Public Access to the Intellectual, Cultural, and Social Record
For the past two centuries, copyright law has promoted broad public access to a vast array of informational works. As James Madison once observed, copyright has been characterized by a happy coincidence of public and private interests. The private interests of authors and publishers in creating and selling copies of works have coincided with the interests of members of the public eager to get access to the ideas, information, or expression in the works.1
Public access has not, however, been achieved solely by virtue of the incentives arising from the grant of rights to authors and the transfer of some of these rights to publishers; a number of other mechanisms have been at work as well. A variety of public policies promote access, including fair use, the first-sale rule, and the copyright registration requirement of depositing a copy in the Library of Congress; some of these accomplish their ends by placing limitations on the rights of authors and publishers.2 The emergence of organizations such as public libraries, universities, and cultural heritage institutions has also added significantly to public access, making a wide variety of works available, by acquiring them and making
1ln Federalist 43, Madison wrote, "The public good fully coincides in both cases [patent and copyright] with the claims of individuals."
2Fair use, the first-sale rule, and the copyright registration requirement are discussed in the Chapter 4 section entitled "The Challenge of Private Use and Fair Use with Digital Information" and in this chapter's sections entitled "Public Access Is an Important Goal of Copyright" and "Archiving of Digital Information Presents Difficulties," respectively.
them available to the public for no direct payment by the user.3 Public access and use occur in a variety of ways, including purchase (new or used), borrowing (from libraries or friends), educational exposure, and reading of commentary that quotes a work. To date, this collection of mechanisms has worked well, providing protection and thus incentive to authors and rights holders while also ensuring wide public access to work through a variety of routes.4
Changes brought about by digital intellectual property (IP) and the information infrastructure are challenging the existing set of policies and practices for public access. This chapter focuses on the implications of those changes for public access, including both the optimistic and pessimistic projections of their possible consequences.
Public Access is an Important Goal of Copyright
U.S. courts and commentators have repeatedly emphasized the fundamentally utilitarian nature of copyright, noting that the Constitution provides for intellectual property protection with the pragmatic goal of promoting the public interest in access to knowledge and innovation. This intent is evident from the Constitution's grant to Congress of power "to promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries." In United States v. Paramount Pictures, Inc., the Supreme Court's decision considered the purpose of this clause:5
The copyright law, like the patent statutes, makes reward to the owner a secondary consideration. In Fox Film Corp. v. Doyal, 286 U.S. 123, 127, Chief Justice Hughes spoke as follows respecting the copyright monopoly granted by Congress, "The sole interest of the United States and the primary object in conferring the monopoly lie in the general benefits derived by the public from the labors of authors." It is said that reward to the author or artist serves to induce release to the public of the products of his creative genius.
Further comment on the constitutional concern with access to infor-
3Of course, costs are involved in operating these institutions, which are funded by government, universities, and other organizations.
4Because the term "public access" has been used historically in a variety of ways, it is useful here to note one way in which the committee is not using the term. It does not include in public access any access to specific copies of a work, especially any unique copies of a work (e.g., the originals of artwork). The public has access to versions of a work that have been published and distributed, placed in publicly accessible collections (e.g., libraries), or otherwise made available through normal channels.
5334 U.S. 131, 158 (1948).
mation is found in the Court's decision in Sony Corp. v. Universal City Studios, Inc.:6
As the text of the Constitution makes plain, it is Congress that has been assigned the task of defining the scope of the limited monopoly that should be granted to authors or to inventors in order to give the public appropriate access to their work product. Because this task involves a difficult balance between the interests of authors and inventors in the control and exploitation in their writings and discoveries on the one hand, and society's competing interest in the free flow of ideas, information and commerce on the other hand, our patent and copyright statutes have been amended repeatedly.
U.S. copyright law has promoted public access to ideas, information, and works of authorship in a number of ways. The grant of exclusive rights to authors promotes public access because it presumesusually correctlythat authors want to make their works widely available to the public and will, in fact, do so as long as they have the right to stop unauthorized exploitations.
Numerous other means to promote public access have also been used. For example, U.S. copyright law, like the copyright law of many other nations, promotes public access by the "first-sale" rule, a limitation on the right of rights holders to control copies of their works that have been distributed to the public. This rule provides that the first sale of a copy of a work to a member of the public "exhausts" the rights holder's ability to control further distribution of that copy. A library is thus free to lend, or even rent or sell, its copies of books to its patrons. Libraries are, of course, not the only institutions that are reliant on the first-sale rule for their operation. Bookstores, art galleries, and auction houses also depend on it, as does the practice of sharing copies of books or magazines with friends or of giving purchased books to friends.
The Congress recognizes that the preservation of historythe record of social and cultural discourse, scholarship, and scientific debate and discoveryis of fundamental importance to society in innumerable ways. Both while these items are still protected as intellectual property and later when they are part of the public domain, they form an essential base upon which new artistic and scholarly work is developed. As a result, Congress has adopted a number of other rules enabling libraries and archives to promote public access to informational works. These rules have, for example, permitted the making of some copies for specified purposes, such as to replace pages from a damaged book or to preserve materials that are deteriorating, to ensure that future access to these items will be
6464 U.S. 417 (1984).
available. From time to time, stakeholders have negotiated guidelines to establish standards for such activities (e.g., the Guidelines for Educational Uses of Music (U.S. Copyright Office, 1998)). Congress has also passed rules limiting copyright in order to promote public access to copyrighted materials in classroom settings, including a privilege to perform or display such materials in the course of face-to-face teaching in the classroom.
Nonprofit institutions, such as libraries and schools, are not the only institutions that have public-access-promoting privileges. Congress has occasionally used compulsory licenses to promote new recordings of copyrighted music, jukeboxes of sound recordings, and cable distribution of broadcast network programming, all of which have public access implications.
Fair use rules also promote public access.7 The most common form of fair use is the ability of an author to quote from previous works (thereby copying a small part of them) in order to comment on them or report news about them. The Supreme Court's decision in the Sony case cited public access concerns as favoring a fair use ruling about making copies of television programs for time-shifting purposes and about the impact of the ban on videotape recorders that certain motion picture studios sought.8
Finally, public access is also promoted by the limited duration of monopoly rights conferred under copyright law.9 Every copyrighted work eventually becomes part of the public domain, making the work even more widely accessible to the general public by virtue of its royalty-free public domain status (e.g., books by Mark Twain, operas by Giuseppe Verdi). Unfortunately, public domain status can have an opposite effect as well. A public domain work may go (or stay) out of print if no one is willing to invest in preserving the work without the ability to assert exclusive rights as a way of recouping the investment. Contractual arrangements can attempt, as a matter of contract between two parties, to create protection beyond the date of copyright expiration.10
7See Chapter 4 for an extended discussion of fair use in the digital age.
8The reading of the Sony case is not uniform among committee members. Some members view Sony's fair use analysis as applying to private home taping of programs from cable and other forms of pay television as well as from free broadcast television, whereas other committee members view Sony more narrowly, as applying to free broadcast television only, which is what was directly involved in this case.
