On May 11, 2012, the federal district court in Atlanta handed down a landmark decision regarding the burgeoning use of copyrighted materials in electronic reserves (e-reserves) administered by college and university libraries. In Cambridge University Press v. Becker, Judge Orinda Evans rejected all but five of 75 claims that the Georgia State University (GSU) Library's system of e-reserves infringed copyrights owned by Cambridge University Press, Oxford University Press, and Sage Publications.1 As the first case to deal with the question, it will strongly influence further judicial considerations of the issue.

In a long, thorough opinion, Judge Evans sustained a relatively broad interpretation of "fair use" in educational contexts and imposed bright-line rules regarding the quantity of material that qualifies for "fair use." As long as an instructor put no more than a chapter of a work with more than ten chapters on e-reserve, or less than 10 percent of a work with fewer chapters, she found the use fair. That conclusion was reinforced if the publisher did not make permissions for the use of digital copies readily available at a reasonable price, but even if it did, as long as the amount used was within the limits, the purpose was educational, and the work being used was nonfiction and factual in nature, the judge found that the "fair use" doctrine precluded a finding of copyright infringement.

Although publishers and academic librarians were well aware of the importance of the case, historians and educators paid little attention, despite the prospect that a victory for the presses would have created "a nightmare scenario for higher education."2 Most observers believe that the decision has averted that "nightmare scenario," but this may be an overly optimistic conclusion.

According to the Copyright Act of 1976, the public can, without securing the permission of copyright owners, make "fair use" of limited amounts of copyrighted material "for purposes such as criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research."3 Another section of the law enacts special provisions for photocopying in nonprofit academic libraries, permitting individuals to make single copies of articles or small portions of books for their own study, scholarship or research.4 However, this section does not permit "concerted" or "systematic" photocopying.5 So when libraries established closed reserves from which students could make photocopies of supplemental course readings, they had to rely on the idea that it is a "fair use" of the original.

Closed reserves remain common, but they soon became secondary to coursepacks of photocopied materials that students were required to purchase. The development of digital copying and transmission has led to the establishment of e-reserves, which are much more convenient than both closed reserves of photocopyable hard copies and coursepacks. Students simply go to the e-reserve website and download the reading. Likewise, teachers can post digital copies of copyrighted material on their own class websites for students to download. But because the protections of the Copyright Act's library-copying provisions do not apply to the "concerted reproduction or distribution of multiple copies,"6 libraries have justified e-reserves as a "fair use" of the originals, making the same argument that has justified closed reserves.

For publishers, especially university presses, widespread photocopying and especially digital copying pose two problems. First, they potentially reduce the sales of hard copies of academic books and periodicals. In response, the presses have slowly, often grudgingly, moved towards replacing part of their book sales income with income from fees they charge to give permission to copy excerpts from their products. They can earn this income through in-house permission-granting facilities or through a service like the nonprofit Copyright Clearance Center (CCC), which collects permission fees and distributes them to rights holders. If educational institutions or their students must pay for permissions, the rise of e-reserves promises university presses a new revenue stream, and a respite from the fiscal constraints imposed by ever-declining subsidies from their sponsoring institutions. Publishing historians also would benefit, seeing some increased income from royalties. But universities and their students face financial constraints of their own, of which academic historians are well aware. Educational institutions claim that the very purpose of "fair use" is to encourage education by reducing the costs otherwise imposed by copyright. These were the issues at play when the Association of American University Presses (AAUP) encouraged Sage Publications and the two university presses to bring the 2008 suit against GSU for massive alleged copyright violations. They were bankrolled by the much wealthier Association of American Publishers (AAP) and the CCC.

GSU had taken an extremely aggressive position on what constitutes "fair use" of copyrighted materials. In establishing its system of electronic reserves, GSU adopted the expansive interpretation of "users' rights" advocated by L. Ray Patterson, a member of the University of Georgia law faculty, who had also been a visiting professor at GSU. Patterson stressed that the U.S. Constitution authorizes Congress to secure copyrights explicitly "to promote the progress of science and the useful arts" (Article I, §8). The purpose was not to protect the property rights of creators but to encourage learning and innovation. "Fair use," as a key concept in the protection of users' rights, must be construed broadly to prevent copyrights from inhibiting rather than promoting the goals specified by the Constitution.7 GSU placed virtually no restrictions on what instructors might post, as long as it was for educational purposes. In their initial filings the publishers alleged that GSU made excerpts from more than 6,700 works available to students in more than 600 courses.

