence Lessig, "The framers [of the U.S. Constitution] never envisioned giving copyright holders perfect control over the use of their writings." Only books, maps, and charts were protected by the first copyright act, in 1790, whereas current law covers virtually all media. Until 1976, copyright had to be registered with the Copyright Office, but now putting hand to pen or keyboard is sufficient. Originally lasting for 28 years (at most) from the time it was issued, copyright now usually extends for 70 years beyond the life of the author; according to Lessig, its term has been increased 11 times in the past 40 years.
Dickens's holy grail, a copyright agreement between Britain and the United States, was signed in 1891, just in time for printing presses to be outshone by electrotechnology newcomers, about whose copyright status the law was initially in doubt: player pianos, jukeboxes, phonographs, and movies.
A hundred years of new media
Changes in copyright law have been incremental, but always in one direction--toward narrowing the public domain. For the past century, they have responded to new technologies, which, in effect, created new classes of publishers where none had existed before.
In the case of music, the odd player piano technology of perforated-paper rolls required that a new distinction be drawn between a performance right and a mechanical right. The perforations directed air currents at levers, depressing them as a pianist's fingers would. But there was no pianist, engendering a conundrum over whether copyright even applied.
The U.S. Congress solved that problem with its first law for mandatory licensing. As in every scheme of its kind, a piano roll publisher of a copyrighted song paid a fixed or negotiated rate to use the work, but did not need permission to use it. The player piano is an historical footnote today, but an astonishing two million were in operation in the first part of the last century.
In the 1940s, a universal licensing scheme was developed for radio and ratified by the courts in the form of a consent decree. Compulsory licensing was again the chosen mechanism in the 1970s, when the cable TV industry outgrew its roots. As Lessig tells the story, cable began as a pirate operation: "community access television," as it was called, consisted of placing "an antenna on a mountain" and running "a cable line down to a community in a valley." Lawsuits eerily like those against Napster followed, but the courts were slower to act than Congress, which stepped in with a compulsory license arrangement in 1976.
New technologies, such as photocopiers and videocassettes, on occasion had built-in safeguards against illicit copying, some of them unintentional. Changes in 1976 to copyright law clarified fair use of printed material, but by and large print publishers could rely on the mediocre quality of the photocopies themselves (especially compared to bound books and glossy color magazines) to deter illegitimate use. For videocassettes, though, the law allowed VCR owners relative freedom in copying TV programs, so film publishers added technology that, mimicking photocopies, artificially degrades a videotape when a consumer copies it. (Today this strategy is belatedly being tried for CDs.)
Digital media, such as music CDs, present an even tougher challenge to existing publishers. Commercial software manufacturers flirted with copy protection schemes in the 1980s, but for entertainment media nothing much changed in law or technology until the Internet and MP3 compression. The first created new distribution channels, and the second enabled listeners to use those channels, even with narrowband connections, to download songs.
The Internet has inspired visions of a universal jukebox, host to every song ever recorded. But that's a lot of songs, most of which are out of print. Then, too, there's the mare's nest of copyright rights going to composers, their publishing companies, rights-collecting agencies, featured artists, nonfeatured vocalists, nonfeatured musicians, their rights-collecting agencies, and so on, all for multiple different countries, performed (streamed) as well as recorded (downloaded).
Napster solved both those problems, the first by offering a peer-to-peer system that inspired millions of people to themselves create and store MP3 equivalents of every piece of music they owned, even old vinyl records and bootleg concert tapes. As for the second, it just ignored the mare's nest, a strategy that brought it millions more users but sat less well with others, such as major record companies and the rock band Metallica, who sued Napster, and Judge Marilyn Hall Patel, whose Federal court in the northern district of California ordered the service off-line. Napster as an epicenter of unpaid file-sharing is gone, though replacements have emerged--by October more people were using Morpheus and KaZaa, two file exchanges constructed by Amsterdam-based FastTrack, than had ever used Napster.
The major content holders have further insulated their copyrights: under the 1998 Digital Millennium Copyright Act (DMCA), no one may create or distribute hardware or software that circumvents a copy-protection scheme. Thus besides suing someone for copyright infringement, encryptors of copy-protected content can sue someone for merely decrypting it.
If earlier technologies and the ensuing changes to copyright law narrowed the public domain, it has been eviscerated by the DMCA, critics charge. The copy protection, and the digital rights management schemes that incorporate it, can take over traditionally fair uses of content, like your ability to copy a song or book onto a portable device for use when away from your computer. Arguably, a work like Great Expectations, which was in the public domain for the entire 20th century, could be combined with some new copyrightable illustrations or an introduction and wrapped in a copy-protection scheme that would prevent a schoolchild from accessing Dickens's no longer copyrighted words.