9Although the Constitution restricts the duration of copyright, a specific term is not specified. The last two major amendments, the Copyright Act of 1976 and the Copyright Term Extension Act of 1998, extended the term of copyright. The constitutionality of the 1998 act is challenged in Eldred and Eldritch Press v. Reno, filed in the U.S. District Court for the District of Columbia on January 11, 1999.
10The proper duration of copyright protection has always been and remains a highly
(footnote continued on next page)
Access: Licensing Offers Both Promise and Peril
Licensing has been widely used in the software world for some time.11 Software developers routinely distribute their products under licenses, rather than by the sale of copies. For example, the ubiquitous shrink-wrap license on a box of mass marketed software is just that, a licenseone that typically recites as one of its first points that the consumer is purchasing a right to use the software and that the software remains the property of the manufacturer.12 A variety of other licensing schemes exist, including site licenses, in which an organization purchases the right for all its members at a particular location to use the software. Licenses may also provide a natural mechanism for use-metered payment for software (i.e., paying by how much you use the program, rather than buying a copy for unlimited use). More generally, licenses for software offer the potential to custom tailor the agreement in ways that are less easily done with outright sale. Within the entertainment industry, licensing is often used (e.g., in the exhibition of movies in a theater). Access to online databases also occurs frequently under a licensing arrangement.
Licensing is a newer phenomenon for some other digital information productsparticularly for those traditionally delivered in physical form. An increasing amount of the information acquired by libraries, for example, is in digital form, and unlike print materials, which have historically been available on a sale-of-copy basis, digital materials are frequently available only through licenses. Licensing can have advantages: The
(footnote continued from previous page)
controversial issue. For the purposes of this report, the question is whether the proliferation of digital IP, networks, and the Web materially change the long-standing debate. The committee believes that the first-order answer is no, because the underlying issues concerning the duration of copyright in the digital world are fundamentally the same as they were in the predigital world.
11For an overview of the licensing of electronic content, see Okerson, Ann, "Buy or Lease: Two Models for Scholarly Information at the End of an Era," Daedalus, 125(4):55–76. Of course, some forms of software (e.g., video games) are often sold rather than licensed. For a discussion of the licensing of electronic content and the implications for libraries, see "Accessibility and Integrity of Networked Information Collections" (Office of Technology Assessment, 1993).
12The law is unsettled on issues concerning the enforceability of shrink-wrap licenses. See, for example, Step-Saver Data Systems v. Wyse Technology, 939 F.2d 91 (3d Cir. 1991) (holding that a shrink-wrap license delivered in a package after a telephone contract was not part of the contract); Arizona Retail Sys., Inc. v. Software Link, Inc., 831 F. Supp. 759 (D. Ariz. 1993) (holding that a shrink-wrap license was enforceable where there had been no prior agreement but was not enforceable where there was a prior telephonic agreement); and ProCD, Inc. v. Zeidenberg (see Footnote 20). See also Vault Corp. v. Quaid Software Ltd., 847 F.2d 255 (5th Cir. 1988) (a state statute authorizing shrink-wrap licenses was preempted by federal law; the lower court had held that the shrink-wrap license would be an invalid contract of adhesion (a standard contract form offered to consumers without affording the consumer a realistic opportunity to bargain) were it not for the statute).
license may provide clarity on terms and conditions of access (e.g., who is included in the authorized user community and how the material may be used);13 it may provide for an increase in the rights for the institution that go beyond those provided under copyright (e.g., the ability to make unlimited copies for local use); and it may limit the organization's liability in the case of misuse by patrons. Considering the degree of financial, ethical, and legal exposure, establishing a written agreement is increasingly endorsed by both librarians and publishers.
Licensing, coupled with a rights management and payment infrastructure that is efficient and easy to use, can facilitate access that has been impractical in the past. It does this by providing consumers with a straightforward way to access large collections of information items, often when the individual items are of low value.14 In the absence of licensing arrangements, publishers likely would not make some of these collections available at all. Such arrangements can also allow consumers to access information at a more specific level; perhaps the day will come when instead of purchasing a book, one can obtain access to the executive summary for a limited period of time (presumably at a much lower price than the purchase of the book).15 Thus, licensing can increase the options for making information available.
However, there are also concerns about licensing as a model of information dissemination, particularly the impact it may have on public access. The trend toward licensing means that (digital) information is in some ways becoming a service rather than a product.16 Buy a book and you own it forever; pay for a service and when the period of service is over, you (typically) retain nothing. The increased use of licensing seems to diminish greatly the public access accorded through the first-sale rule. Consider libraries as an archetypal example. In the print world, a library's failure to renew a subscription or buy an updated version of a book has no effect on the availability to patrons of earlier volumes or editions. In
13A book in a traditional library has a limited user community by virtue of the effort needed to access it (geographic limits) and the inherent restriction on the number of patrons who can read it at once. As noted above, neither of these natural limits exists in the digital world, hence the utility of an agreement on who is to be included in the library's community of users of its digital works.
14For example, consumers could be able to view a digitized photograph for the payment of a license fee (which might be very smallperhaps much less than one dollar). Physical delivery mechanisms and standard payment methods (checks/credit cards and billing departments) will cause such transactions to be cost prohibitive. Licensing with an efficient rights management and payment infrastructure can cause such transactions to become economically viable.
15See CSTB (1998) for a discussion on pricing models for digital information.
16This trend toward licensing is a part of a larger shift from a goods-based to a service-based economy that is enabled in part by information technology (CSTB, 1994).
the world of licensed information, ending a subscription to an electronic journal may mean the end of access to earlier volumes or editions, as well.
A second issue arises from the nature of licenses as contracts. Contracts might not incorporate and indeed may attempt to override the public policy considerations that have been carefully crafted into copyright law. Those who contract for information may find that their access is far more restrictive than what they were accustomed to for print materials, unless fair use and other such considerations are explicitly a part of the agreement.
Some institutions (especially libraries) have worked to negotiate licenses that preserve fair use and other public access features. Publishers are currently experimenting with licensing models to respond to these concerns. Yet the concern remains about the use of a mechanism such as licensing that lacks any of the built-in protections for public access that are embodied in copyright law.
Mass marketed information products raise a more general concern about the proliferation of licenses. Where some institutions are by necessity becoming more sophisticated negotiators, the situation is far less clear for the typical consumer. Do consumers face the prospect of having to scan the tiny print of licenses to see whether fair use and other public policy considerations have been incorporated? Will they even know what such things are? Will contracts simply override some of those considerations? Copyright law currently gives owners of copies of computer software the privilege to make back-up copies; can that privilege be taken away by a shrink-wrap license in the software package that says loading the software onto the computer means one has agreed to the license terms that expressly prohibit making a backup? Does merely designating a transaction a "license" and stipulating that use of the product constitutes the user's agreement to the terms convert a transaction that would otherwise seem to be a sale into a license? If a license term prohibits making disparaging remarks about the licensed information or disclosing flaws to other potential users, should that term be enforced?17 These and related questions have been the subject of heated debate in the past few years and
17A restriction in a negotiated contract prohibiting a licensee from disclosing flaws in the licensed information may, as a matter of law, sometimes be enforceable (even if as a practical matter it might be difficult to monitor and enforce). For example, if as a matter of private contract under a software license, a licensee agreed not to make disparaging remarks or not to reveal publicly any flaws in the software, that licensee's publication of a book or a magazine article criticizing the software and revealing its flaws may constitute a breach of contract and under the contract might entitle the licensor to seek money damages or even to seek equitable relief to enjoin the further publication of the disparaging or flaw-revealing information. There is, however, very little precedent for awarding damages, let alone injunctive relief, in such cases except in special circumstances.
