(Watch Webcast)

Statement of Marybeth Peters
The Register of Copyrights
before the
Subcommittee on Courts,
the Internet, and Intellectual Property,
Committee on the Judiciary

United States House of Representatives
110th Congress, 2nd Session

March 13, 2008

The “Orphan Works” Problem and Proposed Legislation

Chairman Berman, Ranking Member Coble, and Members of the Subcommittee, I am pleased to appear before you today to testify in support of orphan works legislation. Like you, I believe it is important to address orphan works because they are a problem for almost everyone who comes into contact with the United States copyright system. Moreover, they are a global problem. Every country has orphan works and I believe that, sooner or later, every country will be motivated to consider a solution. The solution proposed by the Copyright Office is a workable one and will be of interest to other countries.

In my testimony, I will briefly explain the scope of the orphan works problem and why it is so important to provide relief—important not only to the copyright community but also to the public discourse. I will then turn to the challenge of how best to craft a solution that will move some copyright users forward without moving copyright law and copyright owners backwards. I am certain that this is possible.

The Orphan Works Problem

As you know, in 2005, with direction from this Subcommittee and the Senate Subcommittee on Intellectual Property, the Copyright Office conducted a comprehensive investigation of the orphan works problem. In 2006, we published our findings and recommendations in a study entitled Report on Orphan Works. The Report documents the nature of the orphan works problem, as synthesized from the more than 850 written comments we received and the various accounts brought to our attention during three public roundtables and numerous other meetings and discussions.

We heard from average citizens who wished to have old photographs retouched or repaired but were denied service by the photo shops. Unfortunately, if those photographs were taken by professionals (for example, wedding photos), the photo shops' actions make sense under the current law: they know that the photographer, not the customer, probably holds the copyright in the photograph. They ask the customer to produce evidence that the photographer has agreed to allow the reproduction of the photo (which will be necessary to retouch or repair the photo). But of course the customer has no idea who the photographer at his parents' wedding was, or quickly hits a brick wall when attempting to track that person down. Many other examples were presented to us as well, from museums that want to use images in their archival collections to documentary filmmakers who want to use old footage.

In fact, the most striking aspect of orphan works is that the frustrations are pervasive in a way that many copyright problems are not. When a copyright owner cannot be identified or is unlocatable, potential users abandon important, productive projects, many of which would be beneficial to our national heritage. Scholars cannot use the important letters, images and manuscripts they search out in archives or private homes, other than in the limited manner permitted by fair use or the first sale doctrine. Publishers cannot recirculate works or publish obscure materials that have been all but lost to the world. Museums are stymied in their creation of exhibitions, books, websites and other educational programs, particularly when the project would include the use of multiple works. Archives cannot make rare footage available to wider audiences. Documentary filmmakers must exclude certain manuscripts, images, sound recordings and other important source material from their films. The Copyright Office finds such loss difficult to justify when the primary rationale behind the prohibition is to protect a copyright owner who is missing. If there is no copyright owner, there is no beneficiary of the copyright term and it is an enormous waste. The outcome does not further the objectives of the copyright system.

More than one phenomenon has contributed to the orphan works problem. Digital technology has made it easier for a work or part of a work (such as a sound recording or a “sample”) to become separated from ownership or permissions information, whether by accident or through deeds of bad faith actors. Business practices have furthered the publication of works without any credit of authorship or copyright ownership, as in the publication of photographs in some advertising contexts.

Sweeping changes to copyright law in the past 30 years have also contributed heavily to the problem. On January 1, 1978, the date on which the Copyright Act of 1976 became effective, the United States dramatically relaxed the requirements of copyright protection in order to move to a system that fulfilled the standards of international conventions. In doing so, we moved away from the highly formalistic system we had for the first 188 years of our copyright heritage.

The Copyright Act of 1976 changed several basic features of the law. First, copyright protection became automatic for any work of authorship fixed in a tangible medium (e.g. on paper, on tape, in a computer file) and registration with the Copyright Office became optional. (Registration was retained only as a requirement of filing suit in a U.S. District Court and as a condition of collecting statutory damages and attorney's fees.) To reduce the possibility of a work falling into the public domain because of failure to publish without a copyright notice, the new law contained liberal curative measures.

Second, it changed the term of copyright protection for new works to a period of the life of the author plus an additional 50 years after the author's death. Prior to this change, the term had been bifurcated. An initial term of protection was available for 28 years, then a renewal term was available for another 28 years, but only upon affirmative application to the Copyright Office. In 1978, the renewal term for pre-1978 works was extended to 47 years and in 1998 it was extended again, to 67 years. In 1992, “automatic renewal” was instituted, removing from the law the requirement that renewal claimants file applications with the Office. Until this time, in practice, only a small percentage of copyright claims had ever been renewed, leading to earlier injection of works into the public domain.1 In 1998, under the Sonny Bono Copyright Term Extension Act, we extended term to a period of the life of the author plus 70 years. In practice, for an author who creates while young and lives a long life, this could easily mean 125 years of protection or longer.

