CHAPTER 4

PRESENTATION OF EVIDENCE: EXEMPLARY CASES

Courts may decide infringement cases on many different factors; each case presents its own peculiar issues. Analysis of the music may predominate the inquiry, or the plaintiff may lose his case on a legal flaw. Infringement is difficult to prove. No one paradigm case can adequately explain the legal complexities involved. Lawyers will appreciate the legal intricacies, and musicians will understand how a jury of laymen would have difficulty with the abstract concepts of music analysis.

This chapter attempts to present some of those issues by describing three recent trials in some detail. The cases chosen do not perfectly represent the area of music litigation. Most infringement cases settle out of court, perhaps in recognition of how difficult they are to try. The cases presented in this chapter all involve jury trials, although many cases are tried to the court. Selle v. Gibb, despite the jury, was decided on purely legal grounds. Gaste v. Kaiserman turned almost exclusively on musical issues. Baxter v. MCA involved issues that highlight analytical problems in the Ninth Circuit's "total concept and feel" test. This chapter uses another case, Baron v. Leo Feist, Inc., to provide an historical context for musical analysis by expert witnesses.

This description will omit or seriously condense some aspects of the litigation. For example, it lies beyond the scope of this work to explain all the reasons why Baxter did not come to trial until five years after the lawsuit was filed. Also, some of the issues raised at trial or on appeal do not relate sufficiently to expert testimony. Selle and Gaste are presented roughly in chronological order. Chronology is sacrificed in Baxter because of lengthy testimony from four expert witnesses.

A few preliminaries will place the three exemplary cases in context. Music experts tend to be deposed late in the discovery process, closer to trial. In most cases, however, the hiring attorney will have consulted the expert closer to the filing of the suit. The expert often produces a pre-trial report, a musical analysis, on which he may be questioned at his deposition. He may also have the opportunity to read the opposing expert's pre-trial report, and he may be questioned on his reactions to that analysis as well. The expert's pre-trial report will form the basis of his testimony. For many experts, the analysis will be more detailed than the testimony; the expert cannot explain all aspects of a competent analysis on the witness stand. Some aspects of the presentation of expert testimony have evolved over the years, but many remain stubbornly resistant to change. A brief re-examination of a case described forty years ago will provide context.

Louis Nizer devoted a chapter of his book My Life in Court[1] to his successful prosecution of Maurice Baron's claim of infringement in Baron v. Leo Feist, Inc.[2] In 1906, calypso musician Lionel Belasco composed the song "L'Année passée," which told the story of a girl from a good family of Trinidad who ran away with a "ne'er-do-well" and eventually became a streetwalker. Although Belasco had not filed a copyright registration, trial testimony firmly established the date of composition; the story was based on a notorious local event, and Belasco had taught the song to several of his protégés soon after. Maurice Baron obtained Belasco's permission to publish "L'Année passée" in a 1943 collection of calypso music.[3]

The facts surrounding the creation of the accused work were rather obscure. Morey Amsterdam had performed for U.S. troops in Trinidad in 1943, during which time he said he heard a song called "Rum and Coca-Cola," whose tune resembled "It Ain't Gonna Rain No More." According to the defendant publishing company, Amsterdam wrote his own lyrics based loosely on what he heard, and Paul Baron (no relation to Maurice Baron), head of the music department of Columbia Broadcasting Company, composed music variously described as derived from the Trinidad tune heard by Amsterdam or from other calypso tunes known to Baron. The Leo Feist company recorded and published this American version of "Rum and Coca-Cola," whose ribald text satirized the relations between U.S. soldiers and native women.[4] It was an immediate commercial success.[5]

The plaintiff, publisher Maurice Baron, had no problem showing access. Songwriter Paul Baron had ordered copies of publisher Baron's collection of calypso music before "Rum and Coca-Cola" was recorded.[6] Further, Paul Baron's testimony on independent creation was not credible.[7] He first claimed that he wrote "Rum and Coca-Cola" before Amsterdam demonstrated the folk melody heard in Trinidad,[8] but Baron later admitted that Amsterdam's demonstration came first.

Plaintiff Baron testified as an expert for his own case, and displayed the two songs side by side. For further visual impact, Baron made steel bars on which he hung steel notes. He moved notes of various colors from one song to the other to show the physical act of "substantial taking."[9] He introduced charts employing colored notes as well as solmization.[10] Graphs showed the melodic contour, its ascending and descending motions, in red and black lines.[11] The court overruled the defense attorney's objections that the charts did not reflect precise time values.[12] Of thirty-six notes, Baron showed thirty to be identical.[13] A second expert, composer Normand D. Lockwood of Columbia University, found comparable melodic similarities.[14] The harmonic progression had fourteen of sixteen chords in common.[15] Walter Merrick, a third expert and an authority on calypso, played a hybrid for the court, taking alternately two measures from one song and two from the other, thus demonstrating that they were interchangeable.[16]

At least two features argued strongly for copying. Lockwood noted a weak suspension (the rhythmic figure of short note tied to long) occurring regularly in measures 2, 4, 6, and 8 of both works. Although the weak suspension was a common figure in calypso music, placement at such regular intervals was unusual.[17] More telling, "L'Année passée" employed a chromatic chord to bring out the pathos of a particular textual statement. "Rum and Coca-Cola" duplicated the chord even though it did not belong in the calypso style and bore no relation to that song's lyrics.[18] Furthermore, both songs resolved the dissonant chord to C dominant 7, not the resolution one would normally expect.[19] These "common errors," what might be considered musical oddities or non-idiomatic writing, provided the trier of fact with key evidence‑-musical fingerprints.[20]

Nizer took some delight in describing his cross-examination of defendants' expert, Sigmund Spaeth. On direct examination, Spaeth had refused to discount rhythm and accent, an approach that allowed him to show that only twelve of thirty-eight notes were the same. He called this number of similar notes not substantial. But on cross-examination Nizer read from Spaeth's own books to impeach his testimony. Spaeth, widely known as the "tune detective," often did music analysis for audiences as a kind of vaudeville stunt. His book showed how "Yes We Have No Bananas" was derived from three other works, and Spaeth claimed he could similarly trace many musical origins. Moreover, he had labeled as flagrant plagiarism even similarities that did not match rhythmically. Nizer caught Spaeth following a loose standard in his books and a different, more rigorous standard in court.[21]

Although Spaeth and defense counsel Julian Abeles received bad reviews in Nizer's book, they set the standard for infringement litigation in numerous trials of the 1930s and 1940s.[22] Spaeth continued to serve as an expert witness in cases through the 1950s. Abeles's law partners inherited a preeminent position as specialists in music litigation. Their influence remains strong today.[23] Robert Osterberg of the New York firm Abeles, Clark, and Osterberg often represents one party or the other when infringement of a music copyright is alleged. Experts since Spaeth have adopted many of his methods. Although experts' methodologies differ, many common practices seem to have originated with Spaeth.

Preliminaries to Analysis

An expert must first address the question of which materials to compare. An attorney hiring an expert may dictate that the expert is to compare particular materials. The attorney may even minimize bias in the initial analysis by sending the expert two unidentified pieces of music with no indication of who his client might be or which is the accused work. The expert thus has little opportunity to tailor his analysis even subconsciously.[24] However, a blind analysis is not always possible; for example, one work may obviously predate the other, as where the expert is asked to compare a three-minute song to a sixty-second commercial. In such cases, the expert may infer which work was produced by the lawyer's client.[25] Steps taken to ensure or demonstrate a lack of bias in the analysis will likely prove beneficial at trial.

Any version of the accused work may infringe the original's copyright. What the defendant registered is largely immaterial. The relevant question is whether the defendant copied, and the expert should seek out the best evidence of copying, but only copying by the defendant. Testimony comparing versions of the music that might have been modified by non-parties to the litigation, such as transcriptions, performances, or arrangements, may be irrelevant or accorded less weight. Experts should strive to make comparisons using the version of plaintiff's work that is registered with the copyright office. Of course, because copyright attaches when the work is fixed in a tangible medium of expression, not upon registration, the plaintiff may hold rights in other, non-registered versions that need to be examined as well.

Transcription Techniques

Written music both distorts and reduces; it inadequately represents the experience of music. Presumptions rooted in culture and style often control what aspects of the music are represented in writing.[26] Consequently, the expert should be wary of relying on a writing provided to him for his comparison and be aware of the strengths and weaknesses of whatever writing he uses in his comparisons.

One expert has testified that making his own transcriptions as a matter of course affords several benefits.[27] Analysts generally compare writings, but infringement may lie in a performance made subsequent to the writing. Thus, the expert may miss critical infringing elements by relying on a writing that predates the performance. Furthermore, the expert can always testify as to the authenticity of his own transcription in court and can better defend its accuracy. Expert testimony does not serve the client's purpose if the expert disavows the writing as not his own when it is subject to attack on cross-examination. Finally, by transcribing both works himself, the expert removes stylistic differences in notation.

Transposition of the works to a common key facilitates comparison and is seldom criticized in court. The gestalt psychologist Christian von Ehrenfels established in 1890 that melody is recognizable without regard to key.[28] Although courts apparently have not cited this authority, the expert could easily cite numerous infringement cases in which transposition to a common key was not only allowed but even recognized as essential to comparative analysis.[29]

Transposition of equivalent meters should receive the same treatment but does not. Groupings of pulse in music can be represented by more than one mathematical formula, which is the essence of metric notation. The difference between duple and triple groupings must be acknowledged, but it does not preclude comparison. Some meters are mathematically equivalent, such as 4/4 and 2/2. Although these meters may suggest certain styles and be associated with specific musical traditions, they are too closely related for counsel to raise a meaningful objection to one being transcribed to the other.

Experts themselves have resisted questions predicated on metric alterations and refused to acknowledge strong similarities between certain meters.[30] Harold Barlow took a particularly obstinate stand on the difference between the meters 4/4 and 2/2 while testifying in Ensign Music Corp. v. Avis, Inc.[31] He maintained that he had not transposed a song from 4/4 to 2/2. The registered version was written in 4/4, but Barlow insisted it was performed in 2/2 and that his transcription, made from the performance, had not transposed the meter.[32]

Methods of Presentation

Visual formats can be crucial in explaining technical matters to a lay jury. Some techniques used by Nizer in Baron have been superseded; no other use of steel notes has been found. But colored graphs and charts remain popular devices. The duplication of pitches at equivalent or near-equivalent rhythmic positions, the mainstay of Spaeth's demonstrations, continues to be the predominant focus of infringement analyses. This requires a side-by-side comparison conducive to drawing connecting arrows between notes. Music is most often reduced to the melody, sometimes with chord symbols indicated in lead-sheet fashion.[33] A two-staff system, one work on each, allows the measures to be aligned and similarities highlighted.

Earl Spielman, a frequent expert witness, has further refined this technique in what he calls "chromatranscriptions."[34] By assigning a different color to each pitch and plotting the colors on a horizontal time line, the superimposition of one work over the other clearly shows the temporal coincidence of common pitches. These transcriptions are no less valid than common notation. In fact, the technique tends to depict temporal aspects more accurately than common notation. Most important, chromatranscriptions are easier for the layman to read and more clearly focus on their limited purpose‑-showing similarities. Chromatranscriptions work equally well for showing harmonic progression and might be applied to other parameters as well.

Experts have also represented melodic contour in graphic form and asserted it to be a critical feature. Whatever the potential relevance of this technique, some contour graphs better represent the music than others. A graph in which the expert plots each pitch at its point of attack and then draws connecting lines distorts both duration and motion; it implies continuous glissandi.[35] Horizontal bar graphs generally provide a more accurate picture but, perhaps, with less visual impact.[36] Where general contour alone appears without an accurate plotting of pitch, the graph lacks foundation and its use at trial might be disallowed as being more prejudicial than relevant.[37]

Attorneys frequently ask experts to perform their examples for the court. Electronic keyboards, more portable than pianos, have recently gained favor. Unfortunately, the transcript made of the court's proceedings cannot capture these performances. Clues to what happened may appear in the court record, but as often the record merely states "witness played" or "witness sang." In at least one instance, an appellate court has refused to act on an appellant's assertion of error, noting the lack of an adequate trial record to review.[38] Some attorneys therefore prefer to use pre-recorded examples wherever possible to incorporate the musical recordings themselves into the court record.

Depositions, taken during pre-trial discovery, are now often recorded on videotape. In addition, some federal courtrooms have installed equipment to videotape their proceedings and to supplement the written record. The nature of music infringement cases suggests that a party should request a taped record where possible. Most lawyers probably are not sufficiently knowledgeable about music to make an immediate verbal objection on the court record to inaccuracies in musical evidence. Because a timely objection is necessary to exclude evidence at trial, some undetermined amount of misinformation and prejudicial evidence probably has been placed before triers of fact and entered the court record in a form that does not reflect the error. No doubt the value of musical demonstrations has outweighed the dangers of prejudice and silent records, but it should be unnecessary to balance those factors with the technology currently available.

Exemplary Cases

The three cases that follow are relatively complete examples of infringement litigation. Legal strategies, analytical techniques, and the relative importance between musical and legal issues vary in these cases. The testimony of all expert witnesses in these cases is discussed; the testimony of various fact witnesses has been omitted. Legal issues affecting the expert testimony are explained where necessary. Actual trial exhibits, where available and relevant, are reproduced in the appendices. Although the trials are presented with a minimum of commentary, many of the musical and legal issues raised at trial will be re-examined in subsequent chapters.

Selle v. Gibb

Ronald H. Selle, an Illinois antique dealer with a master's degree in music education, composed popular and religious songs and played locally in a small band. One day in 1978 while Selle was working in his yard, a cassette player belonging to the teenager next door blared out what Selle recognized as the music to his song "Let It End." Since writing the song in 1975, Selle had sent the lead sheet and a home recording to fourteen publishers.[39] Eleven publishers had returned the materials; three never responded.[40] Selle instituted an action against Barry Gibb, Robin Gibb, and Maurice Gibb (a/k/a the Bee Gees), Phonodisc, Inc. (a/k/a Polygram Distribution, Inc.), and Paramount Pictures, accusing them of misappropriating his music in the hit song "How Deep Is Your Love." The case was tried to a jury in 1983.[41]

Selle's counsel, Allen C. Engerman, outlined the plaintiff's case in his opening statement. Experts would testify to striking similarities between the two songs. Of thirty-four notes in the opening eight measures of "Let It End" and forty in the corresponding measures of "How Deep Is Your Love," twenty-four were identical in pitch and thirty-five were identical in rhythm.[42] Because no link was ever established between the Bee Gees and any of the publishers to whom Selle had sent his song, access could not be shown. Instead, the plaintiff would show striking similarities to rule out the possibility of independent creation.[43]

Robert Osterberg, representing the Bee Gees, countered. Copying was the key word, and similarities alone could not show copying. Copyright does not grant a monopoly on a particular musical expression; it only prevents another from copying that expression. Coincidental similarities, attributable to the limited number of notes in the vocabulary of popular music, provided no basis for legal redress.[44] Selle could not establish copying or any opportunity for the Bee Gees to have heard his song. Selle's song was never recorded or performed where the Bee Gees could have heard it, and the Bee Gees had a strict policy of not reviewing unsolicited material.[45] Furthermore, the two songs had little in common. Plaintiff's song, according to Osterberg, was predictable and amateurish‑-the Bee Gees', complex and sophisticated. "How Deep" reflected the musical genius of one of the most outstanding songwriting groups in popular music history.[46] "Let It End" met with universal rejection, "How Deep" with immediate and phenomenal success.[47]

Direct examination of Selle placed the basic facts in evidence.[48] He explained to the jury how he wrote his song and the efforts he made to market it. On cross-examination, Osterberg confronted him with examples of prior art, works which predated both songs at issue.[49] He pointed to similarities between the works at issue and other Bee Gees songs. Selle's song, Osterberg suggested, resembled the Beatles' song "From Me to You." Were these similarities between prior art and your song, Osterberg asked Selle repeatedly after identifying each example, the result of coincidence?

Selle claimed the Bee Gees took two key sections of his work, the opening (theme A) and ending (theme B). He admitted that the middle section of "How Deep" was original. Both key sections involve a motive treated sequentially.[50] Osterberg sought to establish that sequential repetition entails only the mechanical transference of the motive up or down in pitch and, therefore, that each subsequent statement was not an independently significant creative act. Selle characterized sequential repetition as only one option of many. Theme A (opening eight measures of plaintiff's song) employed an ascending sequence while theme B's sequence (last four measures of plaintiff's song) descended.[51] That choice required a creative act. But Selle admitted, inaccurately, that a composer had only twelve options once the decision was made to treat the motive sequentially.[52]

Direct Examination of Plaintiff's Expert

Plaintiff's expert, Arrand Parsons, possessed impeccable credentials as a theorist and musicologist. He had taught at Northwestern University since 1946, chaired the music theory department for five years, and published writings on music analysis.[53] Although he had never made a comparative analysis of two popular songs prior to this case, he believed he was qualified to testify because he was first of all a life-long student and teacher of music theory, a discipline which involves analysis.[54] The analytical process would not vary if the music were pop, country, rock, or classical.[55] Parsons began his direct testimony with a short lesson in music fundamentals. He then proceeded to explain how these two songs were so strikingly similar that they could not have been written independent of one another.[56]

Each juror held his personal copy of Exhibits 18 and 19, a printed card with a plastic mylar overlay.[57] Theme A of each song was written on the card on a separate staff, one above the other. Exhibit 18 covered the first phrase and Exhibit 19 the second. Where the pitches matched at or near the same rhythmic position, those notes appeared in red on the mylar overlay with red arrows drawn between them.[58] The arrows were numbered 1 through 24. Another copy of the exhibits was placed close to the courtroom piano so that Parsons could play and point without having to move. Parsons explained Exhibits 18 and 19, note by note.[59] Exhibits 20 and 21 made a rhythmic comparison of the same material.[60] The corresponding ties at the end of measures 1, 3, 5, and 7 were displayed in turquoise.[61] Thirty red arrows appeared on the mylar overlays of Exhibits 20 and 21.[62] The process was repeated for Theme B. The pitch comparison, Exhibit 23, showed fourteen red arrows in four measures.[63] The rhythmic comparison of Theme B matched at eleven points.[64] At each juncture, Parsons testified that similarities were so striking and vivid that, based on his experience and background in music theory and analysis, independent creation was precluded.