(footnote continued on next page)
are far from resolved.18 In the print world, efforts to impose "license" restrictions on mass marketed copies of copyrighted works were generally treated as ineffective.19 New license models that entirely avoid the foregoing problems could be part of these new paradigms for public access. Some recent case law and proposed legislation would enforce mass market licenses in the case of digital information.20 Nevertheless, the issue remains unsettled.
The mass market issue raises an additional concern if licensing becomes the dominant model of distribution for works that are considered part of our intellectual and social heritage. One could imagine a world in which novels, poems, and plays, for example, are available only (or mostly) by license rather than sale. The consequences of such a world for
(footnote continued from previous page)
The foregoing discussion, regarding the legal enforceability of a contractual waiver of fair use rights by a licensee, supposes a negotiated license agreement at arm's length between sophisticated parties. The use of such prohibitions in mass-market form license agreements, particularly as they may be embodied in shrink-wrap licenses or "point-and-click" licenses on the Internet, may present different issues of enforceability under established doctrines of law regarding contracts of adhesion and unconscionability. Neither the committee report nor the responses indicated above express a view pro or con on whether such waivers would be enforceable in the context of a mass market license.
18A model information licensing law (once known as Article 2B of the Uniform Commercial Code and currently known as the Uniform Computer Information Transactions Act (UCITA)), proposes (among other things) to validate mass market licenses of computer information. It regards licenses or license terms as enforceable unless they were unconscionable at the time they were made. In November 1998, this model law was amended to provide expressly for public policy limitations on the enforceability of license terms in mass market or other licensing contexts. Section 105(b) of this model law provides a basis upon which to make an argument that, for example, the public policy favoring a particular user's interest (e.g., in making a back-up copy) is strong enough to override a mass market license provision purporting to deprive the user of that interest. However, section 105(b) will likely require case-by-case determinations about whether the public policy interest asserted should override contractual provisions limiting it. In July 1999, the National Conference of Commissioners on Uniform State Laws (NCCUSL) passed UCITA. State legislatures must approve UCITA before it becomes functional as uniform state law.
19License restrictions on the resale of books were found to be unenforceable in Bobbs-Merrill Co. v. Straus, 210 U.S. 339 (1908).
20See for example, ProCD v. Zeidenberg, 908 F. Supp. 640 (W.D. Wis.), rev'd., 86 F.3d 1447 (7th Cir. 1996) (holding a shrink-wrap license surrounding the packaging for a CD-ROM to be an enforceable contract and rejecting a claim that the state law breach of contract claim should be preempted by federal copyright law). Mass market licenses of computer information (including software) would be validated, as a matter of contract law, if states choose to adopt the model licensing law presently known as UCITA, described in footnote 18 above. Section 105(a) of UCITA recognizes that federal law and policy may preclude enforcement of some contract terms that are at odds with federal policy. Section 105(b) recognizes that other fundamental public policies may preclude enforcement of certain contract terms.
public access are far from clear. This underscores the notion that intellectual property should not be viewed solely in terms of economic considerations.
Access and Technical Protection Services
Technical protection services (TPSs) are generally infeasible in the print environment. One can keep a manuscript under lock and key, but when the work has been published, further uses of it generally cannot be controlled by the rights holder. But then the need for control is not so great: Large-scale infringement of the reproduction right of a hard-copy work is generally too expensive or difficult for most members of the public, and most infringements significant enough to undermine the copyright owner's economic interests are public enough that they come to the owner's attention, directly or indirectly.
Technical protection in the digital world is far more practical, despite a somewhat checkered history of success.21 Recently, there has been great interest in these technologies, especially among publishers concerned about the vulnerability of digital works to inexpensive, rapid copying and distribution. Many rights holders have high hopes that these systems will provide them with control over unauthorized uses, thereby enabling worldwide markets for a variety of digital information products to emerge and thrive on the Internet. And, indeed, without effective TPSs, rights holders may well choose to restrict greatly the availability of their most valuable digital content.
As with licenses, however, there may also be consequences for public access, including issues such as fair use. TPSs can allow content owners to restrict access to and uses of their works in ways not possible in the print world. One method, relying on a combination of cryptography and special hardware and/or software, would make it possible for a vendor to specify that a consumer could read a document but not print it, save an unencrypted copy, or e-mail a part of the unencrypted text to a friend.22
21The first copyrighted works to employ technical protection services in a mass market were copy-protected software programs. These copy-protected programs proved to be sufficiently unpopular with consumers and sufficiently vulnerable to competition by vendors willing to sell competing unprotected products that the industry ceased using them. TPSs are successfully used in the cable and pay-per-view television businesses.
22The user could of course capture the protected content through low tech means, such as writing it down or recording it with a camera, then transferring the content to a computer readable format. No TPS is perfect; like any security system, it relies on the deterrence that comes from making some actions sufficiently difficult or expensive (try copying an entire database by hand).
Such a transaction would clearly remove much of the public access that has been an inevitable part of more familiar information access methods.
TPSs raise public policy issues that have been the subject of considerable debate. Because these systems rely in substantial ways on encryption, new tools and technologies for decryption become of special interest. Reflecting this interest, the Clinton Administration's white paper Intellectual Property and the National Information Infrastructure (IITF, 1995) argued for new rules to outlaw tools and technologies whose primary purpose was to bypass (or circumvent) technical protection measures used by rights holders to restrict access to or use of their works. The white paper argued that instituting rules against circumvention would have no effect on fair use or on the public domain.
Congress was persuaded otherwise. It heeded the concerns of major copyright industry groups about the dangers to their markets from acts of circumvention and from the availability of technologies to circumvent TPSs. However, it also recognized in the Digital Millennium Copyright Act of 1998 (DMCA) that granting copyright owners the right to control public access to protected works by outlawing the circumvention of technical protection measures was an unprecedented and significant extension of author and publisher rights, one that might have a negative impact on fair use.23 In addition to carving out several specific classes of circumvention activities that it found, on balance, to be socially useful, Congress provided for a 2-year moratorium on the ban on the circumvention of access controls, during which time the Librarian of Congress is supposed to determine whether the use of technical protection measures is having an adverse effect on the ability to make noninfringing uses of a particular class of copyrighted works. If the librarian finds such an adverse effect, he can identify that class of works or users so that the ban on access control circumvention wilt not apply to that class of works or users for the next 3 years. The report of the Librarian of Congress on the impact of circumvention legislation, which will be published in the year 2000, is likely to be of considerable importance in addressing public access concerns raised by these systems.24
In summary, some people are optimistic about the prospects of technical protection services to enable far greater public access to copyrighted works than has ever been possible previously, and they view the
23Even if the circumvention of technical protection measures is deemed ultimately to be legal when done in pursuit of access and fair use, such access is then accorded only to those with the technical skill to circumvent effectively or the luck to be interested in content whose protection has already been circumvented.
