« Week in Review | Main | Week in Review »

Oral Argument in Authors Guild v. Google, 13-4829-cv, U.S. Court of Appeals for the Second Circuit, December 3, 2014

By Robert J. Bernstein

Robert J. Bernstein practices law in New York City in The Law Office of Robert J. Bernstein. He is a frequent author and lecturer on copyright law and litigation, an Honorary Trustee and past President of the Copyright Society of the U.S.A., a member of the Copyright and Literary Property Committee of the New York City Bar Association, and formerly served as Chairman of the Copyright Law Committee of the American Intellectual Property Law Association and as a member of its Board of Directors. Mr. Bernstein has been co-author of the New York Law Journal bi-monthly "Copyright Law" column for 28 years.

For those of you who could not attend last Wednesday's oral argument in Authors Guild v. Google and do not want to await the transcript/CD, or who have neither the time nor inclination to read or listen to the recording of the 75 minute proceeding, here is a summary of how the argument unfolded in chronological order:

Disclaimer: Your author represents as co-counsel, along with fellow co-counsel Peter Jaszi and lead counsel Daniel F. Goldstein, the National Federation of the Blind and four print-disabled individuals (collectively, the "NFB") in the case of Authors Guild, et al. v. Hathitrust, et al., 902 F. Supp.2d 445 (S.D.N.Y. 2012), affirmed 755 F.3d 87 (2014) ("HathiTrust"). In HathiTrust, although Google was not a party and its liability/fair use was not adjudicated, the Second Circuit held that the digitization by Google of the entire contents of several major university libraries for the purpose of enabling university students and scholars to conduct research, in particular searching for books containing the searched terms without displaying any text from the books, constituted a transformative fair use. The NFB had intervened in the district court proceedings in order to obtain full access to the digitized collections for print-disabled students and scholars under both the fair use doctrine and the Americans with Disabilities Act ("ADA"). The Second Circuit rejected the argument that such full text reading of the digitized text was a transformative use, viewing reading as reading whether via print or digitized text-to-speech software. Nevertheless, applying general fair use principles, and relying in part on the U.S. Supreme Court discussion of fair use for blind individuals in Sony Corp. of America v. Universal City Studios, Inc., 464 U.S. 417, 455 n.40 (1984), as well as Congressional recognition in the ADA of the public policy benefits of equal access for those with disabilities, the Second Circuit held that the NFB uses constitute fair use. The Second Circuit remanded certain issues relating to preservation uses to the district court, where the case now resides. The remanded issues do not impact the Second Circuit holding with respect to the NFB.

Comment on Disclaimer: Your author recognizes, both as a lawyer and legal commentator, the obligations to disclose any connection that could reasonably be considered to influence the contents of a published writing. However, the readers should also keep in mind that, in this posting, the author will be attempting to report on the unfolding of the oral argument as it happened, in real time, without editorializing. The transcript/CD will appear in due course, so the readers will be able to judge for themselves whether my attempt to report objectively has succeeded.


The Panel: The Second Circuit panel hearing the appeal consisted of the Honorable José A. Cabranes, the Honorable Pierre N. Leval, and the Honorable Barrington D. Parker. Judge Cabranes, as the only active judge on the panel, would normally have presided, but as he was appearing via video from his chambers in Connecticut, the longest-serving senior appellate judge present, Judge Leval, presided. This may be viewed as fitting in view of the fact that Judge Leval introduced the concept of "transformative use" into the fair use lexicon in his 1990 Harvard Law Review article, "Toward A Fair Use Standard," 103 Harv. L. Rev. 1105, which the Supreme Court adopted four years later in its still-governing, but pre-digitization era, fair use decision in Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569 (1994).

This same panel had also considered the earlier appeal of the class certification issue in Author's Guild v. Google (hereafter the case will be referred to as "Google," and non-italicized "Google" will refer to the defendant-appellee). In its Order remanding the case to the district court for a determination of the fair use defense, the panel, per curiam, directed the Clerk of the Court to assign any subsequent appeal of the case to the same panel. 721 F.3d 132 (2d Cir. 2013).

Two members of the panel, Judges Cabranes and Parker, were also members of the Second Circuit panel that decided HathiTrust, wherein, in an opinion by the Honorable Barrington D. Parker, the court held, inter alia, that the digitization of the entire contents of the university libraries for the purpose of search and without displaying any text constitutes a transformative fair use.

