The Authors Guild v. Google, Inc.

Filing 135

BRIEF, on behalf of Appellee Google, Inc., FILED. Service date 07/03/2014 by CM/ECF. [1263769] [13-4829]

13-4829-cv
UNITED STATES COURT OF APPEALS
FOR THE SECOND CIRCUIT
THE AUTHORS GUILD, BETTY MILES, JIM BOUTON,
JOSEPH GOULDEN, individually and on behalf of all others similarly situated,
Plaintiffs-Appellants,
HERBERT MITGANG, DANIEL HOFFMAN, individually and on behalf of all others
similarly situated, PAUL DICKSON, THE MCGRAW-HILL COMPANIES, INC., PEARSON
EDUCATION, INC., SIMON & SCHUSTER, INC., ASSOCIATION OF AMERICAN
PUBLISHERS, INC., CANADIAN STANDARD ASSOCIATION, JOHN WILEY & SONS, INC.,
individually and on behalf of all others similarly situated,
Plaintiffs,
v.
GOOGLE INC.,
Defendant-Appellee.
On Appeal from the United States District Court for the
Southern District of New York, No. 1:05-cv-08136 (Chin, J.)
BRIEF FOR APPELLEE
DARALYN J. DURIE
JOSEPH C. GRATZ
DURIE TANGRI LLP
217 Leidesdorff Street
San Francisco, CA 94111
[phone]

July 3, 2014

SETH P. WAXMAN
LOUIS R. COHEN
DANIEL P. KEARNEY, JR.
WEILI J. SHAW
WILMER CUTLER PICKERING
HALE AND DORR LLP
1875 Pennsylvania Avenue, NW
Washington, DC 20006
[phone]

RULE 26.1 DISCLOSURE STATEMENT
Appellee Google Inc., by its undersigned attorneys, hereby states, pursuant
to rule 26.1 of the Federal Rules of Appellate Procedure, that it has no parent
corporation and that there is no publicly held corporation that owns 10% or more
of its stock.
Dated: July 3, 2014

/s/ Seth P. Waxman
SETH P. WAXMAN

-i-

TABLE OF CONTENTS
Page
TABLE OF AUTHORITIES ................................................................................... iv
INTRODUCTION .....................................................................................................1
STATEMENT OF THE CASE ..................................................................................4
A.

How Google Books Works ...................................................................4
1.

The digitization process ..............................................................6

2.

Searches using Google Books .....................................................7

3.

The effects of Google Books ....................................................12

B.

Google Books’ Benefits For Authors ..................................................14

C.

Google Books And Licensing Markets ...............................................16

D.

Copies Downloaded By Libraries .......................................................17

E.

Prior Proceedings ................................................................................19

F.

Decision Below ...................................................................................20

SUMMARY OF ARGUMENT ...............................................................................21
ARGUMENT ...........................................................................................................25
I.

GOOGLE BOOKS’ SEARCH TOOL IS FAIR USE .................................................25
A.

Google Books’ Uses Are Transformative ...........................................26
1.
2.

B.

Snippet display serves Google Books’
transformative purposes ............................................................30
The fact that Google is a commercial entity does
not weigh against fair use .........................................................32

Nature Of The Works At Issue: All Of Plaintiffs’ Books
Are Published, And Most Are Non-Fiction ........................................35

- ii -

C.

The Amount And Substantiality Of The Portions Of
Plaintiffs’ Works Used By Google Are Appropriate To
Its Transformative Search Tool ...........................................................38

D.

Google’s Uses Have No Adverse Effect On The Market
For Or Value Of Plaintiffs’ Works......................................................44
1.
2.

Plaintiffs’ arguments about security lack merit ........................50

3.
II.

Google Books causes no harm to the market for
Plaintiffs’ works ........................................................................45

Google Books provides immense public benefits ....................52

PLAINTIFFS’ DISTRIBUTION CLAIM FAILS BECAUSE THERE IS NO
INFRINGEMENT AS TO THE LIBRARY COPIES..................................................53
A.

Google Assists The Libraries’ Fair Uses ............................................53

B.

Google Books Does Not Distribute Scans “To The
Public” .................................................................................................56

CONCLUSION ........................................................................................................58

- iii -

TABLE OF AUTHORITIES
CASES
Page(s)
A.V. ex rel. Vanderhye v. iParadigms, LLC, 562 F.3d 630 (4th Cir.
2009) .......................................................................................................passim
American Broadcasting Cos. v. Aereo, Inc., No. 13-461, 2014 WL
2864485 (U.S. June 25, 2014) .................................................................56, 57
American Geophysical Union v. Texaco Inc., 60 F.3d 913 (2d Cir.
1994) .......................................................................................................passim
Authors Guild, Inc. v. Google Inc., 721 F.3d 132 (2d Cir. 2013) ............................ 20
Authors Guild, Inc. v. Google Inc., 770 F. Supp. 2d 666 (S.D.N.Y.
2011) .............................................................................................................. 19
Authors Guild, Inc. v. HathiTrust, No. 12-4547, 2014 WL 2576342
(2d Cir. June 10, 2014) ...........................................................................passim
Bill Graham Archives v. Dorling Kindersley Ltd., 448 F.3d 605
(2d Cir. 2006)..........................................................................................passim
Blanch v. Koons, 467 F.3d 244 (2d Cir. 2006) .................................................passim
Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569 (1994) ...............................passim
Cariou v. Prince, 714 F.3d 694 (2d Cir. 2013)........................................................ 39
Castle Rock Entertainment, Inc. v. Carol Publishing Group, Inc.,
150 F.3d 132 (2d Cir. 1998) ....................................................................33, 49
Clapper v. Amnesty International USA, 133 S. Ct. 1138 (2013) .......................51, 52
Davis v. The Gap, Inc., 246 F.3d 152 (2d Cir. 2001) .............................................. 26
Harper & Row Publishers, Inc. v. Nation Enterprises, 471 U.S. 539
(1985) ....................................................................................................... 36, 40
Infinity Broadcasting Corp. v. Kirkwood, 150 F.3d 104 (2d Cir. 1995) ................. 55
Kelly v. Arriba Soft Corp., 336 F.3d 811 (9th Cir. 2003) .................................passim
- iv -

Maxtone-Graham v. Burtchaell, 803 F.2d 1253 (2d Cir. 1986) .............................. 34
Perfect 10, Inc. v. Amazon.com, Inc., 508 F.3d 1146 (9th Cir. 2007) ..............passim
Princeton University Press v. Michigan Document Services, Inc.,
99 F.3d 1381 (6th Cir. 1996) ................................................................... 29, 34
Rumsfeld v. FAIR, Inc., 547 U.S. 47 (2006) ............................................................ 27
Sony Computer Entertainment, Inc. v. Connectix Corp., 203 F.3d 596
(9th Cir. 2000) ............................................................................................... 52
Sony Corp. of America v. Universal City Studios, Inc., 464 U.S. 417
(1984) ............................................................................................................. 50
Stewart v. Abend, 495 U.S. 207 (1990) ................................................................... 36
Swatch Group Management Services Ltd. v. Bloomberg L.P., No. 122412, 2014 WL 2219162 (2d Cir. May 30, 2014) ...................................29, 33
Thompson v. County of Franklin, 15 F.3d 245 (2d Cir. 1994) ................................ 27
Twentieth Century Music Corp. v. Aiken, 422 U.S. 151 (1975) .............................. 53
UMG Recordings, Inc. v. MP3.com, Inc., 92 F. Supp. 2d 349
(S.D.N.Y.2000) ........................................................................................29, 55
Video Pipeline, Inc. v. Buena Vista Home Entertainment, Inc.,
342 F.3d 191 (3d Cir. 2003) .......................................................................... 32
DOCKETED CASES
American Broadcasting Cos. v. Aereo, Inc., No. 13-461, Transcript of
Oral Argument (U.S. June 25, 2014) ............................................................. 30
STATUTES
17 U.S.C.
§ 101 .............................................................................................................. 56
§ 106 ........................................................................................................44, 56
§ 107 .......................................................................................................passim
§ 121 .............................................................................................................. 18

-v-

OTHER AUTHORITIES
Egnal, Marc, Evolution of the Novel in the United States: The
Statistical Evidence, 37 Soc. Sci. Hist. 231 (2013) (JA13571380) .............................................................................................................. 13
Leval, Pierre N., Nimmer Lecture: Fair Use Rescued, 44 UCLA L.
Rev. 1449 (1997) ........................................................................................... 33
Leval, Pierre N., Toward a Fair Use Standard, 103 Harv. L. Rev.
1105 (1990) .............................................................................................passim
Rampell, Catherine, The ‘New Normal’ Is Actually Pretty Old, New
York Times Economix Blog (Jan. 11, 2011), http://economix
.blogs.nytimes.com/2011/01/11/the-new-normal-is-actuallypretty-old/ ...................................................................................................... 13
RULES
Fed. R. Civ. P. 23(b)(3)............................................................................................ 19

- vi -

INTRODUCTION
Google Books gives readers and scholars a dramatically new way to find
books located in major research libraries: full text searches using queries the users
themselves formulate. Google Books thus enables users to identify, determine the
relevance of, and locate books they might never have found using old-fashioned
card or electronic catalogs.
Google Books does not replace or supersede printed (or electronic or audio)
books. To read any substantial portion of a book, the reader must obtain it
elsewhere, such as by buying it from a store or borrowing it from a library. What
Google Books does is enable users to find the books they want to read; in doing so,
as the district court observed, it “enhances the sales of books to the benefit of
copyright holders.” SPA25.
Google Books serves not only individuals—anyone connected to the Internet
can now conduct research that used to be simply impossible—but also the
advancement of human knowledge. It enables users to draw new connections
among books and topic areas that might not have been possible before. And it has
already been midwife to new fields of scholarly research—for example, work
based on analysis of the historical usage of ideas, words, and phrases.
Plaintiffs argue that fair use doctrine does not allow any of this, because the
search tool was made possible by Google’s full-text digital scans of about twenty

million books, some of which are still protected by copyrights. They urge the
Court to focus solely on the “mechanical conversion” (Br. 32) of paper book to
digital scan, without regard to the purpose and function of Google’s book search
tool. Plaintiffs contend that search engines may properly search for digital content
in materials already available online, “there to be located and accessed” (Br. 39),
but that it is unlawful to create digital scans that make it vastly easier to discover
printed or electronic books, which were published to be read but may be virtually
undiscoverable on library shelves. And Plaintiffs claim that Google Books harms
their works simply by including them in a search tool, even though the search tool
enables their books to be found and does not serve as a substitute for their books.
All of those arguments are unavailing. As this Court’s decision in Authors
Guild, Inc. v. HathiTrust, No. 12-4547, 2014 WL 2576342 (2d Cir. June 10, 2014),
makes clear, a search tool like Google Books is a fair use principally because it is a
“quintessentially transformative use” that “does not ‘supersede[] the objects [or
purposes] of the original creation.’” Id. at *7 (quoting Campbell v. Acuff-Rose
Music, Inc., 510 U.S. 569, 579 (1994)) (alterations in original). The district court
properly applied the statutory fair use factors to Google Books’ uses, see 17 U.S.C.
§ 107, and likewise concluded that they are fair use. It found the uses at issue
highly “transformative” because they do not “supersede” books but “add[]
something new,” a greatly improved way of finding them. Campbell, 510 U.S. at