We made additional changes to our copyright law when we joined the Berne Convention, which prohibits formalities that interfere with the exercise or enjoyment of copyright protection. In 1989, the United States loosened the requirement that all works be registered as a condition of filing suit, making it inapplicable to foreign works. We also rescinded the condition that a published work must contain a proper copyright notice; thus, a common means of verifying the year of publication and the name of the copyright owner became less available. Finally, on January 1, 1996, under the Uruguay Round Agreements Act, we recognized millions of copyrights in foreign works that had been previously in the public domain because of failure to comply with the formal requirements of U.S. law, such as registration, publication with notice, and lack of copyright relations with the work's country of origin.

The Proposed Solution

In our study of the orphan works problem, the Office reviewed various suggestions from the copyright community. These included creating a new exception in Title 17, creating a government-managed compulsory license, and instituting a ceiling on available damages. We rejected all of these proposals in part for the same reasons: we did not wish to unduly prejudice the legitimate rights of a copyright owner by depriving him of the ability to assert infringement or hinder his ability to collect an award that reflects the true value of his work. We also rejected proposals that would have limited the benefit of orphan works legislation to certain categories of works or uses. Both commercial and noncommercial users made compelling cases; moreover, these parties often collaborate on projects and both need the benefit of the law. Likewise, we concluded that there were significant problems with respect to all categories of works: published, unpublished, foreign and U.S. works.

Instead, we recommended a framework whereby a legitimate orphan works owner who resurfaces may bring an action for “reasonable compensation” against a qualifying user. A user does not qualify for the benefits of orphan works legislation unless he first conducts a good faith, reasonably diligent (but unsuccessful) search for the copyright owner. As defined in our Report, reasonable compensation should be the amount “a reasonable willing buyer and reasonable willing seller in the positions of the owner and user would have agreed to at the time the use commenced.”2 Such a recovery is fair because it approximates the true market value of the work. It allows a copyright owner to present evidence related to the market value of his work and, at the same time, allows the copyright user to more precisely gauge his exposure to liability. Statutory damages would not apply to use of an orphan work. (The Office agrees with copyright owners who have since suggested that an award of attorney's fees might make sense in certain instances where an orphan work user acts in bad faith.)

That said, we stress that statutory damages would not be off the table perpetually. If an owner were to emerge, his legal ownership of the copyright in his work is unchanged. Full remedies, including full statutory damages, would be available against new users and, indeed, against the original user making a new, subsequent use. It is a basic tenet of the proposal that subsequent uses may not be based on stale searches, thereby increasing the probability that an owner may be found.

The Copyright Office proposed one exception to the basic rule of reasonable compensation, which is a safe-harbor for certain limited uses performed without any purpose of direct or indirect commercial advantage. The exception would apply only where the user ceased infringement expeditiously after receiving notice of a claim for infringement. We believe that this provision is a critical piece of the orphan works solution.

In most instances, we expect that the kind of uses that fall within the safe harbor will be made by museums, archives, universities and other users acting for cultural or educational purposes. In order to effectively bring important material to light, these users may need an additional safety net. For example, in the case of a local historical society seeking to make multiple orphan photographs available on its website or in a pamphlet, it is possible that reasonable compensation, in the aggregate, would still prove onerous. Such uses are in the public interest, and they will rarely conflict with the normal exploitation of the work or conflict with the legitimate interests of the copyright owner.

Finally, we note that injunctive relief is limited under our proposal. If a user has added significant new expression, we do not support the availability of an injunction, provided, however that the user pays reasonable compensation. If the user has not added significant new expression, we support the availability of an injunction with the caveat that a court be instructed to account for any harm to the extent practicable, in order to mitigate the harm resulting from the user's reliance.

Response of the Copyright Community

The Office received broad support for its Report and proposed solution, with the exception of photographers and some other owners of visual content. However, despite their opposition to legislation, visual artists have openly acknowledged the magnitude of the orphan works problem in their own community. One concern of photographers is that their works are sometimes perceived to be orphans when they are not really orphans. This is because photographs and other images are often published without credit lines or copyright notices. They do not always have metadata or watermarks. Certain categories of images are not routinely managed or licensed. These are genuine problems, but they are in fact the very essence of the orphan works problem.

