CHAPTER 5

METHODS OF REDUCTION: THE SEARCH FOR MUSICAL SIGNIFICANCE

Music analysis entails reduction and comparison.[1] Neither act is entirely independent of the other. Nevertheless, proper ordering of these steps of analysis must be established. Legal proof of infringement demands some standards, and the real danger remains that similarities may influence analysis more than analysis reveals similarities. Analysis should discover similarities and contrasts and assess their significance. To do this, the analyst must approach the music objectively and not simply construct arguments to support an established hypothesis.

This chapter concentrates on analytical methodologies for defining, through reduction, the material to be compared for similarities. A subsequent chapter will focus on methods of comparison. The initial step of reducing the music to its essential thematic components, the process of segmentation, profoundly affects on subsequent comparisons. Segmentation involves the discovery of musical units such as sections, phrases, and chords. Bad decisions in segmentation lead to erroneous conclusions. Problems lie in articulating appropriate criteria for segmentation and other types of reduction. Many factors, some seemingly based primarily on intuition and experience, influence the choices that must be made.

Because similarities define much of the structure of music and influence its proper segmentation, one may question the apparent tautology of this exercise. Segmentation defines the criteria for comparison while similarities define the criteria of segmentation. Analysts can offer as a disclaimer Schenker's observation quoted earlier: "Music is always an art. . . . Under no circumstances is it a science."[2] However, the expert testimony of Baxter suggests a principled, if not scientific, explanation,[3] and it will be illustrated in other cases as well: Segmentation must be driven by similarities within the individual work and not by similarities between two works. The latter is truly tautological. The former uncovers the elements that coalesce to turn sound into a coherent and identifiable piece of music. Only by determining the boundaries of these significant elements within each work can the court avoid basing its legal decision on haphazard comparisons and coincidental similarities.

The Expert's Methodology

The expert can expect two questions on his methodology: (1) "Do you always use the same criteria when you compare two works?" and (2) "Would your analysis be the same had you been employed by the opposing party?" The second question may be posed as "Did you analyze dissimilarities in addition to similarities?"

In Baxter, one of the experts answered the second question by pointing out the difference between a comparison and a contrast. He had been asked to do a comparison.[4] With this answer, the expert cast doubt on the legitimacy of his conclusions and transferred responsibility for many critical aspects of analysis to his lay employer. Such testimony does not help the trier of fact. It is rather like having a physician testify that he was hired to diagnose a particular disease. Music theory does not distinguish comparative analysis and contrasting analysis. Analysis accounts for both factors. The distinction belongs only to the less demanding process of description.

Experts usually testify that they examine the same factors whenever they compare two works.[5] Standard methodology seems essential if the court is to trust the expert's initial objectivity.[6] But standard procedures must yield at some point to the peculiarities of the individual work.[7] Significant features differ according to the nature of the music. Consequently, some comparisons become more meaningful than others. The final analysis of a fugue does not resemble that of a chorale. The question is this: At what point does the music itself begin to dictate the analytical methodology?[8]

The trained theorist has access to various methods of analysis that may prove useful. The legal process seems to be best served by allowing the expert to employ any legitimate tools of analysis that may be helpful to the trier of fact. No particular school of analysis automatically qualifies as superior. The analyst should choose methods that illuminate the composition process, that disclose the building blocks of the pieces in question and that enable him to compare something more consequential than coincidence. Certain guidelines may help all concerned to assess the trustworthiness of particular techniques and to ascertain whether a given analysis presents a complete or sufficient account of the music.

Similarities are often easy to find. The expert witness is supposed to demonstrate, however, similarities that provide indicia of copying. Although the law has left the term "substantial similarity" undefined, the purpose behind the inquiry is clear. Whether substantial refers to quantity or quality, similarities per se do not prove infringement. A plagiarist takes more than a succession of sounds, more than a certain quantum of notes. In order to infringe, he must take sounds that appear in the same context, that have similar functions, and that follow some of the same musical principles.