Naturally, publishers have a much more restrictive interpretation of "fair use." In their view, which of course also serves their financial interest, copyright is intended to protect property rights in intellectual productions. The Constitution's language merely recognizes that the protection of property rights is essential to progress and enterprise; it does not mean that protection of property rights takes second place to those goals. In the conservative legal and constitutional climate that has developed since the 1980s, the notion that creative work constitutes "intellectual property" has grown stronger, and the more liberal philosophy that underpinned broad conceptions of "fair use" like Patterson's has receded. Consequently, publishers won suits against commercial producers of college coursepacks in the 1990s, with courts rejecting the argument that commercial photocopying, even if ultimately for educational purposes, constituted "fair use."8 Commercial photocopying services now routinely secure permissions for materials included in coursepacks, with the cost passed on to student purchasers. If they are unable to secure permission, they decline to publish the problematic material, with instructors often providing substitutes.

Despite this victory, a number of university presses, among them Cambridge University Press, have been slow to treat permissions as an important potential source of revenue, unwilling to undermine hard-copy sales with too generous a permissions policy. They are particularly reluctant to permit digital copies, which can be broadcast all across the internet much more easily than photocopies can be widely reproduced and distributed. Even those publishers, like Sage Publications (another plaintiff), which see permissions for digital copying as potentially lucrative sources of income, are worried about losing control of their products. Naturally, any press that looks to permissions for future revenue desires a narrow interpretation of "fair use," which authorizes copying without permission.

Armed with their legal victories against commercial photocopiers, the E-Reserves Task Force of the AAP Copyright Committee sent blunt letters to academic institutions whose libraries had, in its opinion, taken too liberal a view of "fair use" in e-reserve procedures. They urged administrators to educate faculty and students about copyright, correct misapprehensions about the scope of fair use, to control access to the materials, and to seek permission from rights holders before posting digital copies exceeding the amounts that the publishers considered legally justified. In negotiations with academic librarians they stressed the significance of a multilateral agreement on guidelines for classroom copying negotiated in 1976 (hereafter referred to as "Guidelines for Classroom Copying"—see sidebar at right for details of this agreement), which provides a "safe harbor" for those who adhere to a minimal interpretation of "fair use." The libraries generally responded by tightening rules for e-reserves; sometimes, as in the case of Cornell University, they issued guidelines worked out jointly with the AAP. Some institutions explicitly barred faculty from posting copyrighted materials on class websites. It became normal for library administrators to appoint copyright specialists to review requests for e-reserve postings. Some universities paid hefty amounts for annual licenses from the CCC to make limited uses of copyrighted material without having to seek individual permissions, even though the CCC is not authorized by every publisher to give permissions for every work. Others have designated funds, usually small, for securing permissions. Some simply discourage instructors from posting materials beyond what the specialists consider fair use.

Committed to its broad interpretation of user rights, GSU did not respond to letters from the Task Force, precipitating the suit. Constitutional doctrines concerning "sovereign immunity" of states from federal lawsuits precluded a claim for damages, which might have been staggering, but they permitted the publishers to seek an injunction to alter GSU's practices. Probably informed by its lawyers that the Patterson position was no longer legally defensible, if it ever was, GSU immediately announced a review of its procedures. In 2009 it revised them to conform fairly closely to those that had been adopted by other universities. Observers expected the presses to declare victory and drop the suit. But instead they pressed on.