-2-

579; see SPA20-21. The full-text scans are “used as raw material, transformed in
the creation of new information,” i.e., a search tool that gives readers a new way to
discover books of interest to them. Blanch v. Koons, 467 F.3d 244, 251 (2d Cir.
2006) (internal quotation marks omitted).
As to harm, there is no serious argument that Google Books “serves as a
substitute for the original work.” HathiTrust, 2014 WL 2576342, at *9. A user’s
search queries will identify every book mentioning a particular topic or using a
particular phrase. A further inquiry about a book will produce up to three short
“snippets” (each approximately one-eighth of a page)—to allow the user to
understand the context in which a search term appears. Like a paper index,
bibliography, or card catalog—but far more helpfully—Google Books enables
users to find books of interest but does not substitute for obtaining them elsewhere
and reading them.
Plaintiffs in fact concede the obvious value of a full-text search tool. The
heart of their position is that if the Google Books search tool were enjoined,
someone else would create a similar tool and pay royalties to copyright owners.
But this aspiration is unsupported by evidence, factually implausible, and legally
irrelevant. There is no reason to believe any creator of a search tool would pay
authors for the “privilege” of helping readers find their published books or that
authors would refuse inclusion unless paid; on the contrary, as the Authors Guild

-3-

has acknowledged, authors benefit from search tools that help readers find their
books. See SPA25. In any event, as a legal matter, “any economic ‘harm’ caused
by [a] transformative use[] does not count” in the fair use analysis, HathiTrust,
2014 WL 2576342, at *9, because a copyright holder cannot preempt a fair use by
creating, let alone by merely imagining, a competitor in that use.
STATEMENT OF THE CASE
A.

How Google Books Works

Digital technology and the Internet now enable users to find an enormous
range of previously hard-to-access information. A large portion of humanity’s
knowledge, however, lies on the printed pages of books that need to be discovered
and located before they can be read. Many of these books are found in major
research libraries.
Before Google Books, finding relevant books within these massive
collections was—even for persons with access—laborious, time-consuming,
inexact and, frequently, impossible. JA223, ¶ 3. Researchers and other persons
looking for books on a particular topic had to rely on card or electronic catalogs.
JA414-415, ¶¶ 40-41. But a catalog, even an electronic one, generally contains
only a small amount of bibliographic information about each book, including about
four subject headings assigned by the librarian that catalogued the book. JA415,
¶ 42. Whether a relevant book turned up in a user’s “subject” search depended on

-4-

whether the assigned subject headings happened to include the user’s search terms.
Id. And if a reader succeeded in locating a book with a relevant subject heading,
she often had to locate the physical book, travel to or borrow from a perhaps
distant library, and examine the book—even to determine its actual relevance to
her interests. JA415, ¶ 43.
Google Books dramatically improves this process. Readers can now search
the full texts of the collections of several major research libraries, using search
terms the readers themselves select rather than the subject fields a librarian has
selected. Suppose, for example, a reader is interested in the role of Archimedes in
the history of the calculus. Searching card catalogs for books with either
“Archimedes” or “calculus” in the title or the listed subject fields would obviously
return many thousands of items; searching for both terms in the title or subject
fields might return some items of interest but would surely miss many, because the
cataloger did not have the same interests as the reader. A text search, designed by
the reader, will find passages that would have been effectively impossible to find
before. The “snippet” feature would then give the reader a bit of context so that he
can decide whether to pursue the book itself. This method is far more
comprehensive, efficient, and precise than those previously available; it allows
discovery of information that would have been effectively impossible to discover
before. JA415-416, ¶ 44.

-5-

1.

The digitization process

Creating a full-text digital search tool necessarily started with scanning the
full texts of physical books. Beginning in 2004, pursuant to agreements with the
University of Michigan, the University of California, and other institutions with
major research libraries, Google began making full text digital scans of the books
in their libraries. JA223, ¶ 5.1 The agreements specified that, after scanning the
books, Google would return the physical books and provide a mechanism for each
library to create its own digital copies of the books scanned from its own
collections. JA223-224, ¶ 5. The agreements also provided that the libraries
“agree[d] to abide by the copyright laws with respect to the copies they make.”
SPA6; JA223-224, ¶ 5. 2
Under the agreements, Google, over time, took the libraries’ books to
Google scan centers, scanned them, used optical character recognition (OCR) on
page images to extract the text and convert it to machine-readable form, and
created an index of the entire text of each work. JA224, ¶ 6; JA396. That index
1

Other participating research libraries include those at Harvard University,
Stanford University, Oxford University, the University of California, Columbia
University, Princeton University, Ghent University, and Keio University, as well
as the Austrian National Library and the New York Public Library. JA208, ¶ 11.
2

The activities described in this paragraph were known as the “Library
Project.” Google Books also contains a separate program, now known as the
“Partner Program,” that involves the display of extensive excerpts from books with
the authorization of their publishers. That program is not at issue here. See infra
p. 14.
-6-

links words or phrases appearing in each book with the locations in each book
where the word or phrase is found. JA224, ¶ 6.
All of the files involved in this process are securely stored on servers that are
not accessible from the public Internet and are protected by the same security
systems that protect Google’s own confidential information. JA393, ¶ 3. Google
retains the original scanned images of book pages because as image-processing,
character-recognition, and indexing technology improve, Google’s systems use the
original images to derive more accurate text and create a more complete and
efficient index of the book’s contents. CA378-379, 381.
Google Books has scanned more than 20 million books, including both
copyrighted works and works in the public domain. JA403. These include almost
every type of book: biographies, treatises, memoirs, textbooks, novels, reference
works, books of poetry, instruction manuals, children’s books, etc. JA396, ¶ 4.
The composition of the corpus is heavily weighted, however, towards the types of
books that one might expect to find in the major research libraries whose
collections make up the corpus. The vast majority of the works in these libraries
are non-fiction, SPA22, and most are out of print, JA396, ¶ 4.
2.

Searches using Google Books

When a user enters a search term on the Google Books website, Google uses
its index to return a list of books in which the term appears. JA224, ¶ 8. As an

-7-

example, the following are search results for the query “Steve Hovley” (a baseball
player):

JA242.
After entering this query, a user can click a particular search result to see the
“About the Book” page for the relevant book. Id. “About the Book” pages include
links to buy the book online and to find the book in a nearby library where that is
possible. JA225, ¶ 9. These pages do not contain advertising (other than
bookseller information), and Google receives no payment in connection with the
“buy the book” links. Id.; JA588, ¶ 3. Below is the first part of the “About the

-8-

Book” page displayed when one clicks on Ball Four in the search results page
pictured above:

JA244.
For many works, the “About the Book” page will also include “snippets” of
text from the book. JA225, ¶ 10. Each “snippet” is about one-eighth of a page in
length and is intended to provide context showing how the user’s search term
appears in the book, enabling the user to determine whether the book is likely to
contain information of interest. Id.; JA208, ¶ 10; JA213, ¶ 25; SPA20. The next
part of the “About the Book” page for Ball Four, shown below, reveals that there

-9-

are 34 references to Hovley in the book, but only three of those 34 references are
shown to the user, as shown below:

JA246.
Google Books restricts the display of snippets in several ways to ensure that
users cannot read or copy any significant amount of a book. JA393, ¶ 4. No more
than three snippets are displayed in response to a given query, even if the same
search term appears many other times in the book. Id. Google always displays the
same snippets in response to a particular search, no matter how many times the
search is run. Id. Only the first responsive snippet on any given page is displayed,
- 10 -

and thus, no matter how many times a search term occurs on a page, and how
many times the user searches, the user is limited to seeing only one snippet from
that page. Id. In addition, snippets are “fixed”—that is, they do not “slide” around
the search term or overlap; a user therefore cannot summon the next “snippet” by
searching for a term that appears near the end of the previous one. Id. Google
Books further limits snippets by “blacklisting” (i.e., making unavailable for snippet
view) at least one snippet per page and one page out of ten per book. Id. Google
Books applies additional restrictions to prevent automated downloading of
snippets. JA394, ¶ 6. These and other measures prevent users from using multiple
searches to read or reconstruct any substantial portion of a book. JA393, ¶ 5.3
Snippet view is not available for all books. Google does not display snippets
of works broken into short “chunks” where even a small snippet could potentially
substitute for the reader’s accessing the book itself—for example, reference works
like dictionaries or cookbooks, or books of short poems. JA225, ¶ 11. Works that
are in the public domain can be viewed and downloaded in their entirety. Id.
Google also excludes any works a rightsholder has asked Google not to display.
Any rightsholder can exclude a book from both search and snippet view by filling
out an online form that has been available since 2005. JA396, ¶ 5.
3

As a result, an “attacker” could not obtain an entire book even if he had a
physical copy of the book in front of him and selected unique words from each part
of each page to use as search terms. JA393-394, ¶ 5.
- 11 -

3.

