`
`No. 18-956
`
`
`
`In the Supreme Court of the United States
`________________
`
`
`GOOGLE LLC,
`
`Petitioner,
`
`v.
`
`ORACLE AMERICA, INC.,
`Respondent.
`
`________________
`
`On Writ of Certiorari to the
`United States Court of Appeals
`for the Federal Circuit
`________________
`
`BRIEF OF AMICUS CURIAE DIGITAL JUSTICE
`FOUNDATION IN SUPPORT OF AFFIRMANCE
`________________
`
`
`
`
`
`Edward F. Cunningham
`Counsel of Record
`LAW OFFICES OF EDWARD F.
` CUNNINGHAM
`62 Cambridge Avenue
`Garden City, NY 11530
`(516) 328-3705
`ed@edcunninghamlaw.com
`
`Counsel for Amicus Curiae
`
`Andrew Grimm
`DIGITAL JUSTICE
` FOUNDATION
`15287 Pepperwood
`Drive
`Omaha, NE 68154
`(531) 210-2381
`
`Gregory Keenan
`DIGITAL JUSTICE
` FOUNDATION
`81 Stewart Street
`Floral Park, NY 11001
`(516) 633-2633
`
` February 19, 2020
`
`
`
`
`
`
`
`i
`
`TABLE OF CONTENTS
`TABLE OF AUTHORITIES ...................................... iii
`INTEREST OF AMICUS CURIAE ............................ 1
`SUMMARY OF ARGUMENT .................................... 1
`ARGUMENT ............................................................... 3
`I. GOOGLE CONFLATES SUBSTANTIAL SIMILARITY
`WITH COPYRIGHTABILITY AND CONFLATES
`NECESSITY WITH CONVENIENCE. .......................... 3
`A. The first question presented pertains to
`eligibility for copyright protection—not
`the scope of that protection. ......................... 3
`B. Google could have enabled Java
`programmers to transition to Android by
`making a code-conversion tool. .................. 13
`II. GOOGLE’S ALL-OR-NOTHING APPROACH TO
`FAIR USE DESTROYS MARKETS AND
`CONCENTRATES WEALTH. ................................... 21
`A. Even if it is seen as transformative,
`Google’s use was not fair. ........................... 21
`B. There are major, and growing, costs to
`an all-or-nothing approach on fair use. ..... 23
`C. A more nuanced approach to fair use
`shows greater fidelity to the Copyright
`Act and its animating purposes. ................ 31
`
`
`
`
`
`ii
`ii
`
`CONCLUSION ......................................................... 38
`CONCLUSION ......................................................... 38
`
`
`
`
`
`
`
`
`
`iii
`
`TABLE OF AUTHORITIES
`
`Cases
`Authors Guild v. Google, Inc.,
`804 F.3d 202 (2nd Cir. 2015). .................... 21, 24, 29
`Campbell v. Acuff-Rose Music, Inc.,
`510 U.S. 569 (1994). ................................... 21, 35, 36
`Camreta v. Greene,
`563 U.S. 692 (2011). ............................................... 33
`Feist Publ’ns, Inc. v. Rural Tel. Serv. Co.,
`499 U.S. 340 (1991). ....................................... 5, 9, 16
`Harper & Row, Publishers, Inc. v. Nation
`Enter’s, 471 U.S. 539 (1985). ................................. 22
`Kingdomware Techs., Inc. v. United States,
`136 S. Ct. 1969 (2016). ........................................... 32
`Obduskey v. McCarthy & Holthus LLP,
`139 S. Ct. 1029 (2019). ............................................. 4
`Oracle Am., Inc. v. Google Inc.,
`750 F.3d 1339 (Fed. Cir. 2014). ............................. 12
`Oracle Am., Inc. v. Google Inc.,
`872 F. Supp. 2d 974 (2012). ............................... 7, 13
`Perfect 10, Inc. v. Amazon.com, Inc.,
`508 F.3d 1146 (9th Cir. 2007). ............................... 24
`Star Athletica, L.L.C. v. Varsity Brands, Inc.,
`137 S. Ct. 1002 (2017). ........................................... 20
`Sun Microsystems, Inc. v. Microsoft Corp.,
`188 F.3d 1115 (9th Cir. 1999). ............................... 18
`
`Statutes
`17 U.S.C. § 101. ................................................ 4, 5, 14
`
`
`
`
`
`iv
`
`17 U.S.C. § 102(a). .................................................. 4, 5
`17 U.S.C. § 102(b). .................................................. 4, 6