Other similarities appear. Theme A occurs at the beginning of both songs, theme B at the very end.[65] Theme B is not related to theme A; as two independent musical thoughts or melodic thoughts, they would require two creative acts.[66] Based on a structural analysis of the two songs, coupled with his detailed analysis of the melodies of Themes A and B, Parsons believed and reiterated that the two songs could not have been independently created. He did not know of any two musical compositions by two different composers that contain as many striking similarities as exist between these two songs.[67] Plaintiff's counsel Engerman tendered the witness.

The transcript suggests that Parsons was lucid and to the point. Going through each exhibit note by note may be somewhat dry, but a transcript fails to convey the witness's animation and the impact of musical examples. Parsons played the piano, sang a little, and played tapes as he pointed to the similarities on the charts. In all, he gave what seems a credible performance.

Cross-Examination of Plaintiff's Expert

Osterberg began his cross by questioning Parsons's choice of source material for his comparisons. Parsons had used the copyrighted lead sheet of "How Deep," but the Bee Gees do not read or write music. They claimed to have written "How Deep" through a process of trial and error while gathered at the piano with a cassette recorder.[68] A transcriber committed the song to writing only after the composition process was completed and a demo tape made. Over objections, Osterberg tried to establish that Parsons had based his comparisons on secondary materials.[69]

Parsons had formalized his opinions after completing his analysis of the two songs a few years before trial. His report, which matched the conclusions stated in his direct testimony, had been made without ever examining other songs written by the Bee Gees.[70] Although Parsons claimed to have listened to some of the Bee Gees' music, Osterberg named eighteen of their albums, asking each time if Parsons had listened to that particular album. In each case, Parsons answered no.[71]

After Parsons reiterated that he had never before compared popular songs, he made a critical admission: he did not know whether or not there is a great deal of similarity between songs in the pop music field.[72] Although certainly qualified to analyze similarities, Parsons's answers raised a question on whether he could judge how striking those similarities were. No other witness would substitute for Parsons on this inquiry; only an expert can testify that similarities should be considered striking in the legal sense. For the plaintiff to prevail, Parsons needed to state that the similarities could result only from copying. If he could not judge the uniqueness of the similarities of the two works at issue compared to the many similarities commonly found in popular music, then he could not testify as to their significance. If he had not taken the time to familiarize himself with the relevant stylistic elements of popular music, he could not know which similarities were unique to the two songs at issue and which were merely idiomatic and unprotected.

Asked to define his use of the term "striking similarity," Parsons stated that it exists where pitches and rhythms coincide.[73] Osterberg continued:

Q . . . [Y]ou have used it several times during the course of your initial presentation. I would like to know whether that's the sole meaning you attribute to the phrase striking similarity.

A Those were the two points I used in my report because the question was one of melodic comparison. . . .

My assignment here was to compare the two melodies under consideration. That is what I did.[74]

Judge Leighton wanted more. At the end of cross-examination, he put the question to the witness himself. Was Parsons familiar with the expression "striking similarity" before he undertook the analysis of these two songs? Parsons responded that he was not previously aware of the legal implications, but he had used the term in his musical instructions.[75]

The judge's question was critical particularly in light of Parsons's hedging during Osterberg's questioning on the existence of copying. For example:

Q Is it your opinion, Mr. Parsons, that the only way the B Theme of "How Deep Is Your Love" could come about was as a result of copying Mr. Selle's song?

A I, I don't believe‑-put it positively. I believe that the Bee Gees' song, with these elements which we have described in common with the Selle song, I believe that the Bee Gees song could not have come into being with the‑-I must correct that. Because that is again dealing with a conjecture.

I believe that the elements, if I may just wipe that away and start again, because it's gotten twisted up.

I believe that the elements which are in common between the two songs in question are of such striking similarity that the second song could not have come into being without the first.

Q By that are you saying without reference to the first? Are you saying without reference to the first? I don't‑-can you explain what you mean by couldn't have come into being with the first?

A Could not have been composed without the first.

Q I am asking whether you are saying that it couldn't have been composed without seeing the first song, without referring to it, without copying it? What is your testimony?

A All of those things you described, copying, having seen it first, I have forgotten the others, I, I have no way, that is conjecture, I have no way of knowing whether it was seen, it was‑-I only know that the two songs have so much in common that it is‑-that the‑-that it precludes‑-this is too long. That the second song has so much in common, that is‑-let's get the names right. That the Bee Gees Brothers' song has so much in common with the Selle song that I cannot see, I cannot believe that they were created independent of one another.

Q Referring to your deposition, Mr. Parsons, page 89, commencing at line eight, I asked you the following question: "Q. Do you think the only way it could have come was a result of being copied from Mr. Selle's song?" "A. I could not answer that because I wouldn't know," end quote, line 11. Is that still an accurate answer today?

A I, I would not know. Yes, I would answer it the same.

Q So you can't answer the question because you don't know?

A No.[76]

Because plaintiff could not show access, his case depended on showing striking similarities such that coincidence, independent creation, and common source were all precluded. Parsons seemed to be using the term "striking similarity" loosely and would not state unequivocally that the similarities could result only from copying. The question of striking similarity goes to the heart of the composition process. In that sense the question is a purely musical one, and the court would have recognized Parsons as qualified to answer. However, without familiarity with the pop music field, that field's standards, and the differences of purpose between classical and pop composers, Parsons could not answer relevant questions on stylistic and economic considerations. Osterberg exploited this lack of knowledge and cast doubt on whether Parsons knew the significance of the similarities he had found. When Parsons failed to state unequivocally that the similarities proved a composition process that relied on copying, his use of the term "striking similarity" became legally meaningless. Without expert testimony on this point, plaintiff could not meet his burden of proof. The judge commented to the attorneys in chambers prior to giving jury instructions that Parsons had not satisfactorily answered his question.

THE COURT: Since he isn't here let me tell you why I asked him. I wanted to know from him whether the expression "striking similarity" is found in the works of analysts of music‑-that's what I wanted to know‑-or in his vocabulary, as I suspect, is an expression that began with his work in this case. He told me, I thought, that the expression is found in the works of music analysis. That's what he said. I wasn't satisfied the way he answered my question, but I didn't want to pursue the matter further so I left it.

MR. OSTERBERG: I thought he said he had used it before, but then I had asked him to define what he meant by striking similarity. His definition does not correspond to the legal definition.[77]

Another expert witness, Harold Gelman, waited to be called in from California. Plaintiff, however, elected to hold him in reserve for rebuttal and called only one more witness, Maurice Gibb. In his deposition, Gibb had been asked to identify a short excerpt prepared by one of plaintiff's experts. It was played again in court, and Gibb again identified the excerpt as from "How Deep Is Your Love." Engerman then read the stipulation into the record that the excerpt was "Let It End."[78] The press considered this a most dramatic event, and it was widely reported.[79] With a favorable impression left on the jury but no definitive testimony on striking similarity, plaintiff rested his case.

Evidence of Independent Creation

The defense concentrated on the work tape that purported to document the composition of "How Deep Is Your Love." The writing session took place in France at the Chateau D'Herouville in January 1977.[80] Barry Gibb was called to authenticate and explain the tape.[81] Albhy Galuten, the Bee Gees' record producer, testified that he was present at the composition session.[82] He had played piano and made minor suggestions.[83] Maurice and Robin Gibb corroborated this account of their song's creation.

Engerman tried to cross-examine Robin Gibb on discrepancies in the defendants' collective account of the tape's creation, but the judge disallowed it on the grounds that such questions went beyond the scope of direct examination.[84] The judge noted that the witness had only been asked whether he was a co-author and whether or not he ever heard about "Let It End" before he participated in the creation of "How Deep Is Your Love."[85] Engerman explained at side bar[86] that, although it was admittedly a work tape, his purpose was to show that the tape might not be the product of the initial creative effort.[87] He argued that there was a question as to when the work tape was created. It might be a work tape merely to refine a song. Engerman wanted to show that the work tape might have been made after the Bee Gees' left the chateau and after the music was first submitted to the Copyright Office.[88] The judge did not change his ruling.

The defense rested. In a surprise move the defense elected to forego their experts. Osterberg had surmised from some of the judge's comments to the attorneys in chambers that the judge did not think Parsons's testimony established the requisite level of striking similarities.[89] If the defense put on its own experts, then the plaintiff could counter with rebuttal experts. Proceeding with the defense would raise issues of fact for the jury to decide; as it stood, one issue of law, striking similarity, was overriding. Nevertheless, the judge refused a defense motion for a directed verdict for the defendant as a matter of law. Striking similarity, he said, was a question for the jury. However, he observed outside the presence of the jury that it was obvious to him that the first eight measures were not strikingly similar.[90]

Plaintiff could not at this point put Gelman on the stand to show striking similarities, because plaintiff had rested his case and there was no defense expert to rebut. Plaintiff considered presenting expert testimony to show that the work tape introduced by defendants did not represent the initial creative effort but apparently realized that such testimony would be ineffective. Composition is a mental process, and the work tape could not prove the composers' mental processes. The jury probably viewed the tape with justifiable skepticism. Unreliable as that tape might be, an expert's assertion that he can prove the tape a fraud would be still more fantastic. In the end, plaintiff never rebutted the work tape's feeble contribution to showing independent creation.

Jury Verdict

Jerrold Gold, plaintiff's co-counsel, summed up the case for the jury. Parsons testified to striking similarities and the defense had presented nothing to refute his conclusions.[91] Defendants' evidence of independent creation was inconsistent and inconclusive.[92] Those in the entertainment world hear many songs; the Bee Gees might have copied Selle's subconsciously.[93]

Osterberg then began his summation. He reminded the jury that Parsons used a definition of striking similarity at odds with the legal definition. Parsons would not rule out the possibilities of independent creation, coincidence, and common source.[94] Although skilled in the analysis of classical music, Parsons clearly knew little about popular music. What was critical about this gap in his knowledge was that he could not judge the significance of the similarities. Therefore, when Parsons said he did not know of any two musical compositions by two different composers that contain as many striking similarities as exist between these two songs, he drew a conclusion beyond his expertise. Osterberg implored the jury to substitute their own superior knowledge of popular music for the opinions of Parsons.[95]

Engerman, in rebuttal, completed plaintiff's summation on a more personal note. Even though the defense experts were seated at counsel's table and introduced to the jury at the start of trial, Parsons was the only expert to testify. His testimony had been eloquent and beyond impeachment. A professional such as Parsons relies on his reputation and would not sell his opinion for an expert witness's fee. His conclusions were dictated by the evidence. When the red arrows are counted up, how could the songs be characterized as other than strikingly similar?[96]

The judge read the jury instructions and sequestered the jury at the end of the day. They deliberated for most of the next day, and at 3:30 p.m. returned a verdict for the plaintiff. "How Deep Is Your Love," they said, infringed Selle's song.

Judgment Notwithstanding the Verdict

The jury decision did not finally resolve the matter. Judge Leighton nullified the jury verdict, granting the defense motion for a judgment notwithstanding the verdict (j.n.o.v). The judge's written opinion explains his reasons. A judgment notwithstanding the verdict is nothing more than a directed verdict granted after the jury has brought in its verdict.[97] To satisfy the criteria for a j.n.o.v. the judge must view all evidence in the light most favorable to the nonmoving party, here the plaintiff. The judge must find that, even with all justifiable inferences, the evidence simply cannot support the jury's decision.[98] It is immaterial that at the close of evidence the court, knowing no more then than now, denied the motion for a directed verdict and sent the case to the jury.[99]

The judge explained that the plaintiff never rebutted defendants' evidence of independent creation, the testimony of witnesses present at the composition of "How Deep Is Your Love" buttressed by the recording of the composition process. Although an inference of copying might be justified had plaintiff shown striking similarities, that inference here would be at war with the undisputed testimony of independent creation. Inference alone cannot outweigh actual testimony. Therefore, the inference could not stand; some evidence, albeit weak, for the defendants trumped no evidence to support the plaintiff's claim. Without at least an inference of copying, there can be no copyright claim regardless of the similarities. Thus, the verdict could not stand as a matter of law.[100]

Selle appealed, claiming the district court misunderstood his theory of proof.[101] The Seventh Circuit began its analysis by stating that copying must be proved. Although direct evidence of copying is often lacking, circumstantial evidence of access and substantial similarities are sufficient. Where there is no evidence of access, striking similarities may raise a permissible inference of copying by showing that independent creation, coincidence, or common source are, as a practical matter, precluded.[102] Striking similarity per se, however, does not suffice. It provides only one piece of circumstantial evidence tending to showing access. It must be viewed with other evidence. For example, if the plaintiff admitted to keeping his work under lock and key, striking similarity could not allow the jury to infer access.[103]

In this case, the possibility that the Bee Gees had access to Selle's song was de minimis.[104] Evidence of access must extend beyond mere speculation.[105] Plaintiff relied on Parsons, who ruled out independent creation but did not state that the similarities could result only from copying.[106] Parsons had not addressed the possibility of a prior common source. Although the burden of showing a common source normally rests with the defendant, a plaintiff attempting to show striking similarity must bear the burden of showing no prior common source.[107]

Judge Leighton's judgment was affirmed. Parsons had shown striking similarities in a non-legal sense only. Plaintiff had failed to provide a sufficient basis for the jury to infer that the Bee Gees had access to Selle's song.[108]

Those not trained in law may find this result remarkable, but part of the trial judge's role is to guard against excesses of the jury. Here, the jury apparently believed that similarities alone could support an infringement claim. They cannot. No evidence had been put before them that the Bee Gees could have had access to Selle's song or that the similarities were so striking that they could result only from copying. This decision did not rest on a legal technicality. Rather, the missing evidence addresses the core of substantive copyright law. Copyright does not grant a monopoly on particular expression; it protects the author only against copying. Plaintiff had not met his burden of proof. One can only speculate as to why Judge Leighton allowed the question to go to the jury rather than granting a directed verdict at the close of evidence.

Gaste v. Kaiserman

Morris Albert (née Kaiserman[109]) composed "Feelings" in Brazil in 1973.[110] First released in the second quarter of 1975, the song earned royalties of $47,242 during the remainder of that year. In 1976, the annual figure rose to $154,331. By 1985, royalties totalled $1,372,571.[111]

"Pour toi" appeared in 1956 as part of the score to the motion picture Le Feu Aux Poudres. Composer Louis Gasté obtained a U.S. copyright in 1957.[112] Because it was an adult film and Albert was a child at the time, Albert would not have seen it when it was first released in Brazil. But the film was re-released in the late 1960s, and Albert might have had the opportunity to see it then. The sheet music of "Pour toi" was available, but the song was not widely known. Its earnings totalled less than $15,000 at the time of trial.[113]

Gasté also alleged a direct link between co-defendant Fermata Music and Gasté's publisher. According to Gasté, Fermata's owner Enrique Lebendiger had received a copy of "Pour toi" in the 1950s when Gasté was trying to market the song and then passed it on to Albert in 1973.[114] Lebendiger denied the allegation.

Direct Examination of Plaintiff's Expert

Plaintiff's counsel, Andrew Bart, called Anthony Ricigliano, chairman of the music theory department of the Manhattan School of Music. Ricigliano compared five elements: melody, chords, rhythm, lyrics, and structure (or form).[115] He defined melody to include interval content, rhythm, basic shapes, and structure.[116] "Rhythm is the organization of time values or duration."[117] "Form refers to the organization of the musical ideas."[118] The expert had transposed "Feelings" from E minor to C minor, the key of "Pour toi," for comparison.[119] Both songs were then transposed to E minor for demonstration at trial.[120]

The court admitted Ricigliano's "Chart 4" into evidence as Exhibit 96.[121] The chart compares the two songs. Where the same note begins on the same beat of both songs, Ricigliano drew a rectangular block encompassing both. Diagonal arrows connect duplicated notes that occur on different beats. Line graphs showing melodic contour are produced above "Pour toi" on the top staff and below "Feelings" on the bottom staff in contrasting colors.

The first eight measures of Exhibit 96 (theme 1, repeated) show few similarities. The contour graphs appear too far removed from each other to be effectively compared. Rectangles showing note duplications occur on the first beat of every other measure. Six diagonal arrows connect rhythmically displaced duplicate pitches. A system of three staves depicts similarities in the chorus. This method is dictated by the fact that "Feelings" repeats the chorus with variations. The expert thus compared "Pour toi," which does not repeat, both to the original statement and to the repeat of "Feelings." The harmonic progression is written in chord symbols above each staff, although the chart makes no attempt to point out harmonic similarities.

Ricigliano turned almost immediately to Exhibit 93, a contour graph showing pitches connected by colored lines and approximate rhythmic positions only.[122] The exhibit discarded common notation and provided an opportunity for Ricigliano to explain scale construction. The chart employs one line for each scale degree, evenly spaced without regard for whole and half steps.