The effects of Google Books

Google Books has transformed the ways readers and scholars find books.
JA415, ¶ 44. Users anywhere in the world can now search the collections of major
research libraries located in New York, Michigan, Massachusetts, and California,
as well as libraries abroad. All users need is an Internet connection. See JA210214, ¶¶ 17-30; JA402-403, ¶ 5; JA414-415, ¶¶ 40-41; JA276-277. Google Books
thus allows users to find books they would otherwise be unable to find. JA210211, ¶¶ 18-19; JA214, ¶ 30; JA338, ¶¶ 12-14.
For example, a researcher interested in “500 Pearl Street” would find no
results in the Library of Congress catalog in response to this search term. See
JA1345. Entering the same search term in Google Books, on the other hand,
returns results pointing, as one might expect, to the U.S. Court Directory and the
page in the Congressional Record containing Senator Schumer’s floor speech
introducing the bill to name the courthouse for Daniel Patrick Moynihan. See
JA1347. But Google Books’ search results also point to other books including
Arthur Bonner’s 1996 Alas! What Brought Thee Hither?: The Chinese in New
York, 1800-1950, which reveals that in the late 1800s, 500 Pearl Street was the site
of a cigar factory operated by William A. Hong, one of three Chinese Civil War
veterans from New York. See JA1349 (showing extended excerpt from Partner
Program display). A further search for “Hong Kee Kang” (Mr. Hong’s Chinese

- 12 -

name, see id.) provides a link to a Library Project book with more information on
Mr. Hong (an out-of-print collection entitled Chinese America: History and
Perspectives 1996) and a snippet from that book. See JA1351. The snippet—
which displays two full sentences and two partial sentences—appears to be part of
a lengthier discussion of Mr. Hong’s life, but the “About the Book” page does not
show the rest of that discussion. Instead, it provides links to locate the book in
library collections and to purchase the used book through Amazon.com. See
JA1353 (snippet view page); JA1355 (Amazon.com listing).
The Google Books project has also enabled new kinds of research. SPA1011. Google itself has used the text from the Google Books corpus as an input in its
Ngram research project, which has opened up new fields of digital humanities
research. JA226, ¶ 15. For example, through the Ngram project, scholars have
studied the evolution of the American novel and how changes in word usage reflect
changes in American society and values.4 Economists have used the tool to
analyze the prevalence of the phrases “new normal” and “new plateau” as a
historical signal of market sentiment.5

4

See Egnal, Evolution of the Novel in the United States: The Statistical
Evidence, 37 Soc. Sci. Hist. 231 (2013) (JA1357-1380).
5

See Rampell, The ‘New Normal’ Is Actually Pretty Old, New York Times
Economix Blog (January 11, 2011), http://economix.blogs.nytimes.com/2011/01/
11/the-new-normal-is-actually-pretty-old/.
- 13 -

B.

Google Books’ Benefits For Authors

Google Books also provides benefits to authors and publishers, who have
long encouraged readers to browse books because they believe that browsing
promotes sales. Historically, browsing occurred in bookstores, where books are
typically displayed on shelves or tables (rather than kept behind the counter) to
facilitate browsing. See JA287-288; JA348, ¶ 15. Today, rightsholders
affirmatively seek to make their works available for online browsing through
Amazon’s “Search Inside the Book” and Google’s Partner Program. JA288-289;
JA348, ¶ 17. Through those programs, publishers or other rightsholders authorize
display of much longer excerpts of books than those available in the Library
Project (for the Partner Program, generally about 20% of a book) and receive no
compensation in return.6 JA226, ¶ 13; JA1609-1610, ¶ 14. They do so because
browsing facilitates readers’ discovery of new books, and “[d]iscoverability is one
of the most important factors in an individual title’s chance of success.” JA345346, ¶ 7. Therefore, as an expert on the publishing industry explained, “[s]earch
tools such as Google Books, which make it easier for authors to be found, benefit
rather than harm authors.” JA329, ¶ 3.
6

Although Google formerly shared advertising revenue with participants in
the Partner Program, that revenue was so low that Google decided to remove
advertisements from the Program in 2011 and phased them out completely by early
2012. JA1603, ¶¶ 5-7. Google has never displayed advertisements on the “About
the Book” page for books in the Library Project. JA225, ¶ 9.
- 14 -

The Authors Guild agrees that online browsing promotes book sales. See
JA297; see also JA212-213, ¶¶ 21-27; JA217, ¶¶ 40-42. It recommends to its
members that they make the entire first chapter of a book freely available on the
Internet because “allowing a book to be browsed in this way promotes the sale of
the book.” JA294.7 William Morris Endeavor, a leading literary agency, instructs
its authors that it “appears to be in an author’s best interest to have their work
come up in a search through [Google Books], just as website rankings are desirable
in connection with Google searches.” JA323-324. As a representative from
William Morris explained in a deposition below: “I think any tool that helps
readers or buyers find your product above someone else’s is beneficial.” JA321.
Many authors also recognize an entirely different benefit from Google Books:
assistance with their own research. See, e.g., JA1386 (testimony of executive
director of Authors Guild); Brief of Amici Curiae American Library Association et
al., at 1-9 (D. Ct. Dkt. No. 1048); JA226, ¶ 15.
By contrast, Plaintiffs have not introduced evidence of a single lost sale due
to Google Books. The individual Plaintiffs have admitted that they have not lost
7

This recommendation is made in connection with the Authors Guild’s “Back
in Print” program. That program allows authors to digitize their out-of-print books
and make them available for sale through a company called iUniverse. JA291-293.
The Authors Guild similarly advised its members with regard to Google’s
Partner Program that “[w]e think [it] will likely prove to be useful in promoting
certain titles. Midlist and backlist books that are receiving little attention, for
example, may benefit from additional exposure in searches.” JA1446-1447.
- 15 -

sales or that they are unaware of any such harm. JA1421 (Plaintiff Miles); JA1434
(Plaintiff Bouton); JA1437-1438 (Plaintiff Goulden). Nor is there evidence that
any other Authors Guild member has lost sales as a result of Google Books.
Indeed, the vast majority of authors do not object to inclusion of their works in
Google Books and do not believe that they suffer harm as a result of Google
Books. See Decl. of Hal Poret in Opp’n to Pls.’ Mot. for Class Certification, Ex. 1,
at 14 (D. Ct. Dkt. No. 1001-1) (survey results).
C.

Google Books And Licensing Markets

Authors and publishers traditionally have not received license fees for the
types of uses Google Books makes of their works, namely indexing and the display
of short snippets. Nor does the evidence in the record provide reason to believe
that copyright holders would ever obtain such license fees. Copyright holders have
long made extended book excerpts (not merely snippets) available for free. JA346347, ¶¶ 10-14. They continue to do so today, as noted above, through programs
such as Amazon’s “Search Inside the Book” and Google’s Partner Program.
JA288-289, 296; JA348, ¶ 17; JA226, ¶ 13; JA1609-1610, ¶ 14. As one former
publishing executive testified, “you don’t try to charge people money for doing
things that might help you make a sale.” JA349, ¶ 18. Plaintiffs offered no
evidence below that Google Books’ indexing and display is any different.

- 16 -

Libraries likewise have not paid license fees in the past to index or display
snippets from books they own, and they are unlikely to do so in the future. JA282;
JA403, ¶ 5(c). As preliminary efforts at digitization demonstrated, libraries lack
the resources to engage in comprehensive digitization and indexing efforts like
Google Books. JA406-408, ¶¶ 13-23.
Organizations such as the Copyright Clearance Center (“CCC”), which
Plaintiffs point to (Br. 15), have traditionally licensed uses that allow reading the
full text of entire works or substantial portions of works—that is, uses that
substitute in whole or in part for reading the original work. See, e.g., American
Geophysical Union v. Texaco Inc., 60 F.3d 913, 930 (2d Cir. 1994). No evidence
in the record establishes that such organizations can establish a market for uses,
like Google Books, that encompass search and only limited display of text, or that
such an arrangement would generate license fees. Plaintiffs’ expert could not
identify a single instance in which the CCC has licensed a book for indexing and
snippet display, or anything similar. See JA1478, 1490-1491, 1501.
D.

Copies Downloaded By Libraries

Pursuant to its agreement with Google, a library that has submitted a book to
be scanned may make and download a copy of the scan of its book using a system
called the Google Return Interface (GRIN). JA396, ¶ 6. The agreement further
requires the library to abide by copyright laws with respect to these copies. SPA6;

- 17 -

JA223-224, ¶ 5. No library may obtain a digital copy created from another
library’s book—even if both libraries own identical copies of that book (although
libraries may delegate that task to a technical service provider such as HathiTrust).
JA397, ¶ 8.
To make this copy using the GRIN system, a library submits a request that
triggers the automatic creation of an encrypted copy of the book on a secure
Google server. JA397, ¶ 8. Each library has a unique encryption key. Id. The
library may then download this encrypted copy. Id. Some but not all of the books
in the Google Books corpus have been copied by the libraries that own them, using
GRIN. JA397, ¶ 9. Where a library takes no action with respect to a particular
book, the GRIN system does not do anything with that book. JA397, ¶ 9.
Libraries have used the copies they downloaded (again, always of books
they themselves already owned) for several purposes. They have jointly created
their own full-text index of the works in their collections to enable full-text
searches by their own patrons. JA278. That search tool, known as the HathiTrust
Digital Library, is similar to Google Books, albeit with somewhat more limited
search functionality. HathiTrust, 2014 WL 2576342, at *1 (noting lack of snippet
view). And as 17 U.S.C. § 121 expressly permits, libraries have made the digital
copies available to the blind and others who cannot read print material, so that such
“user[s] can obtain access to the contents of works in the digital library using

- 18 -

adaptive technologies such as software that converts the text into spoken words, or
that magnifies the text.” HathiTrust, 2014 WL 2576342, at *1; see also JA272.
There is no evidence that any library has reduced its purchase of books as a result
of downloading scans using GRIN. JA281.
E.

Prior Proceedings

In 2005, the Authors Guild and several individual authors sued Google for
copyright infringement. The suit sought injunctive and declaratory relief, and the
individual authors also sought statutory damages. On October 28, 2008, the parties
filed a proposed class settlement agreement, which was later amended by the
parties and resubmitted to the district court for approval on November 13, 2009.
The agreement would have permitted Google to continue the project and, under
certain conditions, to make full electronic versions of the works available online.
The district court rejected the proposed settlement on March 22, 2011. See
Authors Guild, Inc. v. Google Inc., 770 F. Supp. 2d 666, 686 (S.D.N.Y. 2011). On
December 12, 2011, several individual Plaintiffs moved pursuant to Rule 23(b)(3)
to certify a proposed class consisting of natural persons in the United States who
hold copyright interests in one or more books Google has scanned from library
collections. The district court granted the motion for class certification on May 31,
2012. JA42, D. Ct. Dkt. No. 1023. On interlocutory appeal, this Court vacated the
class certification order. The Court explained: “[W]e believe that the resolution of

- 19 -

Google’s fair use defense in the first instance will necessarily inform and perhaps
moot our analysis of many class certification issues, including those regarding the
commonality of plaintiffs’ injuries, the typicality of their claims, and the
predominance of common questions of law or fact[.]” Authors Guild, Inc. v.
Google Inc., 721 F.3d 132, 134 (2d Cir. 2013). The Court remanded “for
consideration of the fair use issues” after finding that such a remand would not
result in prejudice to the parties. Id. at 134-135.
F.