24This report will presumably be written by the staff of the U.S. Copyright Office, a major service unit of the Library of Congress.
anticircumvention regulations as a necessary protection against those who would otherwise bring serious harm to the copyright industries. In contrast, there are those who view TPSs as deeply threatening to public access and to other societal values reflected in copyright law and policy, including promoting freedom of expression, innovation, and competition. This group believes that this threat could be realized as a result of steps toward other desirable goals, such as electronic commerce, yet may have serious consequences nonetheless (this scenario is discussed in Chapter 5). These two views represent the extremes of the spectrum. There is a third group of people who see the likely impact of TPSs to be mixed. This group believes that the implementation of some TPSs will facilitate public access, whereas other uses of TPSs will reduce public access. The perspective of this third group is bolstered by the possibility (see Chapter 5) that TPSs will not be as successful, either technologically or in the marketplace, as their promoters hope.
The committee does believe it naive, however, to think that TPSs, if widely deployed, would have no impact on fair use and, as a consequence, believes that policy makers should continue to monitor that impact and be prepared to act if necessary. How effective TPSs will be and how much acceptance they will achieve with consumers of information remains to be seen. Policy intervention may prove unnecessary, if the market adjusts to provide broad access to consumers while still offering appropriate levels of technical protection.
The New Information Environment Challenges Some Access Rules
The discussion of the exceptions in copyright law for public access is important because there is reason to believe that the change to digital distribution could make a number of those exceptions less applicable and less effective. For example, consider the situation of libraries or other institutions that acquire digital copies of protected works, if the first-sale rule of copyright law no longer provides them with an established legal basis for their lending practices. The white paper expressed the view that the first-sale rule does not apply to digital materials because first-sale limits only the distribution right. Lending digital materials to patrons would, however, inevitably involve the making of copies of the materials, copying that the white paper asserts is not privileged by the first-sale rule. Under this viewwhich some disputelibraries could not lend digital materials to patrons unless they had negotiated a license to do so.25 The
25Some have argued that a temporary copy made to transmit one's copy of a digital work to another person could be justified as a fair use, necessary to enable the first-sale right, as
(footnote continued on next page)
white paper also predicted a diminishing role for the fair use rule in the digital environment.
In addition, such established practices as interlibrary loans, document delivery services by libraries, and copying for course reserve purposes, particularly if in electronic form, are undergoing serious challenges. Publishers assert that circumstances have changed since guidelines permitting these activities were negotiated, making them unsatisfactory. Putting a paper copy on reserve is one thing, making a (widely accessible and easily copied) version available online is, they argue, something else again.
The New Information Environment Blurs the Distinction between Public and Private
Public access to informational works has been promoted by the existence of a relatively clear distinction between works that were published (in the legal sense embodied in the copyright law) and those that were not. This distinction is often critical in determining whether the information in the work (e.g., the ideas and facts in it) is public and can be reused and redistributed without permission of the copyright owner.26 Public access to facts and ideas in published works has been part of the copyright normative structure. As the U.S. Supreme Court said in its venerable Baker v. Selden decision:
The very object of publishing a book on science or the useful arts is to communicate to the world the useful knowledge that it contains. But this object would be frustrated if the knowledge could not be used without incurring piracy of the book. And where the art it teaches cannot be used without employing the methods and diagrams used to illustrate the book, or such as are similar to them, such methods and diagrams are considered as necessary incidents to the art, and given therewith to the public.27
In another famous U.S. Supreme Court decision, Feist Publications v. Rural Telephone Service, the court explained that copyright could not be used to protect facts, because appropriation of them from published works
(footnote continued from previous page)
long as one deleted one's copy after the transmission. Making this argument for the ''lend-in" of digital copies by libraries is, of course, more complicated.
26An author's particular expression is protected under copyright law, but the facts and ideas in the work are not. Whether published or unpublished, those facts and ideas are in the public domain and may be freely reused under copyright law. See for, example, Salinger v. Random House, 811 F.2d 90 (2d Cir. 1987).
27101, U.S. 99 (1879).
was an important means by which the constitutional purpose of copyrightpromoting knowledgecould be achieved.28
Of course, the choice about whether to publish remains with the author. Some may prefer to commercialize their facts and ideas without publishing them (e.g., by licensing them). Public access is not such an important value in the copyright system that the law imposes on authors a requirement that they must publish their works. Nor is it the case that the distinction between "published" and "unpublished" works has always shimmered with clarity. Nevertheless, the accumulated case law aids in drawing this important boundary line. And the general rule applied: The predominant way that an author could exploit a work commercially was to publish it, and, once published, both the law of copyright and the institutional infrastructure for usage of works meant that public access to information in the work could generally be had. In the print world, the distinctions between public and private, and published and unpublished, have been generally clear. Printed information is typically either one or the other, and where it is public and published, public access expectations arise.
Problems arise in the digital environment because it is not always easy to tell when information has been published and when it has not. The extreme cases may be clear: Placing a copy on a publicly accessible Web page would almost certainly be viewed as a publication,29 and the information on the Web page would be considered "public"30 Moreover, this kind of posting is considerably more "public" than a print world act of distributing the same text in a flyer or posting it on a public bulletin board. If instead one loads the only copy of the same text on a password-protected computer system, it is clearly a private and unpublished copy, even if a small number of others are authorized to access it.
But numerous cases also exist in the digital world in which the line between public and private, published and unpublished, is blurred. What if, for example, someone distributes information via a list server (software that distributes information to a specific list of e-mail addressees); is that publication or not? Does it matter how large the list of addressees is or whether outsiders can join the list without permission? What if a user has
28499 U.S. 340 (1991).
29Publication on the Web raises other interesting questions, such as, Where has publication occurred? This is important because copyright laws vary among countries. Suppose a citizen of the United States publishes an article on an AOL-owned (a U.S. company) server based in Germany and that this server has a mirror site in Hong Kong. Which is the country (or countries) of publication? Could one claim that publication on the public Internet is simultaneous publication in all countries of the world?
30In this context, "public" is not equivalent to public domain.
to register with a Web site to get access to its contents and use a password; is that site public or private? What about posting information on a multi-firm intranet? What if a scientific organization establishes a system to post draft papers on a site that is intended to be accessible only by members of the appropriate scientific community, but they are from many different institutions; are the papers posted on that site published or unpublished?