Prior Proceedings: The long-and-winding road leading to Courtroom 1703 of the Thurgood Marshall U.S. Courthouse is beyond the scope of this report. The only issue before the Court concerned the appeal of the grant of summary judgment to Google on its fair use defense. In his summary judgment opinion, the Honorable Denny Chin (who retained the case in the district court after his appointment to the Second Circuit) held that Google's digitization of the university libraries collections, and its own use of that corpus for purposes of enabling users of Google Books to search the database for books containing the searched terms (search function) and for reading "snippets" displayed on the Google Books website (display function), both constitute fair uses; and that Google's provision to the university libraries of copies of the digitized databases from each library's own collection, alleged by the Authors Guild to infringement the distribution right, also constitutes a fair use because it enabled the libraries to engage in the fair uses approved by the Second Circuit in Hathitrust. Google, ___ F.Supp.2d ___ , 2013 WL 6017130 (S.D.N.Y. Nov. 14, 2013).

Further Background: Readers interested in articles summarizing the decisions referred to above may find the following New York Law Journal (NYLJ) "Copyright Law" columns of interest: R. Bernstein and R. Clarida, "Fair Uses of Hathitrust Digital Library," June 18, 2014, p. 3; R. Bernstein and R. Clarida, "Google Granted Summary Judgment on Fair Use Defense," NYLJ, December 20, 2013, p.3; and D. Goldberg and R. Bernstein, "The Supreme Court Balances the Act," NYLJ, March 18, 1994, p. 3.


Judge Leval prefaced the proceedings by stating that due to the complexity and importance of the issues presented, the Court would disregard the 15-minute posted limits and instead allow each side 30 minutes, with the possibility of additional time if needed. As it turned out, the total time elapsed was close to 75 minutes. No other cases were on the docket at this 2:00 p.m. sitting.

The author will now report on the arguments, questions, comments and rebuttals as they unfolded chronologically. The names of the participants are boldfaced to indicate who is speaking. Quotation marks are only used when I am reasonably confident that I was able to capture the actual words in my notes Generally, but with many exceptions, each new paragraph starts when a different speaker enters the dialogue.

For Appellant the Authors Guild (Paul M. Smith, Jenner & Block, LLP [arguing]; and Edward R. Rosenthal, Frankfurt Kurnit Klein & Selz, PC):

Mr. Smith opened with the necessary acknowledgment that HathiTrust is the law of the Second Circuit, while reserving the right to challenge the holding in other circumstance (referring, sub silentio, to the possibility of a request for Supreme Court review). He then immediately introduced the reasons why the Authors Guild ("AG") contended that Google's uses were qualitatively different from those in HathiTrust, and why those differences mandated a holding of infringement rather than fair use. He stated that Google's uses are "quintessentially commercial," engaged in by a for-profit company to preserve its key asset - maximizing visits to its website in order to increase advertising revenues.

Judge Leval commented that, in his view, notwithstanding a distinction between commercial and noncommercial uses, classic holdings of fair use have involved commercial uses (that is, the user was at least significantly motivated by potential profit), and that he "would be surprised if [AG] win[s] on that basis."

Mr. Smith rejoined by stating that this case was different due to the enormity of the commercial use, leading Judge Leval to ask whether his proferred distinction was between "big profits" and "little profits," followed by this query: "If Google does not cause any harm" to the copyrighted works, what difference does it make how much profit it derives from the use?

Mr. Smith then turned to what he argued is the second determinative difference from the uses in HathiTrust: Google displays text (referred to as "snippets" to indicate their brevity), whereas no text was displayed in the libraries' uses. He stated that such displays harm an existing and emerging market to license books for digitization and related uses, which he characterized as "derivative uses" [referring to the copyright holder's exclusive right to create derivative works based on the copyrighted work]. Here, he cited a 1992 decision by then-district-court Judge Leval holding that photocopying of scientific articles by Texaco scientists was not a fair use. American Geophysical Union v. Texaco, Inc., 802 F.Supp. 1 (S.D.N.Y. 1992), aff'd, 60 F.3d 913 (2d Cir. 1994).

Judge Leval stated that he did not consider his opinion in Texaco to be of guidance here due to "big differences" between Texaco's and Google's uses, noting that, in Texaco, 80 in-house scientists made multiple copies of entire articles in order to have convenient access to them rather than purchasing additional journal subscriptions. In reply, Mr. Smith asserted that the test applied in Texaco - whether the use is one for which a license exists or is likely to emerge - is relevant to the emerging digitization marketplace.

