On October 16, 2015, the U.S. Court of Appeals for the Second Circuit took another step towards strengthening the fair use doctrine.
His decision in The Authors Guild v. Google Inc. harmoniously combines theoretical elaboration and practical usefulness. So it can be safely recommended for study to those who are interested in the modern vision of this doctrine in American jurisprudence.
To begin with, in this decision, the court examines in detail the arguments why the GoogleBooks project, which involved the digitization of many books without the consent of the copyright holders, should be considered legitimate. Prior to this, the courts had already approved a university library program that provided books for scanning (Authors Guild, Inc. v. HathiTrust, which I wrote about here and here). Now the turn has come to the direct executor of the project – Google, whose actions are also recognized as fair use. The Court carefully analyzes each factor of the doctrine, revealing their content, significance and mutual connection. The result is a handy guide that clarifies acceptable routes.
In conditions when technologies are constantly changing the space of creativity, and the legal mechanism available to creators, albeit due to its universal orientation, is excessively complex in application, so that it is rarely resorted to and with caution, it is difficult to overestimate the usefulness of such guides. In addition, the court raised a number of thorny legal issues and attempted to resolve them by providing a sort of overview of the cutting edge of legal thought in the area of fair use. For those interested, all these points will be discussed in more detail below.
Secondly, Judge Pierre N. Leval presided over the case and the preparation of the decision. He is well known for his active participation in the formation of the doctrine of fair use, both in practical and theoretical terms. As a judge since 1977, P. Leval has been involved in many of the cases that have shaped the current fair use policy. He embodied his experience in the policy article “Toward a Fair Use Standard”, published in 1990. Thanks to his persuasive arguments, American courts in the analysis of fair use began to give priority to the first factor instead of the fourth. Finally, it was recognized that a transformative, creative subsequent use of a work, regardless of its commercial or non-commercial nature, should be called fair. P. Leval’s article was often cited by American courts. And in the U.S. Supreme Court’s most famous fair use decision, Campbell v. Acuff-Rose of 1994 (the decade of which was recently widely celebrated in American legal science), P. Leval’s approach became part of the guiding clarification. The merits of P. Leval are also appreciated outside the United States. Thus, the Australian Government Commission on Copyright Reform calls his work one of the most influential in the field of fair use.
P. Leval continues to contribute to the theory and practice of applying the fair use doctrine. His latest work, Campbell as Fair Use Blueprint?, is also devoted to this topic, which in many respects has something in common with the solution under consideration. A funny fact comes to mind. In his article, P. Leval raises an interesting problem that has not yet been solved in jurisprudence – how do the subsequent and derivative works relate? It seems that both are the result of a reworking of the original work, but their statuses are radically different. P. Leval then tries to formulate the answer to his question in his court decision. An excellent example of a symbiosis of theory and practice.
Now let’s look at the main elements of the solution.
Let me remind you that Google Books is a project of cooperation between Google and leading universities that provide it with their book collections for processing. Since 2004, more than 20 million books have been digitized, mostly non-fiction, mostly rare, out of print, and in the public domain. Anyone can search the entire array of texts by keywords. And although the issue will show no more than three brief excerpts from the book, the service allows you to conduct in-depth analysis of texts based on information about the frequency of occurrence and the context of the use of the searched words in time and space. Realizing the enormity of the project, the court was quick to point out that in the absence of the Google Books project, which provides instant searches in digital copies of millions of works, researchers would not have had the entire duration of a human life to manually browse a comparable number of paper books. The service can be useful and simple for users to find a wide variety of information that fits in a short passage: remember the date, find out the definition and examples of the use of the term, find a name, position or organization, or simply dig out an interesting historical fact in an authoritative publication.
Copyright holders, of course, see this as a threat to their cars rights, well, or elusive profits. They give five reasons why the Google Books project is not considered fair use:
- copying the full text of books and providing users with excerpts from them leads to the appearance of substitute goods for protected works, which is not a transformative use;
- although Google does not charge a fee for using the search and browsing functions, and does not advertise on these pages, the search engine still pursues commercial purposes, strengthening its global dominant position in the search market through the Google Books project;
- even if Google’s actions do not violate copyright, they still infringe on the derivative rights (derivative rights) of copyright holders, depriving them of the potential profit for issuing licenses in the search market;
- hackers can gain access to digital copies of works that will become freely available on the Internet, destroying the value of copyright;
the distribution of digital copies to participating libraries is not a transformative use, and also poses a risk of reducing the income of rights holders.