The elimination of rhythm from the chart was defended by noting that bar lines remained to keep notes within their proper groupings.[123] Melodic contour matches precisely at the opening figure, a descending perfect fifth (B ‑ E). The succeeding figure rises and falls a third. The "Feelings" figure reads E F# G E; the corresponding figure of "Pour toi" is F# G A G F#. The chart shows the figures occurring at the same point, the only difference being that the "Feelings" figure begins one step lower. Ricigliano pointed out that they both rise and fall the same musical distance.[124] In measure 5, where "Pour toi" moves up to E² and "Feelings" repeats E¹, the graph seems to show the songs mirroring each other as they approach a resolution on B. Neither side commented on this feature. Ricigliano testified that "Feelings'" approach to B (shown as F# G A F# in measure 6) is derived from measures 2 and 4 of "Pour toi" (the figure F# G A G F#).[125] The final B was described as the "goal note" common to the main theme of both songs. While "Pour toi" ascends a perfect fourth in measure 5, "Feelings" restates the descending perfect fifth. "Pour toi" then works its way down to B while "Feelings" works it way back up.[126]

"Pour toi" repeats the theme literally in measures 9 to 16. Variations on the theme appear in "Feelings," principally in measure 13 where the descending fifth of measure 5 becomes a descending major third (B ‑ G). Otherwise, Ricigliano stated, measures 9 through 16 bear similarities equivalent to measures 1 through 8.[127]

Ricigliano testified that the ballad style of "Feelings" accommodated considerable rhythmic variety and rubato in the vocal line. In addition to liberties taken by the performer, some of these variations appear in print. Ricigliano produced a chart comparing five variations of the first eight measures.[128] Many of those variations were attributed to translations of the text into other languages. Plaintiff's counsel Bart stipulated that the purpose of introducing the exhibit was to show that a song is not confined to a particular arrangement, that its essence remained constant despite being arranged in different styles.[129] The exhibit helped to bridge the considerable stylistic gap between "Feelings" and "Pour toi." To further illustrate the point, Ricigliano played "Pour toi" in three different styles: ballad, up-tempo, and waltz.[130]

The choruses bore fewer surface similarities than the main theme. Exhibit 94 (comparing the choruses, measures 17 to 24) employs large and small notes to separate the functional from the ornamental. Ricigliano placed arrows over repeated notes in "Pour toi" to show the predominance of the descending melodic sequence. Functional notes were identified as E D C B. On the staff below, a quasi-Schenkerian graph reduces the chorus to tied whole notes, with stemless filled note heads in parentheses representing the initial note of each sequential statement a third above. Ricigliano laid a foundation for his chart by identifying the technique as common to 20th-century theorists.[131]

Ricigliano explained his choice of E as the functional note of the first sequential statement. The G occurs only at the beginning. The E is repeated and predominates for three quarters of the two-measure statement.[132] Once that decision was made, the sequence dictates the same analysis for each succeeding statement. However, Ricigliano could not resist pointing out that the C in the third sequential statement receives additional emphasis through slight variation of the sequential pattern.[133] More to the point, the obvious cadence on B supports Ricigliano's reduction.

The repeat of the "Feelings" chorus (measures 25 to 31) is a simpler version of its initial statement. The melodic figurations of the initial statement obscure the descending sequence somewhat, but the repeat seems to emphasize it. Ricigliano capitalized on this feature by inverting the organization of staves on his chart. The same quasi-Schenkerian reduction of "Feelings" (minus the parenthetical initial notes) appears on top of the second system, immediately under the similar reduction of "Pour toi." The more obvious sequence, the repeat of the "Feelings" chorus is placed under the reduction, and the initial statement of the chorus appears on the bottom staff. Ricigliano thus charts a progression from "Pour toi," to the underlying sequence of both songs, to the surface manifestations of "Feelings," in their order of complexity.

Between the two systems, chord symbols for the progression of "Pour toi" are shown. Because the "Feelings" progression is omitted, this aspect of the chart could have been challenged as implying that "Feelings" employs the same progression. Although Ricigliano testified to the differences and indicated that the "Feelings" chord symbols had been added to the exhibit at the bottom,[134] they appear as an afterthought. Ricigliano defended his emphasis on the "Pour toi" progression, saying it represents the underlying progression of "Feelings."[135] By omitting the "Feelings" progression, however, the chart fails to reflect the critical harmonic similarities. Example 9 shows the actual progressions.[136]

Example 9: "Pour toi" v. "Feelings," harmonic analysis.

Ricigliano correctly noted that "Feelings" employs the chords C Bm Am at points corresponding to those same chords in "Pour toi." Ricigliano continued though to claim that those chords were followed by "the basic chord in both versions here[,] really . . . the dominant seventh chord."[137]

The underlying progression common to both songs (C, B minor, A minor) was played to illustrate its anticipated resolution to a G chord instead of B7. Ricigliano found significant similarities in this transition back to the "A" section.[138] The two songs also modulate to the relative major key at the same point.[139] Most important, Ricigliano described the method of modulation leading to the chorus as unique. Although the modulation itself is commonplace, the preparation of one resolution followed by an evaded resolution provided Ricigliano with a musical fingerprint.[140] "Pour toi" modulates to G in measures 6 and 7. The dominant chord in E minor follows in measure 8. It suggests a return to the original key, which in fact occurs when the "A" section is repeated. However, the same B7 chord leads to the C-major chord of the chorus. The expected resolution to tonic in E minor is thus evaded; what follows is described as subdominant in the relative major.[141]

"Feelings" does not duplicate this progression, but Ricigliano testified that it employs the same principle of an evaded resolution.[142] The modulation to the relative major within the A section of "Feelings" matches that of "Pour toi," and the first ending follows, in essence, the model of "Pour toi." But unlike "Pour toi," "Feelings" has a second ending which moves to an E7 chord instead of B7. The chorus of both songs begins on a C major chord. "Feelings" evades the anticipated resolution of E7 just as "Pour toi" evades the resolution of B7, according to Ricigliano.[143]

Based on his analysis, Ricigliano found the two songs extraordinarily similar. Every measure of "Feelings" contains references to ideas contained in "Pour toi."[144] The same similarities of both sections could not be found in prior art. No prior art modulated in the same measures. None used the same evaded resolution as a transition to the chorus.[145]

Cross-Examination of Plaintiff's Expert

Osterberg concentrated on surface differences between the two songs and the omission of those differences from Ricigliano's analysis. He wanted to know why Ricigliano had not focused his analysis on the actual duplication of notes. To illustrate the lack of this kind of literal similarity, Osterberg had Ricigliano draw arrows between identically corresponding notes. The result, which matches that of Exhibit 96, was to connect only the melodic B appearing in measures 1, 3, 5, and 7.[146]

Osterberg then questioned Ricigliano's Exhibit 93. Didn't it omit any reference to rhythm? No, Ricigliano countered. It showed bar lines, one aspect of rhythm.[147] But Osterberg was not satisfied that it showed a true picture of how those notes occur within each measures. Argument ensued on whether bar lines were the proper point at which to segment music. Ricigliano maintained that bar lines were inconsequential to analysis, perhaps contradicting his decision to retain that one aspect of rhythm.

Osterberg also criticized the method of designating pitch on Exhibit 93. He suggested that common notation on a five-line staff would show contour just as well, and it would also show rhythm. Ricigliano disagreed. His contour graph showed more accurately the distance between notes, he claimed.[148] Further, melodic contour provides a better analysis than isolating and counting notes.[149]

Other differences were pointed out on cross-examination. The bass line of "Pour toi" was merely accompaniment; that of "Feelings" stated a distinctive melodic idea.[150] Ricigliano would not admit that the bass line was unique or even distinctive. Osterberg could get him to admit only that the bass line descends, while that of "Pour toi" ascends.[151]

Osterberg successfully distinguished overall form. Although the two songs track each other in phrase lengths, points of modulation, repeat of the "A" section, and the appearance of section "B," Ricigliano admitted that "Pour toi" was essentially ternary (A B A) and "Feelings" binary (A B).[152] The repeat of the "Feelings" chorus replaces the return to the A section in "Pour toi."

Osterberg turned to the rests. "Feelings" utilizes prolonged rests between phrases, where "Pour toi" has none. Ricigliano attributed the difference to different notational styles. In the 1950s, rests were often omitted to make the music easier to read. The points at which a performer would have to insert rests in order to breath, Ricigliano explained, would match the rests of "Feelings."[153]

The final demonstration involved asking Ricigliano to play only those notes that had been designated as identical, a performance that must have been marked by considerable silence. Bart sought to overcome this more rigid approach to analysis on re-direct. He asked Ricigliano what would result from starting the composition one beat late and then making a note-for-note comparison. Obviously, nothing would match, but the piece would sound the same.[154]

Direct Examination of Defendants' Expert

Defendant Fermata Music had employed Irwin Coster as its expert. He did not provide any new charts comparing the two songs. Instead, he attacked the validity of Ricigliano's charts and the similarities they implied.

The four-note melodic figure in measure 2 of "Feelings" (E F# G E) could not be equated with measure 2 of "Pour toi" (F# G A G F#). The figures were different in contour and rhythm.[155] Although Coster's point was well taken, he missed an obvious counter to Ricigliano's non-literal analysis‑-the two figures are not functionally equivalent. Still, Coster demonstrated that the element of rhythm was critical by showing that its omission allowed the four-note "Feelings" figure to be equated to the theme from Dragnet.[156]

Both songs begin with a four-measure introduction. Not only did Coster find those introductions dissimilar, but he used them to argue for differences in the bodies of the songs. Introductions function to set the stage for the song and clue the listener to important features. "Pour toi's" introduction highlighted the rhythm and harmony in a French cabaret style; the "Feelings" introduction emphasized lyrical aspects.[157]

Coster also distinguished the harmonic progressions. In measures 3 and 4, "Pour toi" repeats the harmony of measures 1 and 2. It reiterates the motion from tonic to subdominant. "Feelings," on the other hand, progresses without repeat to measure 8. It reaches the subdominant only in measure 5, which sets up a progression to the dominant.[158] The harmony of "Pour toi" is diatonic; "Feelings" adds chromaticism.[159] While the "Pour toi" chorus simply descends harmonically through C, B minor, A minor, to B, "Feelings" follows the circle of fifths, albeit with some chord substitutions.[160]

Where Ricigliano had equated the ascent of a perfect fourth in measure 5 of "Pour toi" to the corresponding descending perfect fifth in "Feelings," Coster argued they were not the same.[161] He provided good reasons. The appearance of the high E in "Pour toi" diminished its effectiveness in the chorus. The high E stated for the first time in the "Feelings" chorus was fresh; its novelty was already expended in "Pour toi." For this reason, Ricigliano should not have analyzed the "Pour toi" chorus as having a functional melodic background of E D C B.

The G on which "Pour toi's" chorus begins is essential.[162] Thus, the sequential figure of the chorus had to be built on an underlying G F# E B in "Pour toi" and E D C B in "Feelings." After discounting background similarities, Coster dispensed with foreground similarities as well. "Feelings" does not employ the same three-note repeated figure appearing in "Pour toi" at measure 18 (C D E C D E) and in each sequential restatement. The corresponding figure in "Feelings" (E D C B C) could not be considered similar.[163]

Finally, Coster argued that the evaded cadence pointed out by Ricigliano was not unique. In fact, Albert had used a similar device in another song appearing on the same album. The E7 chord in "Feelings" sets up the first melodic E of the chorus. In "I'm Going to Love You More," Albert uses the harmonic transition A minor to A major, with an E7 chord going to the melodic E at the start of the melody.[164] Thus, Coster posited that the transition to the chorus of "Feelings" employed a device already within Albert's vocabulary.

Osterberg then had Coster introduce examples of prior art. They included excerpts from Die Meistersinger,[165] a Schumann violin sonata, and the Russian folk song, "Donna Donna."[166] Coster's written example comparing Die Meistersinger to "Feelings" augmented Wagner's note values to make them match those of the accused work. Following an objection to this liberty, the judge asked for an explanation.[167] Coster pointed out that just as a common meter obscures temporal differences between "Pour toi" and "Feelings," the contrasting meter of Die Meistersinger obscures the similarities. The judge, apparently confused by the meaning of meter, finally asked Coster to play Wagner both ways (with augmented note values and as written by Wagner). Unable to detect any difference, he overruled the objection.[168]

Confusion continued with the Schumann sonata.[169] The question was whether Coster changed any notes when he transcribed it.[170] Coster had changed one note by an octave in Schumann "to illustrate that if I made the same octave jump as in `Pour Toi,' this is what it would sound like."[171] The judge ruled on the objection: "You testified earlier that that was a significant difference in changing an octave. . . . Now you're changing an octave in Schumann's work and I'll sustain the objection. What's fair for the goose is fair for the gander."[172]

Bart also objected to the defense playing a recording of the Schumann sonata because it was in 6/8 meter and thus not comparable. The objection was overruled.[173] He then objected on the same grounds to a comparison with a Bach work in 3/4 time.[174] Objections on the grounds of differing meter may seem like good advocacy; counsel should raise reasonable arguments against the admission of potentially damaging evidence. However, just as Coster could not stress the significance of register and then change octaves in the Schumann example, Bart had based much of his case on a witness who claimed virtually all aspects of rhythm could be eliminated from the analysis. Further, Bart's expert had played "Pour toi" as a waltz, in 3/4 time. An astute juror might spot the inconsistency. The court overruled this objection and admitted the prior art.

Cross-Examination of Defendants' Expert

Bart began his cross-examination of Coster on the subject of rhythm. Coster admitted that there were many arrangements of "Feelings," each with rhythmic variations. Yet even with those variations, the song was still recognizable as "Feelings."[175] Bart returned to the subject later when he asked if the singer of "Pour toi" would have to insert rests in order to breathe.[176] Making the witness draw bright lines in a gray area may not be musically sound, but it made Bart's point. Coster stated that inserting quarter rests would not change the melody but inserting half rests would require rhythmic alterations and change the melody. If all of the rhythmic differences between versions of "Feelings" were insignificant, Bart asked, why were they so important between "Feelings" and "Pour toi"?[177]

Bart brought out the point that the "Feelings" bass line, even if distinctive, would not excuse copying. He argued that the bass line alone could not provide an independent basis for protection and that the bass line did not mitigate other similarities.[178] The song's introduction received similar treatment. Bart played Exhibit 39, a recording of "Pour toi" performed by Ario Moreno. Coster noted that the introduction was different and that the first bar sounded three notes instead of two. "So a performer can perform a song with a different introduction, different arrangement, even with a different number of notes in a particular measure and it's still the same song, correct?" Coster responded affirmatively.[179]

Ricigliano had described the phrasing of the two songs as the same. The lengthy rests in the opening theme of "Feelings" clearly demarcated phrases. Ricigliano argued that phrase breaks would occur at the corresponding points in "Pour toi." Thus, measure 1 stated a complete idea, and measures 2 and 3, 4 and 5, and 6 through 8 should be grouped together.[180] Coster, however, chose a different analysis. He maintained that the proper groupings of the "A" section were at bar lines. Measures 1 and 2, 3 and 4, and 5 to 7 represented the most musically logical segmentation.[181] Measure 1 contained a statement, measure 2 the answer. Although he admitted that the lyrics could provide a clue to the phrasing, Coster had not looked at lyrics, because they were not a part of the claim.[182]

Coster gave a confusing account of modulations. First he claimed the song never changed key, and then he said it modulated to the relative major. He described the harmonic progression as moving to the G chord but not modulating.[183] Next, Coster claimed there could be no modulation without a notated change of key signature.[184]

Jury Verdict

Coster had made some strong points on direct but could not maintain them on cross. Ricigliano's testimony had not been free of gaffes either. No witness is likely to endure cross-examination by seasoned attorneys without losing some points. As Osterberg noted in his summation,[185] Ricigliano had to correct mistakes on his charts during his testimony and he could not define the performance marking on "Pour toi"‑-tempo di Batucada.[186] Osterberg read the definition to him from the Harvard Dictionary‑-a minor embarrassment for Ricigliano.[187] Nevertheless, Osterberg later credited Ricigliano with giving a largely unshakable performance.[188] Perhaps of more importance, Coster's failure to provide his own exhibits left the jurors to deliberate with reference to Ricigliano's charts alone.

After deliberating for one hour, the jury returned a verdict for the plaintiff. Judgment was entered. On appeal, Albert asserted, among other claims, that plaintiff had failed to prove copying as a matter of law and that the district court incorrectly instructed the jury on the issues of access and striking similarity.[189] At the heart of these issues was Albert's contention that plaintiff's evidence of access was too remote.[190]

The jury apparently found a sufficient showing of access, and the Second Circuit refused to find the evidence inadequate as a matter of law.[191] The court also found that the trial judge properly instructed the jury regarding access in a case of striking similarities. Arguments based on Selle, that there must be some other evidence of a reasonable possibility of access, were rejected. Not only was the trial judge's statement of law correct, but sufficient evidence existed to warrant the inference of access.[192]

Baxter v. MCA, Inc.

The Baxter[193] case illustrates some of the difficulties encountered in applying the judicial tests for infringement. Following pre-trial discovery, the defendants moved for summary judgment, claiming the plaintiff had failed to present any issue requiring a trial. The trial judge granted the summary judgment. On appeal, however, the Ninth Circuit applied the same tests as the trial judge but reached a different conclusion. Therefore, the appellate court ordered the trial to proceed, and a jury was asked to apply the same tests once more. Fortunately, at trial the issues were submitted to the jury one at a time, a factor that helps to clarify the process.