Decision Below

The parties then filed cross-motions for summary judgment on Google’s fair
use defense. The district court granted summary judgment for Google, denied
summary judgment for Plaintiffs, and dismissed the case. In reaching this result,
the district court considered each of the four fair use factors outlined in 17 U.S.C.
§ 107. The court found that the first factor, the “‘purpose and character of the
use,’” “strongly favors a finding of fair use” because Google’s use is “highly
transformative.” SPA19, 22. The court found the second factor, “‘the nature of
the copyrighted work,’” also favors fair use because “the books at issue are
published and available to the public” and “the vast majority … are non-fiction.”
SPA22-23. The court found the third factor, “‘the amount and substantiality of the
portion used,’” to weigh slightly against fair use. SPA23-24. Finally, the court
found the fourth factor, “the effect of the use upon the potential market for or value

- 20 -

of the copyrighted work,” to “weigh[] strongly in favor of … fair use” because
Google Books does not “negatively impact the market for books” and indeed
“improves book sales.” SPA24-25. The court also found that “Google Books
provides significant public benefits,” and “advances the progress of the arts and
sciences, while maintaining respectful consideration for the rights of authors and
other creative individuals, and without adversely impacting the rights of copyright
holders.” SPA26. Together, the district court held, these factors weighed in favor
of fair use.
The district court also granted Google summary judgment with respect to the
copies of scanned books downloaded by the libraries. SPA26. The court
explained that “Google provides the libraries with the technological means to make
digital copies of books that they already own,” and that the libraries’ activities with
respect to the copies they obtained through GRIN were fair use. SPA26-27. The
court concluded that “Google’s actions in providing the libraries with the ability to
engage in activities that advance the arts and sciences” itself “constitute[d] fair
use.” SPA27.
SUMMARY OF ARGUMENT
The essence of the fair use inquiry is whether the use at issue furthers
copyright’s goal of promoting science and the arts or whether it impedes that
purpose. Google Books provides an unprecedented method for discovering books

- 21 -

relevant to a reader’s interests, revealing stores of human knowledge that would
otherwise be difficult or impossible to find. The Google Books index also enables
new kinds of digital research in fields like lexicography and history. At the same
time, Google Books does not displace book purchases or diminish incentives for
authors to advance human knowledge by creating new works; to the contrary, it
benefits authors by enabling readers to find books and by fostering the advance of
knowledge and scholarship.
Analysis of the statutory factors in 17 U.S.C. § 107 confirms that Google
Books’ uses are fair. The critical question in analyzing the first factor, the purpose
and character of the use, is whether the use is transformative—that is, whether the
use serves a new and different purpose, or whether it merely supersedes the
original use. As the district court correctly concluded, Google Books is highly
transformative. Books are written to be read. Google Books, by contrast, is a tool
for discovering books, not reading them. This Court in HathiTrust held a similar
book search tool highly transformative, and other courts have likewise held that
search tools are transformative, even where they display limited portions of
copyrighted material to enable users to determine whether a search result is
relevant. Further, the fact that Google is a commercial entity has little relevance
because Google Books’ uses are transformative. The first factor therefore weighs
strongly in favor of fair use.

- 22 -

The second factor—the nature of the copyrighted work—generally turns on
two distinctions. The district court correctly found that the first distinction—
whether the works in question are published—weighs in favor of fair use because
all of the works in the Google Books corpus have been published. The second
distinction—whether the works are creative—bears little significance here because
Google Books does not exploit the creative expression of the works in its corpus.
Google Books does not allow users to read expressive works as they would the
original books. Moreover, the vast majority of works in the Google Books corpus
are nonfiction. The district court correctly concluded that the second factor weighs
in favor of fair use.
The third factor also weighs in favor of fair use. This factor concerns the
amount and substantiality of the portion used, and asks whether the amount used is
reasonable in relation to the purpose of the copying. Google Books’ limited
display of snippets serves its transformative purpose of enabling users to discover
books by allowing users to determine whether a book identified in search results is
relevant to their interests. The fact that Google digitized the full texts of the works
in the Google Books corpus has little bearing on the inquiry, because Google had
to do so in order to create its transformative search tool. This Court’s HathiTrust
decision precludes any argument that scanning entire books weighs against fair use
where, as here, it enables the creation of a transformative search tool.

- 23 -

Google Books also has no adverse effect on the potential market for or value
of the copyrighted work (the fourth factor). The analysis of this factor is
straightforward because Google Books serves an entirely different market function
than the original work—it enables users to discover books, not read them.
Plaintiffs have produced no evidence that Google Books displaces book purchases;
instead, by enabling readers and scholars to discover books, Google Books
enhances book sales. Plaintiffs’ arguments regarding a potential licensing market
are irrelevant because “[l]ost licensing revenue counts under Factor Four only
when the use serves as a substitute for the original and the full-text-search use does
not.” HathiTrust, 2014 WL 2576342, at *10. In any event, there is no meaningful
licensing market for uses like Google Books, and Plaintiffs cannot preempt
Google’s transformative uses simply by positing such a market. Commercial
success in the book industry depends on getting books discovered, and it would be
inconsistent with the economics of that industry for authors or publishers to
demand license revenue from services that enable discovery. Indeed, there is no
evidence that any rightsholder has ever been paid for the uses at issue here.
Further, Plaintiffs’ worry about security issues is speculative and lacks any factual
basis.
Finally, Plaintiffs’ claims about the libraries’ acquisition of digital scans of
their books lack merit. The libraries’ uses of these copies—in their separate full-

- 24 -

text search tool and to expand access to print-disabled individuals—constitute fair
use. As the district court concluded, Google provided libraries the technological
means to enable the libraries’ fair use of works they already owned. There is also
no distribution “to the public” because a digital scan can be accessed only by the
library whose collection contains the actual physical book from which the scan was
created.
ARGUMENT
I.

GOOGLE BOOKS’ SEARCH TOOL IS FAIR USE
Google Books’ creation and maintenance of a comprehensive book search

engine that allows readers to discover books relevant to their interests is fair use.
See Authors Guild, Inc. v. HathiTrust, No. 12-4547, 2014 WL 2576342, at *11 (2d
Cir. June 10, 2014) (“the doctrine of fair use allows the Libraries to digitize
copyrighted works for the purpose of permitting full‐text searches”). The nonexclusive statutory factors to be considered in the fair use determination are: (1)
the purpose and character of the use, (2) the nature of the copyrighted work, (3) the
amount and substantiality of the portion used, and (4) the effect of the use upon the
potential market for or value of the copyrighted work. 17 U.S.C. § 107. The
statutory factors are not a “scorecard” but rather “direct courts to examine the issue
from every pertinent corner and to ask in each case whether, and how powerfully, a
finding of fair use would serve or disserve the objectives of the copyright.” Leval,

- 25 -

Toward a Fair Use Standard, 103 Harv. L. Rev. 1105, 1110-1111 (1990). Google
Books’ search tool is fair use because, most importantly, it is a transformative
use—it “adds to the original something new with a different purpose and a
different character”—and does not serve as a substitute for books or in any way
threaten to harm sales of Plaintiffs’ books. HathiTrust, 2014 WL 2576342, at *7.
A.

Google Books’ Uses Are Transformative

“The heart of the fair use inquiry” is the first statutory factor: “‘the purpose
and character of the use.’” Davis v. The Gap, Inc., 246 F.3d 152, 174 (2d Cir.
2001) (quoting 17 U.S.C. § 107(1)); see also Bill Graham Archives v. Dorling
Kindersley Ltd., 448 F.3d 605, 608 (2d Cir. 2006). The “central purpose of this
investigation is to see … whether the new work merely supersedes the objects of
the original creation or instead adds something new, with a further purpose or
different character … . [I]t asks, in other words, whether and to what extent the
new work is transformative.” Campbell, 510 U.S. at 579 (citations and internal
quotation marks omitted). A transformative use “is one that serves a new and
different function from the original work and is not a substitute for it.” HathiTrust,
2014 WL 2576342, at *6.
The first factor weighs heavily in Google’s favor because “the creation of a
full‐text searchable database is a quintessentially transformative use.” HathiTrust,
2014 WL 2576342, at *7. In HathiTrust, this Court held that the libraries’ own

- 26 -

full-text search tool—created using the very Google scans at issue in this case—is
transformative. The results of a full-text search, the Court explained, are “different
in purpose, character, expression, meaning, and message from the page (and the
book) from which it is drawn,” and there is “little or no resemblance between the
original text and the results of the … full‐text search.” Id. Further, the Court
found, the libraries’ full‐text search function does not “‘supersede[] the objects [or
purposes] of the original creation,’” because “[t]here is no evidence that the
Authors write with the purpose of enabling text searches of their books.” Id. That
conclusion applies equally here. As the district court concluded, Google Books is
“highly transformative” because it “transforms expressive text into a
comprehensive word index that helps readers, scholars, researchers, and others find
books.” SPA19. It is a tool for discovering books and the connections among
them—not for reading books—and it does not supersede the original works. See
SPA21.8

8

In HathiTrust, this Court also held that the Authors Guild lacked standing to
seek “an injunction for copyright infringement on [its] members’ behalf” because
“§ 501 of ‘the Copyright Act does not permit copyright holders to choose third
parties to bring suits on their behalf.’” 2014 WL 2576342, at *14. This Court
need not reach that issue in affirming the district court’s decision, because the
unquestioned standing of the individual plaintiffs is sufficient to sustain this
appeal. See Rumsfeld v. FAIR, Inc., 547 U.S. 47, 53 n.2 (2006). In the event of a
remand, however, this Court should direct the district court to dismiss the Authors
Guild’s claims for lack of standing. See Thompson v. County of Franklin, 15 F.3d
245, 248-249, 251-252 (2d Cir. 1994) (obligation to address standing issues
- 27 -

