AGAINST DEFIBRILLATING THE API COPYRIGHT DEAD: A RESPONSE TO ADVOCATES OF COPYRIGHTABILITY OF SOFTWARE FUNCTIONAL SPECIFICATIONS.

• Author: Menell, Peter S.
• Position: Application program interface - Special Issue: Software Interface Copyright

TABLE OF CONTENTS I. INTRODUCTION 653 II. CLARIFYING THE TERMS OF THE DEBATE 655 III. SECTION 102(B), CONTU, LEGISLATIVE HISTORY, AND NINTH CIRCUIT 656 LAW IV. SOFTWARE COPYRIGHT EXCEPTIONALISM IS BAKED INTO COPYRIGHT 659 LAW; OR WHY HARRY POTTER NOVELS ARE NOT TREATED THE SAME AS FUNCTIONAL SPECIFICATIONS FOR MACHINES V. NOT SO FINAL WORDS 663 I. INTRODUCTION

As this Special Issue of the HARVARD JOURNAL OF LAW & TECHNOLOGY confirms, the Oracle v. Google (1) litigation has revived a long dormant battle over the scope of copyright protection for computer software. That struggle, which raged from the mid-1980s through the mid-1990s, brought many of the contributors to this Special Issue together. Thus, this Issue has the feel of a reunion. Pam Samuelson, David Nimmer, and I wrote some of the early articles on these questions. (2) Jonathan Band worked with his then-colleague Michael Jacobs representing Fujitsu in its landmark software arbitration with IBM that focused on copyright and interoperability. (3) They were active in the formation and work of the American Committee for Interoperable Systems ("ACIS"), (4) which advocated for less protectionist intellectual property policies for computer software. Mark Lemley joined the party just as the first API copyright battle was subsiding. (5)

It is also nostalgic to see my good friend Annette Hurst participating in this Special Issue. Annette and I, along with Annette's Orrick colleague Joshua Rosenkranz, collaborated in the epic battle between Mattel and MGA over the Bratz dolls. (6) Opposite to us, Ralph Oman served as an expert witness for Mattel. As lead appellate counsel, Josh played a key role in persuading the Ninth Circuit to reverse the trial court's unwarranted constructive trust against MGA, earning him the noteworthy appellation "The Defibrillator" for his ability to revive companies "that appeared to be at death's door." (7) We proudly celebrated the Ninth Circuit decision vindicating fundamental copyright limitations that promote creativity and competition. (8)

Yet, there is something amiss about this reunion. Michael Jacobs, who participated in the LaST Frontier Conference on Copyright Protection of Computer Software and contended that "copyright law should clearly permit the independent development of compatible computer programs," (9) served as lead trial counsel for the plaintiff in the first Oracle v. Google trial. Sun Microsystems and Oracle, who were among ACIS's founding and lead members, are now pushing for unusually strong and broad copyright protection for computer software. And my comrades from the Bratz litigation, which successfully fended off overbroad copyright claims for doll designs, are now advocating that copyright law robustly protect API declarations as though they were character names and chapter titles of HARRY POTTER novels.

Is this reunion a bad dream? Unfortunately, no. The legal profession has a way of distorting logic and principle in the name of zealous advocacy. In this response, I refute Mr. Oman's (10) and Ms. Hurst's (11) critique of my lead article (12) questioning Oracle v. Google's resurrection of copyright protection for functional features of computer software. (13) Contrary to their assertions, I do not contend that APIs are not copyrightable. My position, grounded in Section 102(b) of the Copyright Act, the legislative history (including the CONTU Report), and Ninth Circuit precedent, is that the functional requirements of APIs, like the internal workings of other machines, are outside of copyright protection even as the implementing code for APIs is protectable.

2. CLARIFYING THE TERMS OF THE DEBATE

   Both Mr. oman and Ms. Hurst construct their critique of Rise of the API Copyright Dead? by erecting a strawman. According to Mr. Oman, I "implicitly embrace[] the view that Congress meant for the protection afforded computer software to be different from the protection afforded other works, because computer software is functional," and this leads me to contend that the Federal Circuit "erred at every turn." (14) Ms. Hurst reads my article to suggest that APIs are not creative, Sun Microsystems authorized companies to use the Java API without a license, and there is no copyright protection for API code. (15) Both Mr. Oman and Ms. Hurst conclude by contending that I predict that the Federal Circuit's decision will cause the sky to fall. (16)

   None of these assertions is correct. My article comprehensively and faithfully presents the legislative process leading to Congress's recognition of copyright protection for computer software. (17) I agree with the Federal Circuit's holding that the Java API implementation attracts copyright protection. (18) I explain the development and coding of the Java APIs and recognize that they entail substantial creativity. (19) I dispassionately present the extensive interplay between Sun and Google over licensing of the Java APIs, including Oracle's trial slides depicting the two faces of Jonathan Schwartz. (20) None of my analysis turns on whether or not Sun consented to unlicensed use of the Java

   APIs. (21)

   Thus, Mr. Oman, Ms. Hurst, and I agree about many of the underlying facts and several of the core legal issues underlying the API copyright battle. Nonetheless, we differ on two key issues: (1) the proper interpretation and application of copyright's limiting doctrines to functional specifications, particularly as construed by the Ninth Circuit; and (2) whether copyright law treats computer software differently than other works of authorship.

3. SECTION 102(b), CONTU, LEGISLATIVE HISTORY, AND NINTH CIRCUIT LAW

   I concur with Mr. Oman and Ms. Hurst that computer software generally falls within the subject matter of copyright protection. (22)

   From there, however, we diverge. Both of them downplay the limitations reflected in Section 102(b) of the Copyright Act (23) and selectively discuss the legislative history referencing computer software and other significantly functional works, such as useful articles. Mr. oman downplays and Ms. Hurst entirely disregards Baker v. Selden, the Supreme Court's seminal decision channeling protection between patent and copyright law. (24)

   Rise of the API Copyright Dead? meticulously examines the pertinent legislative history, (25) including the statement that "Section 102(b) is intended, among other things, to make clear that the expression adopted by the programmer is the copyrightable element in a computer program, and that the actual processes or methods embodied in the program are not within the scope of the copyright law." (26) My article also highlights CONTU's statement that "one is always free to make a machine perform any conceivable process (in the absence of a patent)." (27)

   From these foundational sources, the rift grows wider when we reach the jurisprudence. The copyrightability of the Java API declarations litigated in Oracle v. Google turns on application of fundamental limitations on copyright protection set forth in Section 102(b) as interpreted by Ninth Circuit law. (28) Neither Mr. Oman nor Ms. Hurst take seriously this critical procedural posture. Whereas regional circuit courts of appeals are bound by their own and the Supreme Court's interpretation of applicable law, the Federal Circuit is obliged to operate differently when addressing legal issues outside of its core patent jurisdictional authority. (29) In establishing a specialized national appellate tribunal to harmonize patent jurisprudence and discourage regional forum shopping, Congress constrained the Federal Circuit's...