The Court of Appeal, in turn, fully agreed with the lower court that the activity of the search engine is fair use, i.e., under the Copyright Act of 1976, “is not an infringement” of copyright, and therefore does not need approval or licensing.
To begin with, the court set clear priorities: “The main purpose of copyright is to increase public knowledge and understanding. Copyright seeks to achieve this by giving potential creators the exclusive power to control the copying of their works, thereby financially incentivizing them to create meaningful, intellectually enriching works for public consumption. This purpose is clearly reflected in the constitutional authority of Congress to “Promote the Advancement of Knowledge… by granting Authors for a limited Period of time… the exclusive Right to their Works” (Section 8 § 8, Article 1 of the US Constitution). Thus, while authors are undoubtedly among the intended beneficiaries of copyright, the primary, primary beneficiary is the society whose access to knowledge copyright provides by rewarding authorship.
For nearly three hundred years, since the birth of copyright in England in 1710, the courts have recognized that, under certain circumstances, giving authors absolute control over their works would have limited rather than expanded public knowledge. In the words of Lord Ellenborough, “Though I feel obliged to safeguard the right of every man to use his right to copy, no one should put fetters on knowledge” (Caryv. Kearsley, 1802). To this end, the courts developed a doctrine, eventually called fair use, which allows, under certain circumstances, to be copied without obtaining the consent of the author, in order to fulfill the “primary purpose of copyright” to promote the Progress of Knowledge and the Craft” (Campbellv. Acuff-RoseMusic, Inc., 1994 )”.
The four factors of the fair use doctrine are set out in §107 of the Copyright Act 1976.
Despite the doctrine’s long history of judicial application, it was not until 1994, in the Campbell case, that the US Supreme Court first set standards for factor analysis. And no less important, he pointed out the unevenness of the factors themselves, starting to line them up by rank (recall the article by P. Leval). The most significant was the first factor (the purpose and nature of the further (“subsequent”) use of the original work). The more new, different from the original goals pursued by the one who borrows someone else’s work, the more his actions correspond to the purpose of copyright (to replenish public knowledge) and the less he replaces the original (or derivative) work, and the less it encroaches on the market potential of the protected work. Therefore, in the future, the judicial practice of applying the fair use doctrine was based on two foundations: correct analysis and the correct ratio of factors.
The decision under consideration perfectly demonstrates the successes achieved, which, we note, so far not all American courts can boast of.
1. So, why is the placement of digitized versions of books in an electronic library with the functions of searching and displaying excerpts recognized as fair use.
1.1. The first factor of doctrine is that the purpose or nature of secondary use must be transformative. Since the doctrine itself is intended to expand the space of creative freedom for the benefit of the whole society, only those who, through whose efforts, the original work begins to bring additional benefits to society, can rely on its protection: the transformed work communicates something new to the public, expands the usefulness and applicability of the original, fills it with new meaning. and sound, satisfies some other needs of the reader (viewer, etc.). To do this, the court needs to decide whether the subsequent work (secondary work) serves the same purposes as the original, or pursues new, additional purposes, in any way related to the increase in public knowledge. If new purposes are discovered, the use is considered transformative and therefore fair. Examples might be writing a commentary, critique, or creating a parody of the original work.
We emphasize that the transformed work does not have to be original, creative: the depth of transformation (transformation), the amount of intellectual effort, the novelty of the subsequent work do not play a special role. For example, the Supreme Court has found it transformative to use search engines automatically generate thumbnails of protected images to display search results. There is no outstanding creative contribution to this, but the benefits for society are obvious: the search becomes more convenient, knowledge is more accessible.
