(to be published in the December 1994 issue of Communications of the ACM)
If the National Information Infrastructure (NII) is to achieve its potential as a channel for distribution of a wide range of creative works, says the Report, authors and publishers of those works will need reasonable assurance that their intellectual property rights will be respected. Digital networked environments pose particularly severe challenges for owners of intellectual property rights because digital networks make it so simple for members of the public to make multiple copies of those works and distribute them to whomever they choose at virtually no cost. Left unregulated, this activity would undermine the incentives of authors and publishers to invest in the creation and distribution of creative works, for the first distribution of a digital copy to the public would enable those who receive it to set themselves up as alternative publishers of the work, able to undercut the first publisher's price because they need not recoup any development costs. On this point, the drafters of the Report and I are in agreement.
Where we principally disagree is about the wisdom of making certain changes to copyright law and about the Report's characterization of these proposed changes as "minor clarifications and changes necessary to modernize copyright law for digital networked environments. The Report recommends: (1) making digital transmission of a copy of a copyrighted work an act of copyright infringement; (2) abolishing the "first sale" rule for works distributed by digital transmission (this rule generally permits owners of copies of copyrighted works to redistribute their copies without the copyright owner's permission); and (3) making it an infringement of copyright to construct or distribute any device intended to circumvent copy-protection systems by which owners of the copyright might attempt to protect their work.
As the remainder of the column will demonstrate, the Report misrepresents the current state of copyright law in several important respects. In particular, it overstates the extent to which current law favors publisher interests. It downplays the extent to which the changes it recommends would, in fact, bring about a radical realignment in the historical balance between publisher interests and the public interest in access to information products, pushing the law in a direction that would favor publisher interests to the detriment of the public interest. It would abolish longstanding rights that the public has enjoyed to make use of copyrighted works, rights that have been consistently upheld in courts and in the copyright statute. The Report is full of legalistic terminology that makes it difficult for members of the public to read and comprehend. As a consequence, it doesn't provide an adequate basis from which the public, including the technical community who reads Communications, can make an informed judgment about whether the public should accept this revised copyright law. The remainder of this column will translate the Report and its recommendations into plain English so that readers can understand what is at stake and why I question whether the Report's recommendations would be in the public interest.
To put the point plainly, let me say that not since the King of England in the 16th century gave a group of printers exclusive rights to print books in exchange for the printers' agreement not to print heretical or seditious material has a government copyright policy been so skewed in favor of publisher interests and so detrimental to the public interest.
One respect in which the Report interprets copyright law more expansively than Congress has intended is in its statement that "browsing" a work in digital form is an infringement of copyright (unless authorized by the copyright owner). Neither browsing nor reading a work has ever been regarded as an infringement of copyright. When I go to bookstore or a dentist's office, I can browse a book there without infringing its copyright. If I thereafter buy it or another book, I can lend the book to a friend so he or she can read it. Neither of us has interfered with any exclusive rights of copyright owners. (Although I will have distributed a copy to my friend, this does not violate the exclusive distribution right because the copyright owner is generally entitled to control only the first sale of a copy to the public. My personal property rights in the copy I purchase override the copyright owner's interests in further distributions of that copy.)
So what makes the drafters of the Report think that browsing and reading--or any other use, for that matter--of digital works should be regarded as copyright infringement? It is because, in contrast with printed works, works in digital form can only be browsed, read or used if the machine on which they are displayed makes copies of them. But rather than explicitly recommending that copyright law be amended to make all browsing, reading, and uses of copyrighted works in digital form into acts of infringement--a recommendation likely to be highly controversial--the Report takes advantage of an incidental property of digital works (that they need to be copied in order to be browsed or otherwise used) to assert that existing law already allows publishers to control all uses of works in digital form. This lucky happenstance makes it unnecessary for the drafters of the Report to mention that they are advocating a vast expansion of copyright scope.
AN EXCLUSIVE RIGHT OF DIGITAL TRANSMISSION?