This discussion of Baxter will examine first the summary judgment and the legal standard applied in the Ninth Circuit. The district judge and appellate judges at this stage did not have the benefit of any testimony; they relied on the pleadings, a pre-trial report of plaintiff's expert, and recordings of the works at issue. This discussion will then move on to the trial. The trial was divided so that the jury would reach its first special verdict without hearing testimony; the two pieces of music would speak for themselves. Only after reaching this initial determination would the process turn to the expert witnesses. Then, based on the experts' analyses, the jury would answer three additional questions leading to a verdict on infringement.

The case was initiated when Leslie T. Baxter brought suit against John Williams, alleging that Williams's score to the 1982 film E.T.: The Extra-Terrestrial copied Baxter's "Joy," one of seven songs in a collection called The Passions. The film E.T. enjoyed enormous success, and Williams earned an Academy Award for his score.[194]

Baxter asserted that Williams had performed "Joy" at the Hollywood Bowl in the 1960s, an allegation that, if true, would establish access to the plaintiff's work. The similarities claimed to exist between the two works involved a rather short segment of each. What was referred to as the "B" theme of "Joy" first appears in measure 18 (Example 10). This theme appears subsequently in altered form, descending a perfect fourth to G at the end of the second measure (Example 11). This subsequent, altered statement of the B theme bears a closer resemblance to the principal theme of E.T. (Example 12). Plaintiff limited his claim to the similarities between these themes.[195]

Example 10: "Joy" "B" theme, meas. 18-25.

Example 11: "Joy" "B" theme, varied restatement, meas. 57-58.

Example 12: E.T. Theme, meas. 5-12.

Summary Judgment

The district court granted defendants' motion for summary judgment, determining that no substantial similarity of expression existed between the two works.[196] The court applied the "lay observer" test, as it was required to do, stating that the court's ear was "as lay as they come."[197] The court could not hear any substantial similarity between defendant's and plaintiff's expression of the idea. Indeed, it confessed that it could not even tell what the complaint was about. Although Harvey Bacal had presented a comparative analysis pointing out similarities in "a sequence of notes," the court concluded that, although those similarities might be apparent to an expert, the "total concept and feel" of the works was not sufficiently similar to warrant submission to a jury.[198]

Plaintiff appealed the summary judgment rendered against him. The Ninth Circuit, reviewing the district court decision, reiterated the test for infringement in that circuit. Plaintiff must show: (1) access, which at this point in the case was conceded,[199] and (2) substantial similarity of both general ideas and expression.[200] Williams had also conceded for the purpose of the summary judgment motion that the general ideas of the two works were substantially similar.[201] The only question on appeal, therefore, was whether to uphold the district court's finding, based on the judge's ear, that substantial similarity of expression was "totally lacking and could not be submitted to a jury."[202]

Understanding the precedents established in the Ninth Circuit by the Krofft decision is helpful.[203] Krofft requires the plaintiff to show both substantial similarity of general ideas under the "extrinsic test" and substantial similarity of expression of those ideas under the "intrinsic test."[204] General ideas have been held to encompass the objective, concrete elements of a work. In a literary context these elements include plot, theme, dialogue, mood, setting, pace, characters, and sequence of events.[205] The Ninth Circuit has yet to explain just what elements it considers objective and concrete in the field of music. Nevertheless, it treats as axiomatic the notion that these "general ideas" exist in all artistic creations, and courts must struggle to understand where the "idea" can be found in the two musical works at issue.

In Baxter, because Williams had conceded similarity of idea, the sole issue before the court was similarity of expression, requiring application of the intrinsic test. Application of the intrinsic test does not call for analytical dissection and expert testimony; the gauge of substantial similarity is the response of the ordinary lay hearer.[206] In Krofft, the court had rejected extrinsic analysis of similarities and differences of characters in a television program in favor of asking whether the defendant's work captured the "total concept and feel" of plaintiff's work.[207] The district court in Baxter had properly applied this subjective test in reaching its decision that no reasonable juror could find substantial similarity of expression between the two works at issue.

The Ninth Circuit had to decide if the district court reached the right result or whether reasonable minds could differ as to the presence or absence of substantial similarity of expression. The Ninth Circuit reviewed de novo[208] the total concept and feel of the two works at issue. It expressly recognized that, although it must base its decision on untutored first impressions, the question of substantial similarity of expression entails subtle and complex determinations.[209] The court decided that reasonable minds could differ as to whether "Joy" and E.T. employ substantially similar expression. Summary judgment was therefore improper. The court also rejected defendant's argument that, because the alleged similarity involved only a six-note fragment, it could not, as a matter of law, be found substantial. The court noted that no bright line rule exists regarding the amount of material that can qualify as a substantial taking.[210] Therefore, the Ninth Circuit reversed the district court ruling and, following Williams's unsuccessful application for the Supreme Court to hear the case,[211] the case was remanded to the district court for trial.

Since Baxter, the Ninth Circuit has ruled that it is improper for a court to grant summary judgment where substantial similarities are found through the extrinsic test.[212] The extrinsic test searches for similarity of idea, something Williams had admitted for purposes of his summary judgment motion. Although the first element of Baxter's case was met by this admission, the trial judge determined, by his own listening, that a reasonable person could not find substantial similarities of expression‑-the intrinsic test and second element that plaintiff had to prove. The Ninth Circuit came to the realization in Shaw, which presented a similar situation,[213] that a judge who grants summary judgment after the extrinsic test is satisfied does so on a purely subjective basis. This conflicts with the Supreme Court's ruling that "at the summary judgment stage, the judge's function is not to weigh the evidence and determine the truth of the matter but to determine whether there is a genuine issue for trial."[214]

The holding of Shaw, had it predated Baxter, would have mandated that Baxter reach the same conclusion, denial of summary judgment, but through different reasoning. Under Shaw, if sufficient evidence exists to present a triable issue on similarity of idea, the judge may not apply the lay observer test for similarity of expression before trial in order to grant summary judgment. (This is essentially what the trial judge in Baxter did.) However, if no triable issue exists on similarity of idea, then the judge may apply the lay observer test for similarity of expression. Where he finds no genuine issue of material fact on either question, summary judgment is proper. Shaw makes it very difficult to get a summary judgment in an infringement suit in the Ninth Circuit.

The Ninth Circuit in Shaw seems to confess real misgivings about the subjectivity of the intrinsic test. The inconsistency of its reasoning cannot be ignored. The plaintiff must satisfy both the extrinsic and intrinsic tests in order to prevail. But the Ninth Circuit now says that the case should go to trial even where as a matter of law the intrinsic test cannot be satisfied, as long as the extrinsic test can be satisfied.

Pre-Trial Motions

The case finally came up for trial in August 1988, almost five years after the suit was filed.[215] Attorney John Blanchard represented the plaintiff. He would call two expert witnesses, Robert Winter and Philip Springer. Defendant Williams was represented by Ronald Rosen. Louis Petrich and Edward Ruttenberg represented corporations named as co-defendants. The defense would also call two expert witnesses, Earl Spielman and Harold Barlow.

The parties agreed to trifurcate the trial: the first issue to be submitted to the jury involved the "total concept and feel test" for similarity of expression, phase two involved substantial similarity of idea, and phase three would assess damages. If the plaintiff lost either of the first two phases of trial, the proceedings would end immediately with a verdict for defendant. Phase one would be short. The court would play the recordings of the two works for the jury, and counsel would make opening statements and closing arguments. They would call no witnesses. The jury would merely listen to the two works without benefit of analysis and answer one question: "Is the expression of the musical idea and the music from `E.T.' substantially similar as defined in the [jury] instructions to a qualitatively important music expression in `Joy'?"[216] A negative answer would eliminate the need to proceed to phase two. A positive response would necessitate the presentation of expert testimony and a much more elaborate inquiry in the second phase.

Because phase one entailed no testimony, the most critical questions involved which tapes to play for the jury. These issues were decided in Judge Harry Hupp's chambers before jury selection. Judge Hupp heard arguments concerning the admissibility of various tapes that had been produced by the parties. The expert witnesses testified as to how the tapes were made and whether they were accurate renditions of the music. Counsel argued why each tape should or should not be presented to the jury.

"Joy" had been recorded originally in 1953 by full orchestra for the album released in 1954.[217] Plaintiff had created a synthesized version (Exhibit 36[218]) for the sole purpose of demonstrating the music at trial. Plaintiff's counsel Blanchard called Robert Winter, a professor of musicology at the University of California, Los Angeles. Winter stated that the original recording of "Joy" suffered from inferior technology and would present a less clear version of the music than would the new synthesized version. Winter authenticated both the original recording and the synthesized version as consistent with the copyrighted score, but he admitted some difficulty in reading the original score of "Joy."[219] The copy registered with the copyright office was faint and virtually indecipherable in places.[220]

Rosen, counsel for Williams, cross-examined Winter. He obtained an admission that variances could be found between the dynamics of the original orchestral version and those of the synthesized Exhibit 36, but Winter added that dynamics were relative.[221] Rosen pursued the issue, seeking an admission that because Baxter had conducted the original version that his interpretation should be given greater weight.[222] The point of this colloquy did not surround the aesthetic value of Exhibit 36. Defendants contended that the B theme had been conceived as a background feature, barely noticeable in the original recording. It was marked "piano" and the score directed it to be played on the xylophone with soft hammers. Rosen asked if the B theme hadn't been played louder on Exhibit 36 for the purpose of bringing it out at trial.[223] Winter would admit only that the B theme was clearer in Exhibit 36, not louder.[224] The lack of clarity in the original version could be attributed to the recording technology available in 1953, not to any conscious intent by Baxter to subordinate that particular theme.[225]

In addition to the synthesized recording of "Joy" in its entirety, plaintiff hoped to introduce recordings of short segments and reductions of the work. All of these extracts of "Joy" that plaintiff had synthesized presented the theme in its subsequent, altered form (Example 9). None of these extracts presented the original statement in which the last note of the second measure is A.[226] Rosen suggested that the first statement of a theme was the most important and should be considered its definitive statement. Winter offered to change them all to Rosen's preferred version, intimating that he considered plaintiff's case equally strong in either case.[227] The judge ordered this improper response stricken from the record, and Winter eventually characterized the choice between the two versions as a judgment call.[228] Winter continued:

The subsequent instances, and in my judgment, my musical judgment, the more dramatic, the more important, the weightier instances have the fourth down to G as does "E.T." And I think Mr. Rosen is placing more weight on the first occurrence; I'm placing more weight on the subsequent occurrences.[229]

Winter considered neither the descending third nor fourth to be dictated by the harmony. The post-romantic harmonic vocabulary of "Joy," he said, would support either pitch.[230]

Louis Petrich, counsel for defendant Universal City Studios, also cross-examined Winter. He asked if the changed dynamics in the synthesizer version weren't in fact distortions of the original. Winter reiterated, "I don't think the crux is how loud or soft; both of these themes are very prominent in both works. For that reason, it's not central whether they're mezzo piano, piano, mezzo forte, forte or what have you."[231] Petrich also asked about the reductions entailed in making the synthesized arrangements. Winter responded that, just as piano arrangements were reductions, it would be possible to distort in the resulting reduction the importance of what exists in the original. One could make a bad reduction. But the reductions entailed in synthesizing "Joy" in Exhibit 36, Winter asserted, were well within the range of fairness.[232]

Petrich presented defendants' expert, Earl Spielman, to counter Winter's testimony. Spielman noted a difference between the copyrighted score (Exhibit 412) and plaintiff's recordings, which cast doubt on the authenticity of Exhibit 36. The last note in the trombone part at measure 73 had been "whited out." A second copy of the score showed an ink blob covering both the space and the line above it. Consequently, the ambiguity could not be eliminated by reference to any writing.[233] The plaintiff performed a G on the synthesized recording of Exhibit 36; Spielman stated that it should be A.[234]

Spielman also testified that the synthesized version of "Joy" would be distracting to the jury, who were more accustomed to hearing acoustic instruments. He drew an analogy to the effect of playing a piece on bagpipes or steam calliope. The timbre is so striking that all other considerations tend to fade.[235]

Following the experts' testimony in chambers, the judge observed that even the 1953 recording of "Joy" might be considered a distortion. The judge could hear differences of emphasis in how the two recordings of "Joy" treated the B theme.[236] But over Petrich's objection that plaintiff's were deliberately obscuring the primary melodic line in the voice in order to exaggerate the importance of a subsidiary theme,[237] the judge allowed the synthesized version into evidence.[238]

Plaintiff's counsel Blanchard wanted to play short clips of "Joy" and E.T. for the jury, and then let the defense play each entire work if they wished. The judge, however, intended to play both works all the way through before counsel gave their opening statements. Petrich wanted only the complete works played. He argued that playing clips accomplished dissection and analysis by other means.[239] Clips segmented the music and highlighted features on which the plaintiff wanted the jury to focus, and playing segments ran counter to the "total concept and feel" rationale.

Blanchard countered that playing the works only in their entirety would defeat his theory of the case. Plaintiff's claim, he argued, did not encompass both works as a whole. Either of two theories of infringement could meet the "total concept and feel" test: "fragmented literal similarity"[240] and "comprehensive nonliteral similarity."[241] Plaintiff had to choose one or the other. If he chose "comprehensive nonliteral similarity," he could show less identity between the two works but would have to demonstrate that the similarities were pervasive. However, plaintiff had chosen to base his case on "fragmented literal similarity," which requires virtual identity but of a shorter portion. Blanchard argued that, under that test, he should be able to present the relevant fragments.[242] The judge decided to allow Blanchard to present fragments to the jury after the two works were played in their entirety.

Trial: Phase One

Judge Hupp impaneled a jury and explained the limited inquiry of phase one:

The first phase of the trial is devoted to the question of whether the expression of the musical ideas contained in the works of the two composers is substantially similar to a nonexpert or what we call a lay audience. I have directed that this phase of the trial be tried and submitted to you first because it's necessary as closely as you can give us a nonexpert or lay judgment on the items which I mentioned before you hear the opinions of medical [sic] experts pro or con on the similarities and differences between the two works. . . .

To do this it must be shown that a portion of [the] "E.T." theme is virtually identical to a portion of "Joy," and that the portion of "Joy," which . . . is virtually identical‑-if you find that it is‑-is a significant part of the entire "Joy" composition. . . .

You should determine the probable response of the ordinary reasonable nonexpert or lay audience in hearing the two works as they would likely hear the works in an ordinary setting. Thus you should not try to listen to the works as a musical expert might, nor should you try to dissect or analyze the works in a way different from the average listener.

The plaintiff in this case claims that a small but qualitatively important portion of his song was taken. The test for you to apply is whether a lay or non-expert audience would hear a virtually identical and important part of "Joy" when listening to the music from "E.T."[243]

Both works were introduced by the judge, the original orchestral recording of "Joy" (Exhibit 322) and the film soundtrack of the E.T. theme (Exhibit 323).[244]

Blanchard proceeded with a short opening statement. "Joy" consists of three musical themes which Blanchard labeled A, B, and C. Only the B theme was at issue. He would play portions of E.T. to acquaint the jury with Williams's theme and then do the same with "Joy." Some portions of "Joy" would be played side by side with portions of E.T. to demonstrate the similarities.[245] He pointed out two important questions: whether the music E.T. is substantially similar to a portion of "Joy," and whether that portion of "Joy" is important to the entire work "Joy."[246]

Defense counsel Petrich's opening statement was even more cursory. In addition to the two questions mentioned by Blanchard, Petrich posed a third question as equally important. "Once you've decided those [first two] questions you have to decide whether the average listener would hear it that way."[247] Although plaintiff's counsel would highlight certain portions, it remained important to listen only as a lay person would.[248] Williams's counsel Rosen provided a short addendum to Petrich's statement. "The trick is to keep it simple."[249] If the jury would refuse to listen critically, they would hear no similarity at all. Rosen encouraged the jury not to listen as plaintiff's counsel asked them to but as they normally would listen to a work for the first time.[250]

Blanchard played the taped exhibits.[251] He presented them without commentary, because to do otherwise would amount to the impermissible testimony of counsel. Blanchard had no witness to question; plaintiff's only evidence in phase one was the music, which essentially spoke for itself. Defendants put on no additional evidence. They relied solely on the two complete tapes that the judge had played initially.[252]

After this short demonstration, the court heard closing arguments. Blanchard addressed his first question, whether the music E.T. is substantially similar to some important part of "Joy." He asked the jury just to listen to the music as music, not as a critic or an expert. Exhibits 60 and 61, which compared portions of the works side by side, show, Blanchard said, that these two pieces of music were in fact virtually identical.[253] The second question asked whether the B theme is a significant part of "Joy." Blanchard argued that the B theme's making up about one-third of the total material of "Joy" answered that question.[254] Finally, contrary to what Petrich had stated in his opening, Blanchard told the jury there could be no third question as to how an average listener would hear the music, because all on the jury were average listeners.[255]

Petrich responded in his summation that the average listener doesn't hear two measures at a time, over and over.[256] Similarities could be found between any two pieces if one looks hard enough. The acid test was really quite simple: "If someone has to tell you what the similarities are then the average person just isn't likely to hear them."[257] Rosen continued on this theme. Music is not snippets, he said; it has to go someplace. The examples played by Blanchard were short; they stopped abruptly.[258] Such snippets did not convey the emotion that the composer intended.[259] Rosen again played the two complete versions in order to leave the jury with an impression of the whole.[260]

The jury instructions repeated much of the judge's introduction to the case. He provided some guidelines as to the meaning of substantial similarity:

If you find that an average audience would hear parts of the two works as being virtually identical, those similarities are to be considered substantial only if they relate to material which constitutes a qualitatively important portion of the song "Joy." When I refer to "qualitatively important" part of "Joy," I mean to refer to a part of "Joy" that gives "Joy" its artistic value.[261]

The jury reached a decision on its special verdict quickly. The expression of the musical idea and the music from E.T. was substantially similar as defined in the instructions to a qualitatively important music expression in "Joy."[262] The defense team attributed their loss of phase one of the trial to the ability of the plaintiff to play fragments of the works for the jury. Had the jury heard only the works as a whole, the jury would have experienced more difficulty in ascertaining the precise similarities alleged. Taken as a whole, the works do not share the same "total concept and feel." However, the isolated fragments at issue bear definite similarities. Thus highlighted and without further explanation, the jury could not ignore those similarities. Petrich moved for a new trial based on prejudicial error in allowing the fragments to be played for the jury. His motion was denied,[263] but it preserved the issue for a possible appeal.