On the other hand, there should be a creative relationship between the first and second work. The courts do not seek to remove absolutely all the shackles from subsequent creativity, reasonably fearing that in some cases this may lead to parasitism on someone else’s work. Not all transformative use, even when aimed at increasing public knowledge, is considered to be fair. It is necessary to justify why it was necessary to borrow this particular original. If a subsequent author creates his work by developing a theme, changing and expanding the meaning, purpose, understanding (by the public) of the original material, or relies on the expressive means and artistic finds used in it, the law is likely to classify his actions as bona fide. We can say that in this case the intellectual efforts of the original author are supplemented by the intellectual work of the subsequent one, the new work grows out of the original. This happens when commenting on, critiquing or parodying the original work. Such a contribution to the enrichment of knowledge is approved by the law. If the creative relationship between the two works is absent or weak, if someone else’s content is borrowed only to draw attention to a new object, to save the efforts of the second author, he will have to convincingly prove the existence of good reasons for such behavior. In the Campbell case, the Supreme Court illustrated this reasoning with parody and satire. Parody can be called a type of ironic commentary. It takes as a basis some characteristics of the original in order to deepen it and give us some new knowledge. Satire, as a rule, does not bring anything new to the original. Its author only wants to draw attention to his own idea, and therefore relies not on the content and style of the original, but, for example, only on its popularity. Satire has its own goal, which can be achieved by completely different means. Therefore, the author of the satire must justify why he should be allowed to borrow this particular original.
Of course, the distinction here is very arbitrary. But in the present case, it is of great importance. The creative connection of the Google Books project with the original works is not obvious. Therefore, the court (as in the case of HathiTrust) carefully examines two circumstances: the contribution of the project to public knowledge and the reasons why these particular paper books are used and in such quantity.
The question of creative interconnection leads to the next difficulty, which I mentioned above. The point is that “transformation”, which generates subsequent works, is close in meaning to “recycling”, which creates “derivative works” (derivative works). But if, in the case of a transformative use, all rights to a subsequent work arise with its creator, then the rights of the original author apply to derivative works. Section 106 of the 1976 Law establishes the exclusive right of the copyright owner to “create derivative works based on the protected work.” In a word, we are talking about two similar processes, as a result of which new objects with different statuses arise. How to distinguish between such processes? Unfortunately, the law does not provide a definition of a “derivative work”, limiting itself to listing examples: translation of a book into another language, adaptation of a novel for cinema or theater, remake of a work in the form of an electronic or audio book, etc. The Court in the present case provides the following guidance: “While such changes [translation, adaptation or adaptation] may be called a transformation, they do not pursue any transformative purpose that satisfies the fair use criteria. The statutory definition assumes that derivative works are associated with transformations in the form of a change in form. Whereas transformative use, covered by the doctrine of fair use, entails a change in the purpose, purpose of the work.
One of the district courts (the Seventh) proposes to refer to a subsequent use that satisfies the fair use criteria as a “complementary” use (complementary use) how hammer and nail complement each other in achieving a common goal that they alone will not achieve (Ty, Inc. v. Publ’ns Int’l, Ltd., 2002; Kienitz v. Sconnie Nation LLC, 2014). The Second District Court is sure that the term is not appropriate. It refers to creation of derivative works (change of form) rather than fair use. For example, a book and a film based on it perfectly reinforce each other, jointly achieving a goal that they cannot achieve separately: the insights of the author of the book are complemented by the interpretive skill of filmmakers, creating a more effective joint object in the form of a film. However, the film is purely a derivative object. But when a subsequent work borrows the original to parody it, to demonstrate its inaccuracies or illogicality, when the purpose of the second object is primarily to expose the first, both objects cannot be called complementary for the sake of a common goal that they cannot achieve alone. Such actions are not the exclusive prerogative of the original author, they are transformative, fair use, which may be performed by any other person without his consent. Thus, the concept of “transformative” cannot be understood too broadly, otherwise it may cover a borrowing that falls within the scope of the derivative rights of the original author.
A) Search function.
Earlier in the HathiTrust case, the court had already recognized as profoundly transformative the act of making digital copies of books and then using them to find and analyze which books contain searched words and phrases. The court explained that “the result of a word search differs in its purpose, character, expression, meaning and message from the page (and book) from which the result is obtained.” There are already enough examples in judicial practice when the creation of a digital copy of the entire protected work (and not just its separate part) is recognized as fair use, if the copy performs a function different from the original. In addition, the project is clearly aimed at both increasing the availability of existing knowledge and creating new ones. Thus, digitization, even in the absence of agreement with the copyright holders, is recognized as legitimate.
There are two differences from the HathiTrust case: Google displays excerpts from books, and is a commercial company looking for profit.
B) Viewing excerpts.
The ability to view passages of text in which the search words are found does not change the picture. It adds value to the core transformative function of search. Browsing allows you not only to establish whether a certain term is contained in the book, but also to understand whether it is used there in the desired meaning. As a result, the researcher can evaluate whether the found book falls within his area of interest and decide whether to purchase it. And Google at the same time will tell you where it can be done. At the same time, nothing threatens the interests of the copyright holder. Thus, providing short excerpts can be considered fair use. However, the situation will change if the passages are provided in such a way that they become a competing substitute for the book itself. The court considers this situation more fully when analyzing the third and fourth factors.