Some who oppose orphan works legislation have also objected to the removal of statutory damages, which are available under Title 17 in certain instances. A few have even asserted that statutory damages are an entitlement under the law that cannot be rescinded. We disagree. Statutory damages are an alternative means by which a copyright owner may recover against an infringer in lieu of proving actual damages and lost profits. However, they are only available if the owner has registered the work prior to the infringement or within three months of publication. (While it is possible that a registered work could be an orphan work within the proposed legislative framework, we think this is unlikely to be a common situation, not because the registration is guaranteed to be found, but because an owner who has taken steps to register his work has likely taken other steps to make himself available outside the registration system.) Statutory damages are not an absolute entitlement any more than copyright ownership itself is an absolute right. Just as there are exceptions to, and limitations on, the exclusive rights of copyright owners (for example, fair use), there are exceptions to statutory damage awards. In cases of “innocent infringement,” the court may reduce statutory damages to $200; for certain infringements by nonprofit educational institutions, libraries, archives and public broadcasters, the court may reduce the award to zero.3 The fact remains that the possibility of statutory damages, however remote, is the single biggest obstacle preventing use in orphan works situations. In cases of non-willful infringement, statutory damages may be as high as $30,000 for each infringed work. In cases of willful infringement, they may be as high as $150,000 per infringed work.

We are not suggesting, in general, that the scheme of statutory damages is unjust. On the contrary, statutory damages fulfill legitimate and necessary purposes. That said, we do believe that in the case of orphan works, the rationale for statutory damages is weak. By definition, in the orphan work situation, the user is acting in good faith and diligently searching for the owner, and the owner is absent. The purposes of statutory damages, i.e. making the owner's evidentiary burden lighter and deterring infringement, weigh less heavily here. If the copyright owner is not identifiable and cannot be located through a diligent, good faith search, we believe the appropriate recovery is reasonable compensation. If orphan works legislation does not remove statutory damages from the equation, it will not motivate users to go forward with important, productive uses. On the other hand, the prospect of orphan works legislation may motivate some owners to participate more actively in the copyright system by making themselves available.

Prior Legislative Action

On March 8, 2006, this Subcommittee held an oversight hearing on our Report, followed by a similar hearing in the Senate on April 6, 2006. On May 22, 2006, “The Orphan Works Act of 2006” was introduced in the House by former Chairman Lamar Smith. The bill included revisions to the Copyright Office's original proposal and incorporated a number of changes that were designed to protect photographers and other visual artists in particular. These changes included a requirement that users document their searches, a definition of "reasonable compensation" (taken from the Office's Report), and the availability of attorney's fees under circumstances where a user fails to negotiate in good faith with an owner who has previously registered his work. That bill was later imbedded in H.R. 6052, “The Copyright Modernization Act of 2006.” The 109th Congress ended before the bill could be addressed.

Current Issues

In the two years since our Report was published, the Office has spent a considerable amount of time meeting with stakeholders to understand their concerns and to consider the policy implications of their suggestions. There have been numerous symposia on orphan works, sponsored by bar associations, academic institutions, industry committees and professional organizations. We are grateful for the additional insight such meetings have provided and agree that refinements can be made. Many of our discussions over the past year have been focused on the goal of providing additional safeguards to the legislation, some of which I will now discuss.

The Role of Best Practices

One of the most important challenges in constructing orphan works legislation is creating search criteria that are both strong and flexible. On the one hand, a user must search for the copyright owner diligently. A short-list or static checklist should not suffice. If one step in a user's search leads him to another step, he must follow the trail and explore the facts that present themselves. On the other hand, a user ought not to be required to explore meaningless steps if he has good reason to believe they will be fruitless. For example, it makes no sense to require a user to check an electronic database specializing in contemporary images of American photographers if what he is looking for is the owner of a 1930's photograph of German origin.

One of the suggestions that emerged in the 2006 bill was to incorporate certain established practices (“best practices”) into the search criteria. Such “best practices” would come from the relevant copyright communities—and thus a user who is looking for the owner of a sound recording would look to the recording industry and recording artists for guidance, as well as to other available resources. A book publisher looking for the owner of a photograph would look to the best practices proffered by photography associations and, also, to the professional guidelines proffered by the publishing industry. The most advantageous feature of this approach is that changes can be made easily as practices evolve. Finally, in the past year, some have suggested that the Copyright Office take a more active role in best practices, not only collecting them but also formulating them. If this would better ensure consistency and fairness across owner and user groups, and make best practices most useful to the public, we would not object to taking on this role.

