In section 107 of the Copyright Act, Congress codified the defense known as “fair use,” under which qualifying, unauthorized uses of copyrighted works “for purposes such as criticism, comment, news reporting, teaching …, scholarship, or research [are] not … infringement[s] of copyright.”1 Scholars have described fair use as a remedy for “market failure,”2 and therefore a temporary substitute for functioning markets;3 a doctrine slouching toward irrelevance4 and even death5 as markets become more sophisticated; a tool for advancing social goals when a finding of infringement promises to produce “bad results”;6 and a doctrine that is “empty” of substance7 and therefore “dangerous”8 because it creates the illusion that there are limits to an increasingly unlimited entitlement. In other words, copyright scholars cannot agree on what, exactly, the fair use defense is for. What hope is there for the courts?
Nowhere is this disagreement more apparent than in the dispute over the “Google Library Project,” as a part of which Google has proposed to scan every book owned by four university libraries and the New York Public Library. The resulting digital images would constitute a powerful research tool: Not only would the images be searchable, by keyword, but researchers also could view the results of their searches online, in image form—along with bibliographic information enabling them to purchase (or borrow) the books of the most interest. In short, “[t]he Library Project [would] make it easier than ever before for users to locate the wealth of information buried in books.”9 Some of those books would be copyrighted, and so Google has announced that if a search produced a “hit” on a copyrighted work, researchers could view only a few sentences from that work in the form of “snippets” surrounding the search term. (If a search revealed a work in the public domain, researchers could browse the work in its entirety.)
When Google announced its project, authors and publishers objected, arguing that Google would be engaging in repeated acts of infringement by engaging in the wholesale copying (scanning) of works in which it did not own the copyrights. This was a valid objection, for the Copyright Act, in section 106, gives copyright owners the “exclusive right[ ] … to reproduce the copyrighted work in copies” and to authorize others to do the same.10 The word “copies” is defined broadly, in section 101, to include any “material object[ ] … in which a work is fixed by any method now known or later developed.”11 Because computer memory (whether volatile or nonvolatile)12 is a “material object,” digital scans of books stored in computer memory are “copies” for the purposes of section 106. In response to these objections from authors and publishers, Google proposed a change to its policy under which copyright owners who did not wish their works to be scanned could “opt out” by November 1, 2005, the date on which Google planned to begin the expensive process of digitization. Unsatisfied with this solution, a group of authors sued Google on September 20, 2005. A group of publishers filed a similar lawsuit a month later, on October 19. The cases are pending in the United States District Court for the Southern District of New York.
Although Google might argue that its conduct does not constitute prima facie infringement, the fate of its library project is likely to depend on the application of the fair use defense. Section 107 of the Copyright Act instructs courts to decide whether an accused use is “fair” by evaluating four statutory factors,13 of which the Supreme Court has placed the most emphasis on the first (the commerciality of the use) and the fourth (the effect of the use on the market for the original). It is easy to see why: Commercial uses are more likely than noncommercial ones to compete with sales of the copyrighted work. Any such competition is likely to result in lower prices and reduced market share for everyone, which market effects, in turn, are likely to diminish the profits of those erstwhile monopolists, the creators (and their assigns). Faced with the prospect of earning lower profits, at least some creators are likely to forego the act of creation in favor of other, more profitable pursuits, thus leading to a decline in the number or the quality of works created. Nobody wants that.
The fair use defense would be easy to apply if commercial uses usurped the market for the copyrighted work, while noncommercial (“nonprofit educational”)14 ones did not. But there are two problems with such a test: First, there is no such thing as a truly noncommercial use, making every use commercial and therefore suspect. One might think that scholarly activities, at least, would be comfortably on the gratis side of the line. Yet courts have found even the activities listed in section 107—including scholarship—to be commercial when users gain an “indirect economic advantage” by failing to pay the copyright owner for a license, thus depriving him or her of potential licensing revenues.15 As a number of scholars have noted, this test is circular. If depriving the copyright owner of licensing revenues were enough to make a use “unfair,” then the fair use defense would be no defense at all, for by definition, the fair use defense comes into play only when a defendant fails to pay for a license. Recognizing this,16 courts have asked whether the defendant has deprived the copyright owner of licensing revenues only in “traditional, reasonable, or likely to be developed markets.”17The problem with this inquiry, however, is that copyright owners themselves can define whether markets for their works are “likely to be developed” by developing those markets themselves. In other words, copyright owners themselves can define away the “market failure” for which fair use is the remedy.
