As technology gallops forward to find new products and digital sites for consumers to access copyrighted material, the copyright holders wage wars in court to cry “foul!” in the form of infringement suits. The sellers typically counter that the copying qualifies as  “fair use” under a statutory defense permitting certain infringements under several factors.  In the 1980s the Supreme Court held videocassette recorders are fair use despite enabling the widespread copying of broadcast copyrighted material; Napster later failed with the defense.  In the latest ground war, a federal judge has now ruled[1] that Google’s massive “library scan” project to create digital copies of more than 20 million books is fair use, freeing Google to move forward and continue marketing it to users.  The essential rationale is that the project provides educational and noncommercial (primarily library) uses that greatly outweigh Google’s indirect potential for profit-making. The Guild promptly filed an appeal.  The case illustrates the lightning-like unpredictability of the defense in the courts and the continuing emphasis on fair use protection for so-called “transformative” uses of protected material.

THE METAPHYSICS: THE FAIR USE DEFENSE

In one form or another, “fair use” has been a fixture of copyright law for centuries.  The Supreme Court has recognized the doctrine is a constitutional necessity, imbued with First Amendment concerns.[2]  It is a true defense, absolving the infringer of liability and freeing continued publication.  Like any defense the alleged infringer bears the burden of proving it.  The courts have struggled with a workable definition that produces predictable outcomes.  Congress abandoned any attempt at a definition in the 1976 Copyright Act, but expressly recognized the defense and listed non-exclusive factors to guide its application. Still, courts have varied so much in their applications of it that some have called it “metaphysics”; the Ninth Circuit recently observed dryly:  “We acknowledge the porous nature of the doctrine but we are obliged to make sense of the doctrine and its predicates.”[3]

What is known is the four statutory factors set out by Congress “to be considered” in determining whether a use is “fair”:

              (1) the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes;

(2) the nature of the copyrighted work;

(3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and

(4) the effect of the use upon the potential market for or value of the copyrighted work.[4]

A preamble cites examples of uses that are most appropriate to a finding of fair use: “criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship or research.”  These uses are “illustrative, not limitative.”[5]  Addressing the enumerated factors, the Supreme Court has also held the factors are “general guidance,” are not exclusive but are “especially relevant,” and are to be weighed together in light of the “purposes of copyright.”[6]

The courts are thus given significant leeway in determining what each factor means, and its weight in any given setting. The leading treatise, Nimmer on Copyright, observes:  “[E]ven taken alone, each of the factors is defined only in the most general terms, so that courts are left with almost complete discretion in determining whether any given factor is present in any particular case.”[7]  The same authors elsewhere observe that the factors, when combined with the emphasis on case-by-case-adjudication, “tend to degenerate into post-hoc rationales for antecedent conclusions, rather than serving as tools for analysis.”[8]

GOOGLE’S SCAN PROJECT

In 2004 Google launched its “Google Books Program” which consists of two parts: the “Partner Program,” which allows hosting and display of material provided by publishers and other copyright holders; and the “Library Project,” which eventually entailed the digital (OCR) scanning of books in the collections of huge libraries like the Library of Congress, the New York Public Library, and those of a number of universities.  On the target scan list are written works of all types, including novels, biographies, children’s books, textbooks, instruction manuals, treatises, dictionaries, cookbooks, poetry books, and memoirs. According to Google’s analysis, the overwhelming bulk (93%) of the works are nonfiction.  Google’s court filings also stated the great majority of the works are out-of-print.

The Partner Program involves material the copyright holder has approved for browsing, and only on the parts so approved. The “partner,” however, provides the entire work to Google for scanning. As of 2012 Google had scanned 2.5 million works with the consent of the copyright holders.

The Library Project is far more extensive and is done without the consent of the copyright holders. Google has entered into agreements with the participating libraries which permit the library to obtain a digital copy of each work it allows to be scanned from its collection.  Google in turn retains copies of the scans and will permit searches under separate terms.  Under this project about 20 million books have been scanned in their entirety in multiple copies stored in Google’s servers and back-up tapes. Some libraries limited the scan authorization to public domain works but some allowed scanning of in-copyright material.  A library, however, may not obtain a digital copy of a book scanned form another library’s collection.

The access for search purposes, presumably available to library visitors and to those allowed access to Google’s aggregate database, is limited to providing “snippets” while “blacklisting” certain paragraphs within the search result, thus preventing tricks that would allow the user to “attack” the work and reconstruct it with serial snippets.  The user may access an “About the Book” summary that has links to online sellers like Barnes & Noble and Amazon.com.  According to amicus briefs filed by library and research association groups, the project facilitates access to vast libraries including out-of-print books and assists word searching, sharing of information between libraries, and linguistic analysis of works, among other features not previously available.

Google does not sell the scans or snippets and does not run ads on the “About the Book” pages.  Its commercial benefit is indirect, in the hope that users are drawn to Google’s websites by accessing the Google Books project.  Google apparently never denied a profit motive in the project but did point out the many educational and research benefits of the applications.

THE SUIT

The Author’s Guild and three individual authors filed a class action in 2005 for copyright infringement, alleging Google scanned their works and made them available for search without permission.  Presiding is Judge Denny Chin of the U.S. District Court for the Southern District of New York.  The court rejected a 2011 settlement as unfair in “rewarding [Google]  for engaging in wholesale copying of copyrighted works without permission….”  The court granted class certification, and in an appeal of that order the Second Circuit remanded for consideration of the fair use defense before final ruling on the class issues.  Google promptly moved for summary judgment in the fair use defense.  It filed supporting papers detailing how the scan project works, its search capabilities and limitations, and the lack of direct profit to Google through sales of scanned material or otherwise.  Various library associations and research groups filed amicus briefs urging adoption of the fair use defense, touting the benefits to library work and scholastic research.