Revealingly, all three authors in the 2001 crop of books on digital copyright, Siva Vaidhyanathan, Jessica Litman, and Lessig, advocate that the public intellectual commons needs to be expanded, not narrowed [see "To Probe Further"]. A commons is usually thought of as a physical resource, perhaps a field in which several farmers may graze cattle. These authors analogize this notion to a virtual field where the grasses are the ideas of science and entertainment.
Lessig argues that in some ways the ideas can create more of a commons than the field of grass can. While the value of grassland diminishes if too many cattle graze on it, the value of ideas increases the more they are used.
Is the domain of ideas a commons? Vaidhyanathan, for instance, traces dozens of threads of inspiration, like the one that wends from the blues cotton fields of Robert Johnson to the rock and roll hits of Led Zeppelin. Another thread connects Shakespeare, himself no slouch at borrowing plots and characters, and his Romeo and Juliet to West Side Story and Tony and Maria. Dickens's A Christmas Carol has been retold any number of times, including one modern version where the character of Scrooge is a network television executive.
In creating and appreciating movies, books, and music, Vaidhyanathan argues, we rely on and draw from a common culture, a cultural commons. How then is the commons of ideas to be expanded, instead of shrunk?
License to copy
One answer to the copyright issues of digital music, which might also legitimize exchange services like Napster, is for Congress to order some form of compulsory or blanket licensing for downloaded music. Such a scheme has been proposed by, among others, Jim Griffin, founder of the Los Angeles record label Cherry Lane Digital and the former chief technology officer of Geffen Records, Santa Monica, Calif.
Earlier mandatory licenses for radio, cable television, and satellite television dictated a statutory rate (set by Congress itself), or required that parties negotiate one, with a provision for arbitration. Congress has even mandated compulsory licensing for Webcasting, that is, radio-like streaming music services. (That law doesn't include "interactive" services, such as Napster's music-on-demand service.)
Compulsory licensing is not required for interactive music services in principle, but is in practice, because of the difficulty of obtaining all the licenses that might be involved. According to Brian Zisk, cofounder and technology director of San Francisco's Future of Music Coalition, "In no entertainment businesses I've been involved with have bandwidth costs or broadband access been a fraction of the hassle of licensing issues."
Another San Francisco enterprise, the on-line music service MusicBank, was well-funded, even to an investment from a major record label, and well-intentioned about copyrights. For want of licenses--though it struggled for its entire two-year life to acquire them--it ran out of money last spring before it could launch. But things are looking up, according to entertainment lawyer Whitney Broussard. Recent agreements involving the Recording Industry Association of America, which represents the major recording companies, provide an outline of a legislated compulsory license for on-line music [see "Three Steps to Breaking the Licensing Logjam" ].
Cherry Lane's Griffin thinks a number of on-line music services would spring up quickly, once the license issue was resolved. He envisions access to these universal jukeboxes as monthly subscriptions, either from companies like Napster or from the major record labels themselves. Users would pay a fixed rate for unlimited access to a database that could include every song recording in existence. This approach is already used in cable and satellite TV, satellite radio, and Internet service.
Subscriptions are not the only source of revenue. For example, a small tax is already placed on sales of blank CDs that are used to store digital music; proceeds go to music rights holders. That tax could be extended to memory sticks, data CDs, even hard disks.
After all, if music is easily played, swapped, and transferred to mobile devices, all at reasonable prices, the music industry might well capture more money than it currently sees. Even lower royalty rates and lower profit margins may generate more revenue if they come from more users and greater usage.
Donna Hoffman, a professor of marketing at Vanderbilt University, has studied consumer behavior on the Internet for almost a decade. According to preliminary research done by the school, US $4.95 per month is "a good place to start" for services with a lot less music than the universal jukebox.
Meanwhile, the boost to Internet service providers, consumer appliance manufacturers, cellphone companies, and suppliers of related goods and services would be enormous. It should certainly be of some interest to companies like Sony Corp. of America and Warner Music. The first makes more money on CD players than its Sony Music subsidiary does with CDs, and AOL Time Warner, parent of Warner Music, sells more on-line services than songs. Large Internet service providers, if they paid a dollar per user per month, would create a fund of almost $2 billion per year just for North America. (Internet access costs $20-$50 per month, so this tax would be 5 percent or less.)
Running the meter
In a compulsory licensing scheme, money has to be distributed as well as collected. Lindsay Moir, president of RightsMarket Inc., Calgary, Canada, has pondered the technology behind dividing the artist royalty pool. The goal is to monitor downloads in such a way as to measure the relative popularity of all songs and artists.
For U.S. radio today, it's fairly easy. There are centralized servers for broadcasts, plus the few thousands of radio stations, many of them controlled by a few companies who use centralized playlists. These playlists, or logs, in conjunction with statistical surveys, tell rights-collection agencies how often a song was played, and approximately how many people heard it.