`17 U.S.C. § 106. .......................................................... 6
`17 U.S.C. § 107. ........................................................ 32
`17 U.S.C. § 121. ........................................................ 11
`17 U.S.C. § 121A. ...................................................... 11
`17 U.S.C. §108. ......................................................... 35
`Other Authorities
`4 Nimmer on Copyright § 13.03[A]. ......................... 11
`Amanda Levendowski How Copyright Law
`Can Fix Artificial Intelligence’s Implicit
`Bias Problem 93 Wash. L. Rev. 579 (2018). .......... 37
`Andrew Gilden & Timothy Greene, Fair Use
`for the Rich and Fabulous?, 80 U. Chicago
`L. Rev. Dialogue 88 (2013). ................................... 29
`Andrew Gilden, Raw Materials and the
`Creative Process, 104 Geo. L.J. 355 (2016). .......... 29
`Antonin Scalia & Bryan A. Garner, Reading
`Law: The Interpretation of Legal Texts
`(2012). ..................................................................... 34
`Benjamin Sobel, Artificial Intelligence’s Fair
`Use Crisis, 41 Colum. J.L. & Arts 45
`(2017). ............................................................. passim
`Eman J. Coco et al., JPT: A Simple Java-
`Python Translator, 5 Computer
`Applications 1 (2018). ............................................ 17
`Goldstein on Copyright § 12.2. ................................. 35
`
`
`
`
`
`v
`
`Greg Vetter, Open Source Licensing and
`Scattering Opportunism in Software
`Standards, 48 B.C. L. Rev. 225 (2007). ................. 19
`Jaron Lanier, Who Owns the Future? (2013). ......... 27
`Margaret Radin, Online Standardization, 70
`Fordham L. Rev. 1125 (2002). ............................... 18
`Neil Netanel, Copyright’s Paradox (2008). .............. 28
`Paul Goldstein, Copyright’s Highway (2d Ed.
`2019). ................................................................ 23, 24
`Paul Goldstein, Fair Use in Context, 31
`Colum. J.L. & Arts 433 (2008). ....................... 22, 26
`Paul Goldstein, The Future of Software
`Protection, 47 U. Pitt. L. Rev. 1119 (1986). ............ 9
`Stephen Schaub, IKVM.NET Developer’s
`Guide,
`http://www.ikvm.net/devguide/intro.html
`(accessed Feb. 19, 2020)......................................... 17
`
`
`
`
`
`
`
`
`
`
`1
`
`INTEREST OF AMICUS CURIAE1
`The Digital Justice Foundation is a 501(c)(3) non-
`profit dedicated to preserving individual rights in dig-
`ital spaces. The Foundation has particular interest in
`the impact of digital technologies on civil liberties,
`personal privacy, individual intellectual-property
`rights, and individual economic well-being. The
`Foundation has particular concern for underrepre-
`sented users, artists, creators, employees, and inno-
`vators, especially those with limited access to law.
`
`SUMMARY OF ARGUMENT
`The Foundation writes as amicus curiae because
`many of the Briefs submitted by and in support of
`Google confuse necessity with convenience. They con-
`fuse copyrightability with substantial similarity, i.e.,
`the scope of protection. Likewise, Google and its
`amici wrongly conflate what Google did (appropriate
`a competitor’s platform by copying that competitor’s
`expression) with what everyday programmers do (use
`some of the method names of APIs to write programs).
`That distinction is why many of the policy con-
`cerns raised by Google’s amici are not implicated by
`this case. Despite the importance of those issues, they
`are inapposite here.
`
`
`
`1 No counsel for any party authored this Brief in whole or in
`part, and no person or entity other than the amicus, its mem-
`bers, or its counsel made a monetary contribution intended to
`fund the Brief’s preparation or submission. All parties have con-
`sented to the filing of this Brief.