The Report is more express in its endorsement of another expansion of the exclusive rights of copyright. It would give copyright owners an exclusive right to control digital transmissions of their works. To understand why such a right might be needed, it is necessary to realize that the present copyright statute grants copyright owners an exclusive right to "distribute copies...to the public by sale or other transfer of ownership, or by rental, lease, or lending." The Report would change this phrasing to add "or by transmission" after "lending" in the statute.
The Report recommends this change because current statute is too focused on the distribution of physical objects and transfers of rights in physical objects. The term "copy," for example, is defined as a "material object...in which a work is fixed...." If the statute only gives copyright owners rights to distribute material objects, it may be ill-equipped to deal with digital transmissions, for they are distributions of bit streams, not of physical objects. Posed in this manner, the Report's argument for adding a provision that permits copyright owners to control digital transmissions seems quite plausible.
Yet, by reading the Report as a whole, one might question whether an explicit digital transmission right is really necessary. The Report discusses two recent cases in which judges treated digital transmissions, such as up- and downloading software from a bbs, as violative of both the reproduction and distribution rights of copyright law. In truth, if the courts took the reproduction and distribution rights as literally as the Report sometimes does, it would be hard to argue that *any* digital copy infringes copyright since all digital copies are, by their very nature, immaterial. Despite the sophistical appeal of an argument that digital copies don't infringe because of their immaterial nature, courts have rejected such arguments. This too suggests that no statutory change may be necessary to give copyright owners the right to control digital transmissions.
Before delving into a more subtle reason for questioning the desirability of the digital transmission right, I want to highlight another respect in which the Report takes a more expansive view of the exclusive rights of copyright than Congress intended. The Report endorses the conclusion of some relatively recent cases that digital copies "fixed" only in RAM infringe the reproduction right, notwithstanding language in the statute and the legislative history indicating that Congress intended to limit the scope of the reproduction right to those copies sufficiently permanent or stable to permit the work to be perceived or reproduced for more than a transitory duration. A legislative report about this provision gave as an example of a noninfringing reproduction the temporary display of images on a screen. Proponents of the view that RAM copies infringe copyrights argue that as long as the machine is on--and it can be on indefinitely--a copy of the copyrighted work stored there can be perceived or reproduced, thereby satisfying the "more than transitory duration" standard. (By this logic, holding a mirror up to a book would be infringement because the the book's image could be perceived there for more than a transitory duration, i.e., however long one has the patience to hold the mirror.) Applying the logic of these cases, the Report seems to view any digital transmission as an infringement of the reproduction right because of the copies made during the transmission as well as when the transmission arrives at its destination. This is a questionable interpretation of current law.
The more subtle reason to question the need for and desirability of a digital transmission right is that it would change existing law far more than the Report admits. This change too would favor publisher interests over the public interest. To understand why, it is worth noticing that of the existing exclusive rights of copyright, the one that the proposed digital transmission right most closely resembles is the exclusive right on which broadcasters principally rely for the protection of their products. Broadcasters don't distribute physical objects; they transmit intangible information which the public can view with the aid of television and radio machines. Like broadcast television today, the NII may eventually be used to provide a wide variety of motion pictures and other programs to the public with the aid of satellite technologies. The NII Report invokes the image of a "celestial jukebox" by which consumers might order a particular movie which, with appropriate compensation to the holder of the copyright, could then be received by the consumer in the privacy of his or her home.
Digital transmissions of copyrighted movies frequently violate two of the existing exclusive rights of copyright: those pertaining to public performances and public displays of copyrighted works. If these exclusive rights already provide a means for controlling many digital transmissions, surely it is fair to ask whether copyright owners really need a new exclusive right to control distributions by digital transmissions. Although the Report does not say so, its digital transmission right would rectify what copyright industries today regard as a very serious limitation on the scope of the rights current law gives to rightsholders. Copyright law does not grant owners rights to control all performances and displays of their works, but only *public* performances and displays of those works. (When you and I sit at home and watch a program on television, copyright law considers our viewing as a performance and a display of a copyrighted program. Because it is not a public performance or display of the work, this activity is not a copyright infringement.)
