Application: US State Government Documents (1923–77)

CRMS took the 1923–77 US copyright notice formality as the basis for a systematic copyright review of US state government documents. We consider the review of state government documents to be a valuable, large-scale, and low-risk area for review. There are over seventy thousand state government volumes currently in our candidate pool, making it a substantial body of work to review. We have also received numerous requests from scholars studying state documents and see this as rich territory for future scholarship.

Approximately 70 percent of the state government documents we reviewed did not bear a copyright notice. This implies that many state governments were relatively unconcerned about the copyright status of these works, as the absence of notice on these works injected them into the public domain.[21]

17 U.S.C. § 104A: Copyright Restoration under the URAA

Copyright restoration means that many works first published outside the United States between 1923 and 1989 will be considered to be in copyright, even if the rights holders didn’t comply with US copyright formalities of the time, including renewal and notice.

Copyright restoration is a wrinkle for copyright review systems that base their determinations on the renewal and notice requirements detailed above. The copyright in works first published outside of the United States may be restored, even when rights holders did not comply with US copyright formalities in existence at the time of publication. This can complicate copyright review, because your review system should account for both the non-US authorship of the work and the publication history of the work. These elements require additional time and research.

Restoration will not apply to works first published in the United States, nor to works published prior to 1923. We detail the key elements of copyright restoration below, but you should consider reading the US Copyright Office’s Circular 38b, Copyright Restoration under the URAA in order to understand the contours of restoration.

Per Circular 38b, a work is eligible for restoration provided all the following conditions are met:

1. At the time the work was created, at least one author (or rights holder in the case of a sound recording) must have been a national or domiciliary of an eligible source country. An eligible source country is a country, other than the United States, that is a member of the WTO, a member of the Berne Convention for the Protection of Literary and Artistic Works, or subject to a presidential proclamation restoring US copyright protection to works of that country on the basis of reciprocal treatment of the works of US nationals or domiciliaries.
2. The work is not in the public domain in the eligible source country through expiration of the term of protection.
3. The work is in the public domain in the United States because it did not comply with formalities imposed at any time by US law, lacked subject matter protection in the United States in the case of sound recordings fixed before February 15, 1972, or lacked national eligibility in the United States.
4. If published, the work must have been first published in an eligible country and not published in the United States during the 30-day period following its first publication in the eligible country.[22]

We have not identified robust tools to systematically address the fourth factor, the “simultaneous publication” (within thirty days) question. Instead, we primarily focus on the following questions: (1) is there non-US authorship in the work; (2) was the work in the public domain in its country of origin as of January 1, 1996;[23] and (3) was the work first published in the United States?

We used the Virtual International Authority File (VIAF) to help us identify non-US authors. We also used WorldCat to study the publication history of volumes where copyright restoration was likely.[24] Even with those tools, restoration is complicated territory and we do not have perfect answers for researching every factor. We do have some ideas, but they do not currently work at scale. Researchers wishing to perform additional research into the fourth factor, the “simultaneous publication” (within thirty days) question, may consider reviewing the publication date information discoverable in the Catalog of Copyright Entries and cross-referencing that information with publication notices in past volumes of trade publications.[25]

US Federal Government Documents (17 U.S.C. § 105)

Federal government documents are given a public domain determination. This is based on Section 105 of the Copyright Act, which disclaims US copyright protection for works of the US government: “Copyright protection under this title is not available for any work of the United States Government.”[26]

US federal government documents have not been a focus of CRMS reviews, but we note the following observations in the Compendium of US Copyright Office Practices for any project that encounters questions related to the copyright status of federal government works:

• Works prepared by officers or employees of the US Postal Service, the Corporation for Public Broadcasting, the Public Broadcasting Services, or National Public Radio are not considered works of the US government.
• Works prepared by officers or employees of the Smithsonian Institution are not considered works of the US government if the author-employee was paid from the Smithsonian trust fund.
• The US Secretary of Commerce may secure copyright for a limited term not to exceed five years in any standard reference data prepared or disseminated by the National Technical Information Service pursuant to 15 U.S.C. Chapter 23.[27]

