Leveraging exceptions and limitations for digital curation and online collections: the U.S. case

Patricia Aufderheide, paufder@american.edu

School of Communication, American University, US

Libellarium, IX, 2 (2016): 49 – 58.

UDK: 027:347.78(7/8)=111

DOI: http://dx.doi.org/10.15291/libellarium.v9i2.247

Conceptual paper


Librarians wanting to use digital affordances for their patron’s and public benefit have increasingly found themselves frustrated by copyright law designed for a pre-digital era. In the U.S., this frustration has driven the nation’s most prestigious library group, the Association of Research Libraries, to explore the utility of the major exception to copyright monopoly rights, fair use, in order to accomplish basic curation and collection goals in a digital era. The ARL’s efforts to clarify how libraries can employ fair use has resulted in sometimes-dramatic changes in how work is done, and has permitted innovation at some universities. Its approach demonstrates the power of consensus in a professional field to permit innovation within the law.

KEY WORDS: library, copyright, fair use, user rights, digital, curation, collection, norms


Changes in library practice in the U.S. since the creation of the Code of Best Practices in Fair Use for Academic and Research Libraries in 2012 demonstrate the power of the user right of fair use, the value of education in changing user access to this right, and the importance of consensus within the professional community in legitimating it.

Fair use is structured in copyright law (in section 107 of the Copyright Act) as a general permission to employ others’ copyrighted work without licensing it, if the use generates new cultural expression and uses an appropriate amount of the original to permit that new expression. This “transformative” use thus does not step into the established market for the original work. In this way, a fair use matches up well with the four “factors” the law requires one to consider (as well as leaving the door open for other considerations): the nature of the original work, the nature of the new work, the amount taken, and the effect on the market.

This user right is structured by the purpose of copyright law in the U.S. There, copyright policy is designed to stimulate the production of new culture, both by offering creators a limited monopoly over the expression of their ideas and also by limiting that monopoly sufficiently to permit new culture to be generated on the platform of existing culture (Patterson and Lindberg 1991, Kaplan 2005, Netanel 2008, Hyde 2010, Netanel 2011). More than a century of industry pressure to extend copyright has resulted in monopolies—that is, governmental permission to control access in various ways to a copyright one holds--that are longer and stronger than ever before. While large, content-holding media companies argue that this is crucial to stimulating creativity, there is also much evidence of creative choice that has been limited by extending these monopolies ((Boyle 1996, Vaidhyanathan 2001, Lessig 2004, McLeod 2005, Aoki, Boyle et al. 2008, Boyle 2008, Tushnet 2009, Sinnreich 2010). Thus, fair use has increasingly become important as a balancing feature of copyright monopoly.

The hallmark of the user right of fair use is transformativeness, a key concept for librarians. The centrality of transformativeness was recently made clear judicially. In 2015, Appeals Court Judge Pierre Leval wrote the decision for the 2d circuit in Authors Guild v. Google (https://assets.documentcloud.org/documents/2461545/agvgoogle.pdf), synthesizing current understanding of fair use. The case involved whether Google Books’ copying of sections of books still under copyright was fair use. The court’s decision was unanimously yes. His decision went further to lay out the way judicial reasoning works and should work on fair use. He wrote, among other things, that “The ultimate goal of copyright is to expand public knowledge and understanding…Thus, while authors are undoubtedly important intended beneficiaries of copyright, the ultimate, primary intended beneficiary is the public...” Google’s copying of entire books in order to provide snippets of them on demand to the public, he explained, was a clearly transformative purpose, and transformative purposes meet the standard of serving the public by generating new culture. This transformative purpose does not intrude on the book market for books written by Authors Guild authors.

Applying fair use to libraries in the U.S.

Leval’s concise and elegantly argued explanation, placing transformativeness at the center, built upon three decades of growing consensus in the U.S. about interpretation of fair use. This interpretation was critical to librarians’ use of it.