C) Google’s commercial motivation.
The Court disagreed with plaintiffs that the commercial motivation argument argued against finding the first factor of doctrine. The presumption that any commercial use is considered bad faith, expressed by the Supreme Court in the Sony case, was later refuted by the Supreme Court. As the court explained in Campbell, “the more transformative the [subsequent] work is, the less important are other factors, such as commercial nature, that can be opposed to fair use.” The Second Circuit Court has consistently taken the same position: the defendant’s general profit orientation does not prevent him from relying on fair use doctrine so long as he pursues transformative aims and does not create competition by substantially replacing the original work.
1.2. The second factor is the nature of the protected work. Most often, it is analyzed as part of the first factor: it is impossible to evaluate the purpose of a subsequent work without comparing it with the purpose and purpose of the original. Therefore, the second factor in itself does not play a significant role.
The Court draws attention to an interesting problem. The Supreme Court in Harper & Row noted in passing that “the law generally recognizes a greater need for the dissemination of works of fact than of works of fiction and fantasy.” Factual works are usually referred to as works whose content is predominantly unprotected information about specific facts and ideas. The Supreme Court’s observation is related to the fact that since facts can be distorted without a direct quotation, it is worth making allowances for more frequent quoting of such works in subsequent works. The courts, relying on this observation, have become overly willing to find good faith
the use of actual works when borrowing. However, the Second Circuit Court reiterates that although information and ideas are not subject to protection, the way in which the author expresses them is protected. The factual nature of a work (for example, news) cannot justify free borrowing of its parts by anyone and unrestricted distribution. Therefore, in the case under consideration, where most of the digitized works of the plaintiffs are non-fiction, the court still does not turn the second factor in favor of the defendant. At the same time, one first factor that speaks in favor of the search engine is enough to recognize the use as fair.
1.3. The third factor is the volume and materiality of the borrowed part in comparison with the entire work. He assumes that it is more likely to be conscientious to use a small or insignificant part of a work, rather than a borrowing that is voluminous or includes important parts of the original. The reason for this is the close relationship between the third and fourth factors: the larger or more important the copied part, the more likely it is that the subsequent work will be an effective competing substitute for the original and will reduce the sales and profits of the copyright holder. However, the courts do not take a categorical position that borrowing the entire work cannot be in good faith. Borrowing of the entire original is perfectly acceptable if it is reasonably necessary to achieve the transformative purpose and is done in such a way as not to create a competing substitute for the original. There are a number of solutions with similar conclusions. For example, Bill Graham Archives v. Dorling Kindersley Ltd. (2d Cir. 2006): “sometimes fair use requires copying the entire work.” In Campbell, the Supreme Court emphasized that “the amount of borrowing allowed varies according to the purpose and nature of the use”, so the correct question would be “whether the amount and materiality of the borrowed share is reasonable from the point of view of the purpose of copying”, given that the answer will be influenced by “the degree with which the [subsequent work] can serve as a marketable substitute for the original or potentially licensed derivative works.”
A) search function.
Endorsing the Court’s conclusion in HathiTrust, the court concludes that “copying the entire length of the original is not only reasonably consistent with Google’s transformative goal, it is literally indispensable to achieve such a goal,” otherwise, the search results (how many times the search term occurs in a given or another book) could not be relied upon. It is important that the search only allows you to find out information about the book, but not to get access to its text.
B) Viewing excerpts.
What matters here is the volume and importance of not copied, but provided to users of the part of the work. The more text available to him, the less likely the use of the work was in good faith, as this directly affects the ability of the passages to replace the original in the marketplace.
After considering the features of the Google Books system, the court considered that the excerpt viewing function was not capable of significantly competing with the sale of original books. His decision was influenced by such facts as: a small amount of passages that are not tied to complete sentences or phrases; display no more than three passages from the entire book, no more than one passage per page, and no more than one page in ten; regardless of the number of search queries and device changes for a specific term, the same examples are displayed; excerpts from dictionaries, collections of recipes and similar books are not provided. In any case, 22% of the text of each book remains closed for viewing in excerpt mode. Plaintiffs-initiated studies have shown that all these rules lead to the fact that even deliberate attempts to extract the maximum text by searching search terms for several weeks in a row allow access to an average of 16% of the volume of the book. And the collected passages do not go sequentially, but randomly. As a result, according to the court, the snippet review feature does not preclude fair use. However, the court noted that if the collected, albeit with great difficulty, passages in the amount of 16% of the volume of the book provided a logically consistent text, the solution to the issue could be different.