In addition to the above carve-outs, we observed some confusion among librarians about what constitutes a federal government work. Typically, the answer to this question requires additional research into the agency and the agent that produced the work. Our guidepost for determining whether a work falls under Section 105 comes from the Section 101 definitions found in the Copyright Act: “A ‘work of the United States Government’ is a work prepared by an officer or employee of the United States Government as part of that person’s official duties.”[28]

Territoriality

Put simply, the copyright laws of any one country are not determinative for questions of copyright worldwide. Copyright law in other territories of the world is frequently different from the copyright laws we find in the United States. To cite one example, Canada features a copyright duration of life of the author plus fifty years,[32] which is twenty years less than the term of protection currently offered in the United States and many European countries. The consequences of this difference are very clear—in Canada, a work by an author who died in 1963 is in the public domain as of January 1, 2014, while a work by the same author may be protected by copyright in the UK until January 1, 2034.

Additional coverage of the territoriality principle can be found in Goldstein and Hugenholtz’s International Copyright[33] and Geller and Nimmer’s International Copyright Law and Practice.[34] Your project team should be aware of territoriality and agree on a means for navigating it. CRMS-World determines the copyright term of works published in the UK based on UK copyright law and bases its copyright determinations for works published in Canada on Canadian copyright law.

National Treatment

National treatment means that, by operation of treaty, a foreign author will receive the same treatment as the nationals of the protecting country. In other words, if Spain and the UK have agreed to treat their nationals identically, Spain will grant copyright protection to UK authors for the same duration as Spanish authors. Likewise, the UK will grant copyright protection to Spanish authors for a term equal to UK authors.

As a consequence of national treatment, a Spanish court recently found that the works of G. K. Chesterton remained in copyright in Spain, despite their public domain status in the UK.[35] The Chesterton case is consistent with the Phil Collins case, a German decision in which the courts held that European Community nationals must be afforded national treatment with regard to copyright. Nationals of the UK, for example, would be afforded the same copyright duration in Spain as Spanish-born authors.[36] The German Federal Court of Justice, in a subsequent case, found that the works of Puccini, an Italian composer, were protected in Germany, despite the fact that Puccini died prior to the original 1958 European Economic Community treaty and that Puccini had a shorter term of protection in Italy.[37]

Special Cases

Your project planners should expect to encounter differences from one international copyright regime to the next. This baseline understanding will help to guide your planning, shape your project scoping and access decisions, and inform the ways you communicate with foreign rights holders.

Narrowing your scope is an important means of dealing with the complexity of international laws. Even with a narrow scope, you will need to attend to nuances and differences when developing review projects focused on international works. Below are a few examples of variations in the UK’s copyright regime. These examples reflect some ways in which the international legal landscape does not always match up with a US-centric understanding of copyright law.

King James Bible

CRMS takes an admittedly conservative approach with regard to public domain determinations of versions of the Bible. Within the United States, we consider any version of the Bible published prior to 1923 to be in the public domain.

Outside the United States, we do not open versions of the Bible as public domain, based on UK law. This is largely due to the unique status of the King James Version, as noted on the Cambridge University Press website:

Though it was published in the seventeenth century and is out of copyright, control over the KJV Bible persists as a royal prerogative. While this may seem surprising, please remember the theme of this section: international legal regimes will not always match your understanding of US law.

Peter Pan

Like the King James Version of the Bible, we would consider the pre-1923 publications of Peter Pan to be in the public domain in the United States. However, we wouldn’t apply that determination universally.[39] To understand the unique status of Peter Pan, again look at UK law:

Based on the standard copyright term in the UK, one would expect all J. M. Barrie’s works to have entered the public domain in the UK. We only discover this variation by looking more closely at UK law and the legal commentary surrounding it.