Fair use, however, as a right can always be challenged, and in the U.S. high fines—statutory damages—can at least theoretically be invoked for copyright infringement. Although some libraries in the U.S. are exempt from such fines, as state or federal institutions, with “sovereign immunity,” others are not. Librarians tend to the risk-averse, and are highly sensitive to the judgment of their peers. Therefore, the Association of Research Libraries—an association of 124 research libraries at major research institutions in the U.S. and Canada--believed that, however useable fair use had become, it would be important to establish best practices in employing fair use, to provide assurance to individual librarians and their staffs that they were operating within acceptable field practice. (Its Canadian members cannot apply this logic directly, since Canada does not have this exception; however, Canada’s recent copyright reform included an expansion of fair dealing that amplified the range of practice for research librarians.) This would permit fair use employment to fall more easily within the bundle of expressive practices that librarians were already more comfortable with, such as their First Amendment rights to criticize both private and public officials publically.

The ARL was aware that other professional groups had been able to expand their employment of fair use, and extend their practices to digital environments, with best practices codes (Aufderheide and Jaszi 2011). Such codes are not sets of rules, but guides to sound reasoning about fair use in the most common situations in which fair use is typically employed in the field. These best practices documents have not only proven useful to practitioners, but have never received any formal criticism from large copyright holders. Furthermore, they have an influence on judicial action. This is not only because they demonstrate a link between what the law permits and what the goals and missions of professional practice are, but also because when deciding cases, judges inevitably refer back to the actual cultural practice of a field (Madison 2004). This is simply because making a fair use decision is grounded in specific cultural practice and in particular what new expression you are trying to achieve.

Common problems with copyright law

The ARL, with funding from the Andrew Mellon Foundation, worked with American University to clarify the needs for and best practices in library employment of fair use. Deep interviews we conducted with 65 research librarians, summarized in a report circulated to ARL members, unearthed many ways that librarians found their mission frustrated by not understanding and misunderstanding fair use. They found themselves employing standards for e-reserves that made no sense—for instance, only allowing teaching material to be posted under fair use every third year, or creating arbitrary quantitative limits. They looked helplessly at shelves of videocassettes, knowing that most classrooms no longer had VCRs and many of the videocassettes held material no longer (or never) commercially available in other forms. But they were not sure they could copy them to viable formats under fair use.

They delayed digitization initiatives for anything that was not in the public domain; prioritizing public-domain materials often made no sense from a mission perspective, either. They resisted letting students use copying software for audio and video, even for projects that legitimately employed fair use. They found themselves telling scholars that they had to visit physical sites, even at great cost, in order to consult documents that could have been digitized. They had no idea what to do with scholars who wanted to scan entire databases for patterns—one of several research practices, sometimes called nonconsumptive analysis, only possible because of digital affordances. When they put up exhibits, both in their buildings and online, they agonized when they did not own the copyright to the material, often taking months of precious professional staff time to solve permissions issues, and sometimes simply abandoning projects. They even found themselves giving disabled students suboptimal access to materials, because of their copyright anxieties. Moreover, they were often supposed to be the experts—several bemoaned playing the role of “copyright police”—to people who wanted exact certainty, when they themselves were confused about the possibilities (Aufderheide, Butler et al. 2010).

They treated different media differently—they were most comfortable with fair use in text, least comfortable in video—even though fair use applies to all media alike. Indeed, in the case of video, librarians appeared to have listened most closely to vendors of independent video, who had vociferously and successfully waged a campaign with buyers of video for libraries to accept a high-priced library rate for their product, arguing that it was a way of supporting independent filmmakers and film artists. Library support was weakening for this position, which resulted in a vigorous assertion by vendors of the library rate as a moral imperative. In the process, fair use was again an innocent bystander dragged into the discussion. Vendors feared that libraries’ employment of fair use, including re-use of previously purchased material on e-reserves for teaching, could jeopardize their business model. They pressured librarians to “protect independent film” by not employing fair use, as well as supporting the library rate.

Some librarians held out wistful hope than open-access campaigns could over time convince more and more scholars to work with open-access publishers and donate their work to institutional repositories. But open-access platforms were not standardized, and most scholars worked within commercial publishing to circulate their work and get recognized for promotion and tenure. While moral panic or, to use another term, data moralism (Patry 2009, Peters 2016) was an undercurrent among the research librarians, they were not in the main a sentimental group. They worked with many vendors to assemble their resources, and accepted copyright holders’ claims and the need for copyright balance.