Plaintiffs do not deny that Google Books’ search uses are transformative—
and indeed cannot, given this Court’s recent decision in HathiTrust. They instead
focus on the “mechanical conversion” of the printed book into digital text—the
scanning that makes full-text search possible—without reference to Google Books’
search functions. Br. 32. But as HathiTrust recognizes, such scanning cannot be
assessed in a vacuum and is permissible if the scans are for a fair-use purpose. See
2014 WL 2576342, at *8 (“creat[ing] digital copies of all the books in [the
libraries’] collections” was permissible “[b]ecause it was reasonably necessary …
to enable the full-text search function”). Like the libraries’ copies in HathiTrust,
Google’s scans cannot be separated from the use to which those scans are put—the
Google Books search tool.
There is also no merit to the claim that Google has “merely articulat[ed] a
new ‘purpose’” for a use that amounts simply to copying or repackaging the
original work. Br. 31. HathiTrust decisively rejects this argument, explaining that
a full-text search tool does not “‘merely repackage[] or republish[] the
original[s],’” but rather “adds to the original something new with a different
purpose and a different character.” 2014 WL 2576342, at *7 (quoting Leval, 103

“extends ‘to the prudential rules of standing that, apart from Art. III’s minimum
requirements, serve to limit the role of the courts in resolving public disputes’”).
- 28 -

Harv. L. Rev. at 1111).9 This is not a case involving “untransformed duplication,”
in which “the value generated by the secondary use is little or nothing more than
the value that inheres in the original.” American Geophysical Union v. Texaco
Inc., 60 F.3d 913, 923 (2d Cir. 1994).10 The original use of the books at issue here
was to be read. Google has converted those books into digital scans to build a
book search tool that allows users to discover books, not to provide a different
format for users to read books (which they cannot do using Google Books).
Courts have consistently held such search uses to be transformative. See 2014 WL
2576342, at *7-8. In Kelly v. Arriba Soft Corp., for example, the Ninth Circuit
held that a defendant’s use of “thumbnail” images from various websites in its
search engine was transformative because it served to “improve[e] access to
information on the internet” and did not supersede the “artistic expression” of the
original use. 336 F.3d 811, 819 (9th Cir. 2003); see also Perfect 10, Inc. v.
Amazon.com, Inc., 508 F.3d 1146, 1165 (9th Cir. 2007) (“[a]lthough an image may
9

This Court has likewise rejected the argument made by Plaintiffs’ amici that
a transformative use must result “in the creation of a new expressive work,”
MPAA Br. 6, holding that “a secondary work ‘can be transformative in function or
purpose without altering or actually adding to the original work.’” Swatch Grp.
Mgmt. Servs. Ltd. v. Bloomberg L.P., No. 12-2412, 2014 WL 2219162, at *8 (2d
Cir. May 30, 2014).
10

The other cases Plaintiffs cite on this point are inapposite. See, e.g.,
Princeton Univ. Press v. Michigan Document Servs., Inc., 99 F.3d 1381, 1389 (6th
Cir. 1996) (publication of course packets with “verbatim copies” of book
excerpts); UMG Recordings, Inc. v. MP3.com, Inc., 92 F. Supp. 2d 349, 351
(S.D.N.Y.2000) (conversion of music into MP3 format).
- 29 -

have been created originally to serve an entertainment, aesthetic, or informative
function … a search engine puts images in a different context so that they are
transformed into a new creation.”); A.V. ex rel. Vanderhye v. iParadigms, LLC,
562 F.3d 630, 639 (4th Cir. 2009) (defendants’ “use of plaintiffs’ works had an
entirely different function and purpose than the original works; the fact that there
was no substantive alteration to the works does not preclude the use from being
transformative in nature”).11 HathiTrust relies significantly on these decisions and
makes clear that their reasoning fully accords with this Court’s application of fair
use doctrine.
1.

Snippet display serves Google Books’ transformative
purposes

In light of the HathiTrust decision, Plaintiffs’ only possible argument on this
point is that Google Books’ display of “snippets” somehow renders its search
function non-transformative. The district court rightly rejected that argument. A
single search query on Google Books returns up to three snippets of text, each
11

There is also no merit to the argument, advanced by amici for Plaintiffs, that
the district court’s analysis is somehow inconsistent with the United States’
international obligations. See Int’l Publishers Ass’n et al. Br. 6-19; Int’l Authors
Forum Br. 3-6. The United States has stated the position that “existing U.S.
copyright law properly construed is fully sufficient to comply with our
international obligations.” Transcript of Oral Argument at 25, American Broad.
Cos. v. Aereo, Inc., No. 13-461 (U.S. June 25, 2014). International law thus adds
nothing to the Court’s application of the fair use analysis. If Google is correct that
its uses are fair use under U.S. copyright law, then the United States’ international
obligations are satisfied.
- 30 -

about one-eighth of a page, that provide important context about how search terms
are used in a book. See supra pp. 9-11. As the district court concluded, “[t]he
display of snippets of text for search is similar to the display of thumbnail images
of photographs for search [as in Kelly, Perfect 10] or small images of concert
posters for reference to past events [as in Bill Graham Archives], as the snippets
help users locate books and determine whether they may be of interest.” SPA20.
The snippets narrow a search: they “act as pointers directing users to a broad
selection of books” actually relevant to them, not as a substitute for the books
themselves. Id. A search for all books that mention “Archimedes” may yield tens
of thousands of books, and a user interested only in Archimedes’ explanation of
levers could use snippets to identify those that are most relevant. Even Plaintiffs
testified that they do not consider snippets to substitute for books. See JA1424,
1439.
Plaintiffs nevertheless complain that the display of snippets amounts to use
of “copyrighted language verbatim without incorporating it into any new work.”
Br. 37. But Google Books’ search tool is a transformative new work, and such
display is entirely permissible in those circumstances, as this Court has explained:
An artist may employ copyrighted photographs in a new work that
uses a fundamentally different artistic approach, aesthetic, and
character from the original. An internet search engine can display low
resolution versions of copyrighted images in order to direct the user to
the website where the original could be found. A newspaper can
publish a copyrighted photograph (taken for a modeling portfolio) in
- 31 -

order to inform and entertain the newspaper’s readership about a news
story.
HathiTrust, 2014 WL 2576342, at *5 (citations omitted); see also Bill Graham
Archives, 448 F.3d at 609-611 (inclusion in a biography of unaltered copyrighted
concert photos deemed transformative). There is also no merit to Plaintiffs’
contention that search engine cases like Kelly and Perfect 10 are different because
“webpages are there to be located and accessed” online, while authors do not
“upload the content of their books onto the Web.” Br. 39. Google Books’ search
tool is transformative precisely because it allows a user to search for and find
relevant content in millions of hard copy books that were effectively impossible to
search before. JA210, ¶ 17. Snippet display makes that tool significantly more
valuable without in any way superseding use of the original books.12
2.

The fact that Google is a commercial entity does not weigh
against fair use

The fact that Google is a for-profit business has little bearing on the fair use
analysis, as the district court recognized. “Since many, if not most, secondary
users seek at least some measure of commercial gain from their use, unduly
12

Plaintiffs also claim (Br. 39) that Video Pipeline, Inc. v. Buena Vista Home
Entertainment, Inc., 342 F.3d 191 (3d Cir. 2003), is “far more analogous to this
case” than Kelly and Perfect 10. Not so. Video Pipeline did not involve a search
tool, but rather defendant’s creation and display of its own movie “previews,”
which superseded movie trailers the defendant had previously licensed from movie
studios. Id. at 195. This Court’s HathiTrust decision therefore properly did not
rely on Video Pipeline but rather looked to the analyses in Kelly and Perfect 10.
- 32 -

emphasizing the commercial motivation of a copier will lead to an overly
restrictive view of fair use.” Texaco, 60 F.3d at 921; see also Castle Rock Entm’t,
Inc. v. Carol Publ’g Grp., Inc., 150 F.3d 132, 142 (2d Cir. 1998) (Court does “not
give much weight to the fact that the secondary use was for commercial gain”);
Leval, Nimmer Lecture: Fair Use Rescued, 44 UCLA L. Rev. 1449, 1456 (1997)
(“The heart of fair use lies in commercial activity. Most undertakings in which we
expect to find well-justified instances of fair use are commercial.”). Moreover,
“the more transformative the new work, the less will be the significance of other
factors, like commercialism, that may weigh against a finding of fair use.”
Campbell, 510 U.S. at 579. This Court has accordingly found fair use even in
cases where the defendants received a direct commercial benefit from the use. See,
e.g., Swatch Grp. Mgmt. Servs. Ltd. v. Bloomberg L.P., No. 12-2412-CV, 2014 WL
2219162, at *7 (2d Cir. May 30, 2014) (recording of earnings call disseminated by
Bloomberg “subscription service available to paying users”); Blanch, 467 F.3d at
253 (defendant sold art work found to be fair use for $2 million); Bill Graham
Archives, 448 F.3d at 612 (book containing copyrighted images was “a commercial
venture” but nonetheless fair use).
Plaintiffs’ attempt to depict Google Books as a rapacious commercial
endeavor is also misplaced. The district court noted that “Google does not sell the
scans it has made of books for Google Books; it does not sell the snippets that it

- 33 -

displays; and it does not run ads on the About the Book pages that contain
snippets.” SPA21-22. And users of Google Books—readers, students, and
scholars—are engaged in precisely the type of educational and other uses
historically favored under the first fair use factor. See Maxtone-Graham v.
Burtchaell, 803 F.2d 1253, 1262 (2d Cir. 1986) (noting that the “commercial
nature of a use is a matter of degree, not an absolute,” and concluding that the
“educational elements of” the book at issue “far outweigh the commercial aspects
of the book”). Google Books’ uses have also opened up new fields of scholarly
research using data and text mining. SPA10-11. Given the vastly expanded
opportunities for research and scholarship made possible by Google Books,
Plaintiffs’ attempt to compare Google Books to a copy shop that profits from
selling extensive unlicensed book excerpts is baseless. See Br. 30 (citing
Princeton Univ. Press v. Michigan Document Servs., Inc., 99 F.3d 1381, 1386 (6th
Cir. 1996)).
Finally, Plaintiffs’ contention that Google Books represents an attempt to
“str[ike] back” at Amazon by luring away potential book purchasers (Br. 2-3)
involves a blatant distortion of the record. To support this claim, Plaintiffs
repeatedly highlight material related not to the Library Project at issue here but to
Google’s Partner Program, which involves extensive display of book excerpts
with rightsholder permission. See supra p. 14. The 2003 presentation cited by

- 34 -

Plaintiffs (Br. 2-3, 25, citing CA440) does not even mention library-copy scanning
or snippet display. Rather, as the full context makes clear (see CA438-454), the
presentation discusses the potential full text display of works with rightsholder
permission, in what later became the Partner Program. The document has no
relevance to the purposes of the Library Project, and the district court properly
ignored it.13
B.