Music follows certain principles that distinguish it from noise. Those principles are not usually apparent on the surface of the music.[9] Through the process of reduction one discovers the components of the music and the functions those components serve in various larger contexts. This process of breaking down music follows relatively objective criteria. When making these reductions the analyst should adhere to standard procedures, and the court should judge the expert accordingly. But the nature of the individual work, thus discovered through standard techniques of reduction, may suggest some specific comparisons as more illuminating than others. Guidelines for comparison, somewhat less stringent than those for reduction, will be suggested later.

Objections to Reduction

Knowledge of any subject matter entails knowledge of its component parts. Courts have recognized the necessity of reduction, mandating repeatedly that expert testimony on infringement shall include "dissection and analysis."[10] The message has not always registered. Counsel will naturally resist divisions of the music that do not favor his case and suggest new divisions that do. Defense counsel Osterberg asked the witness Selle:

Q And in the part that follows the introduction, what we'll call Motif A, there are two measures with similarity?

ENGERMAN: I am going to object, your Honor, unless we are talking about the whole eight bars, which we allege are copied from "Let It End."

THE COURT: Mr. Engerman, this is cross examination. Let Mr. Osterberg proceed with his cross examination. The objection is overruled. Can you answer the question?[11]

One wonders if the objection would have been sustained on direct.

Resistance to reduction comes in other forms. The expert witness who examines musical parameters in isolation, who segments according to musical event rather than at the bar line, or who seeks out the motivating function beneath the surface of the music invites attack on cross-examination. Typically, the attorney will seek an admission from the witness that he left something out or that the notes contained in a particular measure contradict the witness's testimony. It is undeniable. The witness's reduction, omission, simplification, or segmentation is not what the attorney's client wrote. To the jury, unschooled in the techniques of musical analysis, the witness appears to have rewritten the music to make it fit his theory of the case. Yet "dissection" would be a very hollow concept if it meant nothing more than divisions made according to the visual criteria of bar lines.

Relevance of Reductions

Music lends itself to division along any of the three lines: (1) temporal segmentation, (2) isolation of parameters, and (3) hierarchical reduction. Temporal or formal segmentation refers to the division of music into motives, phrases, themes‑-all manifestations of a complete musical statement. Temporal segmentations suggest a certain quantity of music; they define thematic units. Isolation of musical parameters might include analyzing melody without harmony or rhythm without pitch. These reductions separate music's constituent elements, those that function simultaneously. They demonstrate various principles of organization and movement‑-ways in which music attains unity and disunity at the same time. Finally, one might use hierarchical reductions to move from surface to background, gradually simplifying the music to reveal its underlying principles and design, its larger aspects, its relationship to the phenomenon of tonality. The average listener, though perhaps unable to explain it, hears music on all of these levels.

Temporal Segmentation

The introduction in the 17th century of rhetorical analogies to music analysis offered particularly useful insight into the nature of music. It described in terms that could be immediately comprehended the pervasive divisions of music into distinct statements. All art has form; music is not merely a succession of notes. Whether one believes music imitates rhetoric or biological functions, or follows some purely musical principle, music is properly understood as containing natural divisions. The mind tends to seek out patterns, repetitions, and resting points within the overall structure as it listens. One of the first acts involved in an analysis of music is the discovery of these natural divisions. Divisions may be determined by many different musical criteria: phrasing, rhythm, harmonic progression, melodic contour, to name a few. Congruence of musical events may point unmistakably to divisions on which all would agree. But at times, some characteristics of the music obscure divisions suggested by others, and even analysts employing their best musical judgment might disagree on where the division can be made most meaningfully.[12] In all cases, the analyst must base his decisions on musical phenomena to discover what exists, and not impose arbitrary divisions or divisions suggested by non-musical criteria.

The art of analysis uses these divisions as reference points. Analysis segments the music into smaller, manageable musical events. Segmentation aims not to eliminate material from the analysis but to determine how each smaller event serves the whole. Each event has a certain motion and direction that functions within a larger event. At every point, the analyst must determine how smaller segments form the building blocks of a larger segment.

Because music uses some of the terminology of rhetoric, the comparison of segmentation to punctuation is apt. Phrases form sentences, sentences form paragraphs, paragraphs form chapters. How the divisions are made affects the meaning of the whole.