A second difficulty arises from the detailed level of control computer systems provide over access to information, making it feasible to provide specific, multiple layers of conditional access. The result is information availability that comes in multiple shades of conditional gray. One class of user may have one level of rights for one set of purposes and another class a different set of rights to other information for other purposes. Whether and under what circumstances each of these situations constitutes publication has significant consequences for public access to the information in the work, but deciding whether the work has been published may be difficult.31
Thus, what appears to be a relatively simple distinction in the print worldpublished vs. privatebecomes complex in the digital world, where the boundaries between public and private are blurred. The result is a challenge to our notion of how to make this distinction. The concern also exists that more informational works will be distributed in ways that will be labeled as private, keeping them inaccessible to the public and out of our cultural and intellectual heritage.
Noncopyrightable Databases Present Access Challenges
IP protection of databases is a special case, because in 1991 the U.S. Supreme Court, in Feist Publications Inc. v. Rural Telephone Service Co., restricted protection under copyright so that straightforward collections of facts presented in the obvious way (e.g., an exhaustive listing of phone numbers, presented alphabetically) would not be protected, rejecting the perspective that "sweat of the brow" was grounds for protection. This ruling left database producers with what they deem to be insufficient legal protection for their products,32 products in which they may have made substantial investments. Legal reforms have recently been pro-
31Also, by making it much easier to distribute portions of works in varying states of completion, the fluid nature of digital networked information further contributes to the blurring between published and private distribution.
32Most databases have a sufficient modicum of originality to be protected by copyright. In addition, contract and unfair competition law, as well as sui generis laws in some countries, provide legal protection to database producers.
posed, both internationally and nationally, to address IP protection of databases, but none has yet been enacted into law in the United States or agreed to by treaty or other international agreements with the United States.
There are many kinds of databases in terms of content, use and source, and the optimum IP policy may differ among them. Government databases, for example, fall into three categories: internal records, reports and background data for reports (e.g., census, budget), and databases created through grants and contracts to nongovernment organizations or individuals. In the nongovernment sectors, numerous types of databases exist, including internal records, records about others (e.g., customers, patients, suppliers), research results, compilations from diverse sources, and so on. The generators of the digital data and information that appear in electronic databases are governments at all levels, the not-for-profit and commercial sectors, and individuals. Some of these entities are both producers and users of databases (e.g., government agencies and researchers), whereas others are primarily producers (e.g., publishers) or primarily users (e.g., private citizens). For both the producers and users of databases, the same advances in information infrastructure that are affecting IP for copyright and patents also provide unprecedented capabilities for disseminating, accessing, copying, and manipulating databases.
In 1996 the European Union issued its Directive on Databases, which requires members to increase significantly the level of protection of databases and explicitly does not provide protection for databases produced in non-E.U. countries that fail to adopt measures affording comparable protection to the E.U. Directive in their laws.33 In 1997 the World Intellectual Property Organization (WIPO), at the European Union's request, initiated a draft worldwide database treaty largely incorporating the E.U. Directive. The treaty would have given database owners an exclusive right to prevent a temporary or permanent transfer of all or a substantial part of a database to another medium without the rights holder's permission, and it limited the right of treaty countries to provide (public good) exemptions that would interfere with the normal exploitation of the database or the legitimate interests of the rights holder. Because of protests by the scientific community, the United States backed away from signing the treaty after initially favoring it. This action and objections from developing countries halted the treaty negotiations. In 1999, WIPO is reconsidering database protection. Because databases are used globally and access via the Internet cannot easily be restricted to national users, continuing to pursue mutually acceptable international agreements that harmonize the IP protection of noncopyrightable databases is important.
33See Directive 96/9/E.C. of the European Parliament and of the Council of March 11, 1996, on the legal protection of databases, 39 O.J.L. 77/20, March 27, 1996.
The U.S. Congress introduced legislation in 1996, 1998, and 1999 patterned on the E.U. Directive, although nothing has yet been enacted into law, because of the conflicting views among the stakeholders.34 The alternatives for IP protection of databases that have been advocated include de minimus changes to existing law, unfair competition/ misappropriation models, sui generis models that are close to the E.U. Directive and provide strong property rights to the producers of databases, and the sort of technical protection mechanisms discussed in Chapter 5 of this report.35
The Information Infrastructure is Changing the Distribution of and Access to Federal Government Information
From the earliest days of our nation's history, Congress recognized its responsibility to inform the American public of the work of the federal government. One way it did this was by establishing the Federal Depository Library Program (FDLP) to provide no-fee, geographically dispersed access to government publications. By designating depository libraries in each state and congressional district, Congress ensured that government information from all three branches would be distributed throughout the country and available at no charge to the user. This system reflects a commitment to broad-based democracy and public accountabilityprinciples that are as important today as they have been in the past. The advent of the information infrastructure and the enormous opportunities it offers for citizens to access information has led to both the expected increase in access and, paradoxically, to some situations of sharply diminished access. It has also raised questions about the future role that the FDLP should play.
Online systems such as the Government Printing Office's Access system, the Library of Congress's THOMAS system, and the Security and
34Efforts to enact legislation to provide protection for databases that do not otherwise qualify for copyright are taking place in the 106th Congress through H.R. 354, the "Collections of Information Antipiracy Act," and H.R. 1858, the "Consumer and Investor Access to Information Act."
35A recently released NRC study, A Question of Balance: Rights and the Public Interest in Scientific and Technical Databases (NRC, 1999), addresses the scientific, technical, economic, legal, and policy issues regarding the creation, dissemination, and use of scientific and technical data for basic research and other public interest uses, with particular emphasis on the recent developments in IP law in noncopyrightable databases and what IP regime(s) could best accommodate them. An earlier NRC report, Bits of Power: Issues in Global Access to Scientific Data (NRC, 1997), provides background for the 1999 report. The recent study addresses database IP issues both in general and specifically as they relate to science and technology from the perspectives of the government, commercial, and not-for-profit sectors and considers the concerns raised by the different science and technology stakeholders.
Exchange Commission's EDGAR system have vastly enhanced public access by providing free, online access to government information. These systems (and many more) are available through the Internet at no extra cost to the user, illustrating how some federal agencies are succeeding in using information technology to enhance public access.
However, in some other agencies, trends are toward the decentralization, privatization, and commercialization of government information. As a consequence, some government information that used to be accessible to the public at no cost through the FDLP might not be accessible any longer (either through the FDLP or the public Internet), with the shift toward government information created or disseminated in digital form. Broad access and use of publicly funded information are substantially impaired when distribution agreements prevent or curtail distribution of information.36 For example, depository libraries can obtain the National Criminal Justice Reference Service CD-ROM for their collections but have to pay subscription costs for access to the Internet database that contains the actual reports.37
The response of government agencies has not been homogeneous because different agencies have different relationships and interests with respect to the various kinds of government information. Agencies vary in the extent to which they seek to protect the information that they generate in carrying out their task, with these differences arising from the nature of the responsibility of the agency and its relationship with the rest of the government, especially Congress. The bottom line is that some agencies are charged with the responsibility of disseminating data and new technical knowledge as widely and inexpensively as possible, while others are sometimes given the authorityand a strong financial incentiveto sell information to others. When information is sold, some agencies must pay high prices for data or products that were originally created by other agencies using public expenditures.