One reason for their doing so could be that they were outraged at what GSU had done, and considered the new policy merely a cunning tactic to escape the injunction.9 Another may have been a desire to demolish the legal standing of the ideas promulgated by Patterson, for whom the American Library Association (ALA) had named its prize recognizing those who "pursue and support the Constitutional purpose of the U.S. Copyright Law, fair use and the public domain." Even worse, perhaps, from a publishers' perspective, the ALA referred to Patterson's book on users' rights as "the definitive book on the constitutional underpinnings of copyright."10

But there may have been yet another and more important reason. The publishers may have been confident that, in light of the law's shift towards a conservative appreciation of property rights, few judges would consider GSU's prior policies to have been undertaken in good faith. They might well be seen as a simple appropriation of the rights holders' property for the benefit of GSU—in essence, a theft. A similar sentiment seemed to have led the Supreme Court to promulgate a very narrow interpretation of "fair use" of unpublished materials some twenty-five years earlier.11 The publishers in the GSU case may have concluded that the case against an unsympathetic defendant presented an ideal opportunity to establish the restrictive definition of "fair use" that they thought just. They argued to the judge that she should base the injunction on the conduct of GSU prior to its policy revision. Even after she disagreed, the publishers persisted, arguing that GSU continued to infringe their copyrights even under procedures similar to those established by other institutions. It seems clear that the presses and their financial backers wanted something more than to rein in a rogue e-reserve system.

What they wanted becomes clearer when one considers the infringements the plaintiffs alleged under the new GSU e-reserve policy. In a world history course in which the instructor required the purchase of five texts, Cambridge University Press claimed that GSU had infringed its copyright by placing one chapter from a 1083-page, 15-chapter, volume of its Cambridge History of China (2008) on e-reserve as a supplemental reading. Another claim of infringement involved a graduate seminar in Issues and Interpretations in American History. Students were required to purchase 14 texts and complete a number of readings posted on GSU's e-reserves. Yet Oxford University Press objected to posting a single essay from its festschrift for C. Vann Woodward, Region, Race, and Reconstruction (1982) and another from Lawrence Levine's The Unpredictable Past (1993). In other words, if the presses had won their case, the most typical uses historians make of e-reserves in their classes would have been ruled infringements of copyright, beyond the scope of "fair use" of copyrighted materials for educational purposes. To remedy these alleged infringements, the plaintiffs sought an injunction that would have enshrined in legal precedent a narrow, rigid interpretation of "fair use" copying that publishers had favored since 1976, embodied in the "brevity" and the "cumulative effect" sections of the "Guidelines for Classroom Copying."

The proposed injunction required GSU to monitor all its faculty, staff, and students for compliance with this limited interpretation, requiring instructors to fill out forms certifying that their e-reserve posts conformed to the limits. Administrators of e-reserves would have had to reject every proposed reading that exceeded the crabbed limits of the "Guidelines for Classroom Copying" or that another instructor had already requested.

Librarians reacted with dismay and anger to the proposed injunction, which set limits sure to be demanded of other colleges and universities. The injunction meant that academic libraries or students would have to secure permission for virtually every copyrighted item posted on e-reserves, as would faculty who wanted to post copyrighted work on their own class or personal websites. This would mean not only transferring untold thousands of dollars from university libraries and students to publishers and the CCC every year, but also imposing onerous efforts to secure permissions from publishers who did not work with the CCC. Moreover, there are publishers, like Cambridge University Press itself, which simply refuse to give permission to make digital copies of some of their lucrative works. Finally, since nonelectronic closed reserves rely on the same "fair use" justification as e-reserves, the precedent might well be applied to them in the future too.

Librarians were particularly incensed that university presses should take the lead in seeking such a radical change in academic library and educational practices. They were just as angry that the CCC, to which many academic libraries paid large annual license fees, was backing the suit financially. As the dean of libraries at the University of Michigan told the readers of Publishers Weekly, it was understandable that university presses and university libraries had different interests when it came to e-reserves, and that the presses might go to court to stop what they saw as egregious abuses, but in framing such an injunction, he objected, they "have taken a position that crosses the boundary from adversary to enemy."12

To the relief of the academic library community, and to most educators aware of the case, Judge Evans rejected the presses' constricted view of "fair use," finding an infringement in only five of the 75 instances the plaintiffs presented. She explicitly repudiated the notion that the Guidelines for Classroom Copying defined "fair use," castigating the presses for trying to turn what Congress had recognized as a minimum statement of its boundaries into a maximum.13 Academic librarians celebrated the result, with only some reservations about Judge Evans' approach. But while there is no doubt that the rejection of the plaintiff's interpretation of "fair use" is a victory for the academic community as educators, although perhaps not as publishing scholars, there are elements of the judge's opinion that may lead to less welcome results.