`
`
`
`
`
`
`
`2
`
`The Foundation also has concerns about the all-or-
`nothing approach to fair use advocated by Google.
`This all-or-nothing approach to fair use destroys mar-
`kets and concentrates wealth. In the Foundation’s
`view, this all-or-nothing approach, if not course cor-
`rected, will have massively negative implications for
`ordinary Americans. Therefore, the Foundation pro-
`poses a doctrinal solution to permit more nuance.
`Numerous issues where the Foundation’s views
`might diverge from Oracle’s—injunctive relief, fair
`use by individual developers, substantial similarity,
`and scope of damages—are not before this Court.
`On the issues that are before this Court, the Foun-
`dation believes that affirmance is vitally important to
`the rights of individuals not before this Court. The
`Java APIs are clearly copyrightable under the stat-
`ute. Google’s uses were not fair uses, even though
`many interoperable programs made by Java and An-
`droid developers would be.
`For these reasons, the Foundation respectfully
`submits this Brief as amicus curiae and urges affir-
`mance.
`
`
`
`
`
`
`
`
`
`3
`
`ARGUMENT
`I. GOOGLE CONFLATES SUBSTANTIAL SIMILARITY
`WITH COPYRIGHTABILITY AND CONFLATES
`NECESSITY WITH CONVENIENCE.
`
`A. The first question presented pertains to
`eligibility for copyright protection—not
`the scope of that protection.
`1. The first question presented asks “Whether cop-
`yright protection extends to [what Google calls] a soft-
`ware interface.” Google. Br. at i. Google has not
`raised the more complicated and difficult questions on
`the scope of protection in APIs. Or, in doctrinal terms,
`Google has not raised whether its appropriation of
`Java APIs in making a competing mobile platform re-
`sulted in a substantially similar, i.e., infringing, work.
`No issue concerning the scope of Oracle’s copyright is
`before this Court.
`2. Being clear that the first question presented is
`a copyrightability question has twofold importance.
`First, ruling for Oracle would not end this lawsuit. It
`would merely reinstate the first jury’s verdict that
`Google infringed. Cf. District Court Docket No. 1089
`(first jury’s verdict that Google infringed but dead-
`locking on fair use). Then the parties would dispute
`damages. After, if it so wishes and did not forfeit the
`issues, Google could subsequently appeal questions
`about the scope of protection for Java APIs to the
`Court of Appeals. As of now, questions on the scope
`of the copyright protection afforded to APIs simply
`have not been raised on appeal.
`
`
`
`
`
`
`4
`
`3. Second, Google’s arguments truly pertain to the
`scope of protection, not the copyrightability of APIs.
`That’s why the distinction between copyrightability
`and scope of protection is essential to the disposition
`of this case. Google relies extensively on Section
`102(b). Yet Section 102(b) determines the scope of
`protection, not the subject matter of copyright. Sec-
`tion 102(b) establishes how far copyright protection
`“extend[s]”—not in what subject matter copyright
`protection “subsists[.]” Compare 17 U.S.C. § 102(b)
`(“extend”) with 17 U.S.C. § 102(a) (“subsists”). The
`question presented is whether copyright protection
`can “subsist[]” in an API.
`4. What is “most decisive” here “is the text of the
`Act itself.” See Obduskey v. McCarthy & Holthus
`LLP, 139 S. Ct. 1029, 1036 (2019). In Section 102(a),
`the Copyright Act states that “[c]opyright protection
`subsists […] in original works of authorship[.]” Then,
`the Act further states that “[w]orks of authorship in-
`clude” eight listed categories. 17 U.S.C. § 102(a)(1)-
`(8). Among them, the first category is a “literary
`work[.]” 17 U.S.C. § 102(a)(1). The Act defines liter-
`ary works as works “expressed in words, numbers, or
`other verbal or numerical symbols or indicia[.]” 17
`U.S.C. § 101.