As happened with other groups, for most of the problems they encountered, librarians themselves were the obstacle. Their colleagues and general counsels would often be willing to work with them to help them reach their mission, but they themselves suffered from the most common consequence of fair use doubt: self-censorship.

Librarians read the alarming report that ARL issued as it circulated not only within ARL but in ARL’s related organizations, library associations representing K-12 librarians and academic libraries. They eagerly embraced the project of articulating their best practices.

Creating a code of best practices in fair use

They used the process employed to make previous codes (Aufderheide and Jaszi 2011). They held a series of small-group, quiet meetings around the country. Each meeting would involve people from different institutions and different parts of library practice. One or all of the three facilitators, Profs. Patricia Aufderheide, Peter Jaszi and Brandon Butler, attended all the meetings, triggering discussions by creating scenarios and hypotheticals grounded in the most common stories they learned from the report. Each meeting built on previous meetings, building upon consensus and refining understanding of where the field professionals believed the limits of fair use would be for any given practice, and how they would know when they had arrived there.

Creating the code was a process tucked into the annual rhythm of convenings and conferences that this collaborative group of professionals already observed. Knowing they were making the code in the shadow of a major lawsuit, between academic publishers and Georgia State University (Cambridge University Press v. Patton) about how much material was appropriate to upload for e-reserves under fair use, only made them more eager to assert their best practices.

The lawsuit concerned whether the university had infringed on copyright holders’ monopolies by uploading excerpts from academic books to electronic reserves. After years, a trip to appeals court, and a remand to circuit court, the case ultimately reinforced the library’s fair use rights and furthermore applied only to Georgia State--which had immunity from penalties in any case (Cox 2016). But at the time, librarians believed that it would be important to assert community best practices.

The result was the Code of Best Practices in Fair Use for Academic and Research Libraries, available at arl.org/fairuse, which identified eight common situations:

  • Posting all forms of media to e-reserves and online course platforms generally;
  • Physical and virtual library exhibits and exhibitions;
  • Digitizing to preserve at-risk items (e.g. VHS tapes);
  • Digital collections;
  • Reproducing material for disabled patrons;
  • Institutional repository copyright practices;
  • Making databases available for non-consumptive (e.g. digital megasearches) research;
  • Harvesting and making available material from the World Wide Web.

For each of these uses, librarians identified both several ways to identify the limit of fair use (“limitations”), and some measures that would not only be best practices but go beyond field-wide standards, showing an excess of care (“enhancements”).1

There was widespread enthusiasm in the community, and the Code quickly drew endorsements from the major library groups in the US: The American Library Association, The Art Libraries Society of North America, The Association of College and Research Libraries, The College Art Association, The Consortium of College and University Media Centers (CCUMC), and The Music Library Association.

The endorsement by CCUMC was particularly notable because this relatively small organization, representing the people who run audio-visual units in higher education, had earlier acceded to an aggressive copyright-holder initiative to create a stringent and unhelpful document that discouraged fair use. The CCUMC guidelines had never had significant input from library stakeholders. The fact that CCUMC members effectively control access to audio-visual resources at many institutions meant that many universities and colleges had, in practice, a no-fair use policy. CCUMC board members contacted ARL once the Code was created, and asked facilitators to keynote their national convention. This prepared their membership for the board’s subsequent decision to drop the guidelines and endorsement of the Code.

Almost immediately, librarians and archivists put the code to work:

  • At the University of California, Berkeley, a diverse and largely unlabeled collection of historic photographs of the region were uploaded to a website. Information flooded in as people identified landmarks and even authors. The library made it easy to send complaints, but so far, according to an online interview with the library’s director Tom Leonard (http://www.arl.org/focus-areas/copyright-ip/fair-use/code-of-best-practices/2927-arl-member-library-directors-on-putting-the-code-to-work#.VyvWdz82vzk), the librarians have only received praise and information useful for metatagging.
  • At the University of Colorado-Boulder, the university decided it would be uncontroversial to copy VHS cassettes to digital-file formats if no commercial version was available (Butler 2014).
  • A task force at University of California, Santa Barbara, created standards for the university system that allows theses with unlicensed copyrighted material used under the terms of the librarians’ code to be included in institutional repositories.
  • University of California, Los Angeles opened up a digital portal to its video news collection (http://newsscape.library.ucla.edu/).
  • American University’s library changed policies regarding how long video clips could be up on an e-reserves site, moving away from employment of the highly limited TEACH Act to an expectation that faculty set the length of time within the course term, as appropriate to learning needs (Nancy Davenport, personal communication, March 5, 2016).
  • CCUMC encouraged its many member institutions to change their policies on audio-visual use, including posting of audio-visual material to course platforms, copying of VHS tapes, and breaking of encryption on DVDs to allow access to material that would be employed under fair use (http://archive.cmsimpact.org/blog/fair-use/ccumc-adopts-librarians%E2%80%99-code-retires-previous-fair-use-guidelines).