The Role of Technology

The availability of technology will be an important aspect of best practices. As with best practices generally, the content owners and users in the respective copyright industries will be the parties most knowledgeable about whether a particular technology product is viable. For example, we are aware of several private sector companies working on tools and services that could help alleviate the orphan works problem by matching users to owners. On December 8, 2007, the Copyright Office organized a briefing and showcase of technology for Congressional staff.4 At the briefing, companies highlighted image recognition, fingerprinting, watermarking, audio recognition and/or licensing features, and discussed their efforts to develop business models and standards, including database control, security, population fees, and allocation of user fees or subscriptions. We are confident that the marketplace offers, and will continue to offer, an array of databases and search technologies that will result in more choices for the copyright owner and more aids for the prospective user. This is a process that is already underway but, certainly, an orphan works amendment would provide additional incentives for copyright owners and database companies to work together.

As a side bar, we believe that the Copyright Office's role in technology should be limited to reviewing best practices that are submitted to us. For example, we would not want to certify databases or other technological tools because we do not have the technical expertise to undertake such tasks. Moreover, we are not persuaded that certification should be a central concern. A user should take advantage of all reasonable tools likely to lead him to the copyright owner, regardless of whether the government has blessed that tool.

There are related questions, raised by some, as to whether the Copyright Office should have a searchable database of visual images; as we understand it, the Office would make copies of deposits that claimants send to us for registration purposes. We think a government database would be wasteful, ineffective and fraught with legal and practical problems. As a policy matter, the Copyright Office has never in its 200-year history made copyright deposits widely available for viewing (e.g. display or public performance). In contrast to registration information, which is made publicly available, deposits (if they have been retained by the Office) may be viewed by others only under very limited circumstances and subject to regulations that are intended to protect the deposits from unauthorized copying. Some copyright owners may be fearful of having their deposits made available to the public in digital form beyond the limited display that has been the practice for many years. Such a proposition could have a chilling effect on registration, which would in turn reduce the number of works that come to the Library of Congress as deposits through the copyright system.

On a practical level, it is difficult to imagine how the Copyright Office or any government office could ever keep pace with the image technology world that exists outside our doors and beyond our budget. In reality, the Copyright Office does not have and is not likely to obtain the resources that would be necessary to build a database of works that are searchable by image, even if there are some copyright owners who would be amenable to such an undertaking. Our point of comparison is the comprehensive reengineering project that the Copyright Office is just now completing. Among other things, this project has made it possible for authors, publishers and other copyright owners to routinely register their copyright claims electronically. Under the “Electronic Copyright Office” (or “eCO”), claimants may complete copyright applications, pay the required fees and submit the appropriate deposit copies of their works—all on-line. The eCO portion of reengineering took five years and has cost $17 million to date. We used off-the-shelf software (in accordance with Congressional directives) and completed the project on time and within the budget Congress appropriated. It represents the single biggest overhaul of the Copyright Office since 1870 and the most significant adjustment to registration practices since 1978. Based on this experience, we believe it would be highly impractical for the Copyright Office to employ cutting-edge image recognition technology.

Finally, the process of searching for a copyright owner is not a function controlled exclusively by the Copyright Office. Although the Copyright Office is one resource, our records will never be a complete resource because registration is a voluntary process and many copyright owners, including photographers and visual artists, choose not to register. Thus it is the case already that when searching for a copyright owner, users look to private databases, websites, publishers, collecting societies, professional organizations, trade associations and many other resources.

Other Issues

In the course of meeting with stakeholders in recent months, we've discussed a few issues related to the application of our proposed solution. For example, it is our view that beneficial owners of copyright, as well as legal owners, should be entitled to recover reasonable compensation from an orphan works user. (Usually, a beneficial owner, often an author, is someone who has transferred the rights in a work to another party but who nonetheless retains an on-going financial interest, such as the right to an on-going royalty.) Since it is currently the case that a beneficial owner has standing to institute a suit for infringement, we see no reason to change this fact under the orphan works framework. Other issues we've discussed have included providing more detail as to the pleading requirements under orphan works legislation and considering possible new enforcement issues related to small claims of copyright owners.


In closing, we note that millions of orphan works are precluded from productive use by authors, publishers, filmmakers, archives, museums, local historical societies and other users, despite the fact that the copyright owners may never be found. The solution that the Copyright Office has proposed reflects the realities of the problem and creates a framework for limited use. It does not create an exception; nor does it rescind an owner's copyright rights. We look forward to orphan works legislation and we are available to assist that goal in any way we can.


1. A 1961 Copyright Office study found that fewer than 15% of all registered copyrights were renewed. For books, the figure was even lower: 7%. See Barbara Ringer, 'Study No. 31: Renewal of Copyright' (1960).

2. See also Davis v. The Gap, Inc., 246 F.3rd 152 (2d Cir. 2001).

3. 17 USC § 504(c)(2).

4. The briefing included the following companies: Copyright Clearance Center, DigiMarc, Google, InfoFlows, PicScout, and PLUS. Audible Magic and Corbis could not attend but contributed materials.