Judge Leval responded that whether a market for the use is likely to emerge "is not a very useful test" with respect to transformative works because, no matter how transformative the use, someone at some price would be willing to license it to avoid the cost of litigation. Mr. Smith averred that a distinction should be made between uses that a copyright owner might never license, such as parody or criticism, and uses that the owner would be prepared to license for the right price, a proposition as to which Judge Leval asked if Mr. Smith had any case authority.

Mr. Smith also argued that it should be a policy decision for Congress whether "new derivative uses," which, in his view, include digitization, should be free of a license.

At this point, Judge Leval introduced two issues that he stated were of concern to him: (i) what limitations, if any, are there on what the libraries can do with the digitized collections provided to them by Google; and (ii) what vulnerabilities do libraries have to hacking and piracy? In this regard, Judge Leval asked: Even if we decide that Google's use is fair use, "if the upshot is the multiplication of copies and greater potential for hacking," how does that impact the analysis? He further asked whether this could be a particular concern with respect to "hot properties", such as a new Harry Potter book.

Rather than respond commenting immediately on these concerns, Mr. Smith initially directed his response to their predicate ("even if we decide Google's use is fair"). Thus, Mr. Smith argued that Google's use was not fair, inter alia, because, in providing copies of the digitized collections to the libraries, Google was using those copies as "currency" - a form of barter in lieu of monetary compensate in exchange for Google's access to and use of the books for digitization. Judge Leval stated that he "was not following" that point.

Mr. Smith than turned to another reason why he contends that Judge Leval's predicate ("assuming that we find fair use") to his stated concerns was faulty, arguing that in evaluating fair use, only the conduct and purpose of the initial user/defendant is relevant, but not those of downstream users or third parties (such as, in Mr. Smith's characterization, the libraries). For that reason, he contended, Google should not be entitled to rely on the fair use holding in HathiTrust to justify its own use.

Mr. Smith then addressed Judge Leval's stated concern about the vulnerability of digitized collections within library databases to hacking and piracy - arguing that in Google's agreement with the Stanford Library (which is not part of the consortium of university libraries constituting Hathitrust and was not a party in HathiTrust), students are not restricted from simultaneously reading entire texts online).

Judge Leval asked: "Is it Google's responsibility to be sure that Stanford's use is fair?" To which Mr. Smith replied: "Google assumed responsibility" for the library uses of the digital databases it provided to them. Judge Leval asked: "What do you mean?," and Mr. Smith responded: "If Google tries to ride on the libraries' uses," they should be responsible for the consequences.

Judge Parker then turned to a different issue, asking: "Are there any provisions in the documents that restrict the contours of what Google is doing? What keeps them from changing?"

Mr. Smith stated that nothing in the arrangements would prevent Google from expanding or otherwise changing its practices, and posited a multiplier effect: "What if everyone could do this? Isn't this a policy decision for Congress?"

Judge Leval observed that displaying 8 pages rather than snippets would be a different case, and Judge Parker posed his own rhetorical question: "You aren't seriously suggesting that we wait for Congress to Act?"

At which point Mr. Smith changed direction to focus on the "limited" remedies the AG was seeking. He articulated a bifurcated approach to Google's past and future conduct. He assured the court that AG was only seeking monetary damages for past alleged infringements, and was not requesting that the already existing database be undone. As to future digitization and related uses, AG seeks the imposition of a licensing requirement, but not injunctive relief.

Judge Leval then asked whether the present arrangements preclude digitization of a hot new book. Mr. Smith replied that most new books are already licensed in Google's Partners Program with publishers, which led Judge Leval to ask: "Are you saying that hot new books aren't part of the case?"]. Mr. Smith then clarified his position by responding that, "on the contrary," because not all publishers are part of Google's Partners Program and because authors can opt out of Google Books, many hot new books could be affected.

Judge Parker then asked: "What precisely is the mechanism" for an author to opt out? Mr. Smith stated that when an author opts out, the work is no longer displayed in snippets, but he understood that the work would still remain in the database, and may (he was not certain) still be available for search.

Mr. Smith then turned to AG's argument that the display of snippets itself causes harm, explaining, by way of example, that certain types of books are often referred to only for limited purposes and discrete information, and that snippet display could satisfy those readers' needs, resulting in the loss of a sale. Yet Judge Leval observed that dictionaries, for example, were excluded from snippet use. Mr. Smith rejoined by stating that other types of books, for example history books, may be used for short bits of information that could be obtained by viewing a series of snippets rather than purchase.