Dear John,

I want a man who knows what love is all about. You are generous, kind, thoughtful. People who are not like you admit to being useless and inferior, John. You have ruined me for other men. I yearn for you. I have no feelings whatsoever when we're apart. I can be forever happy. Will you let me be yours?

Gloria

New divisions yield dramatically different results.

Dear John,

I want a man who knows what love is. All about you are generous, kind, thoughtful people, who are not like you. Admit to being useless and inferior, John. You have ruined me. For other men, I yearn. For you, I have no feelings whatsoever. When we're apart, I can be forever happy. Will you let me be?

Yours,

Gloria[13]

The analysts who looks for a corresponding succession of words in these two letters will find them at every point. If he looks for nothing else, he will discover no distinction between the two. Similarly, if punctuation were ignored and the two examples segmented according to lines of text, they would be seen as exact duplicates. However, if one compares the examples sentence by sentence, similarities are significantly decreased and might not be considered substantial. The second sentence of each, for example, retains relatively few common elements. Thus, the method of segmentation may either emphasize or obscure differences. Divisions should be made intelligently, considering the rules of grammar and the underlying expression, not at arbitrary or expressively irrelevant points.

False Segmentation

To one who does not read, the most appealing point for dividing text would be at the end of each line; but that segmentation is a function of type size, not literary expression. In the same way, music presents a highly attractive point at which to make divisions: the bar line. Although theorists know that music is not presented in neat, regular-sized little boxes, to the layman, bar lines present an irresistible visual reference. There must be some significant reason, the layman thinks, why a block of music is enclosed by these dividing lines.

The analogy of bar lines to lines of text works at least to this extent: both are evenly placed divisions that do not determine the function and expression of what they divide. Bar lines are mere notational devices that mark the passage of time. Without them, the music would sound the same. They provide a useful frame of reference to the performer. But just as children are taught to read without pausing at the end of each line of text, the performer learns not to express the bar line in any way. Bar lines are like yard lines on the playing field: the ball carrier may find the markings helpful, but he does not run in five-yard segments.

The two lines of music in Figure 14, played as written and isolated from any larger context, sound the same. If played within a larger context in which the meter is perceived, they may convey a different impression because they relate differently to the meter. However, the placement of the figure in a different relation to meter will not justify a new segmentation.

Example 14: Brahms' Variations on a Theme by Handel, Var. VI.

Bar lines reflect meter, which is something different from rhythm. The analyst does not ignore bar lines entirely, because the function of meter must be considered in his analysis. But every aspect to be considered in an analysis does not present an opportunity for segmentation. Meter tends to emphasize the first beat of the measure. Consequently, measure and musical segments do not always coincide‑-otherwise, all segments would begin with an accent, which they do not.[14] Bar lines are, perhaps without exception, an improper basis for segmentation.[15] Music is neither conceived nor heard in one-measure segments. Thus, there can be no rationale for analyzing it according to that criterion.

Meter, which bar lines delineate, forms the context in which rhythm is heard. The significance of the two musical statements in Figure 14 lies in their literal sameness and their contextual variation. If the context were allowed to determine segmentation, then the sameness would go unnoticed. Segmented at the bar line, these figures would become simply two unrelated musical events, and the piece from which they are taken, a simple canon at the octave (Figure 15), could not be analyzed meaningfully.

Example 15: Brahms' Variations on a Theme by Handel, Var. VI.

Parties to litigation often argue over the proper segmentation as a prelude to counting similar features. Division by measure more commonly diminishes the apparent quantity of similarities. Earl Spielman, being deposed by the defense, was asked:

Q In the fourth measure of [defendant's song] "Diggin' Up Bones," Overstreet versus Stuckey, isn't there a note that is not in the fourth measure of [plaintiff's song] "Sun Comin' Up"?

A The very last beat of measure four is the upbeat to the second phrase.

Q There's a different note, isn't there?

A We're talking about the first phrase?

Q I'm talking about the first four bars of the melody of the verse.

And the last bar, the fourth bar of that phrase, isn't there a different note?

A Yes. . . .