Like almost every court that has evaluated the issue, Judge Evans rested her conclusions on four factors specified in the "fair use" provision of the Copyright Act: (1) the character of the use, (2) the nature of the work being used, (3) the amount and substantiality of work used, and (4) the effect of the use on the market value of the copyrighted work. She discussed her interpretation of those factors in an important, initial portion of her opinion and then applied them, one by one, to every instance of alleged infringement.

The judge held forcefully that the use of a work for educational purposes weighs "strongly" in favor of a defendant's claim of fair use. Judge Evans's analysis of the second factor is more problematic. Relying on a number of precedents and the language of the statute, she opined that the scope for "fair use" of factual, informative copyrighted works is larger than that of fictional, "creative" works. As a result, she consistently found that the second factor weighed in favor of "fair use." While historians might take umbrage at characterizing their work as less creative than fiction, the more important problem is that Judge Evans did not take into account the fact that most of the works assigned through GSU's e-reserves were specifically aimed at the academic rather than the general market. It is easy to imagine another judge holding that this aspect of the work tends to a more restricted scope for "fair use," while classroom use of a work of fiction is "transformative" (that is, for a completely different purpose than the author envisioned) and thus more likely to be fair.

In her analysis of the third factor (amount and substantiality), Judge Evans arrived at what legal analysts call a bright line. In works of ten chapters or more, the use of one chapter is fair. In works of fewer than ten chapters, it is not fair to use more than 10 percent of the work, counting front and back matter (notes, bibliography, index, etc.) as well as the main body of the work. When instructors stayed within these bounds, the judge found that the third factor weighed in favor of fair use; when they went beyond these bounds, she found the opposite.

Judge Evans applied these standards both to chapters in single-author works and to chapters—that is, individual essays—in compilations. In doing so, she rejected the presses' argument that each essay in a compilation should be considered a single work, and thus subject to only limited excerpts. But while she offered some reasons for rejecting the publishers' position, she relied mainly on the fact that they had not made this argument in their filings, where they had alleged infringement of the entire, compiled work rather than the individual essays. There is no way to know how future courts will respond when rights holders allege infringement of individual essays, as they surely will. Moreover, because no journal was among the plaintiffs, Judge Evans did not have to deal with the question of fair use of articles in periodicals. This is bound to be an important issue in the future, as scholarly journals, especially in the humanities, are as interested as book publishers in getting permissions revenue.

Judge Evans's analysis of the fourth factor—the effect of the use on the value of the copyrighted work—may prove to be the most important element of her opinion. The judge recognized that there was no likelihood that instructors would require students to purchase whole books in order to read a single chapter, nor that the students would buy them. Thus posting and downloading the digital copies had no bearing on the market for the book itself. But permissions revenue was another matter. While the loss of permissions revenue in each case might be small, if multiplied by similar uses across all of academia, it would be substantial, and precedent is clear that judges must take the potential effect on the whole market into account when deciding individual cases. But, Judge Evans reasoned, permissions revenue can be lost only if the presses in fact offer a convenient, reasonably priced system for giving permissions for digital copying of their works. Sage Publications did so, both through CCC and its in-house permissions service. But in the case of many of the works they claimed to have been infringed, neither Cambridge nor Oxford provided that opportunity. When they did provide it, Judge Evans found that the fourth factor weighed strongly against "fair use;" when they didn't, the judge found that it weighed in favor.

When Judge Evans applied her analysis to the individual cases, she found GSU's fair-use defense against infringement persuasive whenever instructors assigned e-reserves in amounts within the bounds that she specified in her analysis of the third factor. With the first two factors—the nature of the use and the nature of the work being used—always favoring "fair use," the fact that the third one did as well was decisive, even when the presses offered a convenient, reasonably priced system to secure permissions. When the amount placed on e-reserve exceeded her limits, however, the factors balanced each other out if permission for digital copying could be conveniently and reasonably secured, and the judge looked more closely to see where the overall weight of the factors lay, sometimes finding "fair use" and in five instances finding an infringement. The court—and academic observers—are still waiting to see what kind of remedy the presses propose in the case of the infringements. However, they are not pervasive and egregious enough to sustain the draconian injunction they initially sought.

