Reframing similarity analysis in copyright.

• Author: Hickey, Kevin J.
• Position: II. Framing Problems in Substantial Similarity B. The Baseline Problem through Conclusion, with footnotes, p. 707-731

2. The Baseline Problem

   This Part explores similarity's baseline problem: how courts determine the benchmark used to measure when similarity becomes "substantial," and therefore an infringement of copyright. In assessing the quantitative significance of what the alleged infringer copied, should we look to how much was copied from the original work, or how much of the accused work is copied material? Or should we instead assess the significance against some absolute threshold, or solely as a qualitative matter--the economic, artistic, or practical significance of the material copied? In contrast to the timing problem, where there are two main camps--evaluating idea-expression at the time of creation, versus the time of infringement--courts take a looser approach to the baseline for infringement. To the extent there is any consensus, the traditional rule is that the appropriate baseline is the original work. (152) However, courts' choice of baseline is often implicit, and courts frequently employ whatever baseline provides greater rhetorical support for their argument.

   Although the case arose in the context of fair use, the Supreme Court's decision in Harper & Row, Publishers, Inc. v. Nation Enterprises provides a simple example to illustrate the baseline problem. (153) Harper & Row addressed whether The Nation magazine's "scoop" of President Ford's unpublished memoir, A Time to Heal, was a fair use. (154) Having received a prepublication copy from an unidentified source, The Nation quoted and paraphrased from the forthcoming memoir to produce an article detailing Ford's decision to pardon former President Richard Nixon. (155) The Nation's article was 2250 words (about five to ten pages), of which about 300 were direct quotes from the Ford memoir. (156) The original work, Ford's memoir, was about 200,000 words. (157)

   The third statutory factor for determining whether a use is fair directs courts to consider "the amount and substantiality of the portion used in relation to the copyrighted work as a whole." (158) This suggests that the proper baseline in the fair use context is the original work, and the majority opinion in Harper & Row pays some lip service to that notion. (159) Taken literally, this measure would suggest that what The Nation took was quantitatively trivial--300 words out of 200,000, or about 0.1%. Justice O'Connor's majority opinion concedes that 300 words is an "insubstantial" part of Ford's memoir. (160) Rather abruptly, however, the majority opinion shifts the baseline and emphasizes that the quotes were "at least 13% of [The Nation's] infringing article"--and, moreover, very significant as a qualitative matter: the "heart of the book." (161) Justice Brennan's dissent, in a footnote, gently takes the majority to task for this switch. (162)

   Substantial similarity, even more so than fair use, is a common law doctrine, and its baseline is not dictated by any statutory considerations. In practice, however, courts tend to follow the course of the Supreme Court in Harper & Row. They pay lip service to the notion that the original work is the appropriate baseline--if they address the issue at all--but then disregard analytical consistency to select whatever baseline strengthens their argument.

   Newton v. Diamond provides an atypically thorough examination of the baseline problem in the substantial similarity context. (163) Newton involved the Beastie Boys' song "Pass the Mic," which prominently featured a sample of "Choir" by the jazz flutist James Newton. (164) Newton's technique in "Choir" was unusual, with the score instructing the performer to "sing[] into" the flute while fingering particular pitches, creating a ghostly effect. (165) The Beastie Boys took a six-second sample from "Choir" and looped it, using the sample as a background element for almost all of the duration of "Pass the Mic." (166) Newton sued for copyright infringement, and the district court granted summary judgment in favor of the Beastie Boys. (167)

   The case is complicated somewhat by the unusual nature of Newton's work, and the fact that the Beastie Boys had a partial license to use it. There are two types of copyrights in a musical recording: one for the composition, and another for the sound recording itself. (168) The former covers the work of the writer of the music and lyrics, which is sometimes embodied in a score; the latter covers the work of the performers and producers of a particular recording. (169) For "Pass the Mic," the Beastie Boys secured permission from the owner of the sound recording copyright (which Newton had assigned), but not from the owner of the composition copyright, which remained with Newton. (170) Thus, some of the debate in the case focused on delineating which aspects of the sample were attributable to Newton's composition, and which to his performance. (171)