Trial: Phase Two

Phase two began the next day before the same jury. The judge explained that phase two covered the remaining issues involving liability: whether the B theme of "Joy" could be copyrighted, whether it was significant enough to enjoy protection or whether it was a mere musical idea, whether there existed striking similarity between the two works or whether there had been access, and whether examples of prior art suggested that "Joy" had not been copied.[264]

Four experts provided lengthy and sometimes confusing testimony. Winter was joined on plaintiff's side by Philip Springer.[265] Although in many ways Springer confirmed the testimony of Winter, Springer's analysis differed in various respects, and the use of two experts may have weakened plaintiff's case. The defense called Barlow as its second expert primarily to match plaintiff's effort.[266] Both parties to the litigation seemed reluctant to rely on fewer expert witnesses than the other. According to one of the lawyers involved, Barlow and Spielman had both appeared initially on the list of defense witnesses. Plaintiff then added his second expert in order to match defendants' two. Because plaintiff actually called both to testify, the defendants also called both. Although putting on two experts had probably produced more problems than benefits for plaintiff, Barlow was able to avoid contradicting Spielman. A close look at the approach that Spielman and Barlow usually take to analysis, however, suggests that they might differ on many key aspects.[267] The avoidance of contradictory testimony by the defense was aided by the fact that the judge limited the time available for examination of Barlow.[268] Consequently, Barlow's testimony was less detailed and Blanchard elected not to cross-examine him.

Access

Plaintiff sought to prove Williams's access to "Joy" through direct evidence. Williams could have heard "Joy" on the radio; a disc jockey would be called to testify that "Joy" had been played. Further, Baxter would testify that Williams had played "Joy" in concert at the Hollywood Bowl.[269] Blanchard also intended to show that the works were so strikingly similar that access could be inferred. Striking similarities would preclude the possibilities of coincidence, independent creation, and common source.

Petrich noted in his opening statement that Baxter could not say when the Hollywood Bowl performance took place; he had named dates ranging from 1955 to 1968.[270] Evidence would show, Petrich said, that Williams had never played "Joy" in concert as Baxter contended.[271]

In chambers, Blanchard and Petrich argued the legal implications of Selle v. Gibb, which established that striking similarities must preclude the possibilities of coincidence, independent creation, and common source. Petrich contended that a showing that precluded independent creation must demonstrate that the composer was not competent to write the passage. Blanchard characterized this as a misreading of the Selle opinion, which says that striking similarities must make it unlikely that "the defendant was able to compose the accused work as a matter of independent creation."[272] Blanchard did not contend that Williams was incompetent, something he might find difficult to prove. The parties would not agree to stipulate Williams's competence to compose the phrase, because Blanchard thought such a stipulation would mislead the jury. So unless Blanchard was willing to drop any attempt to show striking similarity, Petrich felt obligated to prove Williams's competence.[273]

An interpretation of Selle that equates "able" with "competent" seems to read more into the statement than the context suggests, but it would explain why the court devised a three-part test. Blanchard's interpretation makes no distinction between independent creation and coincidence, and if it were correct, the court should have devised merely a two-part test.

Plaintiff's expert Winter testified that the similarities at issue were not coincidental. It would astonish him to find that E.T. had been independently created. Qualities of the B theme in "Joy" were too specific, too explicit in combination, to be independently recreated by anyone else.[274] Plaintiff's second expert, Philip Springer, also provided the basis for a finding of striking similarity. Similarities were not the result of coincidence, independent creation, or common source.[275] Unlike Parsons in Selle, plaintiff's experts provided the necessary testimony to support a claim of striking similarity.

Despite Blanchard's desire to shore up the weakness in his direct evidence on access with a showing of strikingly similarity, Petrich intended to show that Williams's skill was such that he could have written E.T. without reference to "Joy."[276] Part of Spielman's research required him to look into other compositions of Williams. Spielman expressed his opinion that Williams had the musical skill to write E.T. on his own without reference to "Joy." Barlow shared that opinion.[277] Any similarities between the two works could easily be attributed to coincidence.[278]

Protectability

Winter testified that the theme of "Joy" was neither trite nor commonplace.[279] Defense counsel Petrich pointed out in his opening statement that the musical figure which appeared in both works and formed the basis of plaintiff's claim was so common that musicians have a word for it‑-"turn."[280] Petrich suggested that the jury would find the figure to be a musical cliché.

Edward Ruttenberg, defense counsel and partner of Petrich, began his cross-examination of Winter on the subject of turns. Wasn't that what was at issue here? Wasn't it a common Baroque device? Winter denied it. The figure at issue here could not be described as a turn, an ornamental feature. It did not appear in that context.[281] The pattern described by Ruttenberg, note above ‑ note ‑ note below ‑ note (or the notes F E D E), was only one of several variations common in the Baroque period. This figure was not the "original" variation, he said; all variations came into use simultaneously.[282] Ruttenberg asked Winter if he would consider the Harvard Dictionary of Music an authoritative source for a definition of "turn." Winter would not.[283] Ruttenberg had hoped to impeach Winter's testimony on this subject by reading a contradictory definition. The attorney, however, cannot simply read from any source of authority he chooses. The proper procedure involves asking the witness to authenticate the source first, and then to demonstrate that the witness disagrees with the source. A witness has two choices: he can admit that the source is a valid authority but that he simply disagrees, or he can deny the authority of the source, thus preventing its being read.[284] Winter won this exchange. The judge observed at side bar, however, that the jury had been unable to follow this particular line of questioning.

Defining the subject matter at issue

Plaintiff wanted to emphasize the importance of the B theme of "Joy." The B theme had to make up a significant part, a protectable part, of "Joy." Plaintiff's expert Winter provided support for plaintiff's theory. A second theme might be more memorable than the first, a point he illustrated with reference to Schubert's Symphony No. 8 ("Unfinished").[285] Further, the B theme of "Joy" was used and developed in more interesting ways than either the A or C themes.[286]

A question raised in phase one of the trial persisted: which version of the "Joy" B theme provided its definitive statement. The first appearance of the B theme at measure 18 provided fewer similarities to E.T. that its subsequent, varied return. So plaintiff preferred to consider the varied statement to be the definitive one. The difference lay in the last note of the second measure of the B theme, the descent from C to A in the first statement versus the descent from C to G when the theme returns. (See Examples 10 and 11, above.)

Winter testified that the return of the "Joy" B theme in measure 72 denoted the climax of the work. Baxter obviously intended to bring out the B theme at this point.[287] The B theme, sounding in the trombones, dominated the C theme in the voice; Winter pointed out that normally the second trombone would be silent, but here the score was marked "a 2." The soprano singing the C theme could not hope to compete for prominence with two trombones playing forte. The passage was also marked maestoso,[288] and the simultaneous return of the B and C themes highlighted the importance of this passage.[289] Therefore, Winter found it entirely proper to use the statement beginning in measure 72, rather than the opening statement that is less similar to E.T., in all his comparisons at trial.

Expert witness Springer supported plaintiff's theory that the definitive statement of the "Joy" B theme should be its return in measures 57 through 59 and 72 to 73. A theme's first statement is not always the definitive one. Composers, he said, never state a theme and then leave it; they repeat the theme, and one must look to its repetitions before deciding which version should be considered definitive.[290] Besides, Springer observed, the last interval from C to A and its subsequent version, C to G, were nearly identical. To make his point, Springer arranged for the jury to watch his fingers as he played.

First, the way I think it is nearly identical and then I'll show you why (witness playing) [897]. Now, the last two notes that I played were these two notes (demonstrating). Could you all see my fingers? Can you see the distance between my two fingers (demonstrating)?[291]

Thus, Springer demonstrated the "close relation" between the notes A and G.[292]

On cross-examination, Springer was asked to listen to the original orchestral version of "Joy" and to identify the last note in the trombone part at measure 73. Springer heard G clearly.[293] He could not tell from the written score which note had been intended.[294] Ruttenberg asked, "If you were incorrect, if it were A, would that change your opinion as to what Baxter viewed to be the main statement of the B theme?" Springer answered yes.[295] Although defense experts would dispute Springer's testimony that the note played was G, this is not the type of question that the jury could answer independently. Like many technical issues put before laymen at trial, the critical question is who the jury believes.

Spielman testified that the last note of measure 73 in the trombone part was played as A on the 1953 orchestral recording.[296] Although Springer had said that an A would clash with the G played simultaneously in the timpani, bass, cello, and piano,[297] Spielman discounted those concerns. Those instruments played the G an octave below the notes in question; it was simply the common device of pedal point.[298] The voice sings A in the same register as the trombones at measure 73, and the trombone on G would be more likely to clash with the voice.[299]

The initial notes of the "Joy" B theme also presented dissimilarities that plaintiff had to overcome. Plaintiff was comparing the three quarter-note Ds of "Joy" to the half-note C of E.T. Defense counsel Ruttenberg asked if notes 2 through 7 of each theme weren't different. He wanted Winter to account for the three repeated notes at the beginning of the "Joy" theme. Winter considered the repeated Ds at the beginning of "Joy" to be one note for purposes of comparison‑-the second pitch (albeit the fourth impact of "Joy") was the common G. Ruttenberg had Winter play "Joy" first as written and then with a dotted half note D replacing the three repeated quarter notes to demonstrate the difference. Winter disagreed with Ruttenberg on the importance of the repeated notes, saying he found that feature insignificant. Ruttenberg then asked Winter to play the opening motive of Beethoven's Symphony No. 5.[300] The effect of repeated notes, Winter said, depends on context.[301]

In addition to defining the content of the first two measures of "Joy," there were contradictions concerning how much of the theme was at issue. Literal similarities could be found only in the first two measures of the themes, and plaintiff had elected to prove fragmented literal similarity. However, plaintiff wanted to add proof that the entire eight-measure phrase had a similar structure and developed in the same way.

Ruttenberg gained a few simple admissions from Winter on cross-examination. The B theme of "Joy" is a single, eight-measure phrase‑-very common in popular music.[302] "Joy" states these eight measures in their entirety only twice, at measures 18 to 25 and 25 to 32.[303] The eight-measure phrase can actually be divided in half, a four-measure statement with one literal repetition. The four-measure statement can also be divided in half, a two-measure figure repeated with slight variation. Winter added the proviso that those slight variations were musically significant.[304] Thus divided, the similarities in development described by Winter on direct largely disappeared.

The witnesses' approach to analysis

In his opening statement, Blanchard characterized the difference between plaintiff's and defendants' approach to analysis. Two experts would be called by each side. Plaintiff's music experts would analyze the two works, concentrating on the sound of the music. They would also show that the melodic contour of the works had particular importance.[305] Defendant's experts could be counted on to introduce charts and graphs that had nothing to do with how the music sounds.[306] Indeed, neither of plaintiff's experts introduced any exhibits during the testimony. Springer apparently made extensive use of the blackboard, but none of what he wrote could be captured in the written record. Winter, on the other hand, stated that all visual representations were distortions.[307] Consequently, he provided none. He said he preferred merely to talk and demonstrate at the piano.[308] The witness's preferences do not always control, however. Both Winter and Springer were presented with charts made by defense experts and asked to comment on various features that the charts illustrated.

In order to illustrate the similarities and dissimilarities in a more visual format, defense expert Spielman produced a chart (Exhibit 486) referred to as a "chromatranscription." The chart represented pitches by color and presented them on a time line. The juxtaposition of "Joy" and E.T. allowed the jury to see at every point whether the colors matched.[309] Spielman provided the jury with individual copies of the common notation and placed the chromatranscription on an overhead projector.[310] The charts pointed out all "simultaneous pitch occurrences," a term used regularly by Spielman. Certain coincidental similarities could be found outside the six consecutive similar pitches (G F D E D C), but such transitory occurrences could not be afforded significance.[311] Barlow also provided charts to illustrate his points.

Spielman introduced a synthesized tape recording in which "Joy" was played using a harpsichord sound and E.T. was played using a trumpet sound (Exhibit 471). After each was played separately, the two were superimposed‑-played simultaneously, each through one of the stereo speakers. Spielman played the tape, demonstrating that the melodies did not correspond.[312]

The testimony

Winter iterated six "criteria" that determined the similarities of the two works.[313] The first was the similar placement of the "hook" in the first two measures of the phrase of both themes.[314] Winter testified that the phrase structure of the works matches precisely, his second criteria.[315] The themes are divided into four symmetrical two-measure units. Winter played the themes two measures at a time, demonstrating how they followed the pattern "two plus two plus two plus two."[316] Winter found this symmetry "striking" in a musical sense‑-something the listener would perceive at least subconsciously. The average listener would hear the difference between even divisions and odd divisions of the phrase.[317]

Winter stated that the last six measures of both works were developed out of the material stated in the first two measures, a factor he found striking. This manner of development was the third characteristic he found similar.[318]

E.T. duplicates the pitch sequence of the "Joy" B theme five times‑-"not once, twice, but five times." Similarity of pitch met Winter's fourth criteria. The taking appears in the first two measures, the hook, in which pitches 2 through 8 are absolutely identical.[319]

In order to compare rhythm, the fifth criteria, Winter segmented the themes into three parts. Longer note values appear in the first part, followed by a rapid figure in the middle and long notes at the end.[320] Long notes at the beginning and end of the themes framed and highlighted the short notes in the middle where the similarities of pitch were most pronounced.[321]

Winter's final criteria concerned tempo. The two works were performed "within one click" of the metronome.[322]

On cross-examination by Ruttenberg, Winter compared "notes" with the agreed definition that a note designates both pitch and duration. Within the two themes, only the four notes F E D E (the "turn") matched under this approach.[323] Within the first measure alone, none of the notes were the same. Winter made some good points. The pitch G, common to the first measure of both themes, was an important similarity because of its function in the melodies. The fact that a D could be found in the first measure of both themes was not significant; that would be merely "scavenging" for pitches.[324]

Winter seemed less convincing when he analyzed similarity of pitch in terms of intervals. The first note of the "Joy" B theme is D, whereas that of E.T. is C. Both rise to G, outlining the interval of a perfect fourth in "Joy" and a perfect fifth in E.T. This difference is not readily heard, Winter claimed, because a perfect fourth and perfect fifth can sound very similar in certain contexts. Winter gave a short lecture on the overtone series to demonstrate how closely related those intervals are. He played "Twinkle, Twinkle, Little Star," varying the opening interval.[325] He stated that the reason the opening D of "Joy" is not heard as substantially different from E.T. is because lay persons (and even his students) have difficulty distinguishing those intervals.[326]

Ruttenberg took Winter through a note-by-note description of the relevant themes of the two works. Ruttenberg spent time on the detail trying to nail down Winter's six criteria. He asked if those criteria would require the F E D E figure to be duplicated literally while allowing other notes not to be exactly the same. Developing his point, Ruttenberg called Winter's attention to measure 17 of E.T., in which the theme returns with the F E D E figure altered to D F E D.[327] Had Williams written that statement first, would it meet your six criteria? It would not, Winter admitted,[328] but the second statement of the E.T. theme was not the one that the listener would hear and remember. The transcript suggests no hesitation by Winter to argue that the first statement of E.T. must be the definitive one, but that subsequent statements of the "Joy" B theme could outweigh the first. Finally, Ruttenberg elicited Winter's reluctant admission ("if you're simply going to count notes") that the two statements of the E.T. theme, the original statement and its alteration at measure 17 which did not meet his six criteria, were more similar to each other than to the B theme of "Joy."[329]

Winter defended the six criteria themselves. They had not been derived from any particular text but from his twenty years of experience as a professional musician.[330] The six criteria did not relate to general propositions of music, but to the specific similarities of the two works. "I have illuminated, I hope, the criteria that are in common between these two themes."[331] Winter had a peculiar definition of "criteria," because they were not analytical criteria at all. They might in fact be useless to any analysis that did not compare these two particular works.

Ruttenberg asked if these six criteria simply "jumped out" at Winter when he heard the two works at issue. Winter said no. His deposition was then read for impeachment purposes:

Q Were these six criteria designed by you to show similarities between "Joy" and "E.T."?

A They weren't designed by me to illustrate anything. They were the criteria that jumped out to me as a professional musician as I listened to them which they shared.[332]

Ruttenberg continued his questioning of Winter at trial:

Q After you listened to "Joy" did these six criteria jump out to you as the most musically important in "Joy"?

A Since I was making the comparative studies and not trying to make an isolated study I wasn't working on anything jumping out, since I was asked to undertake a comparison, which I did.

Q So these six criteria that you find in "Joy" weren't six criteria that jumped out at you just from "Joy" itself before you heard "E.T.," were they?

A That's correct. I was looking for a comparison, not for an essential characterization.

Was it true that Winter designed these six criteria specifically for the purpose of making the two works seem similar?