The real purpose behind the proposed digital transmission right is to enable copyright owners to control *all* digital performances and displays of copyrighted works, without regard to whether they are public or private. Adoption of the digital transmission right would, in effect, repeal the public performance and display rights of copyright and replace them with exclusive rights to control all performances and displays of copyrighted works distributed in digital form. Had the Report explicitly recommended repeal of the public performance and display rights, its recommendations would provoke controversy. By seeking the repeal indirectly, the Report hopes to avoid this controversy. Perhaps a case could be made for such a repeal, but the Report does not make a persuasive argument on behalf of this vast expansion of the rights of copyright owners.
To understand how fully the NII Report would limit public access to works in digital form, it is necessary to examine not only the proposed digital transmission right, but also the kindred proposals to abolish the "first sale" rule for works transmitted digitally and to ban devices aimed at defeating copy-protection schemes. Especially given the Report's highly constrictive view of the fair use doctrine, adoption of these three recommendations would dramatically change the historical balance of copyright law as between the interests of copyright owners and of the public. ABOLISHING THE FIRST SALE RULE?
The "first sale" rule allows members of the public who have purchased a copy of a copyrighted work to sell it, give it away, lend it, or even rent the copy to other people. (In the United States, only sound recordings and software cannot be rented; in some countries, no works can be rented without permission from the copyright owner.) The first sale rule grew out of judicial decisions holding that Congress had not granted copyright owners monopoly power over all distributions of their works, but only a right to control the first sale of the work to the public. The first sale rule promotes public access to copyrighted works by allowing members of the public to borrow works from one another (and from libraries) without fear of infringement. It is this rule that the NII Report proposes to abolish for works distributed by digital transmission.
The rationale for abolition of the first sale rule focuses attention on a difference between printed and digital works. The first sale rule presumes that when the owner of a physical copy of a work shares that copy with another person, he or she will give up possession of that copy. Although one copy may move from person to person, such a transmission does not result in more copies being made. With digital transmissions, however, someone who shares his or her copy of a work with another person may retain a copy of it as well. A digital transmission may result in a multiplication of copies. This poses a threat to the economic rights that copyright law gives to authors (and through them, to publishers).
Abolition of the first sale rule may, however, be unnecessary to respond to this threat. A narrower approach would be to limit the application of the first sale rule to situations in which the digital transmitter did not delete his or her copy. (I don't know about the rest of you, but I routinely forward information I receive by email to people who would be interested in it, following which I delete the information. In truth, I delete this information less because I am concerned about abiding by copyright law than because I can only manage so much information at a time. Even if I retain a copy, I consider most of the information I forward to another person to be fair use because of its private, noncommercial character.)
However, even without an abolition of the first sale rule, copyright owners can control this kind of potential consumer abuse of copyrighted works by means of the exclusive reproduction right. If the owner of a copy of digitally transmitted work begins transmitting copies of that copy to a thousand of his or her closest friends, that person will be responsible for multiple reproductions of copyrighted works. Since the first sale rule only limits the distribution right of copyright, not the reproduction right, there is way to deal with the multiplication of copies under existing law. (Just because you own a copy of a book, you do not think you are entitled to make a thousand copies of it for your friends. But you can share your copy with others.)
The NII Report does consider either alternative discussed here, but rather recommends abolition of the first sale rule. It does not provide persuasive reasons why the public should not be entitled to continue to enjoy the right to share their copy of a copyrighted work with a friend, regardless of whether it was received by digital transmission or otherwise.
It would be inaccurate to say that the NII Report recommends abolishing fair use law. And yet, it takes such a narrow view of existing fair use law and predicts such a dim future for fair use law when works are distributed via the NII that the Report might as well recommend its abolition. Since the fair use doctrine has been one of the historically important ways in which the law has promoted public access to copyrighted works, the virtual abolition of fair use law for which the Report argues would represent another vast expansion of copyright law in favor of publishers.