We must also wait to see whether the presses appeal the decision. The AAUP's lawyer was furious that she lost the case, and bitterly criticized the decision at the AAUP's recent convention in Chicago. But the attendees are reported to have concluded that they need to mend their relationship with their academic colleagues, and this may portend an end to the litigation.14 Nonetheless, there is enough ambiguity and problematic reasoning in the decision to make it likely that the AAP and CCC, if not the AAUP, will try again, probably in a federal circuit more sympathetic to publishers' arguments, although they can no longer realistically hope to secure the narrow boundaries of "fair use" that they desire. Instead, they can aspire to an interpretation of the second factor (the nature of the work) that will make the issue between infringement and "fair use" a closer one. They can hope that a future judge will count individual essays and articles in compilations and journals as "whole works" subject to Judge Evans's 10 percent rule. If they succeeded in that endeavor, it would negate a good deal of what educators and librarians won under Judge Evans's analysis.

For now, Cambridge University Press v. Becker will lead many e-reserve administrators to tell instructors to post no more than a single chapter of any individual book to e-reserves, and no more than 10 percent of a book of fewer than ten chapters. If other courts follow Judge Evans's reasoning, that advice will provide a solid "fair-use" defense to infringement suits.

However, this outcome presents a "nightmare scenario" for university presses, and reduced royalty income for publishing historians. Even if they establish convenient, reasonably priced permissions systems in-house or through the CCC, they will not be able to secure permissions fees from e-reserves, except when instructors feel compelled to assign more than one chapter from a book. That is not likely to happen very often. Instructors and libraries are more likely simply to adjust their assignments to the legal rule.

This is an ironic outcome, because Judge Evans put so much stress on the availability of such permissions systems in her analysis of the effect of the use on the value of the work. If she had not found the other factors so clearly weighing in favor of fair use, her opinion would have accelerated the transition in scholarly publishing from relying overwhelmingly on book sales for revenue to relying more on permission fees. As a result, it is questionable whether Judge Evans's opinion can remain the final word. In our capacity as teachers, historians can rejoice that we have escaped the "nightmare scenario" presented by the publishers' demands in the Georgia State case. But it will be dangerous to rely on it to establish a right to place massive amounts of copyrighted materials on e-reserve at little cost to ourselves and with no remuneration to copyright owners. In a conservative legal age, judges will be sympathetic to the idea of charging universities and students modest amounts for the systematic use of copyrighted materials aimed at the education market, if the process is fast and convenient. It would be better for historians, other educators, and librarians to work with the AAUP to arrive at a system that serves all of us, rather than to rely on the courts to deliver what we want. And it will be better for publishing historians as well. Hopefully, a sobered AAUP has learned that lesson. Educators and librarians should not wait until a devastating court loss teaches it to us too.

Agreement on Guidelines for Classroom Copying in Not-For-Profit Educational Institutions with Respect to Books and Periodicals

Representatives of publishers and a small number of library and academic groups produced the “Agreement on Guidelines for Classroom Copying in Not-For-Profit Educational Institutions with Respect to Books and Periodicals” after the collapse of efforts to arrive at a broader consensus on language to be included in the 1976 Copyright Act. The House of Representatives committee report accompanying the law, and which included the text of the agreement, alluded to the Guidelines respectfully as “a reasonable interpretation of the minimum standards of fair use.” Its “brevity” section limits copying either to complete articles, stories, or essays of no more than 2,500 words, or to excerpts of no more than 1,000 words from any prose work or 10 percent of its total words, whichever was less. It allows the copying of only one illustration, cartoon, graph, or picture from any book or issue of a periodical. The “cumulative effect” section restricts the copying of any item, even within these limited bounds, to only one course anywhere in the school in any term; allows only two excerpts or only one complete article, story, or essay by any author; and permits only nine instances of such copying per class per term.