`5. Here, copyright protection subsists in the APIs
`at issue. Oracle’s Java APIs meet the definition of lit-
`erary works because they are expressed in “words,
`numbers, or other verbal or numerical symbols or in-
`dicia.” Cf. Developers Alliance Br. at 4 (“part sym-
`bolic logic, part syntax, part symbolic notation, and
`part pseudo-English”).
`
`
`
`
`
`5
`
`6. The Java APIs are original. “Original, as the
`term is used in copyright, means only that the work
`was independently created by the author (as opposed
`to copied from other works), and that it possesses at
`least some minimal degree of creativity.” Feist
`Publ’ns, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 345
`(1991). “To be sure, the requisite level of creativity is
`extremely low; even a slight amount will suffice.”
`Ibid.
`7. The Java APIs are fixed. The Act permits copy-
`right protection for works “fixed in any tangible me-
`dium of expression, now known or later developed,
`from which they can be perceived, reproduced, or oth-
`erwise communicated[.]” 17 U.S.C. § 102(a). The
`Java APIs are fixed in many computers, from which
`they can be “perceived, reproduced [and] otherwise
`communicated.”
`8. Thus, Section 102(a) directs that the Java APIs
`are copyrightable. Copyright protection “subsists” in
`them. Nor does Section 102(b) subtract from this
`reading. Section 102(b) is not an exception to Section
`102(a). For example, Section 102(b) could specify
`where copyright protection “does not subsist” or “does
`not lie.” It doesn’t. Section 102(b) is not the converse
`of Section 102(a).
`9. After all, Congress, knows how to expressly ex-
`clude from a definition. Take the Act’s definition of a
`“work of visual art.” There Congress tells what it is
`and what it isn’t. See 17 U.S.C. § 101 (defining what
`a “‘work of visual art’ is” and what a “work of visual
`art does not include” (emphasis added)); see also ibid.
`(defining what an architectural work “includes” and
`“does not include”).
`
`
`
`
`
`6
`
`10. Rather, Section 102(b) takes as a given “copy-
`right protection for an original work of authorship”
`under Section 102(a). It then clarifies the scope of
`that protection. Specifically, Section 102(b) tells how
`far “copyright protection for an original work of au-
`thorship extend[s.]” Namely, it says that the copy-
`right protection granted in Section 102(a) never—or
`“[i]n no case”—“extend[s] to any idea, procedure, pro-
`cess, system, method of operation, concept, principle,
`or discovery[.]” Yet scope of protection for an API is
`not at issue. Rather, the threshold question of copy-
`rightability is what’s been presented by Google.
`11. The Act emphasizes that Section 102(b)’s limi-
`tation on scope applies “regardless of the form in
`which [the idea, procedure, etc.] is described, ex-
`plained, illustrated, or embodied in such work.” 17
`U.S.C. § 102(b) (emphasis added). Thus, no matter
`how the author has expressed herself, her copyright
`protection will never extend to an idea. Of course, this
`limitation on the scope of copyright protection does
`not mean she will lose the protection that she already
`received via the grant of a property interest in Section
`102(a). For example, she still has the exclusive right
`to reproduce her expression, 17 U.S.C. § 106(1)—just
`not any exclusivity to broader ideas therein, 17 U.S.C.
`§ 102(b).
`12. Yet this Court need not go through the trouble
`of line drawing between the ideas and the expressions
`in the Java APIs here. Because it is clear that the
`Java APIs are “original works of authorship,” they are
`entitled to some copyright protection, however thin.
`The first question presented should be answered in
`the affirmative. Yes, the Java APIs are copyrightable.
`
`
`
`
`
`7
`
`Their scope of protection, however, has not been
`raised here.
`13. The distinction between the existence of a pri-
`vate property right and the scope of that right is no
`illusory technicality unique to copyright—or even in-
`tellectual property. Take a simple real-property dis-
`pute: next-door neighbors fighting over where the
`fence goes. A threshold question would be whether
`the plaintiff even has a property interest at all. If he
`doesn’t, then the judge has no occasion to answer the
`literal line-drawing question regarding where one
`property line ends and the other begins. So too here,
`except that copyright is about art and technology ra-
`ther than lines in the dirt.