Some agencies are explicitly given the job of creating or maintaining important information products. One class of example is the data-based
36The federal government itself cannot own a copyright to works that it creates. However, private sector organizations can take works of the federal government, add value to them, and own the copyright to the resulting work (assuming that the new work satisfies the requirements for copyright).
37Another instance of reduced access occurs when agency CD-ROMs or Web sites are available to depository libraries, but their use is restricted to only one password that must be shared by many people in the congressional district. The Department of Commerce's STAT-USA is an example of an information service created by an agency that operates under a cost recovery mandate. Depository libraries are limited to one password to STAT-USA, a valuable database that contains thousands of titles that are no longer available in print. Depository libraries that need to provide better access to this information must pay for additional passwords.
agency (e.g., U.S. Bureau of the Census). For the most part, the information products are evaluated on the basis of the quality and utility of the information that they contain. These agencies are extremely interested in disseminating these products as widely as possible, as long as the integrity of their information can be assured. The information that is created from such responsibilities is frequently disseminated widely at little or no cost, and it is sometimes posted on well-managed Web sites that provide unprecedented access. The National Archives and Records Administration is an example of a federal agency chartered to maintain federal government records for historical purposes.
Many agencies are responsible for managing the records of government that reflect the fulfillment of their missions (e.g., the FBI or the courts). For the most part, these agencies either have no particular interest in whether the information they possess is easily accessible or prefer that their activities remain confidential unless they are forced to reveal information through litigation or Freedom of Information Act requests.38 Sometimes, these agencies allow private parties to copy their internal records for the purpose of packaging and reselling them, usually at high prices, including to other government agencies. Agencies that perform research and development, especially in information technology, can sometimes have some disincentives to be open with their results in order to have valuable knowledge that can be used to attract private partners.
State and local governments are also greatly affected by the shift to electronic distribution. In general, there is far less precedent for viewing information from state and local governments as being in the public domain, and there is a correspondingly wide variability in policies and practices among jurisdictions.
Archiving of Digital Information Presents Difficulties
Historically, most of the materials preserved in archives are in printed form (including printed transcriptions of spoken words) and in the physical arts (painting, sculpture, architecture, etc.). In the last century, the print record has been supplemented and enriched by sound recordings, photographs, moving images, and most recently various forms of digital
38An example of an important record of government is the contracting information from a federal agency. For the most part, contracts are part of the public record, but agencies often do not facilitate general access to them. In some cases, agencies have allowed others to collect and package contracting information. For example, the quasi governmental, nonprofit RAND Corporation is the sole source of information about defense contracts, which is available to scholars or others undertaking research on defense contracting for a fee in the thousands of dollars.
information, which may have been "born digital" (i.e., created without a physical embodiment, as for example a Web page).
Preservation of these materials has been carried out for society by libraries, archives, and museums and similar cultural heritage institutions, frequently with public support and funding. When published, materials are available for consultation by the public through these institutions even though the publisher may cease to make them available (for any number of reasons, including lack of interest or lack of commercial viability). After they pass into the public domain when copyright expires, materials can be reused or redistributed to make them widely available again.
The keys to making this system work have been the publication and distribution of physical artifacts and the doctrine of first sale. A library, for example, can purchase a book or a journal and then retain possession of that copy of the work. Because of the first-sale rule and other copyright provisions, the institution can preserve it indefinitely and make it available to members of the public. This ability to make the work accessible in turn supports the library, as patrons see the value in supporting the institution.
Overall, the system has worked fairly well for print, despite a number of difficulties. Libraries have had to grapple with the problem of holdings deteriorating because they were printed on acid paper, and they inevitably have to make funding-based decisions about what to preserve.39 Of course not everything is preserved: Not everything published is acquired by libraries, and not everything acquired is retained indefinitely. Nor are all important printed works published for use by the general public; archives play an important role in such cases, particularly in capturing the processes of government. Yet we have been able to retain the essential core of our printed cultural heritage and enough examples of other materials to provide a good sense of what has not been saved.40
39Unlike earlier printed products, acid paper does not thrive on relatively benign neglect; it deteriorates on the shelves. Libraries have had to spend a great deal of money microfilming or chemically treating deteriorating works to preserve them, and the funds available to support this work have been far less than what has been needed.
40It is worth noting, although outside the scope of this report, that the digital environment may place a heavier burden on the government to implement good archival practices. Previously, the FDLP provided a mechanism for the distribution of key government documents to libraries, which then preserved them. Now, with documents available over the Internet, libraries do not add them to their "collections." If these documents are to be preserved, the government will need to assume the archival responsibilities for the electronic documents that it produces, unless other reliable institutions can be found to accept this responsibility.
The archival system has, however, been much less successful with other media of the 20th century. It is important to briefly recount the situation here for the insights it can offer into the fate of the cultural record in the digital era.
Much of the early history of film and sound recordings is lost, although we do have many treasures from this period. Some of the problems were cultural: These new media were not recognized as important parts of the social record until they had achieved a certain level of maturity, popularity, and acceptance and thus weren't saved early on. Part of the problem is technical: Many of the early recordings were made using equipment that is no longer available, and the media on which they were made have deteriorated (in the case of nitrate film stock, it has in some instances spontaneously combusted!). Part of the problem had to do with business models: Until the VCR appeared in the 1980s, most films were simply not available for purchase by cultural heritage institutions. Thus, although they had a huge public impact and were viewed by millions, therefore unquestionably qualifying as part of the social record, they were in effect never published and hence could be preserved only by their owners or by archives if their owners donated them to such institutions.
Historically, copyright deposit has been another important factor in ensuring the archival availability of our cultural heritage, allowing the Library of Congress to amass an extensive collection of printed materials that it has been able to hold in trust for the nation.41 Sometimes, obtaining voluntary compliance with deposit rules proves difficult. However, the digital world vastly increases the problem. Most publishers will understandably object to an open-ended deposit system for electronic copies of their works, because it could enable anyone, anywhere, at any time of day to access the works in an electronic Library of Congress.42 However, if the Library of Congress received no deposits of digital works,
41Although the deposit of a copy of the work in the Library of Congress has been required as a condition for the registration of the copyright, U.S. copyright law does not require such registration as a precondition to copyright protection. Copyright protection automatically inheres in any work that is fixed in a tangible medium of expression, whether it is registered or not. There are, however, a number of practical and legal advantages to registration, including the ability on the part of the rights holder who has registered to receive, under certain circumstances, reimbursement for attorney's fees and statutory damages in a copyright infringement action. Indeed, it is an absolute precondition to bringing a copyright infringement action that the work must first be registered for copyright. So, although copyright registration is not required for protection, it is required if a copyright proprietor wishes to take practical steps to effect such protection, in the form of bringing a copyright infringement action.