A A comparison is different from a contrast. If my instructions had been would you find as many contrasting elements between these themes, which is what you are about, I would have done that. I was asked to compare them, and then where the evaluation came in was to say okay, here are the things, here are the six elements, that are in common.[333]

Ruttenberg continued to emphasize that the six criteria related only to the two pieces that Winter had been hired to compare. Winter's deposition testimony was read again. Winter said he had never used the same six criteria to analyze any other piece. He wouldn't know if he had used as many as three of the same criteria, but he would be very surprised if he had.[334] "They are the criteria I have come up with as being shared closely in common with these two themes."[335]

Defense expert Spielman testified that he had never used Winter's six criteria as a group in any analysis.[336] He had never heard of that group of criteria used to evaluate substantial similarity.[337] Defense expert Barlow confirmed that in his thirty years of testifying on infringement he had never seen Winter's six criteria employed.[338]

Springer, testifying for plaintiff, emphasized melodic shape‑-what he called the gestalt structure of these two pieces of music‑-"its architecture, its lying through space and time as we trace it."[339] He characterized the first four notes of "Joy" (including the repeated Ds) as the "announcement." It "perks your ears up" and tells you something important is about to happen. The second part of the theme is what defendants had called the "turn." Then comes the "final fanfare"; these last two notes are important. The three parts of the "Joy" theme taken together indicated a compositional creation of significance.[340] The first two notes of E.T. corresponded to the announcement in "Joy." The "turn" was exactly the same. The fanfare conclusion was the same as "Joy," only in longer note values. The pitches and expressive qualities of the notes were virtually identical.[341]

If the "Joy" theme went to A in the second measure instead of G it would not be a fanfare-type conclusion, Springer observed; it would not have the conclusive sound that the G does.[342] Where the figure proceeds to G, "we have a concluded piece of music, a concluded musical as it were inspiration, in itself. It comes to a rest."[343] But where the figure goes to A, it seems to want to go somewhere else; so it cannot be evaluated as an entity in itself.[344] Springer seems to refer to open and closed phrase endings, but he failed to use that terminology.[345] However, an analysis along those lines would consider both the G and the A to be open or incomplete cadences‑-in other words, cadences that suggest a continuation, that do not convey a complete sense of rest. In either instance, following Springer's theory, one would have to analyze four-measure segments of "Joy" in order to find a completion of the idea. Springer's distinction did not justify his two-measure segmentation of "Joy." Further, this type of analysis would reveal that E.T. assiduously avoids this sense of completion throughout the work. More dissimilarities than similarities appear through this line of analysis, and Springer seems to use it incorrectly in an attempt to justify two key aspects of his analysis and plaintiff's case: (1) the importance of the two-measure segment where all similarities occur, and (2) plaintiff's choice of the subsequent statement (descending to G) as the definitive one.

Ruttenberg had anticipated Springer's testimony while cross-examining Winter. Winter testified on cross-examination that "announcement" is not a musical term. Winter also stated that a "fanfare" is a short call designed to announce something. If it included only two notes, Winter probably would not call it a fanfare, but that would depend on the context.[346] Ruttenberg read Winter's deposition testimony in which Winter said did he not think a fanfare could consist of only two notes.[347]

Defense expert Spielman had determined the claim to revolve around six notes (G F E D E C)‑-in his opinion, an insufficient amount to determine copying or that the works were substantially similar.[348] Spielman explained that he analyzed certain aspects of the music in all cases. Some aspects could be considered irrelevant when looking for substantial similarities: key, tempo, and meter, for example.[349] Nevertheless, Spielman thought it necessary to take those irrelevant factors into account in any analysis; he considered it his first task to discover every way in which the works are similar and then to determine whether those similarities were relevant.[350]

Five factors comprised the standard for all of Spielman's analyses: pitch, rhythm, melody (the combination of pitch and rhythm), harmony, and formal structure.[351] In this case, only the six pitches mentioned above were the same.[352] Rhythm was defined as the duration of the pitches, how long they were held relative to one another. Some similarities of rhythm appeared along with some very important differences.[353] Spielman demonstrated the difference between duple and triple meter‑-its different feel and different organization. The distinction was an important aspect of the music, one that could be heard and felt.[354]

Spielman found no significant similarity with regard to harmony.[355] Indeed, significant differences can be found in the overall progression of the eight-measure phrases, but the two measures at issue offer little harmonic material of any distinction.

Finally, Spielman compared the formal structure, not only the entire work but also in terms of its internal sections.[356] The only similarity was that the initial statement is two measures in length and develops over eight measures.[357] "Joy" develops through a process of literal repetitions. E.T., however, states its idea at measure 9 and the idea never repeats within the eight-measure phrase. The theme expands using the same rhythmic idea and motivic structure through sequence. Spielman could find no significant similarity of structure to "Joy" within the phrase.

Winter had conceded on cross-examination that E.T. develops through harmonic sequencing, each two-measure figure appearing at a different pitch level. Within the eight-measure theme, E.T. never repeats the first two measures. "Joy" remains static, repeating its material more or less literally.[358] But harmonic sequencing, Winter added, is a very common musical device.[359] Further, although the harmony of the themes differ, harmony was not central to the issues presented by these two works. Both works, Winter claimed, employ very common harmonies.[360] Ruttenberg asked Winter to demonstrate at the piano, to play each work in two-measure segments with five-second pauses between each segment. In this way, Ruttenberg hoped to make apparent to the jury that the two themes did not develop along similar lines. The overall pattern of eight measures divided symmetrically into two-measure units, which had been described by Winter on direct examination, remained as he played each two-measure unit. But it became apparent that the development entailed more than these symmetrical divisions. Winter admitted some differences.[361]

Blanchard cross-examined Spielman. Spielman had not used the subsequent appearance of the "Joy" theme where it descends to G in Joy in any of his exhibits.[362] Spielman conceded that the theme descends to G several times but not in the first full thematic statement. It goes to G in the second statement at measures 25 through 32 and in transitional fragments at measures 57 through 59.[363] It goes to G four times not counting the disputed statement in measure 73.[364] But Spielman used rhythmic dissimilarity to support his contention that, even where "Joy" goes to G, it did not correspond to the G of E.T. Rhythmic juxtaposition would relate "Joy's" G to the last half of E.T.'s C (the seventh note), so that similarity would not be relevant.[365]

Defense expert Barlow produced a chart which compared the B theme of "Joy" as it appeared in measures 18 through 25 to E.T. measures 9 through 16.[366] Barlow had drawn connecting red lines between similar pitches. Forty-two notes appear in the "Joy" theme, thirty-three in E.T. The total number of notes in common between the two, according to Barlow, was seven. Of those seven only five were consecutive. Two were isolated, the result of pure chance.[367] The difference between Spielman's six consecutive notes and Barlow's five can be attributed to the grace note figure in "Joy," which Barlow assigned greater significance. Grace notes, he testified, are not inconsequential; the composer wanted them there. When he didn't want them, as in "Joy's" subsequent statements, he took them out.[368]

Barlow examined the works for similarities both as to quantity and quality. The quantity he found small, perhaps negligible.[369] Quality, he explained, was like having two lines of literature: "once upon a time there was a princess," and "once upon a time there were three bears." "Once upon a time" is a quantity of words. But the significance is small because they exist in hundreds of fairy tales. In other words, they may have quantity but no quality. That is what Barlow thought the five notes in question had.[370]

Barlow insisted that the repeated Ds at the beginning of "Joy" be taken into account. "You can't confuse knocking on a door four times with knocking twice."[371] Thus Barlow explained the element of rhythm. Nor could one confuse a perfect fourth with a perfect fifth. He had never heard a child make that error in singing "Twinkle, Twinkle, Little Star."[372]

Independent creation

John Williams qualified as an expert, but his task was not so much to analyze the music as to give an account of how he wrote it‑-to demonstrate independent creation. Lawyers often characterize this aspect of any defendant's case as one of the most critical. Music analysis aside, the defendant needs to explain to the jury how the creative process works. If the defendant's account is credible, the testimony will offset some of the perceived similarities. If he cannot explain his thought processes, admittedly a very difficult thing to do, his account may be suspect.

Williams, who was called as the first defense witness, apparently performed his task well. He played the piano continuously, to a greater extent that any of the hired experts. More importantly, he was able to describe the composition in extra-musical terms. After Williams's initial testimony had authenticated the film, Williams showed E.T. in its entirety. Williams was able to relate all of his compositional decisions to the images on the screen. Concrete visual images provided him with a tool most defendants lack.

A brief summary of his account here will have to suffice.[373] The similarities, Williams testified, had been taken out of context. The "Joy" theme's definitive statement sounded the notes D D D G F E D E C A. Only the six notes beginning on G were the same in E.T. Williams played only the common notes; he showed that they were insufficient to identify either work. He provided a literary analogy:

She writes a book, page 230, the first paragraph she writes: I took the bus to Bakersfield.

He writes a book, page 390, and he starts his paragraph: We took the bus to Toronto.

What's common is "took the bus." That's what you heard as being similar. . . . It's not "Joy," it's not "E.T.," it's been‑-it's been microscoped out of context.[374]

The opening notes sounded the clarion call, the battle horn. Played in short repeated notes, as in the opening notes of the "Joy" theme, that image is not conveyed. The clarion call comes from antiquity; it had to be the interval of a perfect fifth, not a perfect fourth as in "Joy."

The psychology of the fifth belongs to our collective human memory. It's a call to battle. . . .

I first wrote the clarion call, which is the first thing you are going to hear in the film and the last thing. The clarion call had to do with a little boy looking out of a window and he's trying to make contact with a creature from the other world. I want to compose first a signal, not a melody. The clarion call is the basis for the "E.T." theme.[375]

Baxter's work portrays the emotion of joy. Its mood is happy and light. E.T., however, needed to portray exaltation. The theme appears when the character E.T. and his friend were racing bicycles, straining with great effort to break gravity and fly.[376] Expression of exaltation requires some weight and tension. Opening long notes, not present in "Joy," provided that weight. The meter of E.T. (3/2) was also essential to this purpose; it lacks the bounce of the 4/4 meter of "Joy."[377]

Winter, called later in rebuttal, questioned the significance Williams assigned to meter. Williams's characterization of the 3/2 meter of E.T. may have been accurate‑-"it floats."[378] However, for a different reason, "Joy" has that same quality in 4/4 meter. The eighth-note figure (the turn) comes once at the beginning of the bar (second measure of the theme) and then at the end of the succeeding measure. The "Joy" theme's asymmetrical rhythm thus tends to obscure the regular pulse of the underlying 4/4 meter.[379]

Williams also described how the E.T. theme starts on the tonic harmony, whereas the "Joy" B theme begins on dominant and progresses to tonic. The two composers took opposite approaches. Baxter resolved to tonic in the middle of the second measure; Williams began on tonic and propelled his character through the tension into the exaltation of flying. The harmonic psychology of "Joy" would not work in E.T.[380]

Prior art

Winter searched prior art using his six criteria and found no musical examples that sounded as similar as "Joy" and E.T.[381] He used thematic dictionaries, which index tunes according to their incipit.[382] He considered this method of searching proper, because the most important part of a theme is its beginning.[383] Winter contradicts his own testimony at this point. Because the first note of E.T. and "Joy" are different, he would have to conduct two searches of incipits; and if the beginning of the theme is most important, it would suggest that E.T. and "Joy" are less similar than he claimed.

Spielman conducted an exhaustive search for examples of prior art in all genres.[384] Because the defendant has the burden of proving copying from a common source,[385] it stands to reason that defense experts would put more effort into this search. Spielman suggested similarities with the theme to the TV show Dallas.[386] Blanchard objected that Dallas had been composed after 1953 when "Joy" was written. The objection was overruled; similarities between E.T. and Dallas remained relevant. Other examples introduced include Neil Diamond's "Song Sung Blue,"[387] Schubert's Rondo in A for Violin and Orchestra,[388] J. S. Bach's Partita No. 3,[389] Bizet "Farandole" from L'Arlésienne Suite No. 2,[390] and the film score to Young Mr. Lincoln by Alfred Newman.[391] Exhibit 492 demonstrated how "Joy," E.T., and these examples of prior art were all melodies built around a C-major chord.[392] The figure at issue was so common, Spielman noted, that he had even found it in a teaching exercise.[393]

Barlow also went through the similarities between E.T. and prior art examples mentioned above. He was careful not to draw connecting lines showing similarities where the accents did not correspond. "It's like the rapist and therapist; same letters, different word because the accents aren't the same."[394] Whereas "Joy" and E.T. had five notes in common, Bizet's "Farandole" had all eight in common with "Joy,"[395] Schubert's Rondo had seven,[396] Mozart's Serenade had six,[397] Young Mr. Lincoln had five,[398] and Milhaud's L'Homme et son désire had seven.[399] Barlow concluded that these similarities showed only that the structure exists in music literature, not that they result from copying.[400] The defendant rested.

Winter was recalled by the plaintiff in part to rebut the testimony on prior art. Most of his comments related to the tempo and feel of the works; they were not as similar in that regard as "Joy" and E.T.[401] Dallas employed a disco beat.[402] "Song Sung Blue" had a swing feel.[403] Milhaud had been attempting to depict the primeval jungles of Brazil; he used exotic, ethnic idioms.[404] The bouncy and jaunty flavor of Schubert's Rondo could not be compared to either "Joy" or E.T.[405] The Bach example was not really a theme at all but just a transitory statement.[406] Bizet's "Farandole" appeared consistently in the minor mode except for the brief statement taken by the defense for their example. Further, the similarities appeared only when "Joy" was compared to the last two measures of Bizet's theme followed by the first two notes of its restatement.[407] Young Mr. Lincoln was the wrong tempo, almost like a chorale.[408] None of these similarities, Winter testified, came close to matching those between "Joy" and E.T.[409]

Springer's testimony concerning Williams's previous compositions damaged plaintiff's case. Petrich had explained in his opening statement that Williams, who had composed numerous highly successful film scores, repeated in E.T. some of the compositional devices found in Star Wars and Superman.[410] Springer admitted that E.T. sounded sort of like Williams's score to Star Wars, but only if one were to eliminate the upbeats and "play little tricks with it."[411] Ruttenberg read his deposition testimony:

Q Do you think that the "Star Wars" theme and the "E.T." theme sound different from one another?

A I think that to a layman they would. Yes. To me they sound like each other.[412]

Example 13: Star Wars Theme.

The Star Wars theme (Example 13) was compared to E.T. In fact, the similarities are remarkable, though probably not substantial. Ignoring rhythmic details, the same prominent ascending perfect fifth is followed by a rapid figure of F E D (compared to F E D E in E.T.). Springer's "fanfare" ending is the same but appears an octave higher in Star Wars. Nevertheless, Springer believed that Williams had not been influenced by Star Wars when writing E.T. and that it was more likely for him to have been influenced subconsciously by his memory of "Joy."[413]

Defense counsel Rosen took up this subject with Springer where Ruttenberg left off. Rosen read from Springer's deposition:

Q You think Mr. Williams was kind of repeating himself when he wrote the "E.T." theme after the "Star Wars" theme?

A I never tell a lie, Mr. Rosen. Yes, I do.

Q Many composers do repeat themselves, don't they?

A Yes, but not with quite such banal material.

Q You consider Mr. Williams's material to be banal?

A I consider the "Star Wars" theme to be banal, yes. . . .

[But] I think corny is a better word.[414]

Rosen continued, "`Banal' means commonplace, right? Springer said yes.[415] (Themes that are commonplace can be denied legal protection.) Later, Springer said the following:

A In my class at that time, I was showing how "Star Wars," "Born Free" and the thing that Vanjelis wrote, "Chariots of Fire" were basically all‑-basically all the same musical sort of an idea. And, in fact, what I did with the class was I played them all simultaneously, all three of these themes, and they went together. . . .[416]

They grow out of two ideas. One is the idea of the open fifth, or the perfect fourth, intervals which sound very much like each other.

They start with an open fifth, then subdominant chord with conjunct motion beginning with either the 6th or 7th scale degree, then back to the fifth. It's a sure-fire formula.[417]

E.T., Springer maintained, did not follow this formula. The harmonic progression did not go to subdominant after the opening fifth. Although the written music indicates an F chord (subdominant) over a G in the bass at measure 9, that, he said, was a mistake in transcription.[418] The harmony in measure 9 of E.T. is, in fact, rather ambiguous. Printed piano scores with chord symbols often prove somewhat unreliable, especially to an analyst who would rather reach his own conclusions. The score submitted into evidence further contributes to the ambiguity by labeling chords inconsistently at what are virtually literal repeats of the same material.[419]

Jury verdict

Less confusion arose in the jury instruction conference of this case than in Selle v. Gibb. Petrich didn't like the wording of question one of the special verdict. It asked the jury to decide if "Joy" contains original and protectable material. The claim involved only a portion of the work, so Petrich believed the question should be limited appropriately.[420] Phrasing the question as "does the `B theme' contain original and protectable material" was still too broad for Petrich.[421] He wanted to keep the focus on the two-measure fragment from the start.[422] Blanchard thought focus on the B theme was proper.[423] They arrived at a compromise and proceeded to closing arguments.[424]

The judge read the questions to the jury:

The form of special verdict reads as follows. In fact it's quite similar to the form you had in the first phase of the case, except this has three questions.

Question No. 1: "Does the portion of the `B' theme of the song `Joy' claimed to have been copied contain original material which may be protected by copyright." "Yes" or "No." Then there's a set of instructions. If you answer yes you go on to Question 2; if you answer no, you sign and return the verdict.

Question 2: "Does the musical score from `E.T.' contain music which is substantially similar to the original protectible material from `Joy'?" Again, answer yes or no. If you answer, and there's a space to write in that answer, if you answer it yes you go on to Question No. 3; if you answer it no, you sign and return the verdict.