As with its treatment of the browsing issue, the Report attempts to constrict user rights by acting as though this constriction has already occurred, rather than by admitting that the Report is coming down on one side of, at best, a debatable issue. Without even admitting that any controversy exists about fair use law, the Report purports to resolve definitely one of the pressing controversies of U.S. copyright law today: whether private, noncommercial copying of copyrighted works is noninfringing under fair use law or otherwise.
On this issue, the public and the publishers could hardly have more different ideas. (On this issue, as on most of the rest of the copyright issues discussed in this column, I believe that authors are generally closer to the general public's view because so many of us rely on private noncommercial copying in the course of our research.) The public generally thinks that private noncommercial copying of copyrighted works is not, and should not be, copyright infringement. Publishers, however, regard all reproductions of copyrighted works as infringing. Publishers argue that private noncommercial copying cannot be justified as fair use because it provides a consumer with the benefit of a copy for which the consumer has not paid and usurps a sale that the publisher should have made if the consumer wanted a copy of the work.
The NII Report comes firmly down on the publishers' side in this controversy and fails to mention that the Supreme Court's *Sony Betamax* decision told courts to *presume* that private noncommercial copying is fair use. Only if there is some meaningful likelihood of economic harm to the copyright owner arising from the use should the presumption of fair use be overcome. (The only fair use issue for which the Report cites the *Sony* case is for its statement that commercial uses of copyrighted works should be presumed unfair. Interestingly, the Report neglects to mention that this second *Sony* presumption was repudiated by the Supreme Court this spring in *Campbell v. Acuff-Rose* in which 2Live Crew claimed fair use for the groups' rap parody of "Pretty Woman.") The Report also neglects to mention other sources and precedents that would support the Supreme Court's view that private noncommercial copying should be presumed to be fair use.
Another major fair use controversy concerns the extent to which it is fair to copy portions of copyrighted works for research or educational purposes. As with the private noncommercial copying issue, the Report cites cases that favor the publisher position on this issue without mentioning cases that do not favor the publisher position. For example, the Report mentions the *Basic Books v. Kinko* case in which publishers successfully sued a copying center for making and selling multiple copies of coursepacks to students without being sure that the professors submitting the coursepacks had gotten permission from copyright owners to make them. However, the Report fails to mention the *Williams & Wilkins* case in which a research library persuaded an appellate court that it had made fair use of articles from medical research journals when copying them for research scientists doing work in that field.
As with the private noncommercial copying issue, the Report does not acknowledge the existence of genuine and principled differences of opinion on this issue. It simply acts as though the rule already is what the publishers want it to be. Although the Report says that the Working Group will hold a set of workshops to discuss educational fair use issues, it does not admit any educational use to be fair except if it meets a set of guidelines adopted some years ago that allow teachers to make photocopies of short articles pertinent to their classes that are published during the school term.
The Report also predicts that fair use defenses will be unsuccessful when controversies arise in digital networked environments because it will be so much easier for consumers in these environments to license additional uses if they think they need them. The Report fails to mention two recent appellate decisions that prefigure a broader potential for fair use defenses in dealing with digital data and new technology issues: *Galoob v. Nintendo* in which a fair use defense was successful because kids using Galoob's Game Genie had already tithed to Nintendo by buying its games, and *Sega v. Accolade* in which an appellate court ruled that a competitor's disassembly of a Sega game in order to determine how to make its game cartridges compatible with the Sega machine was fair use. By not acknowledging the existence of these cases, the Report underestimates the potential for fair use to remain a viable defense in disputes erupting in digital networked environments.
To remedy this problem, the Report recommends enactment of the following provision: "No person shall import, manufacture, or distribute any device, product, or component incorporated into a device or product, or offer any service, the primary purpose or effect of which is to avoid, bypass, remove, deactivate, or otherwise circumvent, without authority of the copyright owner or the law, any process, treatment, mechanism or system which prevents or inhibits the exercise of any of the exclusive rights [of copyright]."