42That is, publishers would likely object in the absence of technological protection mechanisms that limited the access, copying, and redistribution to a level that would not have an adverse impact on publishers' ability to derive a reasonable profit from their works.
it could offer far less public access to works in the future than it does at present.43
Copyright has played an important role in the preservation of many earlier nonprint materials. In the case of films, for example, series of still frames were deposited because film was not yet covered by copyright. These printed strips of still frames turned out to be the only extant copies of some early movies. As these works have moved out of copyright protection and into the public domain making them available to the public again as part of the nation's cultural heritage has been possiblein the case of film, by turning the sequences of still frames back into film. No longer required as a condition of obtaining copyright, copyright deposit can no longer be relied on by society as a means of ensuring the preservation of unpublished materials (which are still protected under copyright). Although it is certainly true that the Library of Congress can, and probably should, be more aggressive in demanding deposit of copies of published works, a tremendous amount of digital material exists that is never "formally published" (to the extent that this definition is understood in the digital environment) and for which it would be difficult to compel copyright deposit.
As the line between publication and limited distribution becomes ever more ambiguous in the digital information environment, the prospect grows of an ever larger mass of technically unpublished material that isn't subject to deposit. In particular, there will be a vast range of ephemera and of new genre material that will not be subject to deposit regulations if the criterion is one of publication. There is also the issue of what can be done with material that is deposited, at least while it is still under copyrightwho, if anyone, can view these materials? Finally, in a global world of information, deposit policies tied to the nation of publication are an increasingly awkward proposition.
Fundamental Intellectual and Technical Problems in Archiving44
Fundamental technical and intellectual problems are involved in extending the process of archiving beyond printed materials. Once pub-
43The developments thus far do raise concerns, considering that the deposits of digital materials to the U.S. Copyright Office have been modest at best.
44The discussion of archiving in this report is not intended to serve as a comprehensive review of the topic but to highlight that subset of archiving issues germane to a discussion of digital IP. Many challenges exist in the general area of archiving. Some interesting work is currently ongoing, such as the Intermemory Initiative (see <http://www.intermemory.org>, which aims to develop highly survivable and available storage systems composed of widely distributed processors. Discussion of the digital archiving issue will be included in the
(footnote continued on next page)
lished, printed works are static, and, unless printed on acid paper, they will last a long time in a reasonable environment with little attention or effort. Sound, film, and similar materials are also fixed once published, but they are on media that have variable and often rather short lives. In some cases, works cannot be directly viewed in any practical way because they require equipment that has grown increasingly complex over the years as it has evolved and which, in some cases, is no longer available (e.g., wire recordings and early video recordings for which the recording format is no longer known).45
Although offering the possibility of perfect preservation, digital information also raises many pragmatic barriers to long life. It is often stored on media with short life spans; it may require reading equipment that has an even shorter life span; and even transferring the data to another medium is not enough: Software may be needed to interpret and/or view the data. Thus, unless the reading software is also preserved (which may involve in effect preserving entire computing environments), some digital information cannot be meaningfully archived for long periods. Added complexity is associated with some of the new digital works that contain dynamic, interactive digital documents, because they are not fixed in form at publication; they evolve and change. As a result, exactly what the appropriate archival practices are for capturing the essence of these new genres is unclear. Indeed, these practices likely will vary from genre to genre and, in some cases, unless an archival function is designed into the digital object explicitly, fully archiving a record of its evolution over time may be impossible. One effortThe Internet Archiveis capturing wholesale snapshot copies of substantial portions of the Web (on what many view as very shaky copyright grounds), which will undoubtedly become important archival records.46 Yet the data collection tools used by the Internet Archive to traverse the Web take weeks to capture a single snapshot; so, for some archival purposes, this method may not record
(footnote continued from previous page)
forthcoming CSTB report of the Committee on the Information Technology Strategy for the Library of Congress. Additionally, the CSTB proposes to undertake a separate study on digital archiving in the near future.
45Of course, building new wire recording players is possible, although at considerable expense. One can also envision a time in the not too distant future when turntables for records will share a similar fate. As players for various recording technologies involve complex magnetic sensors, integrated circuits, embedded software, semiconductor laser technology, and other specialized parts that require an enormous and costly infrastructure to produce, building new players for obsolete media at any reasonable cost becomes less likely.
46See <http://www.alexa.com> or "The Digital Attic: Are We Now Amnesiacs? Or Packrats?" available online at <http://www.around.com/packrat.html>. The Internet Archive Web site is at <http://www.archive.org>.
sufficient detail in the evolution of volatile, dynamic Web sites. And there is a great deal of material on the Web or accessible via Web interfaces (e.g., databases) that is not subject to capture by the Internet Archive, either for technical reasons or because the site owners have opted out.47
In the case of various types of sound and moving image recordings, one strategy for overcoming media deterioration and the obsolescence and ultimate unavailability of playback devices is to periodically and systematically copy the recordings from older media to newer ones. However, this in turn requires that the owner of the copy have legal authority under the copyright law to take this action for preservation purposes. In fact, the copyright laws have permitted limited copying for preservation purposes, such as the microfilming of books. For the first time, the DMCA permits digitization as a means of preserving printed materials; until the DMCA, digitization was generally considered reformatting, rather than copying, and thus outside of permitted action under the copyright laws.
For digital materials, a similar strategy needs to be followed of copying data from one storage medium to another, as storage media deteriorate and technology changes. This objective is similar to the practices that large data centers have carried out for the last few decades in copying tapes every few years, both to protect against media deterioration and to shift from older to more modern generations of magnetic tape technology. Clearly, if digital information is to be preserved, such copying will need to be permitted under the law.
A number of strategies have been proposed to circumvent the obsolescence of document formats and unavailability of software to interpret older formats. One approach is to reformat documents periodically, converting them from older to newer formats so that they can be read by currently available software. This process, which goes beyond simple copying and might be considered the creation of a derivative work, will also need to be legal if the meaningful preservation of digital information is to occur.
The technical problems of managing digital information into the future are formidable and have been well documented elsewhere; they are not the primary focus of attention here.48 The key points are that if archives cannot obtain the digital materials that need to be archived and if they cannot obtain clear legal authorization to manage them across time,
47There is evidence that the search engines used by most people capture a relatively small percentage of the public Internet. For example, Lawrence and Giles (1999) found that no search index captures more than about 16 percent of the Web, which represents a substantial decrease in percentage terms from December 1997.
48See, for example, the 1999 NSF Workshop on Data Archival and Information Preservation (Rothenberg, 1999; CPA/RLG, 1995).
then no amount of progress on the technical problems will make any difference. Even if archives can legally obtain digital materials and can manage them (including reformatting as necessary), they will surely find it difficult to obtain the economic support for doing so for a century or more if they cannot provide the benefits that were formerly available to society in terms of limited consultation under the doctrine of first sale.49
Intellectual Property and Archiving of Digital Materials
Large-scale archiving of the cultural record requires resolution of two key legal issuesthe ability to make copies when migrating from one storage technology to another, and the ability to reformat, thereby creating derivative works when moving from one software technology to the next. But even if these issues could be settled, another issue remains: Libraries and similar cultural heritage institutions continue to be dependent on the framework of publication and first sale for the acquisition of most materials.