Crown Copyright

Figure 2 War Office seal, from the front pages of A collection of minor wartime government publications, https://babel.hathitrust.org/cgi/pt?id=uc1.b3039799;view=2up;seq=6;size=175

Crown copyright encompasses works produced by government agencies in Commonwealth countries. Like works of the US government, the copyright term for works covered by Crown copyright is not based on the life of the author of the work. Instead, copyright in a Crown work is typically held by the government for a period of years after publication.

For the three countries encompassed by our CRMS-World project, Crown copyright terms are as follows:

• Australia: Year of publication + 50 years[41]
• Canada: Year of publication + 50 years[42]
• United Kingdom: Year of publication + 50 years[43]

To identify Crown copyright works, we instruct CRMS reviewers to look for indicia that a work from a Commonwealth country was prepared or published by or under the direction or control of Her Majesty or any government department. Frequently, Crown copyright works bear a “Crown Copyright Reserved” notice or carry some other indicator of government publication (such as the Royal Coat of Arms, above).

Inserts

Inserts are third-party content incorporated into a work. When we talk about an insert, we may be referring to a range of materials—to cite a few examples, inserts include individual photographs, illustrations, and articles or chapters previously published in other works. The inserts issue greatly complicates copyright review. The issue is similar for both US-based copyright determinations and copyright determinations for international works. At its most fundamental, the insert issue is an information problem. We often can make a copyright determination for a given volume, but the copyright status of component parts may be impossible to determine or require extensive research.

For US books published 1923–63, a copyright determination for a book may be based on the presence or absence of a copyright renewal record in the Stanford Copyright Renewal Database. However, imagine that the book was not renewed but features fifty-three photographs, licensed from more than one photographer, for the purpose of providing illustrations for the book. We would treat those photographs as inserts and typically end the review with an und/nfi (undetermined/needs further investigation) determination, subject to future research.

Our inserts policy takes a conservative approach, one that has allowed us to move quickly through hundreds of thousands of reviews even though it may keep many works closed that may be properly in the public domain. While we do not review works with photograph inserts, we know that very few 1923–63 photographs were renewed. Inserts represent a very difficult information problem and our conservative stance is one approach to this problem.

Currently, the registration and renewal status of an individual photograph is not easy to determine. Registrations and renewals for individual photographs are findable in the Catalog of Copyright Entries. However, to our knowledge, no one has yet created a visual inventory of all renewed photographs that would allow a reviewer to cross reference a photograph contained in an otherwise public domain volume with the renewed photographs listed in the Catalog of Copyright Entries. Arguably, such an inventory could be created but, without some image search functionality, its usefulness is an open question.

There may be alternate ways to address this problem. We recognize that the concern for possible copyrights in a relatively small number of possible inserts results in a large number of closed (primarily und/nfi) works—over 46,000 volumes in CRMS-US alone. To illustrate the likely mismatch between our concern for inserts and the number of works likely to contain renewed, in-copyright insert material, consider the following additional data points:

• Very few photographs and illustrations published between 1923 and 1963 were actually renewed and would be still in copyright. In 1955, there were only 216 renewals for artwork and photographs.[44] In 1956, there were 256.[45]
• The renewal rate for these types of works was low; therefore, most are likely to be in the public domain.[46]

Based on the relatively small number of likely in-copyright inserts, others may choose to take a different approach.