The Report further recommends making manufacture or sale of such devices or services into an act of copyright infringement. It also recommends that any copyright owner whose works *could* be infringed by such a device should be able to sue the maker or seller of such a device or service for copyright infringement, regardless of whether anyone had ever used the device or service to infringe that owner's copyrights. (Sellers of the technological device being circumvented would not, however, be able to sue those who unlocked the device for copyright infringement.)
The Report admits that these recommendations would overturn Supreme Court caselaw under which it does not infringe copyright to distribute a technology that can be used to infringe as long as the device is *capable* of substantial noninfringing uses (i.e., because videotape machines could be used for noninfringing purposes, the Supreme Court decided that Sony was not liable for copyright infringement despite the fact that some consumers might use Betamax machines to infringe copyrights in Universal or Disney movies.)
The Report is not clear about whether adoption of its recommendations would overturn the *Vault v. Quaid* case in which an appellate court ruled that sale of a program to "unlock" the copy- protection program sold by the plaintiff was not copyright infringement because copyright law gave owners of copies of copyrighted software rights to make backup copies of their software. Since the unlocking software gave owners of copies of software an opportunity to exercise their rights to make backup copies, the court thought that the sale of this software promoted copyright policy, not undermined it.
The drafters of the NII Report would probably say that their recommendation would not undo this case because that lawsuit was brought by the maker of the locking software, not by software publishers who had made use of the locking software. The ban the Report recommends would give rights to sue only to software publishers. Yet the issue of whether selling a product or service that would undo a technological lock on a copyrighted work so that a user could exercise fair use or backup copying rights is not addressed by the report. Given the publisher bias that pervades the Report, it seems likely that the drafters intend to restrict user access in this respect, although they do not say so directly.
Nor does the Report address the question as to whether distribution of programs in object code form should be regarded as a technological means for protecting software, such that tools or services that would be useful in disassembling or decompiling object code would be within the scope of the ban. For those who are concerned about the future of interoperability, it should be of especial concern that the Report does not mention the caselaw favoring fair use to achieve interoperability and speaks only in vague terms about the value of interoperability.
The NII Report acknowledges that its recommended ban on technological "keys" may restrict public access to both copyrighted and uncopyrighted works (the latter are as likely as the former to be distributed in encrypted form on the net). Although the Report expresses some regret that such restrictions may occur, it concludes that, on balance, such "incidental" restrictions on public access are necessary and that the public interest in access is outweighed by the countervailing need to protect the interests of copyright owners. The Report hopes that the "primary purpose and effect" language of the ban will provide a proper balancing of interests. This, of course, depends on the willingness of information providers to encrypt uncopyrighted materials with a different encryption algorithm than they use to encrypt copyrighted works. If the same encryption scheme is used for both, any unlocking technology can be kept off the market until a court rules that the primary purpose or effect of the technology would not be to promote copyright infringement.
While I might be able to support a more narrowly drawn provision aimed at dealing with the problem of technological circumventions of technological strategies for protecting copyrighted works in digital networked environments, I cannot support the proposed provision. As WIRED magazine recently pointed out, the proposed ban is so broad, publishers could probably use it to ban sales of photocopy machines. And they wouldn't even have to prove that *any* of their copyrights had been infringed; it would be enough that the machine *could* infringe their copyrights.
The growth of the Internet has been one of the phenomenal success stories of our time. People have flocked to the net by the hundreds of thousands not because their favorite movies or books may be available there in another five to ten years, but because a wide variety of resources are available there already. Since its inception, the Internet has greatly facilitated and enhanced communication and learning of the very sort that copyright law is supposed to promote. It has enabled researchers to gather and share data more easily, to engage in collaborative work at remote locations, to criticize and refine one another's work, and to make research results and the like available at ftp sites, thereby enabling those interested in these results access to them. A large number of newsletters, journals, and listservs have sprung up and serve as forums for discussion of public policy and research issues in a wide variety of fields. Debate on the Internet could hardly be more robust.