   Putting that issue aside, the judges in Newton struggled to define the baseline against which to measure whether the Beastie Boys' use of the sample was substantial enough to be an infringement. (172) On the one hand, the sample was a limited part of Newton's work (only a few seconds), but on the other, the sample was a significant part of the infringing work (the main background element). The majority in Newton cited the traditional rule that the plaintiffs work should be the baseline, and suggested that the copying was insubstantial as a result. (173) It emphasized that the sample "appears only once in Newton's composition," and that six seconds was merely "two percent of the four-and-a-half-minute" running time of "Choir." (174) But what of the fact that the Beastie Boys looped the sample, such that it constituted a large part of the accused work? Strikingly, because of its chosen baseline--and in contrast to the approach in Harper & Row--the majority found that the looping was entirely "irrelevant in weighing the sample's qualitative and quantitative significance." (175)

   It is worthwhile to trace how Newton reached its conclusion to disregard how the defendants actually used the sample when assessing similarity. The rationale for assessing infringement against the baseline of the original work derives--as does a surprising amount of copyright doctrine--from an opinion by Judge Learned Hand. In Sheldon v. MetroGoldwyn Pictures, Corp., Judge Hand addressed whether a play and a movie--both loosely based on the same real-life story of Madeleine Smith, a cause celebre of the day--were similar enough in plot, character, and incident such that the movie infringed the earlier play. (176) Although the movie took no dialogue directly from the play, Judge Hand thought the characters and certain scenes were so alike that the movie was an infringement. (177) Rejecting out of hand the defendant's attempt to rely on the many differences between the movie and picture, Judge Hand held that this was "immaterial" because "no plagiarist can excuse the wrong by showing how much of his work he did not pirate." (178)

   Judge Hand's pithy line--which I will call "the Sheldon rule"--has become something of a mantra in copyright infringement cases, being "wildly over-cited" in subsequent decisions. (179) Interestingly, the Sheldon rule was originally designed to benefit the copyright holder by ignoring any new material that the infringer introduced. The concern motivating Judge Hand seems to be that an infringer could escape liability simply by "tacking on" a large amount of unrelated new material. The majority in Newton, however, used Sheldon in quite the opposite way, for the benefit of the defendant. It discounted how the Beastie Boys actually used the sample, focusing on the fact that the sample was merely a small part of the original work. Dissenting in Newton, Judge Graber felt that the majority had unfairly dismissed the qualitative importance of what was copied from Newton, noting that much of what made Newton's work so distinctive was the playing technique, a part of the composition and of the sample. (180)

   The recent case of Antonick v. Electronic Arts Inc. (181) suggests that a literal application of the Sheldon rule can sometimes produce troubling results. Robin Antonick, pursuant to an agreement with Electronic Arts ("EA"), designed and wrote source code for 1988's John Madden Football, a video game for the Apple II computer. (182) That game was the first in a phenomenally successful franchise; later games for systems like the Sega Genesis, Xbox, and PlayStation have sold over 85 million copies and grossed around $4 billion for EA. (183) Antonick's contractual agreement with EA entitled him to royalties on any "derivative works" using his code, and--after discovering similarities between the first game and later entries in the Madden series--Antonick sued EA for royalties based on the sale of the later games. (184) EA claimed that the code for the later games was developed independently--it had, falsely, told Antonick the same thing for years. (185) The jury, however, found that EA copied Antonick's code for the football plays and formations and used this code in the later games; it awarded Antonick $11 million in damages. (186)

   EA sought a new trial, arguing that the first Madden game was not substantially similar to the later versions as a matter of law--and therefore no royalties were owed. (187) Judge Breyer agreed that the jury was entitled to find that EA's code for the later Madden games governing plays and formations was, in fact, copied from Antonick. (188) But he ruled that this copying was not substantial when comparing the original and accused games "as a whole." (189) The code that EA had copied from Antonick was only one part of a larger video game. The jury was required to find that the later versions were similar enough on the level of the entire work--not just in the infringing module. (190) This, Judge Breyer claimed, no reasonable jury could find based on the evidence presented. (191) Confusingly, although Judge...