The last note in measure four of "Sun Comin' Up," the last two notes of measure four in "Sun Comin' Up," are the upbeats to the fifth measure to the second phrase. And they're not heard as part of the fourth measure. They're heard as part of the next phrase. They're upbeats in the same way as it is in the first two notes coming before the first measure lead-in to the first measure.

If you're talking about measures as ruled lines, your ear does not hear that. This is only a means of notating it, putting it down on paper. If you're talking about what exists in music, you talk about phrases. And the first phrase ends the third beat of measure four.

Q The only thing I asked you is: Are there different notes in the fourth measure of these two songs?[16]

This type of false segmentation may also be used to inflate the significance of similarities that may be unrelated and legally inconsequential:

Q How many bars of "Feelings" contain references to musical ideas contained in "Pour Toi"?

A Well, I think every bar does.[17]

References to musical ideas do not connote copying. References to "Pour Toi" contained in "Feelings" may not correspond in a way that suggests infringement. Yet the testimony conveys the impression that the composer of "Feelings" copied "Pour Toi" measure for measure.

Other types of segmentation may be equally wrongheaded or disingenuous, but segmentation by measure presents a pervasive flaw in infringement litigation. It provides a good example of how analysis might be skewed to produce the desired result. The danger is not unique to music; the court commonly accepts into evidence materials that can be interpreted in a variety of ways. The fault lies not in music analysis but in meaningless divisions that masquerade as music analysis. The mis-characterization often arises in the questions asked by counsel.

Appropriate Divisions in Formal Analysis

Opportunities for appropriate segmentation present themselves at all levels of analysis, from the smallest figure to entire movements within works of heroic proportions. The analyst must resist any tendency to draw lines according to their effect on similarities. An internal principle must justify whatever segmentation is made. Divisions must be expressively significant with regard to the individual work, and they must be based on established analytical principles. The smaller divisions present the greatest opportunity for abuse. In general, the court should require that divisions be made according either to musical phrase or to musical figure.

A figure is a small structural unit, usually three to eight notes in a distinctive rhythmic pattern.[18] "Motive" applies to a figure which, through repetition and variation, generates some significant portion of the work.[19] Motives serve a unifying function;[20] they tend to be thematic and structural. "Figure" commonly refers to a more subordinate unit of music. Infringement suits rarely involve a musical figure alone.[21] Where the complaint alleges similarities in a small amount of unrepeated material, it will be difficult to show that the similarities are substantial. A motive, however, may be highly significant. The theorist Rudolph Reti posited motives as the primary unifying element of music. He explained the compositional process behind entire works as the development of small thematic cells.[22]

Phrase lengths are defined by their conclusions‑-the cadence.[23] A cadence serves to separate ideas;[24] it usually consists of the last two or three tones of a phrase that direct it toward a resting place.[25] Cadences serve a highly functional purpose, punctuating the end of a musical statement. Various endings are described as complete, incomplete, or even deceptive. Because phrase endings tend to be more functional, music, unlike poetry, "rhymes" at the beginning of phrases.[26]

The cadence ending the phrase may be found with reference to actual or implied harmonic progressions. The phrase may develop a particular figure or motive and often contains some type of repetition.

Example 16: "Joy" B Theme.

Baxter provides a useful example. The relevant phrase of the "Joy" B theme (Figure 16) begins with a motive spanning the first two measures followed by a varied repetition of that motive. Where the melody comes to rest on the tonic note (C) in measure 4, the phrase ends with an authentic cadence. The literal repeat of this material in the succeeding four measures should be considered a second phrase. Had the second cadence of measure 8 attained a greater sense of repose than that of measure 4, even if attained through a factor other than melody, then measures 5 through 8 might be seen as a continuation of the phrase or as the second half of a musical period.[27]

Example 17: "E.T." Theme.