THE RULING: WHEN IS A “WHOLESALE SCAN” OF TWENTY MILLION BOOKS A FAIR USE? 

 Judge Chin granted summary judgment to Google and found fair use as a matter of law.  Addressing the statutory factors, he first addressed the “purpose and character of the use” and noted  that commercial uses that merely replicate the original in content, function and market are more likely to weigh against fair ones, but if they are “transformative” of the original, they can tilt this factor the other way. The authority cited, however, was not closely in point—the Supreme Court’s Campbell opinion finding fair use in rap group 2Live Crew’s replication for parody oflarge parts of the original of Roy Orbison’s song “Oh Pretty Woman.”  Judge Chin noted the important “transformative” research tools like “data mining” and “text mining” enabled by the scan project, among others educational applications.  He quickly downplayed Google’s acknowledged commercial motive, saying the scans were not directly sold and that the most consumer-friendly aspects—the book summaries—are not adorned with ads.  He concluded the first factor “strongly favors a finding of fair use.”

The court found the second factor – the nature of the copyrighted work– also weighed in support of fair use.  Although works of fiction are entitled to greater protection than nonfiction, here the vast majority of the scanned works are nonfiction, and the protected works are already available to the public (and are not, for instance, unpublished manuscripts).

On the third factor—the amount and substantiality of the portion used—Google’s position was weakest.  Admittedly it scans the entirety of the books verbatim.  Judge Chin found this is not always dispositive, but simply another factor to consider. Further, he noted that full-text scanning is critical to the search aspects of the project, and that Google limits the amount of text displayed in response to a search.  On balance, he found this factor weighs “slightly against a finding of fair use.”

The fourth and final factor is the effect of the use upon the potential market for the copyrighted work.  The court rejected the plaintiffs’ argument that the scans would serve as a “market replacement” for the books, noting that the scans merely supplement the original works that are already in the library in question, since that library already provided it to Google to scan.  To the argument that clever users could string search terms together to collect snippets that could reconstruct the entire original, he pointed out the implausibility that anyone would go through that kind of effort just to get a scanned work, and cited the Google technology that blocked certain paragraphs from view in search snippets to ensure users could not “attack” the scanned work to rebuild it with snippets.  Instead, in the court’s mind the scans would be predicted to enhance the value of the originals by making them more widely known to users who then may be impelled to purchase them from a bookseller.  Judge Chin observed: “Many authors have noted that online browsing in general and Google Books in particular helps readers find their work, thus increasing their audiences.” He thus concluded the fourth factor “strongly favors fair use.”

In the overall assessment including consideration of non-statutory factors, Judge Chin summarized the many salutary features of the project and the new tools it presents to researchers who now have full-text searchability available to them for the first time.  It also can make works better known and allow them to be accessed in areas that currently may not have access to the printed works.  The court then went further and cast the positive features as universal: “Indeed, all society benefits.”

He then extended the fair use finding to the acts of making the scanned works available to libraries, as the libraries’ use is transformative and they use the scans in entirely different ways (for text-searching, primarily) than they do the originals.  This also absolves the libraries of copyright infringement, and Google for secondary liability for the libraries’ actions.  Google was thus granted summary judgment on all claims.

COMMENT: WHAT HAVE THE AUTHORS LOST?

The fair use analysis as always is difficult and here the many new uses offered by the scan technology readily supports a conclusion that the uses are “transformative.”  The courts need to be cautious in focusing exclusively on whether the new use changes the form of the original without adding new functionality to the material.[9]  Here the issue was not so much that the scan project “transformed” the books, but that they did so in a way that makes the new form of the works usable by others in ways that are entirely different from, and far more versatile than, the original.  Google prevailed because the scans not only are themselves different (of course they are—they are in digital format and the originals are not), but provide users with search capabilities and specific educational and library applications that the originals could never provide.  It is this new functionality of the form, and not merely its different shape, that makes the new use “fairer” especially if Google has lessened its profit-making potential by refraining from direct sales and product-attached advertising.

Nimmer has pointed out that making a film that is entirely based on a copyrighted novel “transforms” the original into an entirely different medium and art form, but nonetheless it will generally be an infringement unprotected by fair use because it has deprived the owner of the derivative right to adapt the original to that form.[10]  Here the authors in this sense have been deprived of their right to have scans made of their works and provide them to libraries and others on their own terms.  Judge Chin, singing the praises of the purely noncommercial benefits of the project, may have found fair use in any event, but to imply the authors have lost nothing is inaccurate.   New functionality combined with limited direct profit-making potential can amount to fair use but the analysis should identify and address the full rights and interests of both sides.

In the end Google hopes its scan project follows the same fair use path as VCR taping: in the Sony case the Supreme Court found fair use while evading the “transformative” analysis by characterizing the tapers as acting more commonly for the convenience of “time-shifting” viewings of programs rather than creating copies that displaced the originals free of royalty charges (pushing aside that the devices enabled the building of tape libraries that deprived copyright holders of rental royalties).  The more Google and the courts follow that lead and depict the scanned books as allowing new, different, non-commercial, and beneficial uses of the scanned books rather than merely enabling free replacement reads, the closer to “fair” the use becomes.

So far, Google’s arguments have won the day for its massive and unprecedented scan project.  With the ruling the parties might renew settlement discussions and the court ostensibly will be more open to approving a settlement than it was in 2011 when it rejected a proposed settlement as “unfair” because of the “wholesale copying” it permitted; if the scans are immune from infringement liability, those concerns are no longer in play.  In the meantime an appeal is imminent.  Regardless of outcome it is not expected to provide much doctrinal clarity in this troubled niche of copyright law.