Published versus Unpublished

The published/unpublished divide is an important distinction in US copyright law, with implications outside of the United States as well. A work published in the United States in 1960 may be in the public domain in the United States due to failure to comply with US copyright formalities from that time period, such as registration, renewal, and copyright notice. However, if the work was not published and remained unpublished after 2002, the work would be “in copyright” for the life of its author plus seventy years or 120 years from the date of its creation, depending on facts related to its authorship.[47] In the UK, to cite just one international example, many unpublished works will be in copyright until 2039 or later.[48]

A key case articulating the published versus unpublished distinction in the United States is Estate of Martin Luther King, Jr., Inc v. CBS, Inc.[49] In that case, the court articulated the difference between publication and “non-divesting limited publication,” which would not constitute publication for the purposes of US copyright law: “Only a general publication divested a common law copyright. A general publication occurred ‘when a work was made available to members of the public at large without regard to their identity or what they intended to do with the work.’ Conversely, a non-divesting limited publication was one that communicated the contents of a work to a select group and for a limited purpose, and without the right of diffusion, reproduction, distribution or sale. The issue before us is whether Dr. King’s delivery of the speech was a general publication [internal citations omitted].”[50]

This distinction is important for archives. If a work was not published, which is the case for most archival collections, a copyright review will typically involve researching the death date of the author of the work. An unpublished letter, written in 1957 by an author who died in 2002, would be in copyright until 2073.[51] In contrast, a book published in 1957, and one that did not conform to copyright formalities of the time period, would be in the public domain in the United States today.

Application: Dissertations and Theses

Dissertations and theses are a rich resource housed in the collections of many libraries and archives throughout the United States. CRMS takes a conservative stance and does not currently review dissertations or theses because of the unsettled question of their publication status. Ultimately, your institution’s position on whether a given dissertation or thesis volume was published or unpublished will drive your copyright determinations for these types of works.

If published between 1923 and 1989, a dissertation would have required a copyright notice; otherwise the work entered the public domain. A key question then becomes, was this dissertation in fact published? There is a spectrum of opinion on the publication status of dissertations—two noteworthy examples are Gail Clement and Melissa Levine’s Copyright and Publication Status of Pre-1978 Dissertations: A Content Analysis Approach[52] and Peter Hirtle, Emily Hudson, and Andrew Kenyon’s case study, “Dissertations, Theses, and Student Papers,” found in Copyright and Cultural Institutions.[53]

Clement and Levine’s article identifies instances where dissertations are likely to be considered published, such as past publication in microfilm form through University Microfilms International (UMI).[54] Hirtle, Hudson, and Kenyon’s study notes that the University of California, Berkeley, in its guidance on “Publishing your Dissertation,” takes the following position: “The Attorney for the Regents has advised that shelving the dissertation or thesis voids the common law copyright.”[55] In other words, Berkeley’s view is that a dissertation shelved at Berkeley has met the requirements of publication.

At the very least, there is agreement that the publication status of a dissertation is a fact-specific inquiry. Any project that wishes to make public domain determinations for dissertations, based on publication and lack of notice (for dissertations published 1923–77) or failure to renew copyright (1923–63), will need to first take a position on the publication status of (1) dissertations that were placed on a library shelf and accessible to the general public and (2) dissertations that were distributed via microfilm through companies like UMI.

Some commentators feel that asking authors for permission to make dissertations available is the best route, regardless of the possibility that a given work may be in the public domain. In support of this approach, Kevin Smith has noted a recent case involving a student’s dissertation, Diversey v. Schmidly, in which Andrew Diversey sued the University of New Mexico for copyright infringement.[56] Smith draws parallels between the Diversey case and retrospective digitization projects aimed at doctoral or masters’ theses and dissertations. In doing so, he adds an additional factual question to those noted above: how do we know that the author has authorized publication of their dissertation?[57]

Libraries and archives serve a special societal function, and copyright favors uses that promote progress. Decision makers at institutions have a range of options for addressing the dissertation question at their respective institutions. They may (1) bear the costs, complexity, and potential dead ends of seeking permission from dissertation authors; (2) bear the cost of a public domain determination for these works, along with the possible cost of error (note that dissertations may be a particularly sensitive topic for authors); (3) articulate a strong fair use argument, consider bolstering it with a public domain determination process, and filter out all works that are likely “in copyright”; or (4) do nothing. In some cases, a combination of these approaches may be the most appropriate path forward.