`14. Perhaps another analogy is edifying. The dis-
`trict court quoted the Copyright Office regarding its
`policy for registration of recipes and cookbooks. See
`Oracle Am., Inc. v. Google Inc., 872 F. Supp. 2d 974,
`984 (2012) (“Listings of ingredients, as in recipes, la-
`bels, or formulas [are not copyrightable]. […] [T]ext
`directions may be copyrightable, but the recipe or for-
`mula itself remains uncopyrightable.” (quoting U.S.
`Copyright Office, Circular 34)).
`15. Although the Java APIs are dramatically more
`complex than a cookbook and cookbooks do not “oper-
`ate a machine,” see EFF Br. at 20, the analogy to a
`cookbook is still useful to distinguish eligibility for
`copyright protection, Section 102(a), from the scope of
`that protection, Section 102(b).
`
`
`
`
`
`
`8
`
`16. In 1961, Julia Child imported French cuisine
`into kitchens across America when she published her
`cookbook, Mastering the Art of French Cooking. Like
`any cookbook, the authors might choose which recipes
`to include; which ones to exclude; how to organize and
`categorize the individual recipes among types (entrée
`vs. appetizer); how to pair wines and various entrees;
`and maybe some discussion of French table manners
`to boot.
`17. Presumably, Julia Child intended her audi-
`ence to actually cook these recipes (and wouldn’t sue
`them if they did). If Ms. Child opined on the French
`style to prepare a broth, dumplings, vegetables, and
`beef, she might not be surprised to hear one of her
`readers made a nice stew. If she opined on the perfect
`soufflé and a delectable vanilla sauce, a vanilla soufflé
`should not shock her. Likewise here, Sun Microsys-
`tems (and now Oracle America) clearly intend that
`programmers use the Java APIs to mix and match
`among the thousands of methods as they create pro-
`grams.
`18. Now, suppose that Ms. Child learned that an
`enterprising chef published a competing cookbook,
`copying hers. The cookbooks had differences, but his
`also included 37 chapters of the same recipes—6000
`total—reworded to reach the same results. The com-
`petitor grouped the recipes the same; ensured that the
`reworded recipes always led to the same result; added
`some new recipes he liked; and, used the exact same
`6000 titles as Ms. Child so that her audience could
`quickly recognize and use his competing cookbook in
`lieu of hers.
`
`
`
`
`
`9
`
`19. Further, suppose Ms. Child sued the author of
`the competing cookbook for copyright infringement.
`Hopefully, no one would be brazen enough to suggest
`that her cookbook is not copyrightable just because it
`is full of expression intended to facilitate functional
`results. After all, the “same statute that protects the
`finest literary, musical and artistic expressions—and
`some not so fine—has also from the beginning pro-
`tected such functional works as maps and charts and,
`later, directories and instructional texts, and tech-
`nical drawings, diagrams and models.” Paul Gold-
`stein, The Future of Software Protection, 47 U. Pitt.
`L. Rev. 1119, 1121 (1986).
`20. Ms. Child’s lawsuit (Oracle) against the brazen
`chef (Google) would be a far cry from an outlandish
`lawsuit against a home cook for the preparation of a
`recipe in one of the two cookbooks. (Similarly, this
`lawsuit is a far cry from a lawsuit against an individ-
`ual Java or Android developer for programs made us-
`ing Java or Android APIs.)
`21. Although Ms. Child’s hypothetical (and Ora-
`cle’s actual) lawsuit would implicate complicated
`questions about the scope of copyright protection, at
`least one aspect of the lawsuit is clear: that she (and
`Oracle) have a valid copyright (whatever its scope).
`Google’s position that APIs are not copyrightable is
`akin to saying that no cookbook could be copyrighted.
`That’s a bridge too far.
`22. For Ms. Child or Oracle to “establish infringe-
`ment, two elements must be proven: (1) ownership of
`a valid copyright, and (2) copying of constituent ele-
`ments of the work that are original.” Feist, 499 U.S.
`at 361.