"E.T." also begins with a motive stated in the first two measures (Figure 17). Measures 3 through 4 restate the motive with varied sequential treatment.[28] The sequence is real rather than tonal (F‑sharp retains the original intervals rather than the diatonic relationship), but the first and last notes of the sequential statement have been altered. The cadence in measure 4 suggests less repose than measure 2; the melody is on scale degree 7, the leading tone, and the supporting harmony suggests a change of key.[29] If this point had been reached without the stronger close at measure 2, the four-measure statement so far might be considered a phrase ending on a half cadence. The resulting analysis would yield a period of two four-measure phrases. But the sequence trumps any division into four-measure units.[30] Therefore, measure 4 does not end the phrase. The sequential device continues through measure 6, although the melodic content is further varied. At measure 8, the harmony has returned to the original tonic (C). The 7th scale degree in the melody presents some ambiguity; Williams, as he testified, did not want to reach a definitive sense of repose while trying to depict the feeling of flying. But the tonic in C clearly provides a greater sense of completion than the harmony of measure 4. Further, because the next event of the piece is a repeat of the material of these eight measures shown in the example, the phrase must be considered at an end.

When the experts compared the first two measures of each work, they compared motive with motive. But plaintiff's experts, who suggested similarity of phrase structure, went on to make two incorrect assertions. The phrases of the two works do not develop using the same device, as they argued, and the phrases are not of equal length. A natural division occurs at the end of eight measures in each work, and this may have justified comparing eight-measure segments; but in the case of "Joy," eight measures marked the end of a two-phrase thematic statement while eight measures marked the end a single phrase in "E.T."

Sequential treatment of a figure creates differences among experts over how to characterize its originality. The defense in Selle argued that originality could be found only in the first two measures of plaintiff's song. The sequence that followed was labeled pure mechanics.[31] The confusion seems to rest on a failure to distinguish "original" from "new." The requirement of novelty belongs to patent law; copyright requires only originality in the sense that the material must be the composer's own.[32] The debate over the originality of sequential treatment thus has no relevance. Each sequential repetition lacks novelty; but if the composer bases the sequence on an original figure, the work will retain its originality (in the legal sense) throughout the sequence.

Sequence provides the composer with a useful and uniquely powerful developmental tool. Elements of a composition must be related in a way that unifies the work. A work that constantly presents new material will not be understood. Consequently, composers in all musical genres employ many common devices to develop a limited amount of material.[33]

Expert witnesses rarely compare segments larger than phrases. Once the expert accurately segments motives and phrases, analysts are likely to have fewer differences regarding larger divisions. Differences in the larger divisions more often center on terminology and the significance of the segment as it relates to paradigmatic formal structures. Although the divisions themselves are important to internal structure, relationships to paradigmatic forms involve external criteria that have no bearing on substantial similarities between two works.

Indicia of Trustworthiness

Music analysis requires segmentation according to legitimate musical criteria. From the legal perspective, false segmentation by the expert should be inadmissible as more prejudicial than probative.[34] However, the court lacks the knowledge necessary to weigh the validity of the segmentation. If a qualified expert bases his analysis on false segmentation, his opponent will probably have to rely on cross-examination to expose the flaw.

Faulty analysis, however, eventually reveals itself. By presenting rather simplified propositions alone, counsel and expert are both able to alter the criteria in a manner that appears to be merely a question of judgment. The error can be exposed by following it to its logical conclusion, but this consumes time and taxes the jury's ability to follow a complex argument in an unfamiliar field.

The solution to the dilemma lies in the conduct of discovery. Any expert who has had time to prepare for trial and produce a written report should provide an analysis that is sufficiently complete to be tested for accuracy. This entails examining the complete subject matter of each work at issue and explaining the function that each identified segment performs. False segmentation will not yield a coherent analysis on this scale.[35] Opposing counsel will have at his disposal the means necessary to question the witness on his choices of segmentation without having to construct a lengthy argument that the jury cannot follow. At the same time, the expert will have presented an analysis with some indicia of trustworthiness. With analyses from both sides identifying what they consider to be relevant dissection, the judge will have a basis for sustaining objections to questions that elicit unfairly prejudicial testimony. The judge can limit questions to those based on segmentations that the attorney's own expert is prepared to defend.

Example 18: "Let It End."

Figure 18 provides an example of how relevant segments might be identified in a pre-trial report. The pre-trial reports of most experts omit this information. Spielman regularly provides this same type of material in his reports but does so in chart form.[36] The format suggested in Figure 18 may present the relevant segmentation in a format more easily understood by the layman.