Nature Of The Works At Issue: All Of Plaintiffs’ Books Are
Published, And Most Are Non-Fiction

The second factor is “the nature of the copyrighted work.” 17 U.S.C.
§ 107(2). This factor recognizes “that some works are closer to the core of
intended copyright protection than others, with the consequence that fair use is
more difficult to establish when the former works are copied.” Campbell, 510 U.S.
at 586. The district court correctly concluded that this factor “favor[s] a finding of
fair use.” SPA23.
As this Court has noted, “‘[t]wo types of distinctions … have figured in the
decisions evaluating the second factor.’” Blanch, 467 F.3d at 256. The first is
“whether the work is published or unpublished, with the scope for fair use
13

Plaintiffs also selectively quote from objections by competitors such as
Yahoo to the rejected class settlement, in an attempt to show that Google Books’
purpose is commercial advantage. See Br. 26-27. These objections were
principally addressed to the terms of the settlement, which permitted Google to sell
books in certain circumstances, not to the basic text search functions of the Library
Project. See JA137-138; see also, e.g., JA64.
- 35 -

involving unpublished works being considerably narrower.” Id. In other words, a
particular use may be permissible as to “a published work” even though it is not
permissible as to the same work “before its release” the public. Campbell, 510
U.S. at 586; see also Harper & Row Publishers, Inc. v. Nation Enters., 471 U.S.
539, 564 (1985) (rooting distinction in an “author’s right to control the first public
appearance of his expression”). All of the works involved in this case were
previously published. See JA1429. The district court accordingly found that “the
books at issue are published and available to the public,” which weighs in favor of
fair use. SPA22-23.
The second distinction is “whether the work is expressive or creative, such
as a work of fiction, or more factual.” Id. “[F]air use is more likely to be found in
factual works than in fictional works.” Stewart v. Abend, 495 U.S. 207, 237
(1990). However, the creativeness of a work “has limited weight in [the fair use]
analysis” where—as here—the use is transformative. Bill Graham Archives, 448
F.3d at 612; see HathiTrust, 2014 WL 2576342, at *8 (concluding that second
factor was of limited significance with respect to the HathiTrust search engine
created by Google’s library partners).
In this case, the district court correctly observed that “the vast majority of
the books in Google Books are non-fiction.” SPA22; see JA418, ¶ 3. Plaintiffs
argue that the Google Books corpus nonetheless contains some fictional works

- 36 -

among “a vast number of non-fiction works.” Br. 41-42. But the creative nature
of a work bears little significance where the fair use does not seek to “exploit [the
work’s] creative virtues.” Blanch, 467 F.3d at 257. Thus, the fact that some of the
works involved were creative in nature did not weigh against fair use where the
defendant created a database of student papers for the purpose of detecting
plagiarism. See iParadigms, 562 F.3d at 641-642. In that case, the creative nature
of any included works (which included fiction and poetry) did not favor either
party because the fair use was “unrelated to any creative component.” Id. at 642.
The same applies here. Google Books can help a reader find novels that the
reader would not otherwise find, but it provides no way to obtain the pleasure of
reading it. If a reader finds a book he wants to read, he must still purchase it from
a retailer or borrow it from a library. As a result, Google Books’ use of any
creative works in its corpus does not “exploit” their “creative virtues.” Its function
is only to enable readers to discover books relevant to their interests—a purpose
that applies equally to creative works and factual ones. The existence of some
creative works in the Google Books corpus therefore does not alter the district
court’s conclusion that the second factor favors fair use.

- 37 -

C.

The Amount And Substantiality Of The Portions Of Plaintiffs’
Works Used By Google Are Appropriate To Its Transformative
Search Tool

The third factor is “the amount and substantiality of the portion used in
relation to the copyrighted work as a whole.” 17 U.S.C. § 107(3). The relevant
question under this factor is whether the amount used is “reasonable in relation to
the purpose of the copying.” Campbell, 510 U.S. at 586-587 (emphasis added).
This inquiry “must take into account that the ‘the extent of permissible copying
varies with the purpose and character of the use.’” Bill Graham Archives, 448 F.3d
at 613; see also Leval, 103 Harv. L. Rev. at 1123 (“[T]he qualitative aspect of the
third test … may be more important than the quantitative.”).
Google Books’ “About the Book” page displays only up to three “snippets”
per search—each snippet limited to about one-eighth of a page—so that the user
can determine whether the book is relevant. See supra pp. 9-11. These snippets
are reasonable in relation to Google Books’ transformative use and include the
minimum of text necessary to achieve the function of Google’s book search tool.
As with the thumbnail images in Kelly and Perfect 10, the snippets in Google
Books “allow users to … decide whether to pursue more information about”
particular search results. Kelly, 336 F.3d at 821. The snippets include only a small
fraction of a book and are of the minimal “size and quality necessary to ensure the
reader” can determine whether the book is potentially relevant to his or her

- 38 -

research. Bill Graham Archives, 448 F.3d at 613. Eliminating snippet display
would greatly “reduc[e] the usefulness of the … search engine,” Kelly, 336 F.3d at
821, because a user would have little context to determine whether a book is even
relevant to the user’s interests. The third factor accordingly weighs in favor of fair
use.
Plaintiffs argue that Google’s snippet display is “unnecessary,” Br. 40,
because the search tool in HathiTrust “does not display to the user any text from”
the relevant book and, “[c]onsequently, the user is not able to view either the page
on which the term appears,” HathiTrust, 2014 WL 2576342, at *1. There is no
merit to this contention. The issue under the third factor is not whether a search
engine would be possible without snippet display, but whether snippet display
serves the transformative purpose of Google Books’ search tool and is “reasonable
in relation to th[at] purpose.” Campbell, 510 U.S. at 586-587; see also Cariou v.
Prince, 714 F.3d 694, 710 (2d Cir. 2013) (“[T]he law does not require that the”
person making the fair use “may take no more than is necessary.”). Moreover,
while a search tool without snippet display may be helpful to users who, as in
HathiTrust, have ready access to university libraries and can obtain books listed in
search results comparatively easily to determine whether they are relevant, such a
tool is much less useful for the broader public, for whom it is more difficult to
obtain and sift through the physical books to determine their relevance. Google

- 39 -

Books’ snippets provide important context that allows users to refine their searches
and identify relevant books before making the effort to obtain them, but that does
not substitute for reading the books themselves.
Plaintiffs also suggest that the amount used is substantial because some
snippets might happen to display the “heart of the book” for some books in the
Google Books corpus. Br. 44 (quoting Harper & Row, 471 U.S. at 565). Plaintiffs
have offered no evidence that snippets somehow appropriate the “heart” of any of
their books, and the very idea is implausible: Even a great passage does not
capture the “heart” of a published full-length book in the sense Plaintiffs intend,
obviating reading the book.14 Moreover, Plaintiffs’ comparison to Harper & Row
is inapt. There, the defendant published the most newsworthy portions of an
upcoming memoir in order to “scoop[]” the authorized publication and thereby
“supplant[ed] the copyright holder’s commercially valuable right of first
publication.” 471 U.S. at 562. It was only in that context that the Court found the
defendant’s use of 300 words (obtained by reading the book and selecting the most
newsworthy portions, and not through search) constituting the “heart of the book”
to weigh against fair use. Id. at 565-566. Courts have reached a different
conclusion where the copying or display served a transformative use, and did not
14

Google disables snippet view for works in which a small snippet could
potentially substitute for the work itself, such as books of poetry and cookbooks.
See supra p. 11.
- 40 -

seek to exploit the expressive value of the original work. For example, Bill
Graham Archives, Kelly, Perfect 10, and iParadigms all upheld the copying of
works in their entirety—a process that necessarily entailed copying the “heart” of
each work—yet concluded that the third factor did not weigh against fair use
because the copying was reasonable in relation to the transformative purpose at
issue.
Plaintiffs also repeatedly contend that Google “display[s] … 78% of the
verbatim text of millions of in-copyright books” to the public. Br. 37; see also,
e.g., Br. 11, 23, 42, 44. That contention—apparently referring to the fact that 22%
of each book is “blacklisted” entirely from outside access—is unsupported by the
record, which makes clear that no plausible user could actually obtain anywhere
close to 78% of a book. Google Books displays no more than three snippets from a
book in response to a search query. See supra pp. 9-11. In other words, for a book
of 500 pages, a query will display less than one-tenth of one percent of the book.
Plaintiffs contended below that “users could put in multiple searches, varying
slightly the search terms, to access an entire book.” SPA24. But the district court
rejected that argument, and Plaintiffs wisely do not renew it here. As the district
court explained: It is implausible “that someone would take the time and energy to
input countless searches to try and get enough snippets to comprise an entire book.
Not only is that not possible as certain pages and snippets are blacklisted, the

- 41 -

individual would have to have a copy of the book in his possession already to be
able to piece the different snippets together in coherent fashion.” Id.15
Plaintiffs argue that the third factor should nonetheless weigh against fair
use because Google digitized entire works to create its search tool. But this Court
has “concluded that … copying” even the entirety of a work “does not necessarily
weigh against fair use.” Bill Graham Archives, 448 F.3d at 613. That is because
“it may be necessary to copy the entire copyrighted work, in which case Factor
Three does not weigh against a finding of fair use.” HathiTrust, 2014 WL
2576342, at *8. Thus, this Court in HathiTrust found that the third factor favored
the libraries “[b]ecause it was reasonably necessary … to make use of the entirety
of the works in order to enable the full-text search function.” Id. Similarly, Bill
Graham Archives found that the third factor did not weigh against fair use even
though the defendant’s book about the Grateful Dead reproduced concert posters in
their entirety. See 448 F.3d at 613; see also Perfect 10, 508 F.3d at 1167-1168
(noting that “use of the entire photographic image was reasonable in light of the
purpose of a search engine”).
In any event, this Court’s cases addressing the third factor have looked
principally to the amount of the original work ultimately displayed to the public,

15

Google Books also employs several other security measures to prevent users
from obtaining any substantial, contiguous portion of a book. See supra pp. 10-11.
- 42 -

and not to the amount copied during any intermediate step needed to enable the fair
use. In Blanch v. Koons, 467 F.3d at 247-248, for example, the artist Jeff Koons
made a digital copy of a copyrighted photograph and then incorporated portions of
it into a new work. This Court concluded that the third factor “weighs distinctly in
Koons’s favor” because Koons’s ultimate artwork included only part of the
original; the Court did not find it relevant that Koons copied the entire original as
an intermediate step in the process. Id.; see also Bill Graham Archives, 448 F.3d at
613 (evaluating reduced-size images ultimately published in historical work, rather
than copying of full-size original). Focusing the inquiry in this way makes sense
because the third factor goes to whether the defendant’s use is likely to “fulfill[]
demand for the original.” Campbell, 510 U.S. at 587-588. Whether Google Books
supplants market demand for the original depends on the amount displayed to
users, not the amount scanned—and the amount displayed is minimal.
Finally, Plaintiffs take issue with Google’s retention of digital scans of their
books, but that point does not weigh against fair use. The unrebutted evidence
below demonstrated that the digital scans are retained by Google to allow it to
maintain and improve its book search index. As technological improvements are
made (or errors detected), Google’s systems refer back to the original scanned
images in order to derive more accurate text and a more complete and efficient
index of the book’s contents. CA378-379, 381. These facts distinguish this case

- 43 -

from HathiTrust, in which the libraries’ digital scans played no part in their search
tool and the libraries acknowledged that there was no need to “retain … copies to
enable the full-text search use.” HathiTrust, 2012 WL 4808939, at *8 n.5. That
was so because the libraries’ search index is based on the text files that Google
extracts and refines through the digital scans it retains; the libraries are able to
obtain those refined text files for use in their search index. CA358, 381. By
contrast, Google’s retention of digital scans is critical to enhancing quality and
correcting errors in its search tool, and thus does not weigh against fair use.16
D.