`
`
`
`
`
`10
`
`23. Reasonable minds might disagree whether
`Section 102(b)’s limitations on the scope of infringe-
`ment pertain to the second element of infringement or
`are an affirmative defense. Regardless, Section
`102(b) certainly does not govern the first element,
`ownership of a valid copyright. The first element,
`about a valid copyright, is where copyrightability is
`decided. Thus, Section 102(b)—whether about an af-
`firmative defense or the second element of infringe-
`ment—is not at issue here.
`24. Yes, this case has thorny doctrinal questions,
`but they do not bear on copyrightability. Those com-
`plicated issues do not make the issue of copyrightabil-
`ity itself any more complex or different. Ms. Child
`clearly has a copyright in her cookbook whether or not
`we think the competing cookbook infringed. Like-
`wise, Oracle clearly has a copyright whether or not we
`think Google infringed.
`25. Many of Google’s amici are confused on this
`doctrinal point. One amicus repeatedly admits it is
`arguing about the “scope of protection[.]” Risch Ami-
`cus Br. at 1, 2, 7, 13, 24. Others import their scope
`arguments more subtly, i.e., by making assumptions
`about the scope of protection if copyrightability is af-
`firmed.
`26. For example, the Center for Democracy and
`Technology presumes that a copyright in Java APIs
`will necessarily lead to a parade of horribles, even if
`this Court nowhere opines on the scope of the copy-
`right. See Br. at 4 (“fragmentation, higher prices, and
`
`
`
`
`
`11
`
`frustration”); ibid. (purported effects on the blind);2
`ibid. at 6-9 (loss of universal remotes); ibid. at 9-10
`(fed-up parents of diabetic children disenfranchised);
`ibid. at 11 (“mercy of device makers”).
`27. This all sounds very worrying until one recog-
`nizes that Section 102(a)’s broad copyrightability is
`the foot in the door to asserting infringement, not a
`judgment of infringement itself. Yes, a runaway
`scope of copyright (by ignoring Section 102(b) in the
`infringement analysis) and amnesia as to copyright’s
`multitude of affirmative defenses (fair use, §§108-122,
`copyright misuse, implied license, etc.) would be prob-
`lematic. Deciding that highly expressive APIs are
`copyrightable is not.
`28. After all, one of Google’s own amici, Professor
`Nimmer, admits that the issue of a copyright’s scope
`is not readily susceptible to generalities. Professor
`Nimmer writes that the “determination of the extent
`of similarity that will constitute a substantial, and
`hence infringing, similarity presents one of the most
`difficult questions in copyright law, and one that is
`the least susceptible of helpful generalizations.”
`4 Nimmer on Copyright § 13.03[A] (italics in original,
`bolding supplied).
`29. Many of Google’s amici make just these gener-
`alizations without considering that a protection might
`be narrow. Merely saying the Java APIs are copy-
`rightable under Section 102(a) is a threshold ques-
`tion. Whether Google has infringed and whether an
`
`2 But see 17 U.S.C. §§ 121, 121A (exceptions to copyright spe-
`cifically for blind persons).
`
`
`
`
`
`12
`
`affirmative defense applies are separate questions, as
`the second question presented shows. Section 102(a)
`was not designed to be a standalone panacea for all of
`amici’s fathomable policy concerns.
`30. Section 102(b) is not at issue here because it’s
`either an affirmative defense or an aspect of infringe-
`ment analysis, not implicated in the question pre-
`sented. Although Google and some of Google’s amici
`write powerfully and eloquently about the meaning of
`Section 102(b), see generally, e.g., EFF Brief, their ar-
`guments are inapposite. They should have saved
`their thunder. These amici, like the district court,
`have “failed to distinguish between the threshold
`question of what is copyrightable—which presents a
`low bar—and the scope of conduct that constitutes in-
`fringing activity.” Oracle Am., Inc. v. Google Inc., 750
`F.3d 1339, 1354 (Fed. Cir. 2014).
`* * * * *
`The first question presented is whether the Java
`APIs are copyrightable at all. They are.3
`
`
`
`
`3 The Foundation also concurs in the analysis of the Copy-
`right Thought Leaders Brief as to copyrightability.
`
`
`
`
`
`13
`
`B. Google could have enabled Java
`programmers to transition to Android by
`making a code-conversion tool.