Isolation of Parameters

In addition to looking at music in shorter segments, the analyst may choose to examine certain musical parameters in isolation. It should be clear that taking the music apart by parameter represents one method of reduction; it can be justified by the same reasoning that supports any other reduction. The fact that music as a whole or some aspect of music in particular may be defined as including a certain combination of parameters should not preclude omission of any parameter in the course of analysis. Most infringement cases have indulged heavily in this approach to analysis; the parameters that make up melody receive enormous attention at the expense of other parameters. A party cannot reasonably present melody alone and simultaneously object as a matter of principle to the elimination of parameters. Counsel may legitimately argue the relative importance of various parameters, but such arguments go only to weight, not admissibility. The only criterion to be applied to the isolation of parameters is good musical judgment. The rules of segmentation continue to apply to parameters in isolation, although the divisions appropriate to one parameter may contradict those of another.[37]

Each parameter plays a part in the overall structure of music. Some styles, however, emphasize or sublimate certain parameters. For instance, certain jazz forms place a high value on rhythmic variety, often challenging the listener's capacity to relate the rhythms to their underlying metrical pattern. On the other hand, rock music and marches tend to sacrifice rhythmic interest and even variety of tempo in order to thrust regular metrical patterns into the foreground.[38] The analyst must consider how these parameters function internally, in relation to every other parameter and to the work as a whole.

Some musical parameters are susceptible to objective measurements, but all parameters produce psychological effects that outweigh their scientific aspects. Music theory differs from the study of acoustics in this significant regard: the science of acoustics studies how musical sounds are generated; music theory studies how those sounds are perceived.

Music is a highly abstract art, and the parameters that operate within it share that attribute. Only the static phenomena yield precise, objective definitions. But music moves in time, and static phenomena cannot begin to explain music's more significant aspects. Musical motion needs to be understood, whether one is discussing rhythmic motion, harmonic motion, or melodic motion. Motion, more than any other factor, produces the psychological effect. Motion triggers memory and associations; it enables the listener to identify a particular work. Static similarities cannot convince a lay listener that two pieces of music sound alike.[39]

Motion involves more than the mere succession of sounds. Motion requires direction. The listener perceives the direction of musical motion largely through a series of comparisons: consonance versus dissonance, stress versus repose, unity versus disunity. In addition to these comparisons, the listener has certain expectations. Musical motion that confirms or contradicts the listener's expectations may account for many of music's psychological effects. The emotional response to music results from the frustration of expectations.[40]

Music theory uses many terms of art derived from rhetoric and other disciplines. Participants in copyright litigation tend to use musical terms loosely, a factor which contributes considerable confusion. Experts often employ musical terms that connote, but do not entail, the psychological factors that determine perception. Musicians share a terminology in which the abstract and psychological features of music are assumed to be understood. The lay trier of fact may easily misconstrue or discount those connotations.

Parameter Defined

"Parameter" may be generally defined as "any set of physical properties whose values determine the characteristics or behavior of something."[41] The term, borrowed from mathematics, acquires some vagueness in music. Theorists apply the term to any musical variable, such as pitch, rhythm, volume, timbre.[42] Objective measures can quantify some of those variables; frequency, intensity, and timbre, none of which entail musical motion, can be explained in acoustical terms. But some parameters depend on or relate to others. Rhythm and harmony require reference to some constant. Melody represents a particularly complex combination of factors that cannot be adequately described by its components.

Defining those parameters that arise most frequently in litigation might alleviate some of the confusion that surrounds these aspects of music. Definitions provided in theoretical texts, although often inadequate in a legal setting, are preferable to ad hoc determinations by counsel and judge during the jury instruction conference.[43]

Pitch

Pitch relates to the frequency of vibrations made by a sonorous body. Frequency is one of the three measurable qualities of a musical tone, the other two being loudness and timbre. Unlike frequency, which can be expressed in vibrations per second, pitch refers to the subjective classification of tones. A musical tone consists of several frequencies: its fundamental tone and a group of overtones of higher frequency. Usually, but not always, the listener hears the fundamental tone as predominant and classifies the pitch accordingly.[44]