Although copyright owners have argued that the “Guidelines” indicate the boundaries of “fair use,” as the accompanying article indicates, they identify the minimum rather than the maximum use that can fairly be made of copyrighted work for educational, critical, and other purposes. See House of Representatives Report No. 1476, 94th Congress, 2d Session (1976), 68–71. The text of the Guidelines may be found also at copyright.gov/circs/circ21.pdf. The signatories to the agreement were:

• Ad Hoc Committee on Copyright Law Revision: Sheldon Elliott Steinbach
• Author-Publisher Group--Authors League of America: Irwin Karp, counsel
• Association of American Publishers, Inc.: Alexander C. Hoffman, chair, Copyright Committee

Michael Les Benedict, emeritus professor of history at The Ohio State University, is a member of the AHA Task Force on Intellectual Property and is the author of the pamphlet, A Historian's Guide to Copyright, published recently by the AHA.

Notes

1. Cambridge University Press v. Becker, 2012 U.S. Dist. Lexis 78123 (N.D. Ga. May 11, 2012). The opinion as filed, identified as Civil Action No. 1:08-CV-1425-ODE may be found online at <http://www.nacua.org/documents/CambridgeUPress_v_Becker_051112.pdf> (last consulted June 29, 2012). Also available at 2012 U.S. Dist. LEXIS 78123. The plaintiffs named the president and other officers of the university and library as defendants because the Supreme Court has held that the Eleventh Amendment to the Constitution affords state institutions sovereign immunity from such suits. The judge held that the plaintiff’s suit qualified under an exception to that rule that allows a narrow range of suits against state officers.

2.Kevin Smith, “A Nightmare Scenario for Higher Education,” May 13, 2011, Scholarly Communications @ Duke, Duke University Libraries <http://blogs.library.duke.edu/scholcomm/2011/05/13/a-nightmare-scenario-for-higher-education/> (Last consulted June 29, 2012). Smith is Scholarly Communications Officer, responsible for copyright and scholarly publication advice at Duke University.

3. 17 United States Code (USC) §107.

4. 17 USC §108(d).

5. 17 USC §108(g)(1), §108(g)(2).

6. 17 USC §108(g)(1).

7. L. Ray Patterson and Stanley W. Lindberg, The Nature of Copyright: A Law of Users’ Rights (Athens: University of Georgia Press, 1991).

8. Basic Books, Inc. v. Kinko’s Graphics Corp., 758 F. Supp. 1522 (S.D.N.Y. 1991); Princeton University Press v. Michigan Document Services, 99 F.3d 1381 (6th Cir. 1996).

9. See Sanford G. Thatcher, “Georgia State and (Un)fair Use: A Rebuttal to Kenneth Crews,” Against the Grain, Mar. 29, 2010 <http://www.against-the-grain.com/2010/03/v-22-1-from-the-university-presses-georgia-state-and-unfair-use-a-rebuttal-to-kenneth-crews> (last consulted June 26, 2012). Thatcher was long-time director of the Pennsylvania State University Press and served as president of the AAUP.

10. ALA, “L.Ray Patterson Copyright Award” <http://www.ala.org/advocacy/copyright/pattersonaward> (last consulted June 30, 2012).

11. Harper & Row v. Nation Enterprises, 471 U.S. 539 (1985). In this case the Nation magazine published excerpts from the manuscript of Harper & Row’s diaries of President Gerald Ford before its publication, scooping Time Magazine, to which Harper & Row had sold pre-publication rights. The moral indignation of the Court permeates its opinion.

12. Paul N. Courant, “Adversary or Enemy: A Publisher Lawsuit Crosses the Line,” Publishers’ Weekly, June 20, 2011 http://www.publishersweekly.com/pw/by-topic/columns-and-blogs/soapbox/article/47639-adversary-or-enemy--a-publisher-lawsuit-crosses-the-line.html (consulted June 26, 2012).

13. Cambridge University Press v. Becker, 2012 LEXIS 78123, at 85-88; original opinion as filed at pp. 56-59.

14. Steve Kolowich, “Mending Fences,” Inside Higher Ed., June 21, 2012 <http://www.insidehighered.com/news/2012/06/21/university-presses-debate-how-reconcile-libraries-wake-georgia-state-copyright#ixzz1ygOFWZwy> (last consulted July 1, 2012).