This dependence is becoming increasingly unworkable because more and more information is being provided under license. Rights to archive can be negotiated by libraries as part of a license for information; many research libraries are starting this process. With scholarly publishers, who share a common interest with the library community, authors, and readers in ensuring the archival viability of their electronic publications, these negotiations seem to be reasonably successful. With mass market publishers, the likelihood of success is less clear.50 The information supplier is under no obligation to agree to these termscompare this to the print world, where one literally could not publish and sell a book commercially without having the material considered for archival preservation in libraries.
Also, as indicated earlier, a significant amount of material is now being made available under apparently ambiguous terms with limited distributions. It is not necessarily published and placed on the market under the first-sale doctrine, but it is available for viewing on the Web
49The focus of this section is on archiving by cultural heritage and research institutions. However, the issues concerning archiving are also applicable to the consumer realm. For example, consider a consumer's LP record collection today and, perhaps, audio CD collections in the future.
50Some mass market information suppliers control the availability of their content to the public to maximize revenue. Disney and others have done so with their films for decades; other mass market content suppliers want to resell the same content repeatedly as popular media changewitness the shift from LP to CD and now to DVD. The replacement of the LP by the CD contributed considerable revenue to the music industry by permitting it to resell essentially the same content.
and on television and for listening on the radio. It is far from clear that libraries or other cultural heritage institutions have the right to capture this material, much less copy it across generations of media, reformat it, and make it available to the public for reference. Unless some process is initiated to permit this material to be collected by our cultural heritage institutions, such material could be omitted from the heritage of our society.51
Information distribution through sale and the first-sale doctrine has been an important framework for ensuring the preservation of that information for future generations. The information need not even be sold for it to be preserved: There is a significant body of ephemera that libraries have acquired; once the library had a legal copy, it did not have to worry about who held the rights to the underlying intellectual property but could simply keep and circulate the artifact. In the digital world one must always go to the expense of sorting out and clearing rights, even for ephemera.
In addition, is there a point at which a work becomes sufficiently public, even if not published in the existing legal sense, that it should become part of the collections of our cultural heritage institutions, to ensure both continuous availability to the public (at least on some limited basis, as libraries' holdings are available today) and preservation for access by future generations?
Complex and difficult balances will need to be established, as information increasingly becomes an event to be viewed or experienced, rather than packaged as an artifact to be kept and archived. Rights holders clearly should be able to limit the distribution of their content without running the risk that their works fall victim to some sort of intellectual eminent domain; creators should be able to make works available as "events" and to withdraw access to these events; and rights holders should be able to engage in revenue optimization strategies in marketing materials. This degree of control implies that authors and rights holders will be able to prevent the passage of some works into the cultural heritage. Yet there also needs to be a place for the capture of cultural history and social memoryat some point events are both common enough (widely enough available to the public) and important enough to become part of that record. Social agreements need to be developed about the status of information as event, and these agreements need to be reflected in the law and in practice; the legitimate interests of artists, commerce, and society must all be balanced.
The conflicting forces here are profound. The nightly news on a major television network is seen by millions of people and may play a
51For an in-depth discussion, see CPA/RLG (1995).
significant role in shaping public opinion. One could argue that such a broadcast should be available for study by future scholars and indeed by the public at large, at least under some constrained circumstances (e.g., viewing in a library). On the other hand, a concert or theatrical performance is a performance (i.e., an event intended to convey a one-time experience to a specific audience). The performers have no obligation to preserve their event for future scholars, and they might reasonably expect that no recording of the event be made without their permission.52 It is not clear how to draw distinctions and boundaries between these cases: One factor might be the public interest; another might be the scope of the audience (limited attendees versus unlimited availability to the public).
Technical Protection Services and Archiving
As indicated, preservation of digital materials is a difficult and far from fully solved technical problem, and technical protection services for intellectual property make this problem even more challenging.53 If a digital object can be read only through some type of proprietary, secure content-distribution software, this software must be migrated from one generation of hardware to the next; if the system really is secure, then the ability to preserve the content (even if legally permitted) depends totally on the continued viability and commercial availability of the secure content-distribution software. Even with legal authorization, a library could no longer be confident of its ability to migrate digital content protected by such a system. Similarly, new levels of complexity are added by any type of protection mechanism that tries to limit use of a digital object to a single specific machine, tries to report use of that object over the Internet periodically (e.g., in conjunction with a metering or payment system), or gets authorization over the Internet for each viewing. This complexity is likely to be at odds with the long-term ability to preserve or access the digital object, particularly as it becomes of little or no commercial value, causing the publisher to lose interest in preserving its end of the metering or reporting system. Designing into technical protection services features that facilitate migration (e.g., moving from an older
52The Vanderbilt University Television News Archive holds more than 30,000 videotapes of individual network news broadcasts and more than 9,000 hours of other news programming. The archive is developing a plan for digitizing its holdings so that researchers can watch recordings via the Internet. The distribution of these broadcasts using the Internet would amount to a retransmission, argues the broadcasters, and therefore would require a license, possibly accompanied with a fee. Of course, the broadcast networks have the discretion to grant a licenseor not (Kiernan, 1999).
53See the Chapter 5 section entitled "Technical Protection" and Appendix E for an in-depth discussion on technical protection services.
version of secure content-distribution software to newer versions) is certainly possible, but one has to worry about whether software with these capabilities will be made available in a competitive marketplace without a legal requirement.
Existing technical protection services do not, as far as the committee has been able to determine, include an express "self-destruct" mechanism triggered when the copyright on a work expires and it finally enters the public domain. The design of such a mechanism would be tricky, because this date is set at life of the author plus a fixed number of years (or at publication plus a fixed number of years for works of corporate authorship). In the former case, the protection software would in principle have to be able to determine when the author died in order to decide when to cease operating.54 One can envision situations in which works in the public domain remain entangled with protection software that still attempts to limit access to the work. This largely unforeseen consequence of the recent legislative endorsement of technical protection services may, in the long term, run counter to the public good of a healthy and accessible public domain (Lynch, 1997). An assessment is needed to consider whether the requirements for disengagement of a TPS should be part of the legal and social constraints on the deployment of such protection services in the first place.
Although the focus here has been on the potential difficulties raised by technical protection services, they can also have a positive effect on archiving. For example, the lack of effective TPSs may cause content owners to avoid making their works available to the public in digital formats in the first place. This lack of availability could have serious implications for the public good. The committee acknowledges that the foregoing discussion of archiving and TPSs presumes the existence of works and focuses on the issues for archival institutions and archiving processes. Clearly, authors and publishers need to have appropriate incentives and protections so that they create and distribute content in the first place.
54The details of authorship are part of the copyright management information protected under the DMCA. Formally recording this information may ease the transition of works into the public domain on several levels, First, it may make determining when a work enters the public domain less costly, and second, it may make automatically disengaging technical protection feasible.