For many institutions, taking no action would be considered poor stewardship of their collections. Others will adopt a “wait and see” approach to this question, learning from the successes and failures of other institutions. As institutions take a stance and work toward developing robust processes for larger community adoption, we anticipate there will be some lessons learned, and we hope that these are shared broadly.

Additional Authors

For most pre-1978 books published in the United States, the publication date of the book is central to its copyright duration. Additional authors do not typically factor into the copyright duration calculation. A coauthored work published between 1923 and 1963 and not renewed will be in the public domain in the United States regardless of the life-spans of the coauthors.[58]

When working in any copyright regime that bases the duration of copyright on the life of the author plus some number of years, you must calculate the term using the death date of the last surviving author. This can result in some peculiar consequences for copyright review. For example, sometimes we are able to locate the death dates for three of four authors but the fourth is difficult or impossible to ascertain. This may be a modest contributor who died at a much later date than the lead authors, yet it can result in the entire work remaining closed.

Translations

The important thing to remember when working with translations is that there are at least two separate rights holders to consider when making a copyright determination. There will be a copyright in the underlying work, the source of the translation. There will also be a copyright in the translation itself. Therefore, your reviews should take into account both sets of rights.

To give a real-world example, suppose you are reviewing a modern translation of Don Quixote. The underlying work, written by Miguel de Cervantes, entered the public domain long ago—the work was published in the early seventeenth century, and Cervantes died in 1616. If a recently deceased or still-living author, on the other hand, wrote a modern translation, then that translation may still be under copyright.

Multipart Monographs

Multipart monographs are works published over a span of years rather than a single year. This issue is of concern when the copyright in a particular set of volumes is based on publication date, rather than based on the life of the author plus some number of years.

An example of this issue would be a monograph first published in the United States in four parts on the following dates: 1922, 1925, 1927, and 1930. The first part would be in the public domain in the United States, based on its pre-1923 publication. The remaining volumes would be subject to registration, renewal, and notice requirements, so they may or may not be in the public domain. Further research would be required.

Similarly, a four-part monograph published in the United States in 1960, 1965, 1970, and 1979 would be subject to different sets of requirements. The 1960 part would have required copyright notice, registration, and renewal. The 1965 and 1970 volumes would have absolutely required notice, and the 1979 volume would have required notice or, in the absence of notice, registration within the subsequent five years.

Outside the United States, copyright determinations will be based on the death dates of the authors whose work is in the volume. When dealing with multipart monographs, you should watch for changing authorship over time.

The Importance of the und/nfi Category

One of the fundamental elements of the CRMS review process is the und/nfi (undetermined/needs further investigation) category, which is a decision-making outlet for reviewers who encounter works that present more complex issues of copyright. As an example, suppose you encounter a book first published in 1952 in the United States for which there is no copyright renewal record. Is that entire work in the public domain?[59] Does your answer to that question change if it contains illustrations or photographs?

A rights holder may have failed to renew copyright for a book published in 1952, effectively placing the book in the public domain, and yet component parts of the book may not be in the public domain. There may be a photo or illustration in the work that was individually registered and renewed. As an example, Bessie Pease Gutmann’s Love’s Blossom was registered on April 20, 1927, and was renewed on March 4, 1955.[60] If this image were incorporated into a book published in 1952, and the author of that book failed to renew its copyright, that failure would not have thrust Love’s Blossom into the public domain.

Figure 3 Bessie Pease Gutmann’s Love’s Blossom. Image included here as an exercise of fair use.

CRMS is a production-oriented project, and our team did not have the time to research the copyright status of every individual image. When we encounter a work that includes credited content, we mark the work as und/nfi (due to inserts) and set it aside for determination at a future date.