Google’s Uses Have No Adverse Effect On The Market For Or
Value Of Plaintiffs’ Works

The fourth factor is “the effect of the use upon the potential market for or
value of the copyrighted work.” 17 U.S.C. § 107(4). For this factor to weigh
against fair use, a copyright holder must identify market harm that is
“‘substantially adverse.’” Campbell, 510 U.S. at 590. Moreover, as the Court
explained in HathiTrust, this “analysis is concerned with only one type of
economic injury to a copyright holder: the harm that results because the secondary
use serves as a substitute for the original work.” 2014 WL 2576342, at *9. “[A]ny

16

In addition, nothing in copyright law requires Google to destroy digital scans
that were lawfully created. The Copyright Act does not make retention of a copy
an exclusive right. See 17 U.S.C. § 106. Google Books’ retention of a digital scan
therefore cannot by itself amount to an act of infringement.
- 44 -

economic ‘harm’ caused by transformative uses does not count because such uses,
by definition, do not serve as substitutes for the original work.” Id.17
In HathiTrust, this Court held that full-text search is just such a
transformative use, and therefore that any potential revenue from licensing works
for full-text search is irrelevant under the fourth factor. That conclusion also
applies here. And even if such licensing markets were cognizable, plaintiffs have
not demonstrated any market harm; if anything, Google Books increases authors’
revenues by facilitating discovery of new books. The fourth factor therefore
weighs in favor of fair use.
1.

Google Books causes no harm to the market for Plaintiffs’
works

Plaintiffs’ principal argument on the fourth factor is little different than that
rejected by this Court in HathiTrust. They contend that Google Books “will …
undermin[e] existing and emerging licensing opportunities,” Pl.’s Br. 47, 50—in
other words, that someone else would pay for the uses Google is making. In
HathiTrust, the Authors Guild and its co-plaintiffs similarly argued that the
HathiTrust search engine harms “a market for licensing books for digital search …
17

HathiTrust’s conclusion on this point reflects prior Second Circuit precedent
holding that “copyright owners may not preempt exploitation of transformative
markets” even if those transformative markets would otherwise generate license
fees. Bill Graham Archives, 448 F.3d at 615; see also Campbell, 510 U.S. at 591
(where the original work and the second use “serve different market functions,” “it
is more likely that the new work will not affect the market for the original in a way
cognizable under this factor”).
- 45 -

because it allows patrons to search books without any need for a license.”
HathiTrust, 2014 WL 2576342, at *9. This Court categorically rejected that
theory, explaining that “it is irrelevant that the [defendants] might be willing to
purchase licenses in order to engage in this transformative use (if the use were
deemed unfair)” because “[l]ost licensing revenue counts under Factor Four only
when the use serves as a substitute for the original and the full-text-search use does
not.” Id. at *10.
The same reasoning fully disposes of Plaintiffs’ claims of harm under the
fourth factor as a matter of law. As explained (supra pp. 26-32), Google Books is
a transformative use and does not substitute for the original works; this Court has
already concluded that “full-text search … does not serve as a substitute for the
books that are being searched.” HathiTrust, 2014 WL 2576342, at *10. Potential
revenues from a transformative use are irrelevant to analyzing “harm to the
market” for the original use.
HathiTrust did not involve snippet display, but that does not alter the
analysis because snippet display is part of the transformative use and does not
substitute for the original works because it does not allow users to read books,
except by borrowing them from a library or purchasing them from another source.
Plaintiffs produced no evidence that any reader has read Google Books snippets
instead of buying a book, and Plaintiffs themselves testified that they do not

- 46 -

consider snippets a substitute for books. See JA1424, 1439. Because Google
Books is not a “market substitute for the original” and does not “usurp[] the market
of the original work[s],” Plaintiff’s allegation that a licensing market might
develop for transformative uses like Google Books is legally irrelevant.18
Moreover, even a legally relevant market has to be actual or plausible, and
Plaintiffs’ hypothetical licensing market is neither. This Court considers only “the
impact on potential licensing revenues for ‘traditional, reasonable, or likely to be
developed markets.’” Bill Graham Archives, 448 F.3d at 614-615. Plaintiffs have
offered no plausible evidence of any such market in which they might collect
potential licensing revenue for uses like Google Books. Plaintiffs point to
Amazon’s Search Inside the Book program and Google’s Partner Program as
evidence of the existence of a licensing market. But those programs are
substantially different from Google’s Library Project because they involve the
display of far more of the given works. The Partner Program, for example,
generally involves the full display of at least twenty percent of the pages of the
book. JA226, ¶¶ 13-14; JA250. Licensing arrangements for these programs—

18

For the same reasons, this case bears no resemblance to American
Geophysical Union v. Texaco Inc., 60 F.3d 913 (2d Cir. 1994), on which Plaintiffs
heavily rely. In Texaco, “the dominant purpose of the” copying was simply to
“multiply available copies while avoiding payment,” “thereby serving the same
purpose for which additional subscriptions are normally sold, or … for which
photocopying licenses may be obtained.” Id. at 920, 924-925.
- 47 -

which actually permit the reader to read a significant portion of the book—
therefore have little significance in evaluating potential license fees for a use like
Google Books, which entails the display of snippets of only an eighth of a page
each, three at a time.19
Moreover, these services actually demonstrate that it is unlikely if not
impossible for copyright holders to generate license revenue from uses like Google
Books. Neither Amazon’s Search Inside the Book program nor Google’s Partner
Program pays copyright holders anything for including their works. JA219, ¶ 49;
JA295-296; JA1603, ¶¶ 4-8. Google’s Partner Program once paid a share of its
advertising revenues, but it has not done so since 2012. JA1603, ¶¶ 4-8.
Publishers in the Partner Program today receive no payments whatsoever from
Google, and their participation is undiminished as a result. Id. (declaration of
Google product management director responsible for Google Books). In fact,
authors sometimes themselves pay third parties to place their works in Amazon’s
Search Inside the Book or Google’s Partner Program. JA219-220 n.61.

19

Plaintiffs also identify the rejected settlement in this case and full-text
digitization projects in other countries as models for a potential licensing market.
See Br. 16, 49-50. All of these projects, including the rejected settlement, involved
the full text display of whole books or substantial portions of books, thereby
providing a full or partial substitute for the original work, with its own market
value to would-be readers. None of these projects is limited to indexing and the
display of snippets. JA132-133, 216-218.
- 48 -

The common-sense economics of book publishing make any license revenue
from such uses extremely unlikely. As an expert on the book publishing industry
explained below:
[I]t seems to get the economics of book publishing backward to think
authors would or could charge a royalty for things that make it easier
for readers to find their books, learn about them, and hopefully buy
them. The market in which you pay me money to help potential
buyers find my book is not a reasonable or realistic market.
JA334; see also JA221, ¶ 54. Given this context, it is implausible that
rightsholders would be paid anything at all for the search and snippet uses of the
Library Project. The fact that Plaintiffs would like to be paid for these uses is
irrelevant: a rightsholder cannot preempt a transformative use like Google Books
simply by offering to sell a license for that use. See Castle Rock, 150 F.3d at 145
n.11.
Instead, as the district court concluded, “a reasonable factfinder could only
find that Google Books enhances the sales of books to the benefit of copyright
holders.” SPA25. That is because discovery—the ability to find a book—“is one
of the most important factors in an individual title’s chance of success.” JA345.
The purpose of Google Books is discovery. As the district court found, “both
librarians and their patrons use Google Books to identify books to purchase,” and
“[m]any authors have noted that online browsing in general and Google Books in

- 49 -

particular helps readers find their work, thus increasing their audiences.” Id.
“[T]here can be no doubt but that Google Books improves books sales.” Id.20
Ultimately, the fourth factor is focused on “the author’s incentive to create.”
Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417, 450 (1984). “[A]
use that has no demonstrable effect upon the potential market for, or the value of,
the copyrighted work need not be prohibited in order to protect” that incentive. Id.
Plaintiffs have offered no meaningful evidence that any author would be deterred
from creating a new work as a result of Google Books’ fair use, and there is no
reason why a service that only enables readers to discover books would deter
authors from writing them.
2.