Mr. Smith then turned to an evidentiary point, contending that Judge Chin articulated no evidentiary basis for his rejection of the AG's argument that snippet displays cause market harm. [Questions of the proper allocation of burden of proof on this point were not addressed in any detail at the argument, but merely averted to].

Judge Cabranes posed a procedural question concerning the reasons why certain portions of the appendix were redacted. Mr. Smith referred to the protective order entered in the district court which required confidential treatment, but in response to further queries by Judge Cabranes and Leval, the parties agreed to advise the court whether there was any portion of the redacted materials which the parties still requested be redacted from any published opinion.

At this point, Google's counsel presented his case.

For the Appellee Google, (Seth P. Waxman, Wilmer Cutler, Pickering, Hale and Dorr LLP [arguing]; and Joseph C. Gratz, Durie, Tangri LLP):

Mr. Waxman commenced by enumerating what he characterized as three fundamental points: (i) Google's and the libraries' uses "quintessentially" promote progress in the arts in furtherance of copyright's constitutional purpose; (ii) all of these uses are products of a collaboration between Google and the libraries which allow each to engage in transformational uses that neither one could accomplish alone; and (iii) there is no evidence in the record of any market harm to authors.

Judge Leval immediately asked whether, even if there is no evidence of harm to any particular books, there is harm to a market for licensing digitization rights? In response, Mr. Waxman made two points: (i) HathiTrust, which is Second Circuit law, rejected that argument; and (ii) in any event, there is no evidence that any such market exists or is emerging.

Judge Leval then asked about record evidence of licensing digitization rights in Europe, to which Mr. Waxman replied: the European licenses grant the right to reproduce and display the entire book which is a materially different type of market.

Judge Parker asked how much Google has spent on the project, to which Mr. Waxman replied, over $120,000,000. In further colloquy it was stated that there is nothing in the record demonstrating the extent to which Google may have derived profits (direct or indirect) from the uses.

Judge Cabranes then asked a series of questions about the commerciality (or not) of the uses: "Is this only a charitable enterprise by Google? Is it solely designed for eleemosynary purposes?" "Is any part of the Google project for profit?" Judge Leval supplemented this inquiry by asking whether advertising is displayed in connection with the uses?

In response, Mr. Waxman noted that no ads are displayed on pages responding to search requests or on pages displaying snippets, and that Google makes no money from referring sales to third parties.

Judge Parker asked: "What keeps Google from changing the way it operates," such as by lengthening its display of text? Mr. Waxman replied that [presumably significant] changes would have to be evaluated separately under the fair use factors.

In further response, Mr. Waxman argued that if Google derives profit from the use, it would not distinguish this case from numerous fair use cases. He continued: "If Google didn't expect to make some return on this, it may never have engaged in this incredibly transformative use."

Judge Parker noted that in the class certification appeal, Google argued that fair use must be evaluated as to each individual book, but that on this appeal Google argues that fair use may be considered as to the entire corpus of digitized works, suggesting an apparent inconsistency between these two positions. Mr. Waxman replied that there was no inconsistency because in the class certification appeal Google did argue that the entire Google Book Project should be found to be fair use, but that, if the court did not conclude fair use as to the entire project, then questions of infringement would have to be evaluated one book at a time.

On the subject of the libraries' uses, Mr. Waxman noted that Google's agreements with each library required that any use by the library be consistent with copyright law. Judge Leval queried whether, even if the libraries complied with copyright law, weren't the digital copies stored on the libraries' databases vulnerable to hacking and the resulting infringing conduct of others?

In response, Mr. Waxman noted again that HathiTrust is Second Circuit law and that it determined that AG's security concerns did not make the uses unfair. Judge Leval stated that the finding in HathiTrust was merely that, on the record there, the possibility of a security breach was speculative.

Mr. Waxman replied that there is no evidence in the Google record to support AG's security risk argument. He asserted that there is no record of any instance of hacking into the libraries' digitized collections in the years since they were created. He further argued that there is good reason for a record devoid of hacking into the library corpus - digital copies of more recent and commercially popular books are already available on the Internet through, for example, Amazon. He asserted that, in contrast to Google's agreements with HathiTrust member libraries, the standard Amazon contract with publishers does not require Amazon to take security measures.

Judge Leval then asked: Are you saying that digital copies are more easily available elsewhere than from within the libraries' digital corpus? Mr. Waxman replied that, as a matter of policy, Google does not digitize books for search and snippet display until two years after their initial publication, which further reduces any possible substitutive effect.