Question 3: "Was a significant amount of substantially similar original protected expression from `Joy' copied by the composer into the musical score of `E.T.'?" Answer yes or no, and a place for the jury to write its answer. A place for the verdict to be dated and signed by your foreperson.

You will note that Questions 1 through 3 call for yes or no answers. The answer to each question must be the unanimous answer of the jury. Unless all of you agree, you may not return an answer to any question.[425]

Approximately twenty-four hours later the jury answered the first question no. The jury never reached the remaining two questions. Judgment was entered on 20 September 1988 for defendants.

Appeal

On 30 September 1988, plaintiff filed a motion for a new trial, which the court denied. Plaintiff then filed an appeal, contending that the jury verdict was not supported by the evidence. The defendants' appellate brief asserted in part:

This case exemplified the duplicitous strategy often used by a copyright plaintiff. He willingly narrows his claim to the point of unprotectibility in order to get a favorable jury verdict on the issue of substantial similarity, but, after receiving such favorable verdict, he then argues a far broader claim on the issue of protectibility, hoping that no one will notice the inconsistency of his strategy. In this case, the jury noticed, and returned its verdict in the Second Phase of trial that the only substantial similarity between Joy and E.T. that it found in the First Phase of trial was not protectible by copyright. Because lengthy expert testimony during the Second Phase (consuming four of its five trial days) presented evidence (in fact, substantial and overwhelming evidence) in support of this finding, the jury's verdict must be affirmed.[426]

Because plaintiff did not move for a directed verdict at the close of evidence, the Ninth Circuit would examine the record only to determine "whether there is such plain error on the face of the record that failure to review would result in a manifest miscarriage of justice."[427] The question was whether there was any evidence to support the jury verdict, not whether the evidence was sufficient.[428] The Ninth Circuit ruled:

The jury could have found the B theme fragment was not protectible if it found any of the following: (1) the fragment is not original; (2) it is a mere idea; (3) it is trite or commonplace; or (4) it is too short. Experts for defendants testified that the fragments of Joy's B theme played to the jury by Baxter were commonplace. Viewing the evidence in the light most favorable to appellees, we must assume the jury found the Theme from E.T. substantially similar to one of these fragments of the B theme. This evidence supports a finding Baxter's work was not protectible by copyright.[429]

Baxter lost his appeal and had court costs assessed against him.

---

[1] (Garden City, N.Y.: Doubleday, 1961).

2 78 F. Supp. 686 (S.D.N.Y. 1948), aff'd, 173 F.2d 288 (2d Cir. 1949).

[3] Calypso Songs of the West Indies by Massie Patterson and Lionel Belasco, free transcription by Maurice Baron, English version by Olga Paul. See Nizer, supra note 1, at 236.

[4] Excerpts of Feist's published music to "Rum and Coca-Cola" appear in Appendix H, at 1278.

[5] Leo Feist had already been found liable for infringement of the lyrics. Kahn v. Leo Feist, Inc., 70 F. Supp. 450 (S.D.N.Y.), aff’d, 165 F.2d 188 (2d Cir. 1947). The case established that Rupert Westmore Grant wrote the words in 1943 to the Belasco tune taught to him by one Monrose (now deceased), who was one of Belasco's protégés in 1906. Grant changed a few notes to accommodate the new lyrics. See SeeBaron, 78 F. Supp. at 687.

6 Nizer, supra note 1, at 260.

[7] 78 F. Supp. at 689.

8 Nizer, supra note 1, at 264.

9 Id. at 235.

[10] Solmization or solfege refers to the assignment of vocal syllables to degrees of the scale: do, re, mi, etc.

11 Nizer, supra note 1, at 248.

12 Id. at 248-49.

13 Id. at 248.

14 Id. at 259. Because more than one version of "Rum and Coca-Cola" was involved, the number of identical notes varied.

[15] Id. at 249.

16 Id. at 258.

17 Id. at 259.

[18] See "Rum and Coca-Cola" (Appendix H, at 1278), measure 9. The chord was described as a 13th, a diminished seventh on C# against A in the melody. Nizer, suprasupra note 1, at 252. It emphasizes the tragic "fille" (streetwalker). But in "Rum and Coca-Cola," it accompanies "they make you feel so very glad." Nizer, supra note 1, at 253.

19 Nizer, supra note 1, at 253.

20 Id. at 249.

21 Id. at 277-80.

[22] See, e.g., Darrell v. Joe Morris Music Co., 113 F.2d 80 (2d Cir. 1940); Wilkie v. Santly Bros., Inc., 91 F.2d 978 (2d Cir.), cert. denied, 302 U.S. 735 (1937); Allen v. Walt Disney Prod., Ltd.Allen v. Walt Disney Prod., Ltd., 41 F. Supp. 134 (S.D.N.Y. 1941); Foreign & Domestic Music Corp. v. Twentieth Century-Fox Film Corp. 19 F. Supp. 769 (S.D.N.Y. 1937).

[23] See, e.g. "Abeles-Fox Main Guns on Copyright Infringement," Billboard, 27 May 1957, 207; "New Infringement Victory for Abeles," Billboard, 29 June 1959, 3.

[24] See Anheuser-Busch, Inc. v. Elsmere Music, Inc., 633 F. Supp. 487 (S.D.N.Y. 1986), Record at 629.

[25] SeeSee 250 East Music Corp. v. D'Arcy MacManus Masius, Inc., [settled] (S.D.N.Y.), Deposition of Spielman taken on behalf of plaintiffs October 9, 1985, at 10-45, for a discussion of these issues.

26 Nicholas Cook, A Guide to Music Analysis (New York: Braziller, 1987), 227.

27 Anheuser-Busch, 633 F. Supp. 487 (S.D.N.Y. 1986), Record at 690 (Spielman testifying).

28 The New Grove Dictionary of Music and Musicians (1980), s.v. "Analysis," by Ian Bent, 354.

[29] See, e.g., Nom Music, Inc. v. Kaslin, 227 F. Supp. 922, 927 (S.D.N.Y. 1964) (transposition of one composition to the key of another for comparison purposes has been recognized as a valid technique), aff'd, 343 F.2d 198 (2d Cir. 1965); Dorchester Music Corp. v. National Broadcasting Co., 171 F. Supp. 580, 586 (S.D. Cal. 1959). See alsoSee also, Alfred M. Shafter, Musical Copyright, 2d ed. (Chicago: Callaghan, 1939), 203.

[30] See, e.g. See, e.g., Anheuser-Busch, Deposition of Spielman, at 171ff. The witness was asked to double note values in one work in order to compare it with another. He resisted, but justified his position only by arguing tempo variances.

[31] Case No. 80 Civ. 7346 (KTD) (S.D.N.Y. 1980).

[32] The entire exchange is reproduced at Appendix F, at 1240.

[33] A lead sheet generally provides the barest necessities, one line of musical notation‑-the melody. Chord names alone indicate the harmony. The performer is expected to improvise everything else.

34 For an example, see Appendix C, at 898.

[35] A sliding movement between pitches. See e.g., Gaste v. KaisermanGaste v. Kaiserman, Case No. 86 Civ. 5671 (S.D.N.Y. 1986), aff'd, 863 F.2d 1061 (2d Cir. 1988), Exhibit 93. Appendix B, at .

[36] See, e.g., Gaste, Exhibit 97 (Appendix B, at ). See alsoSee also, contour graph of David Greitzer's pre-trial report in Ensign Music. Appendix F, at 1221.

[37] See GasteGaste, Exhibit 96. Because contour here appears along with common notation, it poses little likelihood of misleading the layman and may be helpful. Without easy reference to the common notation, however, the contour itself would be largely meaningless. See Fed R. Evid. 403.

[38] See Schultz v. HolmesSchultz v. Holmes, 264 F.2d 942 (9th Cir. 1959). Transcripts revealed that the witnesses frequently supplemented their answers by vocal or instrumental demonstrations. Witnesses sang, hummed, and danced to demonstrate the rhythm or syncopation of the music then under discussion. Witnesses also played piano and violin. The appellate court had no way to determine just what parts of the music were being referred to during the testimony. The court concluded, "[E]ven if the written testimony and exhibits raised a question in our minds as to the correctness of the findings, we would not feel free to overturn findings based only on that evidence without the accompanying demonstrations." Id. at 944-45. Generally, an appellate court reviews only conclusions of law by a trial court, not findings of fact.

[39] Selle v. Gibb, 567 F. Supp. 1173 (N.D. Ill. 1983), aff'd, 741 F.2d 896 (7th Cir. 1984).

40 The three were Van McCoy Music, Mother Bertha Music, and Music, Music, Music. Selle, Record at 36.

[41] Questions posed to potential jurors on voir dire (jury selection) revealed the jury's level of musical knowledge and the factors counsel considered relevant. Of the twenty-five prospective jurors, none had composed music but two had relatives who did, four had played instruments as children, three read music "a little," about 90% claimed to have attended a concert, 85% listened to popular radio stations frequently, eight purchased recordings frequently, one read entertainment magazines, eight had belonged to a choir or other music group, seven had heard "How Deep Is Your Love," one thought she had heard "Let It End," and none had been to a Bee Gees concert. One voiced concern that she could not distinguish one rock song from another. An undetermined number had attended a classical music concert, one of whom owned a few classical recordings.

None of these potential jurors were challenged or disqualified, and the six jurors were selected by lot. They included a retired schoolteacher, a retired distribution manager, a travel agent, and a department supervisor for the Catholic Church Extension Society. Only the schoolteacher had a college degree, and none claimed to read music. The distribution manager had played clarinet in the high school band. See See Selle, voir dire transcript.

[42] Record at 6-7.

[43] Record at 8.

[44] Record at 9.

[45] Record at 11.

[46] Record at 10.

[47] Record at 12.

[48] Record at 21-36.

49 Similarities to works predating the plaintiff's composition suggest that plaintiff may have copied and that his work therefore is unprotectable because unoriginal. Similarities may also suggest that defendant copied from someone other than the plaintiff or that both plaintiff and defendant may have copied from a common source. The significance of prior art as a defense to claims of infringement is discussed in Chapter 2, supra page 68.

[50] A sequence is the immediate restatement of a short musical phrase or motive beginning on a different pitch.

[51] Record at 109-12.

[52] Record at 116-17. The response ignores at least three factors: the amount of stated material to repeat, the number of repetitions, and the choice between statements that are tonal (entailing the retention of the original key and requiring some adjustment of melodic intervals to maintain diatonic relationships) or real (the literal restatement of the original melodic intervals, requiring accidentals and implying a change of key).

[53] Record at 187-89. Parsons had held the rank of full professor since 1964. His degrees included a Bachelor of Music from Columbia University, Master of Education from Winnetke Graduate Teachers College, Master of Composition from Northwestern University, and a Ph.D. in theory from Northwestern. Record at 187. Relevant courses he had taken included harmony, counterpoint, form and

analysis, orchestration, fugue, composition, and music theory. Record at 189. As program annotator for the Chicago Symphony, his writing included analysis. Record at 191. He co-authored Creative Analytical Theory of Music, used as a textbook at Northwestern and elsewhere. He wrote an article for the Encyclopedia Brittanica and published articles in Polish Music QuarterlyPolish Music Quarterly. Record at 195. For eight years, he had broadcast the radio program Chicago Symphony Notebook, which, he said, included analysis. Record at 197.

[54] Record at 198.

[55] Record at 199.

[56] Record at 202.

57 See Appendix A, at 613 and 614.

[58] Record at 222-23.

[59] Record at 224-30.

60 See Appendix A, at 615 and 616.

[61] Record at 232.

[62] Record at 232-37.

63 See Appendix A, at 617.

[64] Record at 243-48.

[65] Record at 249.

[66] Record at 249.

[67] Record at 249-50.

[68] This "work tape" appears in Appendix I.

[69] Record at 254-62.

[70] Record at 266.

[71] Record at 267-70.

[72] Record at 266-67.

[73] Record at 278.

[74] Record at 279. Parsons sought cover with this answer again when asked why he didn't check to see if the similar rhythm might be duplicated by other songs. Without that knowledge, how could he know if the rhythmic similarities were significant? Parsons replied, "My assignment was to compare these two songs in question and that is what I did." Record at 309. The tendency to defend oneself is understandable, but the effect is to transfer blame to the party on whose behalf the witness is testifying.

[75] Record at 355.

[76] Record at 286-88.

[77] Record at 818.

[78] Stipulations are agreements between counsel concerning facts incidental to the proceedings or not controverted, such as, here, the identity of the taped excerpt. Lawyers are not supposed to waste the court's time by requiring proof of uncontroverted facts. A stipulation read into the record comprises part of the evidence.

[79] See, e.g., "Lawyers Play `Name That Tune' in the Bee Gees Case," Variety, 23 February 1983, 2; United Press International (Domestic News), 17 February 1983. The excerpt, taken from Theme B of "Let It End," was, in fairness to Gibb, the point of strongest identity to his own song and played out of context.

[80] Record at 410.

[81] Record at 414ff.

[82] Record at 571.

[83] Record at 573.

[84] "Cross-examination should be limited to the subject matter of the direct examination and matters affecting the credibility of the witness. The court may, in the exercise of discretion, permit inquiry into additional matters as if on direct examination." Fed. R. Evid. 611(b).

[85] Record at 620.

[86] The judge may periodically confer with the attorneys at side bar, out of the hearing of the jury.

[87] Record at 621-22.

[88] Record at 621-23.

[89] Robert C. Osterberg, interview by author, 26 October 1990, New York.

[90] Record at 637-39.

[91] Record at 959-60.

[92] Record at 960-62.

[93] Record at 963 and 966.

[94] Record at 977.

95 See Record at 973.

96 See Record at 1039-47.

[97] Selle v. GibbSelle v. Gibb, 567 F. Supp. 1173, 1179 (N.D. Ill. 1983), aff'd, 741 F.2d 896, (7th Cir. 1984).

[98] 567 F. Supp. at 1179.

99 Id.

100 Id. at 1182-83.

101 Selle v. Gibb, 741 F.2d 896, 900 (7th Cir. 1984).

102 Id. at 901.

103 Id.

[104] Id. at 902.

[105] SeeSee Ferguson v. National Broadcasting Co., 584 F.2d 111, 113 (5th Cir. 1978) (Access must be more than a bare possibility and may not be inferred through speculation or conjecture).

[106] 741 F.2d at 905.

107 Id.

108 Id.

[109] Although the name Kaiserman appears on the style of the case, the defendant is referred to throughout the trial as Albert.

[110] Gaste v. KaisermanGaste v. Kaiserman, Case No. 86 Civ. 5671 (S.D.N.Y. 1986), aff'd, 863 F.2d 1061 (2d Cir. 1988). "Feelings" appears in Appendix B, at 708.

111 See Deposition Exhibit 52.

112 Gaste, 863 F.2d at 1063. "Pour toi" appears in Appendix B, at 706.

113 Id.

[114] Id. at 1066.

[115] Record at 456.

[116] Record at 457.

[117] Record at 468.

[118] Record at 466.

[119] Record at 467.

[120] Record at 468.

121 Record at 473. See Appendix B, at 722.

122 See Appendix B, at 724. As explained above, this graph employs lines connecting points of attack and thus incorrectly implies glissandi throughout.

[123] Record at 480.

[124] Record at 485.

125 See Appendix B, at 722. Ricigliano's source for the last F# in measure 5 of "Feelings" is not stated. The figure varies in printed copies and on Albert's recording between a simple stepwise ascent (F# G A B) and the same figure with a lower neighbor (F# G A G B). No version in the printed trial record employs the last F# on Ricigliano's charts.

[126] Record at 487.

[127] Record at 490.

[128] See Exhibit 97.

[129] Record at 495.

[130] Record at 490.

[131] Record at 502.

[132] Record at 503.

[133] Record at 504.

[134] Record at 507. Note also that the chords written in for "Feelings" are incorrect.

[135] Record at 508.

[136] This example appears only as a guide to the reader. It was neither a trial exhibit nor a creation of either expert witness. Chord symbols are shown in preference to a harmonic analysis in order to reserve judgment on those issues. Measures 25 to 32 of "Feelings" are compared to "Pour toi" at the asterisk even though the latter is not repeated.

[137] Record at 508-09.

[138] Record at 509.

[139] Record at 517.

140 See Record at 518.

[141] Record at 518.

[142] Record at 519.

[143] Record at 519.

144 Record at 490 (referring to the A section) and at 510 (referring to the chorus). See also, Record at 527.

[145] Record at 527-30.

[146] Record at 561-63.

[147] Record at 568.

[148] Record at 570. The graph suffers from the same shortcoming of common notation: it shows only diatonic relationships. Thus, it is unclear what advantage Ricigliano thought he had achieved.

[149] Record at 571.

[150] The "Feelings" bass line (section A) descends chromatically from E. Subsequently, the same line is stated in other voices.

[151] Record at 578-79.

[152] Record at 581.

[153] Record at 586.

[154] Record at 592.

[155] Record at 671.

[156] Record at 565.

[157] Record at 661.

[158] Record at 667.

[159] Record at 666.

[160] Record at 677.

[161] Record at 672.

[162] Record at 673.

[163] Record at 674-76.

[164] Record at 682-83.

[165] A two-measure excerpt from Act 2, Scene 4.

[166] Record at 684-85.

[167] Record at 688.

[168] Record at 688-90.

[169] Record at 692.

[170] It is not clear why Coster did not first introduce a published version, which he could authenticate, and then the comparison chart.

[171] Record at 692.

[172] Record at 693.

[173] Record at 694-95.