`1. Google repeatedly conflates two different acts
`when arguing it was necessary to copy Java APIs:
`• Using Java APIs to write non-API pro-
`grams in Java, and
`• Copying Java APIs to supplant Java APIs
`with Android APIs.
`2. The district court similarly conflated coding a
`program using the name of methods from an API
`(what individual programmers do to write programs)
`and making a new set of APIs (as Google did). The
`district court said:
`[S]ince there is only one way to declare
`a given method functionality, everyone
`using that function must write that
`specific line of code in the same way.
`872 F. Supp. 2d at 979.
`3. That’s true if you’re writing a program using
`Java APIs. That’s not true if you’re copying APIs into
`a supplanting set of APIs, as Google did. From the
`perspective of the individual Java programmer, tell-
`ing her she cannot use Java APIs is akin to telling an
`English speaker she cannot write using English
`words. If the programmer writing a Java program
`tries to determine the maximum of two numbers other
`than via java.lang.Math.max(x,y), it won’t work.
`Likewise, speaking Mandarin Chinese to someone
`who only speaks English won’t work. That’s the
`
`
`
`
`
`14
`
`concept of a programming language. To express her-
`self using an API, the programmer must necessarily
`employ the API’s syntax.4
`4. The same does not hold for Google. Google has
`copied and incorporated the Java APIs as a part of its
`Android APIs. Google could have changed the APIs’
`structure and renamed them to, for example,
`android.program.Arithmetic.most(x,y). In fact, Google
`could have named any particular Android method
`something entirely arbitrary, including gibberish
`such as ooo.rtun.atjk.imsw(x,y). Google had unlimited
`options in the naming and grouping of its methods for
`the Android APIs.
`5. Critically, a single method name like max is not
`copyrightable, but the names, selection, structure,
`and taxonomy of 6000 of them is. The Act is clear
`that, even if the constituent methods’ names are not
`individually copyrightable, their grouping would be
`copyrightable as a compilation. See 17 U.S.C. § 101
`(defining “compilation”).
`6. It is true that a new naming and grouping of an
`API’s methods would be less familiar to individual
`programmers, until they became acquainted with the
`new APIs. Yet widespread familiarity with a work
`has not, and should not, count against that work’s
`copyrightability.
`
`
`4
` Nor would her programs be anywhere near substantially
`similar to the APIs themselves even if the scope of protection (not
`at issue here) were deemed to be broad. Programmers wouldn’t
`copy the 6000 methods verbatim in their programs.
`
`
`
`
`
`15
`
`7. The Java APIs do not become ineligible for cop-
`yright merely because so many programmers prefer
`and are accustomed to the way that Sun expressed its
`APIs. By analogy, a rival cookbook author cannot
`challenge the copyrightability of Julia Child’s cook-
`book merely because many French-food enthusiasts
`might prefer her book because they’re familiar with it
`already.
`8. Google equivocates to blur this clear distinction
`between its competing APIs and non-API programs
`that use the method names in APIs. Yes, Java APIs
`are “essential tools for building new computer pro-
`grams” but not for creating a new API. See Google
`Br. at i. Yes, a “call will not work correctly unless it
`corresponds precisely to instructions called ‘declara-
`tions’”—but not if the new API-maker (Google) had
`decided to pick different names for its own API’s
`methods. Ibid. at 4. Google could have done differ-
`ently here.
`9. Google acknowledges that Oracle has not made
`the broader claim that someone writing a program us-
`ing Java APIs is in the same circumstances as Google
`copying Java APIs to make Android APIs. See ibid.
`at 22 (“It also is significant that Oracle claims no cop-
`yright interest in the calls that Java developers use to
`invoke the methods.”). Google’s error, however, is to
`conflate itself and its acts with those of individuals
`writing ordinary (non-API) programs using the Java
`APIs.
`
`
`
`
`
`
`
`16
`
`10. Truly, Google’s argument is about convenience,
`not necessity. It’s arguing that Java APIs are not
`within the subject matter of copyright because Java
`programmers could more easily switch existing Java
`programs into Android if Google copied certain Java
`APIs into Android APIs.
