Oracle v. Google and the Scope of a Computer Program Copyright
Dennis S. Karjala
Arizona State University School of Law
ORACLE V. GOOGLE AND THE SCOPE OF A COMPUTER PROGRAM COPYRIGHT
Dennis S. Karjala*
I. Introduction...............................................................................................2
II. Computer Programs are Technology...............................................4
III. Oracle v. Google.......................................................................................7
IV. Patent and Copyright..............................................................................9
V. Application to Programs......................................................................10
VI. The Case Law...............................................................................................18
VII. Application to the Facts of Oracle v. Google..........................22
VIII. Summary........................................................................................................31
[Page 3]
Copyright law protects the expression of an idea but not the idea expressed. Everybody knows that. The problem has always been that nobody knows or can articulate what parts of a work constitute "expression" and what parts constitute "idea." For traditional copyright subject matter like books, art, and music, courts and scholars typically look to the "creative" aspects of the work and the availability of a variety of ways of expressing the same idea.1 Decisions are largely ad hoc2 and nuances have developed depending on the type of work at issue.3
Many courts have addressed the problem of determining the scope of copyright protection in a computer program in the same ways they have long used for traditional copyright subject matter.4 In fact, however, computer programs offer at least one example in which copyright doctrine, statutory language, and fundamental notions of intellectual property policy all lead to a clear and consistent result that gives less protection to program structure than, say, copyright in a fictional story: Copyright in a computer program should be limited to literal program code and close paraphrases of program code. The reason is that computer programs are technological subject matter masquerading as copyright-protected literary works.5 Intellectual property law has always treated art and technology differently, and maintaining patent's general role in protecting technology while accomplishing the anti-misappropriation congressional purpose of placing program code under copyright can only be achieved by appropriate limitation of the scope of the program copyright.6
[Page 4]
Computer programs are the technology that causes computers to perform their functions.7 The word "technology" is important, because the now-centuries-long dogma of copyright law developed for the non-technological subject matter of art, literature, and music.8 With its low threshold for and very long period of protection, copyright is relatively generous with the exclusive rights it affords to copyright owners. Patent law, on the other hand, has always provided for the possibility of exclusive rights in technology but is stingier to technology developers. Patents require formal applications and examination, precisely worded claims, and, even if issued, expire only twenty years after filing.9 Congress's decision to break from tradition in the 1976 Copyright Act by bringing computer programs under copyright protection—without explicit recognition that such a break from tradition was being made—can only be sensibly interpreted in the light of the long history of Intellectual Property Law in which copyright protected nonfunctional works of information and patent protected functional works of technology.10 Copyright doctrine (idea/expression, substantial similarity, fair use) is notoriously vague, and valid doctrinal arguments can be mustered for either a broad or a narrow scope of copyright protection in computer programs. Without reference to policy, therefore, especially the policy underlying the patent and copyright subject matter distinction, no argument is determinative. Recognition that we have two statutes for protecting the fruits of intellectual creativity that operate in now overlapping but fundamentally different spheres, however, leads naturally and almost inexorably to a clear resolution: Copyright for computer programs protects the literal program code; beyond program code, the program copyright protects, at most, close paraphrases of literal code. The program copyright does not protect interfaces generated by the program code, nor should it protect the variety of technological means by which the program code is structurally organized.
[Page 5]
Probably the most important notion to keep in mind in determining the scope of computer program copyrights is that we have two major intellectual property statutes—patent and copyright.11 Traditional copyright law applied to nonfunctional art, music, and literature, while traditional patent law applied to functional works of technology.12 That to some extent we now protect technology in the form of computer programs under copyright law does not and cannot mean, until Congress instructs otherwise, that all of the protection we have recognized for nontechnological novels and plays must carry over in full force to computer program technology.
A second point, grounded in both policy and doctrine, is that "creativity" alone does not equate to copyright-protected "expression."13 A third and related point is that the availability of a wide range of choices—"absence of merger"—does not equate to copyright protection for any particular choice.14
A fourth point is that computer programs, by their nature, are methods of operation15 (they are quite literally the technology for operating computers to bring about specific results). Any particular program is a made up of series of steps that achieve the result effected by that program. It necessarily follows that, given Congress's near verbatim adoption of the CONTU16
[Page 6]
recommendations, § 102(b) has been limited in its application to computer programs. The question is, how far does this limitation go—does § 102(b) not apply to computer programs at all, or is the limitation itself more limited in applying mainly, if not solely, to program code?
A fifth point arises directly from the definition of a computer program as "a set of statements or instructions to be used directly or indirectly in a computer in order to bring about a certain result."17 This definition expressly distinguishes between the code (the set of "statements or instructions") and the "certain result"18 brought about by that set of instructions. It is the set of statements or instructions that constitutes the computer program. The "certain result" brought about by execution of those instructions in a computer is not a part of the computer program and, logically, is not protected by the copyright in the computer program. Some such "certain results," of course, will be independently copyright protectable as, for example, pictorial, graphic, or literary works,19 but such copyrights exist independent of the copyright in the computer program itself.
Sixth, and finally, even if we were to apply traditional copyright doctrine for literary works to computer programs without regard to their sui generis nature as works of technology, we should still conclude that the program copyright covers primarily, if not solely, the program code and certainly not program structure, sequence, and organization (SSO), let alone the functional interfaces generated by the program code. Some traditional literary works have a broad scope of protection (novels and plays), but others have a very narrow scope of protection—protection that is limited virtually to their verbatim language (insurance contracts).20 Anyone demanding computer programs be given the broad scope of protection afforded to novels and plays should supply a policy-based reason for analogizing computer programs to such works, as opposed to
[Page 7]
more technical works like dictionaries, legal forms, and scientific works in which the scope of protection is much narrower.21
The ineluctable conclusion that must be drawn from these basic points is that computer programs are a sui generis type of copyright subject matter (even though falling within the classification of "literary works"). The program copyright covers the program code and close paraphrases of such code, for the purpose of preventing cheap, rapid and exact copying. However, the copyright does not extend to functional aspects of program structure; it does not extend to any aspect of program structure, because all such structure is intended to achieve a functional result in an efficient way.22 A fortiori, it does not extend to any aspects of the program interfaces, which exist at an even higher level of abstraction than the program SSO.23
This interpretation gives full copyright protection to that aspect of program technology that is most in need of protection from verbatim copying, namely, electronic representations of program code.24 It leaves other aspects of program technology to develop under the rules of patent and trade secret, which are designed for the protection of technology. No one has ever offered a compelling policy-based reason for treating noncode program technologies like SSO differently from any other technology, for which patent and trade secret have been the sole choice for centuries, and Congress has given no hint, let alone explicit direction, that copyright protection for technology should extend so far beyond computer program code.
The need to consider the sui generis nature of computer programs as copyright subject matter has been brought into stark relief by the conflicting district court and Federal Circuit decisions in Oracle...