They joined other archivists, who were already using best-practices codes in other fields. For instance, dance archivists had created a code of best practices to address, primarily, the challenges of exhibiting and making digitally available to researchers material—especially audio and visual material—on the history of dance. One of the first projects was the interactive online dance archive created by Jacob’s Pillow (http://danceinteractive.jacobspillow.org/). This interactive archive allows any viewer to access clips from historic dance trends, with accompanying information about both the dancer/choreographers and the trends. The viewer can search by trend, name, or type of dance.

An archivists’ code

The libraries’ code raised new questions for collections managers and archivists, who had specific questions about the many orphan works in their collections. Archivists believed strongly that the point of having a collection was to enhance knowledge, and at the same time were acutely aware that they held in trust collections given to them for safekeeping. In the past, archives have simply let on-site visitors look at such material, without concern about copyright. But the digital environment was different, because of the need to make a copy of the work to get it into digital formats, and by making it accessible online, creating the possibility that, one way or another, someone else could copy it.

A group of archivists, curators, museum staffers and collections managers, who called themselves collectively “memory institutions,” gathered together at American University, with help from The Sloan Foundation. They launched a nation-wide process of deliberating in small groups, which resulted in the Statement of Best Practices in Fair Use of Orphan Works for Libraries & Archives (http://archive.cmsimpact.org/fair-use/best-practices/statement-best-practices-fair-use-orphan-works-libraries-archives).

There has been no pushback from corporate copyright holders to either the terms of or the employment of this or other codes. Furthermore, related litigation has been enormously supportive of the logic of the fair use best practices codes. Several non-library fair use cases re-emphasized the applicability of fair use in situations where transformativeness might be in dispute and/or complete copies were made. A decision about artistic appropriation (Cariou v. Prince) was decided in favor of the appropriating artist, who refused to suggest that he was in any way providing a critique or commentary on the previous work. In Swatch v. Bloomberg LP, courts decided that merely replaying a private conference call for business-news subscribers was fair use. And in White v. West Publishing Corp., the court found that copying and making available for subscribers briefs submitted in lawsuits added value. He also noted the policy of the U.S. Patent and Trademark Office, which consider inclusion of relevant documents, including published work, in a patent application as a fair use (Butler 2015).

The Authors Guild v. HathiTrust decision was also resoundingly pro-fair use. In HathiTrust, publishers sued the makers of an online library, including universities and Google, for threatening the markets of their authors. Courts found instead that limitations HathiTrust had provided did not exceed fair use, and that libraries could make complete copies of everything in their entire collection for the purpose of search and for disabled access; it did not address other kinds of access. The decision was grounded, crucially, in the transformative argument for fair use:

In sum, the court found that a use is transformative if it has a different purpose or function from that of the original work. If the use is transformative, any economic harm such as lost licensing revenue is irrelevant for purposes of the fair use analysis. Further, the user may copy as much of the work as necessary to achieve his or her transformative purpose. (Band 2015)

Finally, there was the previously discussed Google Books settlement, which established with exceptional clarity the right to make available contextualized material from digitized collections openly on the Web.


The U.S. library experience with fair use is only beginning, but it has already made a difference in the practices of major research libraries, which are standard-bearers in the U.S. library community. This approach could provide inspiration for librarians in other countries.