1.4. The fourth factor is the assessment of the impact that the use of someone else’s work has on its value or potential market. Can a copy become a competitive substitute product in the market for the original work (or works derived from it) in such a way that it causes a significant drop in the income of the right holder due to the fact that potential buyers choose a copy instead of the original.
The fourth factor is next in importance after the first. The court recalled that copyright is still “a commercial doctrine designed to stimulate the creativity of potential authors by receiving remuneration for their work.” U.S. Supreme Court in Harper & Row recognizesshaft this factor is the most important. In Campbell, he already gives priority to the former, but points to their close relationship: the more different the goals of the original and subsequent works, the less likely the copy will in any way replace the original in the market.
In the HathiTrust dispute, the court found a fourth factor in favor of fair use, since the process of automatically searching a book for a specific term does not replace reading the book itself. But what about the search result? In Campbell, the Supreme Court clarified that even the transformative purpose of borrowing can be detrimental to the rightholder if it is accompanied by the disclosure of a significant portion of the original work, i.e. creates a publicly available competitive substitute. As shown earlier, this does not happen in the Google Books project. The cost of acquiring books is far less than the cost to invest in acquiring a small set of unrelated excerpts. Therefore, the project does not deprive copyright holders of income and does not reduce the value of their work.
The court acknowledged that a certain decline in sales could still occur.
Some researchers will be satisfied with the results of the search and will not go to the store or library for a book (and those, in turn, will not purchase additional copies of popular publications). “However, the possibility, the likelihood, or even the predestination of some reduction in sales does not turn the copy into an effectively competitive substitute that would turn the fourth factor in favor of the owner of the rights to the original. This requires a serious, significant impact on the potential market or value of the protected work”, as provided by §107(4) of the law.
An important point is that the mentioned decrease in sales, if it happens, will mainly affect interests that are not protected by copyright. Information from the book may be needed to verify the historical fact. For example, in what year did an event occur. The passage is more than enough here, there is no need to buy or find the book itself. Thus, the passage satisfied the need for knowledge of facts, which, as you know, are not covered by copyright. Google may respond to a user’s question by extracting information from a protected book. Displaying some redundant information (three lines of text instead of one date) will still not be a copyright infringement. If the researcher was interested in the protected aspects of the work (in the features of the author’s creative expression), then due to the specifics of the service, when the selection of the necessary passages is burdensome, disconnected and incomplete, this interest will obviously not be satisfied. That is, viewing excerpts cannot replace the purchase of the book. Accordingly, the fourth factor also speaks in favor of recognizing the project as fair use.
2. The Google Books project’s next allegation concerned derivative rights to search and view passages.
Copyright holders demanded that their derivative rights be recognized as violated. Derivative right in American jurisprudence is understood as the exclusive right to “create derivative works based on protected works” (§106(2) U.S. Code). Plaintiffs contend that they own derivative rights to search and view excerpts on original works. The market for these objects should be exclusively owned by the owners of the rights to the works themselves. However, Google, they say, has captured this market, depriving them of the opportunity to develop similar services of their own.
The Court rejected these claims because no such derivative rights exist. Copyright does not include the exclusive right to provide information about works collected by searching through digitized copies. The logic behind legislators generally protecting derivative works is extremely simple. As we remember, a derivative work in American jurisprudence means the embodiment of the original creative content in a different form. All protected elements of the original remain unchanged, but they are presented in a new form. Copyright is meaningless if it can be easily circumvented by changing the form of the work. Let the novel be translated into another language or made into a movie, but in a modified form the novel remains the same. The copyright of a song extends to its arrangement for other musical instruments and to the recording of the performance. The form changes, but the work is used in exactly the same way and satisfies the same user needs. Therefore, if Google digitized the works and made them available to the public, i.e. they could be read in exactly the same way, albeit in electronic form, then the claims of the plaintiffs would have been justified. However, in the Google Books project, the creative content of the books remains closed to the public. Therefore, digitized works in this case are not derivatives, they are created for a different purpose and satisfy completely different needs – not to get aesthetic or intellectual pleasure from