Returning to a point raised earlier by Judge Parker, Judge Leval asked whether that policy could change? Mr. Waxman did not, at that point, address the possibility of change, but instead emphasized that the point of the Google Books is to allow users to find books, "and nothing else."

Judge Leval then asked: "Is the two-year blackout in the record as a characteristic that the court can base its fair use analysis on," and, if so, requested a record cite. Judge Leval later repeated this query, and Mr. Waxman undertook to reply "anon" (the archaic form of "soon").

Judge Leval then returned to his question regarding the impact of Google's agreement with Stanford, which contains no restrictions on the Stanford library's full text display of digital copies which could be read by students simultaneously. Mr. Waxman replied that under that agreement, Stanford is still obligated to only make such uses as are consistent with copyright law, and, in any event, even if Stanford were to breach its agreement with Google, that would not render Google's own use unfair.

Mr. Waxman then focused on the "copy-shop" cases relied on by AG, in which commercial copy shops created and sold coursepacks to students. He argued that this use was the polar opposite of Google's transformative use because the copy shops packaged and sold the coursepacks to students as substitutes for their purchase.

Judge Cabranes then stated: "Let us follow the dollars (even if ultimately it may not matter)". Toward that end, he engaged in an extended colloquy with Mr. Waxman that ended with this stipulation: Mr. Waxman: "I will stipulate, arguendo, that Google had [has] a profit motive because [the Google Books Project] would bring more eyes to Google." However, he qualified this stipulation in two ways, stating that (i) there is no evidence in the record that Google provided digital copies to the libraries to avoid paying for access to the libraries' hard copies; and (ii) that, on its most fundamental level, Google and the libraries were engaged in a collaborative project designed to make millions of books available for search by the public. With respect to the latter point, Mr. Waxman quoted from the agreement between Google and the University of Michigan which is prefaced with a statement of their joint purpose.

The Authors Guild Rebuttal (by Mr. Smith):

In a brief rebuttal argument, Mr. Smith made the following points:

(i) this case was decided by the district court on summary judgment, and therefore a holding of fair use cannot be based on any genuinely disputed material facts;

(ii) Google has the burden of proof on its fair use defense, and therefore, where proof is absent or insufficient, inferences must be taken in AG's favor;

(iii) the court should weigh in the balance the allegedly anti-competitive conduct of Google in allegedly using the digitization of library collections to maintain and increase its dominant market position in book digitization vis-à-vis Amazon and others; and

(iv) it would be inappropriate to impose on AG an impossible burden of proving that the display of particular snippets led to lost sales of particular books.

At this point in the rebuttal, Judge Cabranes interjected with this question: "What is the decree of court that you are seeking from us?" He observed that, in its brief, AG's requests broad relief, but that, in the oral argument, it appears that AG is seeking less.

In response, Mr. Smith repeated the remedies he outlined during his main argument, while elaborating by citation to the Supreme Court opinion in eBay, Inc. v. Mercexhange, L.L.C., 547 U.S. 388 (2006). [There, the court held that before issuing injunctive relief, a court must make findings of fact on four separate factors, rather than merely apply presumptions. For an excellent discussion of the impact of eBay in fair use cases, see R. Dannay, "Copyright Injunctions and Fair Use: Enter eBay -- Four-Factor Fatigue or Four-Factor Freedom?" 37th Annual Brace Memorial Lecture, Journal of the Copyright Society of the U.S.A.,Vol. 55, No. 4 (Summer 2008), p. 449-468)].

Summarizing, Mr. Smith stated that in light of (i) eBay; (ii) the passage of time; and (iii) the value of the database, "we do not request an injunction but monetization and licensing for future scanning."

Judge Cabranes then asked: "Do you think the case could be resolved by a sum certain?" Mr. Smith replied; "Yes," suggesting that the numbers could be worked out with the aid of expert testimony and computer software, and proffering the example of the recent settlement between Viacom, et al. and YouTube, in which he represented the plaintiffs against Google's YouTube subsidiary.

Judge Cabranes asked whether discovery is over, to which Mr. Smith replied "yes."

As a final point, Mr. Smith revisited the alleged risk of a security breach, citing by way of example the massive hacking of the JSTOR database of scholarly articles and journals by a proponent of the "research-must-be-free" point of view.

Judge Leval then declared the case submitted for resolution by the panel and adjourned the proceedings.

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)


This page contains a single entry from the blog posted on December 10, 2014 10:44 AM.

The previous post in this blog was Week in Review.

The next post in this blog is Week in Review.

Many more can be found on the main index page or by looking through the archives.