Plaintiffs’ arguments about security lack merit

Having failed to identify any actual harm to their works, Plaintiffs offer a
speculative one—that hackers may someday breach Google’s security and steal
copies of their work. The Court rejected this argument in HathiTrust, explaining
that the “showing of the security measures taken by the Libraries is essentially
unrebutted” and that there was no basis “to conclude that a security breach is likely
to occur, much less one that would result in the public release of” the plaintiffs’
works. 2014 WL 2576342, at *11. The same is true here.
20

Even if a book is not in print, Google Books may reveal interest in the book
sufficient to warrant its republication. That is the entire premise of the Authors
Guild’s “Back in Print” program.
- 50 -

In the district court, Google presented evidence of the comprehensive
security measures it has taken to protect the information in Google Books from
theft. Google Books’ digital scans are stored on computers that are not connected
to the public Internet. JA393, ¶ 3. The scans are protected by the same security
systems Google uses to protect its own confidential information. Id. Plaintiffs’
own security expert praised these systems, explaining that “Google is fortunate to
have ample resources and top-notch technical talents” that others do not enjoy.
JA1558, 1570. As discussed above, Google employs a host of measures to prevent
users from reading or downloading any significant portion of a book. See supra
pp. 10-11. Any would-be thief who sets out to pirate certain books would find it
simpler to scan the physical books himself rather than attempt to hack Google’s
systems. JA394, ¶ 5. Plaintiffs have identified no thefts of works from Google
Books, and Google is aware of none. JA394, ¶ 7.21 In light of this evidence, there
is “no basis in the record on which to conclude that a security breach is likely to
occur, much less one that would result in the public release of the specific
copyrighted works belonging to any of the plaintiffs in this case.” HathiTrust,

21

It is also worth noting that some of the Plaintiffs make their works available
through other online services like Amazon’s Search Inside the Book, see, e.g.,
JA1589 (Plaintiff Bouton’s book Ball Four); JA1591 (Plaintiff Miles’ book The
Real Me), even though the standard contracts relating to those services provide for
no specific security protections. JA1593-1594.
- 51 -

2014 WL 2576342, at *11 (citing Clapper v. Amnesty Int’l USA, 133 S. Ct. 1138,
1143, 1149 (2013)).
Indeed, Plaintiffs cannot identify a single case in which a court has rejected
a fair use defense based on mere speculation that a thief might somehow find a
way to steal the fair use copies. Were it sufficient for a plaintiff to merely assert
that possibility, without regard to evidence of the quality of the defendant’s
security measures or whether any breaches have actually occurred, cases like
Blanch (scans of photographs), iParadigms (a database of student papers), and
Sony Computer Entm’t, Inc. v. Connectix Corp., 203 F.3d 596, 601 (9th Cir. 2000)
(computer software extracted from a computer chip), might well have turned out
differently.
3.

Google Books provides immense public benefits

The fourth factor “requires a balancing of ‘the benefit the public will derive
if the use is permitted and the personal gain the copyright owner will receive if the
use is denied.’” Bill Graham Archives, 448 F.3d at 613; cf. Blanch, 467 F.3d at
253 (“[C]ourts are more willing to find a secondary use fair when it produces a
value that benefits the broader public interest.”). Google Books offers readers and
scholars the possibility of discovering books through a method that was previously
impossible: searching the full text of all the works housed in some of the world’s
largest research libraries. Many of the works in the Google Books corpus, and the

- 52 -

knowledge contained therein, would be lost to would-be users without a tool like
Google Books. By facilitating the discovery of these works, Google Books greatly
improves the ability of researchers to build on the base of current human
knowledge. Google Books also makes possible original research through uses like
the Ngram project, which enable users to not only find but also create new
knowledge.
These public benefits are particularly relevant to the fourth factor because
they directly serve the purposes of copyright law. Copyright seeks to “promot[e]
broad public availability of literature … and the other arts,” Twentieth Century
Music Corp. v. Aiken, 422 U.S. 151, 156 (1975), and is “designed … to stimulate
activity and progress in the arts for the intellectual enrichment of the public,”
Leval, 103 Harv. L. Rev. at 1107. Plaintiffs have not offered any plausible
evidence that Google Books diminishes rather than facilitates that activity. The
fourth factor weighs strongly in favor of fair use.
II.

PLAINTIFFS’ DISTRIBUTION CLAIM FAILS BECAUSE THERE IS NO
INFRINGEMENT AS TO THE LIBRARY COPIES
A.

Google Assists The Libraries’ Fair Uses

The district court correctly held that Google was entitled to summary
judgment as to the scanned copies acquired by the libraries. As Judge Chin
explained, Google merely “provides the libraries with the technological means to
make digital copies of books that they already own,” with the purpose of
- 53 -

“advanc[ing] the libraries’ lawful uses of the digitized books consistent with the
copyright law.” SPA26-27; see also SPA6 (“The libraries agree to abide by the
copyright laws with respect to the copies they make”). The libraries’ uses include
creating a full-text search tool, maintaining copies for preservation, and expanding
access to books for print-disabled individuals. SPA27. Consistent with this
Court’s subsequent holding that the libraries’ search and print-disabled uses are
fair use, see HathiTrust, 2014 WL 2576342, at *13, the district court concluded
that these uses are “protected by the fair use doctrine” and that “the fair use
analysis set forth … with respect to Google Books applies here as well to the
libraries’ use of their scans.” SPA27-28.
Plaintiffs no longer seriously argue that the activities of Google’s library
partners are not fair use; nor could they, in light of the recent HathiTrust
decision.22 Their principal argument is instead that the libraries’ uses are
“irrelevant” to whether Google’s conduct in assisting them is fair use. Br. 36. The
suggestion is that even if the libraries’ uses, with all their significant public

22

Plaintiffs do assert that the libraries’ conduct “is in direct contravention of
the restrictions in Section 108 of the Copyright Act.” Br. 36 n.10. HathiTrust
expressly rejected the claim that § 108 restricts the availability of a fair use
defense. See 2014 WL 2576342, at *4 n.4 (“[W]e do not construe § 108 as
foreclosing our analysis of the Libraries’ activities under fair use, and we proceed
with that analysis”). Plaintiffs’ further suggestion (Br. 56-58) that § 108 of the
Copyright Act should be read to preclude Google’s fair use defense to scanning
works for its own book search tool is likewise meritless.
- 54 -

benefits, are fair use, that Google’s assistance in those uses is not—the libraries
evidently must find a way to do it on their own. That proposition finds no support
in the law. Plaintiffs rely (Br. 36) on Infinity Broadcasting Corp. v. Kirkwood, 150
F.3d 104 (2d Cir. 1998), but there is a wide difference between that case—in which
the defendant retransmitted radio broadcasts by telephone to the general public and
argued fair use based on the hypothetical motivations of those listeners—and
Google’s assistance in the libraries’ transformative search tool and other lawful
uses. The defendant’s retransmissions in Infinity Broadcast were simply another
way for the general public to hear the original broadcast, only in a new medium—
over the telephone instead of the radio. Id. at 106, 108; see also UMG Recordings,
Inc. v. MP3.com, Inc., 92 F. Supp. 2d 349, 351 (S.D.N.Y.2000) (copying served
only to allow users to listen to music in a different digital format). The
retransmissions served as a substitute for listening to the original copyrighted
transmission, regardless of the listeners’ hypothetical motivations. Google, on the
other hand, is not providing full-text digital scans to the general public and then
claiming transformativeness based on users’ hypothesized special purposes in
reading books in digital rather than printed form. Google allows the libraries to
acquire digital scans made from their own physical books, not to allow the libraries
or their general patrons to read the scans, but to advance the libraries’ own fair
uses of creating a search tool and expanding access to books for print-disabled

- 55 -

individuals. Google’s technological assistance directly advances those fair uses,
and indeed makes them possible.
B.

Google Books Does Not Distribute Scans “To The Public”

Plaintiffs’ distribution claim also fails for an independent reason: the
exclusive right of distribution extends only to distribution “to the public” and there
is no “public” distribution here. See 17 U.S.C. § 106(3) (exclusive right “to
distribute copies or phonorecords of the copyrighted work to the public by sale or
other transfer of ownership, or by rental, lease, or lending”). A library may
download a digital copy only if the copy was created by scanning the physical
book provided by that library; no library may obtain a copy created from another
library’s book. JA396-397, ¶¶ 6, 8; SPA6. There is therefore no distribution “to
the public” because Google Books only allows a library to obtain a digital scan that
was made from the library’s own physical book.
This interpretation of “to the public” in § 106(3) accords with the meaning
of “to the public” in the transmit clause of the Copyright Act’s public performance
definition. See 17 U.S.C. § 101 (“[t]o perform or display a work ‘publicly’ means
… to transmit or otherwise communicate a performance or display of the work …
to the public” (emphasis added)). The Supreme Court recently explained that a
transmission is not “to ‘the public’” if recipients receive the transmission “in their
capacities as owners or possessors of the underlying works.” American Broad.

- 56 -

Cos. v. Aereo, Inc., No. 13-461, 2014 WL 2864485, at *11 (U.S. June 25, 2014).
Whether recipients of a transmission “constitute ‘the public’ often depends upon
their relationship to the underlying work,” the Court reasoned, and an entity that
“transmits to large numbers of paying subscribers who lack any prior relationship
to the works” does so “to the public.” Id. By contrast, “an entity that transmits a
performance to individuals in their capacities as owners or possessors does not
perform to ‘the public.’” Id.
The latter conclusion applies here. Google does not allow members of the
general public to access digital scans of books—that is, “individuals who lack a
pre-existing relationship to the” books. Aereo, 2014 WL 2864485, at *11. Rather,
it allows access only to the library whose collection contains the actual physical
book from which the digital scan was created. See SPA26 (Google “provides the
libraries with the technological means to make digital copies of books that they
already own”). In that circumstance, there is no distribution “to the public”
because the library accesses the digital scans in its “capacit[y] as owner[] or
possessor[] of the underlying works.” Aereo, 2014 WL 2864485, at *11.

- 57 -

CONCLUSION
For these reasons, the Court should affirm the district court’s decision.
Respectfully submitted.
/s/ Seth P. Waxman
SETH P. WAXMAN
LOUIS R. COHEN
DANIEL P. KEARNEY, JR.
WEILI J. SHAW
WILMER CUTLER PICKERING
HALE AND DORR LLP
1875 Pennsylvania Avenue, NW
Washington, DC 20006
[phone]

DARALYN J. DURIE
JOSEPH C. GRATZ
DURIE TANGRI LLP
217 Leidesdorff Street
San Francisco, CA 94111
[phone]

Dated: July 3, 2014

- 58 -

CERTIFICATE OF COMPLIANCE
Pursuant to Fed. R. App. P. 32(a)(7)(C), the undersigned hereby certifies
that this brief complies with the type-volume limitation of Fed. R. App. P.
32(a)(7)(B).
1.

Exclusive of the exempted portions of the brief, as provided in Fed. R.

App. P. 32(a)(7)(B), the brief contains 13,968 words.
2.

The brief has been prepared in proportionally spaced typeface using

Microsoft Word 2010 in 14 point Times New Roman font. As permitted by Fed.
R. App. P. 32(a)(7)(C), the undersigned has relied upon the word count feature of
this word processing system in preparing this certificate.

/s/ Seth P. Waxman
SETH P. WAXMAN
July 3, 2014