Most other nations do not have fair use, nor is their copyright policy grounded in the same Constitutional logic that copyright is in the U.S., to spur the creation of culture. But copyright regimes internationally all contain some user rights--exceptions and limitations to the copyright monopolies prescribed by law. Many of them are untested by any significant body of case law. Some are in the general realm of fair use, such as right of quotation—a broad exception permitting the borrowing of some limited material for a wide variety of purposes. Others are specific to certain practices, such as education or librarianship.

In all cases, however, articulating in what way existing law does or does not serve the core mission of librarians would be useful documentation both for the library community and for policymakers. In some cases, such research could discover a gap between what the law actually permits and how people are using, or not using, it. Closing that gap could expand possibilities of functioning in a digital environment, while librarians continue to defend, to policymakers in charge of copyright reform, the public and the future publics for whom they preserve and maintain archives of material.


Aoki, K., J. Boyle, and J. Jenkins. 2008. Bound by law? : tales from the public domain. Durham, NC: Center for the Study of the Public Domain, Duke University Press.

Aufderheide, P., and P. Jaszi. 2011. Reclaiming fair use: How to put balance back in copyright. Chicago: University of Chicago Press. https://doi.org/10.7208/chicago/9780226032443.001.0001

Aufderheide, P., B. Butler, and P. Jaszi. 2010. Challenges in employing fair use in academic and research libraries. Washington, D.C.: Association of Research Libraries, Center for Social Media, and Program on Information Justice and Intellectual Property: 23.

Band, J. 2015. "What does the HathiTrust decision mean for libraries?" Research Library Issues 285: 7-13.

Boyle, J. 1996. Shamans, software, and spleens : law and the construction of the information society. Cambridge, Mass., Harvard University Press.

Boyle, J. 2008. The public domain : enclosing the commons of the mind. New Haven, CT, Yale University Press.

Butler, B. 2014. "UC Boulder takes VHS tapes back to the future with the #Librarianscode." ARL Policy Notes http://policynotes.arl.org/?p=213

Butler, B. 2015. "Fair use rising: Full-text access and repurposing in recent case law." Research Library Issues 285: 3-6.

Cox, K. 2016. "A deeper dive into the New Georgia state decision." ARL Policy Notes http://policynotes.arl.org/?tag=georgia-state

Hyde, L. 2010. Common as air : revolution, art, and ownership. New York, Farrar, Straus and Giroux.

Kaplan, B. 2005. An unhurried view of copyright, republished (and with contributions from friends). Newark, LexisNexis Matthew Bender.

Lessig, L. 2004. Free culture : how big media uses technology and the law to lock down culture and control creativity. New York, Penguin Press.

Madison, M. J. 2004. "A pattern-oriented approach to fair use." William and Mary Law Review 45, 4: 1525-1690.

McLeod, K. 2005. Freedom of expression : overzealous copyright bozos and other enemies of creativity. New York: Doubleday.

Netanel, N. 2008. Copyright's paradox. Oxford; New York: Oxford University Press. https://doi.org/10.1093/acprof:oso/9780195137620.001.0001

Netanel, N. 2011. "Making sense of fair use." Lewis %26 Clark Law Review 15, 3: 715-772.

Patry, W. 2009. Moral panics and the copyright wars. New York, Oxford University Press.

Patterson, L. R., and S. W. Lindberg. 1991. The nature of copyright : a law of users' rights. Athens: University of Georgia Press.

Peters, J. 2016. The idealist : Aaron Swartz and the rise of free culture on the Internet. New York: Scribner.

Sinnreich, A. 2010. Mashed up : music, technology, and the rise of configurable culture. Amherst: University of Massachusetts Press.

Tushnet, R. 2009. "Economies of desire: Fair use and marketplace assumptions." William and Mary Law Review 51: 513-546.

Vaidhyanathan, S. 2001. Copyrights and copywrongs : the rise of intellectual property and how it threatens creativity. New York, New York University Press.

1 Unfortunately, the result of creating enhancements was not always what they intended. They had hoped to include some practices that would strengthen even further a fair use claim; but the enhancements have sometimes confused people about best-practices expectations.

Article Metrics

Metrics Loading ...

Metrics powered by PLOS ALM


  • There are currently no refbacks.

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Libellarium (Online). ISSN 1846-9213 © 2008


Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.