175 Record at 704-05. See Exhibit KQ, Appendix B, at 729.

[175] Record at 715-22.

[176] Record at 770.

[177] Record at 771-72.

[178] Record at 725.

[179] Record at 773-75.

[180] Record at 489.

[181] Record at 761.

[182] Record at 762-63. It is hard to understand why, if measure 7 belonged with 6, measure 3 did not go with 2, and 5 with 4.

[183] Record at 752-55.

[184] Record at 756.

[185] See Appendix B, at 765.

[186] Record at 897.

[187] SeeSee Harvard Dictionary of Music, 2d ed. (1969), s.v. "Batuque."

[188] Robert C. Osterberg, interview by author, 26 October 1990, New York.

190 Gaste v. Kaiserman, 863 F.2d 1061, 1063 (2d Cir. 1988). Jury instructions appear in Appendix B, at 773.

191 Id. at 1066.

192 Id. at 1067. A decision to overturn a jury verdict depends on whether the evidence is such that, without weighing the credibility of the witnesses or otherwise considering the weight of the evidence, there can be but one conclusion as to the verdict that reasonable persons could have reached.

192 Id. at 1068.

[193] Baxter v. MCA, Inc., 812 F.2d 421 (9th Cir.), cert. denied, 484 U.S. 954 (1987), on remandon remand, Case No. 88-6660 (C.D. Cal. 1988), aff'd, 907 F.2d 154 (9th Cir. 1990) [unpublished opinion].

[194] Baxter v. MCA Inc.Baxter v. MCA, Inc., 812 F.2d 421, 422-24 (9th Cir.), cert. denied, 484 U.S. 954 (1987).

[195] Excerpts of the copyrighted score of "Joy" appear in Appendix C, at 783. Excerpts of a piano arrangement of E.T. appear in Appendix C, at 795.

[196] 812 F.2d at 423. The district court decision (C.D. Cal. Case No. CV 83-7081 HLH (JRX)) is unpublished.

[197] 812 F.2d at 424.

198 Id.

199 Id. at 426. Summary judgment cannot be granted if there exists a genuine dispute as to a material fact. Fed. R. Civ. P. 56(c). A dispute over access would present the court with a triable issue of fact and preclude summary judgment. Therefore, defendant conceded access, arguing that the similarities were so insubstantial that plaintiff's case should fail for that reason alone.

200 812 F.2d at 426 (citing Sid & Marty Krofft Television Productions, Inc. v. McDonald's Corp., 562 F.2d 1157 (9th Cir. 1977)).

[201] 812 F.2d at 426.

202 Id. at 427.

[203] Sid & Marty Krofft Television Productions, Inc. v. McDonald’s Corp.Sid & Marty Krofft Television Productions, Inc. v. McDonald's Corp., 562 F.2d 1157 (9th Cir. 1977). The case was introduced in Chapter 2, supra page 70.

[204] BaxterBaxter predates the cosmetic changes put into effect in the Ninth Circuit by Shaw v. Lindheim, 919 F.2d 1353 (9th Cir. 1990); therefore, the terminology of extrinsic and intrinsic tests must be used.

205 Narell v. Freeman, 872 F.2d 907, 912 (9th Cir. 1989).

[206] Baxter, 812 F.2d at 429 (citing Arnstein v. PorterArnstein v. Porter, 154 F.2d 464, 468 (2d Cir. 1946), cert. denied, 484 U.S. 954 (1987).

207 Krofft, 562 F.2d at 1167.

[208] A ruling that as a matter of law there is no substantial similarity between two works is subject to de novo review. 812 F.2d at 424-45 (citing Berkic v. Crichton, 761 F.2d 1289, 1292 (9th Cir.), cert. deniedcert. denied, 474 U.S. 826 (1985)). A de novo review is one in which the appellate court re-examines and re-weighs the evidence that led to the district court's decision.

[209] 812 F.2d at 428.

[210] Id. at 430. A spurious "six bar rule" has been traced to Marks v. Leo Feist, Inc.Marks v. Leo Feist, Inc., 290 F. 959 (2d Cir. 1923). Sherman, Music Copyright Infringement: The Requirement of Substantial Similarity, 22 Copyright L. Symp. (ASCAP) 81, 104 (1977).

211 Williams v. Baxter, 484 U.S. 954 (1987) (denial of certiorari).

212 Shaw v. Lindheim, 919 F.2d 1353, 1359 (9th Cir. 1990).

213 Shaw alleged misappropriation of the pilot script to the television series The Equalizer. The district court awarded summary judgment solely on the basis of a subjective judgment under the intrinsic test. 919 F.2d at 1357.

[214] Id. at 1359 (citing Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 249 (1986) and Celotex Corporation v. Catrett, 477 U.S. 317 (1986)).

[215] Baxter filed suit on November 2, 1983. 812 F.2d at 423.

[216] Record (8/30/88) at 80-81.

[217] Record (8/30/88) at 7.

218 See Appendix I.

[219] Record (8/30/88) at 7.

[220] A better quality copy does not exist.

[221] Record (8/30/88) at 8-9.

[222] Record (8/30/88) at 9.

[223] The importance of a theme within a musical work cannot be tied to loudness. However, because phase one of the trial tested the jury's ability to discern similarities without benefit of analysis, the plaintiff would want to be sure the relevant theme was heard prominently.

[224] Record (8/30/88) at 12-13.

[225] Record (8/30/88) at 14-15.

[226] Record (8/30/88) at 16-17.

[227] Record (8/30/88) at 17-18.

[228] Record (8/30/88) at 19.

[229] Record (8/30/88) at 21-22.

[230] Record (8/30/88) at 22.

[231] Record (8/30/88) at 42.

[232] Record (8/30/88) at 44.

[233] Although two versions of the "Joy" score were available, Exhibit 412 was admitted as the more legible. The other version, which had been marked as plaintiff's Exhibit 101, would have been introduced had any confusion arisen by virtue of its having been used during the taking of deposition testimony. Portions of Exhibit 412 (Appendix C, at 783) are very difficult to read. There remained a dispute over whether the last note of the trombone part in measure 73 is A or G. Record (8/31/88) at 87-93.

[234] Record (8/30/88) at 54. The score reproduced in Appendix C, at shows this note whited out. The fact that the top line of the staff has been removed in the process might suggest that the original note was A rather than G, but this is not dispositive.

[235] Record (8/30/88) at 58-59.

[236] Record (8/30/88) at 121-22.

[237] Record (8/30/88) at 123.

238 See Record (8/30/88) 132-34 (Appendix I).

[239] Record (8/31/88) at 7.

240 3 M. & D. Nimmer, Nimmer on Copyright, § 13.03[A][2] (1989).

241 Id. § 13.03[A][1].

[242] Record at 11-13.

[243] Record at 32-34. It may be difficult for a musician and a layman to understand how the other listens to music. The court seeks an intuitive response from the jury, which, it would seem, is the only kind of response a jury of laymen is capable of rendering. Thus, admonishing the jury to avoid analytical listening may have no effect. However, the admonishment may actually encourage more passive, casual, or inattentive listening by the jury than what would be normal for the average listener. The instruction seems to imply that experts listen in a cold, mechanical fashion and that lay listeners hear music only subliminally, through a fog of ignorance and faulty mental associations. Perhaps the court should credit lay listeners, who, after all, seek out most of the music they hear, with being somewhat more careful and discerning in their listening.

[244] Record at 35.

[245] Record at 37-38.

[246] Record at 37.

[247] Record at 39.

[248] Record at 40.

[249] Record at 40.

[250] Record at 41.

[251] Tapes were played in the following order: Exhibit 59, 56, 60, 61, 54, 55, 36, 57, 58. Record at 43-45.

[252] Record at 48.

[253] Record at 52-53.

[254] Record at 53.

[255] Record at 54.

[256] Record at 55.

[257] Record at 56.

[258] Record at 59.

[259] Record at 62.

[260] Record at 63.

[261] Record at 75. The full jury instructions appear in Appendix C, at 913.

[262] Record at 80-81.

[263] Record at 3-6.

[264] Record at 11-12.

265 Springer had earned his Ph.D. from U.C.L.A. but had never held a full-time faculty position. He worked as a composer on the west coast. He estimated that he had published some 150 to 200 songs, some of which had been performed by well-known entertainers such as Frank Sinatra, Elvis Presley, Judy Garland, Madonna, Frankie Laine, Ertha Kitt, and Steve Lawrence. He had also written the theme to the television show Cross Wits. He was testifying for the first time. Record at 5-7.

[266] Harold Barlow lists his occupation as full-time plagiarism expert. He holds a Bachelor of Music from Boston University and completed a year of graduate work at Juilliard. MCA, Inc. v. Wilson, 425 F. Supp. 443 (S.D.N.Y. 1976), aff’d and modifiedaff'd and modified, 677 F.2d 180 (2d Cir. 1981), Record at 11-19. He co-authored A Dictionary of Musical Themes (New York: Crown, 1975).

[267] Both are frequently called as experts in copyright litigation. Barlow is perhaps the dean of experts, a position held since Sigmund Spaeth left the scene.

[268] The judge has this discretion within certain limitations. This case had taken longer than the attorneys had promised, and there were severe time constraints because of other cases on the docket. Barlow testified last, and the judge believed that the testimony was becoming duplicative.

[269] Record at 26.

[270] Record at 46.

[271] Record at 47.

272 Selle v. Gibb, 741 F.2d 896, 904 (7th Cir. 1984).

273 See Record at 205-11.

[274] Record at 103.

[275] Record (9/2/88) at 46-48.

[276] Record at 43.

[277] Record at 1056.

[278] Record at 974-75.

[279] Record at 1103-04.

[280] Record at 42.

[281] Record at 109.

[282] Record at 111.

[283] Record at 111-12.

[284] The witness who has no support for his proposition must hope that the attorney doesn't have a whole stack of authorities that the witness will have to denigrate seriatim. In such cases, the jury will soon get the impression that the witness holds a peculiar view.

[285] Record at 97.

[286] Record at 98.

[287] Record (8/30/88) at 49.

[288] Majestic.

[289] Record at 78-79.

[290] Record (9/2/88) at 15-20.

[291] Record (9/2/88) at 11.

[292] The relation demonstrated, however, is purely physical, not musical.

[293] Record (9/2/88) at 78. The author hears A.

[294] Record (9/2/88) at 85-86.

[295] Record (9/2/88) at 79.

[296] Record at 967.

[297] Record (9/2/88) at 455.

[298] A sustained note in the bass sounding with changing harmonies in other parts.

[299] Record at 964-67.

[300] The motive also begins with three repeated notes.

[301] Record at 176-79.

[302] Record at 117.

[303] Record at 121. The two statements overlap in measure 25.

[304] Record at 122.

[305] Record at 32-33.

[306] Record at 30.

[307] Record at 263.

[308] Record at 265.

[309] Exhibit 486 is presented in Appendix C, at 898.

[310] Record at 894-95. The comparison chart appears in Appendix C, at 896.

[311] Record at 895-96.

[312] Record at 904. See also, Northern Music Corp. v. King Record Distributing Co., 105 F. Supp. 393, 397 (S.D.N.Y. 1952) (both songs played simultaneously).

313 See Record at 89.

[314] Record at 63.

[315] Record at 65.

[316] Record at 66-67.

[317] Record at 67.

[318] Record at 68-69.

[319] Record at 70.

[320] Record at 71-72.

[321] Record at 72-73.

[322] Record at 74-75.

[323] Record at 144.

[324] Record at 180.

[325] One can assume, although it is not clear from the written record, that he played "Twinkle, Twinkle, Little Star" beginning on the second scale degree (supertonic to dominant).

[326] Record at 87. It probably would have been more useful to describe the similarities in terms of functional scale degrees (tonic to dominant versus supertonic to dominant) rather than as isolated intervals. It is unclear why Winter failed to describe similar motion to the dominant (G) rather than indulge in highly questionable speculation concerning the listeners' ability to distinguish a perfect fourth from a perfect fifth.

[327] Record at 167.

[328] Record at 168. The similarity of pitch would be lost almost entirely, and Winter's fourth criteria would fail.

[329] Record at 170.

[330] Record at 181-82.

[331] Record at 182.

[332] Record at 183-84.

[333] Record at 184-86. Once again, the witness had sought to protect his own reputation by blaming flaws in his analysis on the instructions he was given. See supra note 74.

[334] Record at 188-89.

[335] Record at 189.

[336] Record at 972-73.

[337] Record at 973.

[338] Record at 1055.

[339] Record (9/2/88) at 34.

[340] Record (9/2/88) at 25-26.

[341] Record (9/2/88) at 32-33.

[342] Record (9/2/88) at 39.

[343] Record (9/2/88) at 39.

[344] Record (9/2/88) at 39.

[345] See, e.g., Cedric Thorpe Davie, Musical Structure and Design Musical Structure and Design (London: Dobson, 1953; repr., New York: Dover, 1966), and Douglass M. Green, Form in Tonal Music, 2d ed. (New York: Holt, Rinehart, and Winston, 1979).

[346] Record at 190.

[347] Record at 191.

[348] Record at 836-37.

[349] Record at 881.

[350] Record at 882-83.

[351] Record at 889.

[352] Record at 884.

[353] Record at 885.

[354] Record at 887-88.

[355] Record at 920.

[356] Record at 920.

[357] Record at 921-22.

[358] Record at 134-35.

[359] Record at 136.

[360] Record at 145-46.

[361] Record at 165.

[362] Record at 983.

[363] Record at 983.

[364] Record at 984.

365 Record at 985. See chart in Appendix C, at 896.

[366] Record at 1023.

[367] Record at 1024.

[368] Record at 1032.

[369] Record at 1025.

[370] Record at 1025-26.

[371] Record at 1033.

[372] Record at 1034.

[373] Williams played so many short excerpts that they often tend to form the nouns in his sentences. The transcript, always inadequate in many senses, fails entirely at many points, particularly where Williams apparently improvised illustrations at the same time that he spoke.

[374] Record at 693.

[375] Record at 685.

[376] Record at 697.

[377] Record at 698-99.

[378] Record at 1102.

[379] Record at 1102-03. Winter makes a valid point here, but the reader should remember that the asymmetry does not appear until the third measure of the "Joy" "B" theme.

[380] Record at 705-09.

[381] Record at 94.

[382] Incipit simply refers to the first few notes. Harold Barlow's A Dictionary of Musical Themes (New York: Crown, 1975) provides a good example. Themes are transposed to the key of C (major or minor). The succession of pitches, letter names of the notes without regard to rhythm, are then alphabetized. For example, the second theme of Schubert's Symphony No. 8, D. 759 ("Unfinished") is stated in the key of G, beginning G D G F# G, etc. Barlow indexes the theme as C G C B C D C B. The number of pitches comprising the incipit for this purpose is the minimum necessary to distinguish it from any other. The incipit appearing immediately after Schubert's reads C G C B C D C E. It refers to the Corelli Concerto Grosso in B-flat, op. 6, no. 11 ("B" theme of the Allemande), which bears absolutely no perceptible resemblance to Schubert's theme. Incipit indexes are useful tools when one hears a theme he cannot identify, but a mere succession of pitches presents too little data to help an analyst discover similar themes.

[383] Record at 95-96.

[384] Record at 925-26.

385 This burden was shifted to plaintiff in Selle v. Gibb, 741 F.2d 896, 905 (7th Cir. 1984), only as it related to the issue of striking similarities.

[386] Record at 931-36.

[387] Record at 937.

[388] Record at 940.

[389] Record at 945.

[390] Record at 947-48.

[391] Record at 949.

[392] Record at 954-58. Exhibit 492 appears in Appendix C, at 895.

[393] Record at 959. The exercise, A. E. Müller's Theme and Variations, appears in Appendix C, at 897.

[394] Record at 1041-42.

[395] Record at 1051.

[396] Record at 1048.

[397] Record at 1046.

[398] Record at 1043.

[399] Record at 1041.

[400] Record at 1052.

[401] Record at 1081-1103.

[402] Record at 1097.

[403] Record at 1090.

[404] Record at 1082.

[405] Record at 1085.

[406] Record at 1088.

[407] Record at 1089.

[408] Record at 1093.

[409] Record at 1098.

[410] Record at 42.

[411] Record (9/2/88) at 100.

[412] Record (9/2/88) at 101.

[413] Record (9/2/88) at 109-10.

[414] Record at 489-90.

[415] Record at 490.

[416] Record at 494.

[417] Record at 495-96.

[418] Record at 502.

[419] Compare, e.g.Compare, e.g., measures 9, 17, 34, 55, and 63 of E.T.

[420] Record (9/12/88) at 6-7.

[421] Record (9/12/88) at 8.

[422] Record (9/12/88) at 10.

[423] Record (9/12/88) at 11.

[424] The complete transcript of closing arguments and jury instructions appears in Appendix C, at 952 ff.

[425] Record (9/12/88) at 86-87.

426 Joint Appellee's Brief at 6, Baxter v. MCA, 907 F.2d 154 (9th Cir. 1990) (No. 88-6660). The complete brief appears in Appendix C, at 1061.

427 Baxter v. MCA, 907 F.2d 154 (9th Cir. 1990) (No. 88-6660) [Unpublished opinion at 2 (citing Williams v. Hughes Helicopters, Inc., 806 F.2d 1387, 1391-02 (9th Cir. 1986); accord Herrington v. County of SonomaHerrington v. County of Sonoma, 834 F.2d 1488, 1500 & n.11 (9th Cir. 1986), cert. denied, 489 U.S. 1090 (1989))]. The unpublished Circuit Court opinion appears in Appendix C, at 1118.

428 Id. at 3.

[429] Id.