`11. Google’s argument is tantamount to a reverse
`“sweat of the brow” theory. In Feist, this Court re-
`jected the idea that someone can obtain a copyright as
`“a reward for the hard work” that went into making
`the expression. 499 U.S. at 352 (emphasis added).
`Today, Google argues that Oracle’s copyright protec-
`tion should disappear if it would take too much sweat
`equity, too much investment, to be as creative in the
`naming taxonomy for Android’s API. However, this
`new-fangled reverse-sweat-of-the-brow theory fails
`for the same reason its counterpart did in Feist: it
`strays from Section 102(a)’s plain meaning.
`12. Moreover, Google’s argument about conven-
`ience is undermined by the existence and possibility
`of making code-conversion tools. Google could have
`made a code-conversion tool to help Java programs
`transition to its Android platform.
`
`
`
`
`
`
`
`
`
`
`17
`
`12. For example, suppose you had existing pro-
`grams written using Java APIs and wanted them to
`work in one of Microsoft’s programming environ-
`ments, called the .NET applications. You’re in luck.
`There’s an app for that: IKVM.NET. It permits de-
`velopers to write in Java APIs and will then trans-
`late the code into Microsoft’s .NET languages (such
`as C# or VB.NET)—or vice versa:
`IKVM.NET makes it possible both to
`develop .NET applications in Java, and
`to use existing Java API’s and libraries
`in applications written in any .NET
`language. This guide includes infor-
`mation of interest to developers who
`want to use Java as their preferred
`.NET development language, as well as
`for those who wish to use Java libraries
`in their C# or VB.NET applications.
`Stephen Schaub, IKVM.NET Developer’s Guide,
`http://www.ikvm.net/devguide/intro.html
`(accessed
`Feb. 19, 2020).
`13. Code conversion is not a rare phenomenon. In
`fact, there are excellent tools to convert from Java to
`Python, which is arguably a much harder task than
`the tasks needed to convert between APIs’ method
`names. See, e.g., Eman J. Coco et al., JPT: A Simple
`Java-Python Translator, 5 Computer Applications 1,
`1 (2018) (“present[ing] an approach for programming
`language inter-conversion which can be applied to all
`types of programming languages”).
`
`
`
`
`
`
`18
`
`14. Other programmers have developed code-con-
`version tools to translate Java code into a variety of
`other formats. Google could have done the same.
`And, it’s not as though Google is lacking in sufficient
`engineering talent to make a Java-to-Android code-
`conversion tool.
`15. So why didn’t Google make a code-conversion
`tool? Business strategy. Google didn’t invent this
`strategy either; it copied this too. Microsoft tried
`something similar with Java in the 1990s. Cf. Sun
`Microsystems, Inc. v. Microsoft Corp., 188 F.3d 1115,
`1116 (9th Cir. 1999) (copyright suit by Oracle’s prede-
`cessor). Microsoft tried to “capture an open standard”
`and partially succeeded:
`In a dispute involving Microsoft and
`Sun Microsystems, Inc. (“Sun”), for ex-
`ample, Sun promulgated the Java lan-
`guage as an open standard so that eve-
`rybody would be able to write programs
`in Java (the better to compete with Mi-
`crosoft); but Microsoft changed Java a
`little bit so that those who wanted to
`write programs in Java compatible
`with Microsoft products would have to
`use Microsoft’s proprietary version of
`Java.
`Margaret Radin, Online Standardization, 70 Ford-
`ham L. Rev. 1125, 1134 (2002). Contrary to many
`amici, Sun’s copyright (and copyrightability in APIs
`generally) is essential to ensuring that an open-source
`platform stays open.
`
`
`
`
`
`
`19
`
`16. Google here (and Microsoft before it) used an
`opportunistic strategy called “embrace, extend, extin-
`guish” to take other APIs and make them obsolete:
`In this gambit, a firm [Google] adopts a
`Software Standard [Java APIs], em-
`bracing it, but adds capability to the
`standard [Android APIs], extending it.
`It keeps that additional capability pri-
`vate [or at least not interoperable with
`the Java APIs], and hopes that its ver-
`sion of the implementation will become
`favored a