Although one can express frequency in absolute terms, pitch requires relation and classification. Pitch refers to frequency in relation to other frequencies. Music currently employs the constant A = 440 vibrations per second. The average ear can discern frequencies between roughly 20 and 20,000 vibrations per second. The range of the piano extends from low A at 27.5 vibrations per second, to a C vibrating 4,186 times per second. Western music classifies these frequencies within the range of the piano into a mere eighty-eight pitches. Each octave, defined by the doubling of frequency, is divided into twelve pitch classes[45]; traditional western music provides only twelve labels for all frequencies between A = 440 and A = 880. The A‑sharp one semitone higher that A = 440 has a frequency of about 466 vibrations per second.[46]

In spite of these gross generalizations of frequency, however, different tuning systems require distinctions of 1/50 of a semitone, and scientists have documented the ability to hear as little as 1/120 of a semitone.[47] One may accurately speak of twelve different pitches within the octave, but the ear hears distinctions far more subtle than the terminology suggests.[48] Some such distinctions are regularly performed and perceived even in popular forms, and some have even become idiomatic.[49]

Thus, pitch can be seen as a highly generalized measure of a static musical phenomenon that describes only a very limited aspect of music.[50] Yet limitations of pitch have been advanced as inhibiting the entire expressive content of music.

To draw an obvious comparison, language has several hundred thousand words composed of any of the twenty-six letters comprising the alphabet; music has only thirteen [sic] tones. . . . The limit of musical expression . . . lies in the thirteen tones, their octaves, and their variations.[51]

It is hard to imagine a more naïve and inaccurate characterization of music.[52] Unfortunately, although the statement did not originate with Shafter, Shafter's treatise was accorded some respect, and this statement gained credibility. It appeals to defendants seeking to prove that similarities are coincidental. The statement has been adopted in numerous commentaries and court opinions and comprises part of the folklore of infringement litigation.[53]

Anyone who has listened to a reasonable amount of music should recognize the fallacy of Shafter's equation: musical expression is enormously subtle and varied.[54] Music does not yield the coarse, blunt statements that are suggested by Shafter's reference to a vocabulary limited to twelve tones. Other factors expand the vocabulary of music geometrically, which suggests that pitch is not the central defining element.[55] If pitch so inadequately defines the phenomena of music, then the analysis of pitch alone can reveal very little regarding substantial similarities.[56] By the same token, analysis of pitch alone will yield many similarities attributable only to its unduly narrow approach.[57]

Yet many analyses given as evidence in copyright litigation focus on the number of pitches that the two works have in common. Several factors in this type of analysis mislead the trier of fact. First, a number of pitches in common seems to imply a mere sharing of an unordered set of pitches. The analyst usually refers to a succession of pitches common to both works, something that begins to take motion into account. This qualification, where it exists, is not always apparent. Second, the analyst normally points out similarity of pitch only when it appears in a similar temporal context, but, without more, this does little to refute the possibility of coincidence. Because pitch is a static measure, the significance of similarities of pitch must depend on other factors. Unfortunately, some experts limit their analysis to similarities of pitch, an approach that is wholly inadequate. In such cases, the limits of terminology combine with an undue fixation on the parameter of pitch to produce testimony regarding apparent but musically meaningless similarities.

Most listeners cannot identify pitch as an isolated phenomenon.[58] Students of music learn to identify pitch relationships. Those with "perfect pitch" usually learn that "relative pitch" is a more efficient method of analyzing what one hears.[59] Music is determined by its relations, not its static absolutes. For this reason, music sounds essentially the same when transposed to a different key. The absolute value of pitch is changed, but the relational values are maintained. Music analysis properly examines those relationships and leaves most absolutes to the study of acoustics.[60]

Finally, a distinction should be made between "pitch" and "note." In some instances, the similarities are couched in terms of notes, which would seem to be a refinement of the term "pitch" but is not. True, a note generally designates both pitch and relative duration, thus adding a defining element beyond that of pitch alone.[61] But the term "note" adds confusion, because in idiomatic usage the durational qualification often is not intended. The term is used so commonly as a synonym for pitch that the other factors it designates may be forgotten. Taking into account the full implications of "note," Figure 19 demonstrates that the two works in Baxter share only four notes. Yet even the defense conceded that six "notes" were the same.[62]

Example 19: "Joy" v. "E.T."