The only exception to this rule appears to be that “‘[c]opyright owners may not preempt exploitation of transformative markets,’”18 which reveals the second problem with using commerciality as a test of fairness: As the Supreme Court warned in Campbell v. Acuff-Rose Music, Inc.,19 “the mere fact that a use is educational and not for profit does not insulate it from a finding of infringement, any more than the commercial character of a use bars a finding of fairness.”20 In other words, commerciality does not appear to be dispositive of fairness either way. The “purpose and character” of the use in Campbell was a Two Live Crew parody of the Roy Orbison classic “Oh, Pretty Woman,” but it was commercial nonetheless: The song was being offered for sale in music stores nationwide, and Two Live Crew (or more likely, its record company) was making money. Whether the rap group also was competing with Roy Orbison and his music publisher was another question. Although it remanded the question to the district court, the Supreme Court did not seem to think so. “As to parody pure and simple,” it wrote, “it is more likely that the new work will not affect the market for the original in a way cognizable under this factor ...”21 Why not? Because a parody is transformative, not competitive. It does not “supersed[e]” the original, but instead, it “adds something new, with a further purpose or different character, altering the first [work] with new expression, meaning, or message.”22 According to the Court, “the goal of copyright, to promote science and the arts, is generally furthered by the creation of transformative works”23—even commercial ones.
What does this mean for Google? Under existing law, Google must show either “market failure” or transformation in order to prove its entitlement to the fair use defense. This is likely to be a considerable challenge. Consider market failure first: Knowing that the district court is likely to ask whether Google has entered a market that is “likely to be developed” by the plaintiffs,24 at least one of the plaintiffs has hastened to develop that market by announcing a searchable library of its own. On December 12, 2005, HarperCollins issued a press release in which it stated its intention to “create a digital warehouse for all of its content” that would both “satisfy[ ] the demands of the marketplace” and “allow[ ] the publisher to remain in control of its digital files and intellectual property.”25 Because HarperCollins also has stated that one of its goals for the project is the “monetization” of its content,26 the district court would be perfectly justified in holding that Google is depriving the plaintiffs of potential licensing revenues by providing the public with the means to search that portfolio for free. Google used copyrighted works to enter a market; it did not pay to license rights in those works; and the market it entered was “likely to be developed” by the owners of copyright because at least one owner actually developed it.
Google faces a similar challenge in defining (or redefining) “transformative.” While a handful of lower courts have stretched the definition to include mere reproductions,27 the Supreme Court in Campbell seems to have meant “transformative” to apply to those derivative works in which a user takes expression from a copyrighted work and adds expression of his or her own—creating, in the end, “something new”28 like a biography or an editorial or a parody (but not a satire!).29 At the very least, according to the Court, a transformative work must “alter[ ]” a copyrighted work with “new … meaning, or message,”30 and it is difficult (although not impossible) to imagine doing such a thing without adding new expression, too. Judge Pierre Leval, whose article on fair use appears to have inspired the Supreme Court, probably had very much the same thing in mind. As he wrote a few years before Campbell, a transformative work is one that “adds value to the original … in the creation of new information, new aesthetics, new insights and understandings.”31
Google is not doing any of these things; it simply is proposing to reproduce books in digital form. It is engaging in unauthorized, untransformative, unquestionably beneficial copying.32 The benefits for researchers are obvious, but the publishing industry stands to gain as well. If scanning books and making them searchable online leads researchers to find books that they otherwise might not have found, then a searchable library might help to expand the readership of books, thus leading to increased sales and increased borrowing from libraries (which, in turn, leads to increased sales). Google also is providing these benefits at very low cost to the public, in significant part because nothing that Google proposes to do has the least chance of preventing writers from writing, or publishers from publishing. Google is not competing with the publishing industry—as if, for example, it were enabling consumers to read copyrighted books online. In fact, it is not engaging in any activity in which the publishing industry has shown the remotest interest (until now). Unfortunately, however, none of these observations is particularly relevant to the inquiry into fair use today.
The late Professor L. Ray Patterson once observed, wisely, that “[m]ost discussions of the fair use of copyrighted works provide answers without ever asking the right question. That question is not ‘what is fair use?’ but ‘what is copyright?’”33 If fair use has become “a bizarre, occasionally tolerated departure from the grand conception of the copyright monopoly,”34 the fault is not in the defense itself, but in the “grand conception”: an increasingly proprietary copyright law that overwhelmingly equates the public interest with the private interests of copyright owners. Those private interests, in turn, are held by the very companies who control the market for copyrightable works. Is it any coincidence that the new economics of copyright places so much emphasis on private ordering,35 in which the government has no part to play?
Professor Julie Cohen has argued that “we need a theory of the ordinary user: a theory of what conduct is private.”36 I agree. Because any such theory must “promote the Progress of Science,”37 I begin, in Part I, by locating the strands of the public interest in copyright, which strands include the inducement of creation but also include other values, such as the existence of open and populous markets. In Part II, I describe a copyright law in which the creation of property rights is not an end in itself, but a means of identifying and punishing methods of unfair competition in the relevant markets for copies of copyrighted works. In these markets, acts of copying are not “unfair,” in themselves, because they do not inflict significant competitive harms. Accordingly, I argue that copyright owners should enjoy only the exclusive right to distribute copies (and perhaps, derivatives)38 of their works to the public. In Part III, I reveal how this theory might help to resolve a few of the issues at the very center of the copyright debate: the pervasiveness of copying; the rise of contractual and technological access controls; and, of course, the “death”39 of the fair use defense