Notwithstanding the occasional "pirate" bulletin board on which commercially distributed software is posted for unauthorized copying and the pronouncements of some who would abolish copyright law, the Internet has promoted public access to information far more than it has promoted copyright infringements. I believe that the vast majority of net users are law-abiding citizens who generally make no more than fair and reasonable uses of copyrighted works.
The NII Report does not recognize that there are already both formal and informal ways in which denizens of cyberspace are influencing one another about copyright concerns and the ethics of making certain kinds of uses of other people's work. Policies that actively discourage copyright infringement are one means by which bbs operators have an influence on the practices of those who use their systems. Violation of bbs policy may result in being kicked off the system, a punishment more feared by many users than being sued for copyright infringement. But if this is an effective sanction, this should be appreciated by drafters of an NII Report on intellectual property issues.
Informal exchanges about copyright issues also occur in electronic newsletters, listservs, and on bbs's on the net. If one person makes an unauthorized use of another's writing, a third person may well question the fairness of this conduct and start a dialogue on the issue. The result of this dialogue is discouragement of unfair postings. "Netiquette" limits the extent to which users of the net appropriate other people's work. It simply isn't fair to repost someone else's message on another bbs or insert it into a newsletter without asking that person's permission. However, merely forwarding the message to one or a small number of people who would find it especially interesting is regarded as fair conduct, just as a telephonic exchange of the same information or photocopying a short article from a newspaper or magazine to mail to one's colleagues would be.
The NII Draft Report should acknowledge and build upon the strengths of existing digital networked environments. Its policy recommendations should permit exchanges that promote the learning function of copyright law without having harmful effects on the economic interests of copyright owners. Before recommending dramatic changes to copyright law that would favor those who want to use the NII, the drafters of the Report should consider what effect those policies will have on existing user communities. It should seek to adopt solutions that would improve the lot of those who want to enter the net without harming the lot of the millions of people who now use the net. (Economists speak of this as the search for "Pareto optimal" solutions.) We can only hope that this omission will be cured in the Final Report of the NII Working Group on Intellectual Property Rights.
As the NII Report observes, owners of very valuable copyrights, such as motion picture producers, recording studios, and publishers of books, are unlikely to want to distribute their works via the NII unless they have reasonable assurance that their intellectual property rights will be respected. One can commend the drafters of this Report for tackling a very difficult problem and for offering recommendations that would overcome some of the fears that owners of valuable copyrights have about digital networked environments without approving of the strategy employed to achieve the Report's objectives and without concurring in its judgment about where a proper balance lies between the interests of copyright owners and the public.
I remain unpersuaded that copyright owners really need the dramatic expansion of rights which the NII Report would give them. I believe this proposal would restrict public access to information far out of proportion to the harm likely to result to copyright owners, and that existing law provides plenty of ammunition with which publishers can attack infringers. But I admit the issue of what is proper copyright policy in the coming age of digital networked environments is a subject on which reasonable people can disagree. If the Report had been explicit about attempting to achieve a radical transformation of copyright law so that each and every use of a copyrighted work is infringing unless authorized by copyright owners, then at least there could have been public debate on the issues.
The most objectionable aspect of the NII Report is, in my view, lies in its effort to avert the hard issues and controversy that a plain statement of its intentions would engender. It is simply not true that the Report recommends only minor clarifications and changes to copyright law, even though the press coverage of the Report dutifully echoed the Report's statements that they were. (Where are the investigative reporters when we really need them?) This column aims to provide readers with enough information about the policy issues raised by this Report so that they can begin the policy debate that is so sorely needed in this area and so that they can contribute their views about a solution that will achieve a fair balance between the public interest and the interest of copyright owners.
 Pamela Samuelson, Legally Speaking: Copyright's Fair Use Doctrine and Digital Data, Comm. ACM 37: 21 (Jan. 1994).
|Virtual School||Middle of Nowhere||Brad Cox|