Initially, these radio logs, though a poor measure of Internet downloading, could serve on-line services as well. Moir said that metering technology could be added to the handful of central media servers that will exist--the future Napsters of the world. Not all file exchange services have central servers, though. Thus metering would have to be added to the media-playing software in computers, MP3 players, and other devices. Only actual downloads in the aggregate would be needed, sidestepping concerns about privacy.
While the metering process would need to be proofed against tampering, presumably with cryptography, the media itself might not need proofing. That would avoid problems presented by the encryption upon which digital rights management solutions are based. For example, a digital signature, which is a number unique to a digital song and based on its mathematical properties, might be calculated for more accurate identification of individual songs. Moir noted it is counterintuitive to hide, that is, encrypt, something as public as a song, when you only want to count how often it is copied.
Moir's company, RightsMarket, has already come up with some of the end-user technology for the metering scheme he envisions. A software rights client needs to be downloaded by everyone with a media player. It receives data from the music-playing software, letting it know, by means of the digital signature, what song is to be played. The rights client informs a central counting server, storing the information locally until the user is on-line. Such a system might involve some changes to media-playing software, but many programs, such as the Netscape and Internet Explorer browsers, Adobe Acrobat, and Microsoft Office applications, already have built-in hooks that let them work with rights clients.
Consumers have an incentive to participate in metering. It provides "votes," hence money, for the artists they favor, at no extra cost to themselves, since metering is used only to divide the revenue pool. Indeed, some might try to rig the system by downloading the songs of their pet artists over and over. Electronically, however, this sort of anomalous behavior is like the fraud that credit card companies are on guard against, so, Moir said, this is a problem with a known solution. Moreover, while credit card companies need to find fraud exactly where it occurs, metering only needs to be accurate enough to compensate artists fairly.
Books and movies, too
A compulsory license scheme, with pooled royalties and monitored usage, could also work well in other media, especially e-books. The public's lack of interest in reading off a computer screen suggests that print sales would be unaffected, while new markets would develop among those for whom lower prices would be a decisive attraction, or who consume books more easily in digital form, magnifying the text on-screen or being read to by text-to-speech software. Open and easy on-line music browsing and previewing has been one explanation for Napster's possible contribution to CD sales; one can easily imagine it helping those of traditional printed books.
Movies are less easily exchanged on-line because of their gargantuan files, but will soon be napsterized anyway. The college campus, where wide network pipes feed even dorm rooms, was the birthplace of Napster. In the last year many institutions have noticed movies being exchanged across their networks; several have instituted bandwidth quotas for the first time. In October 2001, SONICblue, Santa Clara, Calif., released a new version of its ReplayTV personal video recorder with a feature that allows one user to send recorded TV transmissions to another ReplayTV recorder. With the fairly simple addition of software to tell one user which other users have which shows, plus a way to pull, instead of push, them across the network, this would be a complete peer-to-peer video service. (As soon as the new recorders were released, SONICblue was sued by the major movie studios, more for its ability to bypass advertisements than for the peer-to-peer feature.)
The future of all media, not just music, is more, not fewer, copyright rights complications. Vaidhyanathan's examples of rights problems extend far beyond rap jazz riffs and rap music sampling. He notes that from the 1909 BenHur to 1997's Amistad, movies have had rights problems; in an especially egregious example, in 1995 a film was held up for weeks by an injunction obtained by a visual artist who recognized a chair design of his. (The chair was completely incidental to the film's story.)
As digital media move from $500 000 studio production systems to $500 PC editing programs (much as MIDI and other technologies have moved much of professional music production to the desktop), we can expect that the lingua franca of tomorrow's emerging artists will be multimedia synthesis. (With its usual foresight, the Motion Picture Experts Group, creators of the MP3 format, has designed the next MPEG standards to tag and track the individual elements in a multimedia work.)
We can insist on individually compensating the hundreds of rights holders whose work is included in even a two-minute rock video, or the thousands for a full-length movie, which will stymie the creativity of the next generation just as lawsuits over sampling in the early 1990s defused the rap music revolution. Or we can work toward a more rational system, one that recognizes the necessary role played by the primordial soup of mass culture, out of which all new cultural works emerge.
To Probe Further
Three remarkable books on copyright published in 2001 are Digital Copyright, by Jessica Litman (Prometheus Books, Amherst, N.Y.), Copyrights and Copywrongs: The Rise of Intellectual Property and How It Threatens Creativity, by Siva Vaidhyanathan (New York University Press), and The Future of Ideas, by Lawrence Lessig (Random House, New York). Actually, Lessig's is on the future of innovation, in which copyright plays but one critical part. All three authors made themselves available to IEEE Spectrum in interviews as well as through their writings.
Jim Griffin expressed his ideas on compulsory licensing in April 2001 testimony before the U.S. Senate Judiciary Committee. It is posted as a white paper, "At Impasse: Technology, Popular Demand, and Today's Copyright Regime," on the Web at http:// www.evolab.com/at_impasse.html.
Spectrum wrote about digital music and digital rights management services in "Making Music Pay," October 2001.