This stance is a mix of risk assessment, copyright law, and pragmatism. Your project may consider alternatives that do not involve performing copyright determinations on each individual component part. However, we would caution against extensively and fully reviewing every insert; the process would quickly become bogged down. This is the value of the und/nfi category: it allows us to disregard excessively complex copyright objects and instead focus our energies on works that are much more likely to lead to a conclusive determination. At the same time, the und/nfi category is ripe with opportunity deferred.

Notice and Takedown

In Copyright Risk Management: Principles and Strategies for Large-Scale Digitization Projects in Special Collections, Kevin Smith notes that a good strategy for mitigating the risk associated with any digitization project is to have a takedown policy for materials that become subject to complaint.[61] The same may be said for a copyright review management project. While notice and takedown does not eliminate the possibility of a rights holder bringing suit, it does reduce the possibility and helps to avoid escalation of any issue that may arise. Taking a work down does not preclude the possibility of requesting permission to provide access to the work or studying the issue further and concluding that the work is properly in the public domain. A responsive takedown policy provides time to consider future actions without the additional stress of a pending complaint.[62]

Role of an Advisory Working Group: Oversight from Copyright Experts

To the extent that you are planning a long-term copyright review project and intend to review a broad range of material, you should consider forming a copyright advisory group for ongoing informal or formal consultation. If your project is small in scale, narrow in scope, and of limited duration, then an advisory working group may not be necessary. Given the complexity of copyright and the possibility that the legal landscape may evolve over the duration of any given copyright review project, it is worthwhile to have experts available to help with both predictable and unforeseen challenges or opportunities that may arise.[63]

An advisory working group will provide support when your project team faces difficult and legally complex questions. A mechanism for reaching out to experts and drawing on their expertise will benefit any large-scale project. CRMS had the advantage of being able to draw on the expertise of several copyright scholars and practitioners (see acknowledgments).

Your institution’s legal counsel should either be directly involved with a proposed copyright review project or help identify experts to participate in advising your project team. The advisory group should include some participants from outside the institution who can provide fresh eyes for those times when your team needs an objective vantage point. If you are considering working with materials that implicate international legal regimes, consider identifying and collaborating with experts who have experience working with the laws of the relevant country.

We found that in-person meetings with our advisory group were an important way for us to check our processes and recalibrate practices as needed. While other forms of communication are often necessary, hashing out the details of a large-scale project benefits from in-person group discussion. Be prepared to hear a spectrum of opinions on any given topic and understand that you must ultimately decide which path makes sense for your institution. Your advisory group can provide good, meaningful feedback for your project, but issues related to legal liability, public scrutiny, and future relationships with rights holders ultimately begin and end with your own institution.

Partnership and Collaborative Work

Collaborative work offers many advantages. CRMS has benefitted greatly from the contributions of nineteen partner institutions and over sixty copyright reviewers. In a given week, we collectively perform several thousand copyright reviews. In isolation, a single institution could not have accomplished the same outcome. By distributing the labor of copyright review, CRMS was able to accomplish over 300,000 copyright determinations for books published in the United States in less than six years’ time, as well as over 172,000 determinations for books published in the UK, Canada, and Australia. This is a tribute to the individuals who contributed their time and energy to this process. It is also a testament to the power of distributed work.

While the above is a testimonial for cooperative partnership in your copyright review, keep in mind the legal and financial implications of working with a range of partners. Institutions engaging in copyright review projects cannot eliminate the risk of mistake—copyright is far too complex to ever design a completely error-proof system. This risk of error in copyright review projects should not be taken lightly, and the costs of mistakes can range from the institutional costs of remedying an error to the more profound consequences of a lawsuit.

CRMS worked to mitigate the risk of error by instituting double reviews, selecting expert reviewers who are fair but conservative in their adjudications, and managing our partner reviewers through training and regular feedback. However, as the number of reviewers increases, the time commitment of managing the activity also increases. Large-scale copyright review requires continued oversight and guidance. If you plan on performing this work on a large scale, be prepared to invest significant resources in its oversight.