More important, notes are not musical events; they are mere symbols of musical events. The expert is called to analyze the music itself, how it is heard, not the manner in which it was transcribed to paper. Unless notation in a particular case can be shown to have relevance beyond the music it represents, the court should not indulge a hypertechnical approach to similarity of notes.

Rhythm

The subject of rhythm tends to produce some of the least accurate definitions of musical parameters:

Q Now, rhythm does not indicate the sound of the note, but how long or the duration that the note will be played for, is that correct?

A That's correct.[63]

Here, rhythm is inaccurately equated with duration.[64] In the same case, the court defined rhythm differently in the jury instructions:

The music of a song consists of rhythm, harmony and melody. Rhythm is simply the tempo in which the music is written.[65]

Tempo denotes pace,[66] the speed at which the regular pulse[67] proceeds. Rhythm is not tempo, nor can it be defined in relation to tempo.

Plaintiff's expert in Gaste came closer to a coherent definition of rhythm: "Rhythm is the organization of time values or duration."[68] He added the element of organization. Harold Barlow improved on this definition somewhat: "Rhythm has to do with the duration quality involved in music, being in a relationship of relatively short and long notes."[69] Still, these definitions do not suffice. Rhythm is a highly expressive element of music and, as such, it must be understood in terms of motion rather than as a static or absolute phenomenon.[70]

Because common notation expresses a primary aspect of rhythm in terms of relative duration, one may be tempted to equate duration with rhythm. But duration provides a rather ineffective measure of rhythm and ignores the more relevant relationships. If one were to accept as given that each tone to be considered in a rhythmic pattern began at the point that the preceding tone ended, then knowing the duration of each might still provide insufficient information to reconstruct the rhythm of the passage. Rhythm involves more than duration‑-more even than durational relationships between tones. It involves organization and emphasis,[71] and its perception is dependent on certain psychological expectations.[72]

The perception of rhythm entails two primary comparisons. First, the listener hears a pattern of sounds in which some are accented and some not. This involves a comparison of one sound with those surrounding it in which relative weight is assessed. Accents stress or emphasize some sounds over others. These stress points may result from a tone's longer duration (agogic accent), from greater intensity (dynamic accent), or from differing pitch (tonic accent).[73] Duration, therefore, does not account for all accents. Second, the listener compares the pattern of accented and non-accented sounds with the psychological expectations generated by meter.[74]

As explained above, meter provides the context in which rhythm is heard.[75] The mind tends to organize regular pulses into patterns, usually of three of four.[76] Even where a mechanical device such as a metronome generates pulses with no variation, the mind will supply perceived accents in order to organize the pulses into a regular pattern. Once established, the pattern tends to remain fixed in the mind in spite of transitory contradictions. If this regular pattern is changed, the mind will require several repetitions before it accepts the new grouping. Accented beats in meter are those marked for consciousness by the listener; they are psychological. Rhythmic accents result primarily from physical forces.[77] The unity and disunity of the two make rhythmic patterns interesting and determine how they are heard.[78]

The two themes at issue in Baxter illustrate one simple way in which rhythm relates to meter. The rhythmic motive of "E.T." repeats three times within the eight-measure phrase; the last two measures of the phrase bear strong resemblance to the rhythmic motive. Rhythm remains in the same relation to meter, focusing interest on the melodic sequence. The "Joy" B theme, however, presents a simple rhythmic variation. The second two measures restate melodic material from the first two in a different relation to the meter (Figure 20). The internal relationships of rhythm, the value of the notes, remains the same as does the pitch of those notes. But the entire figure has been shifted to a different metrical position.[79] Winter brought out this feature when he rebutted Williams' testimony that the 4/4 meter of "Joy" was unsuitable to Williams' purpose of conveying flying. Williams considered 4/4 to have too much "bounce."[80] Winter testified that the rhythm of "Joy" obscured the regular 4/4 pattern, because rhythm was stated in two different relationships to meter.[81]

The relationship of rhythm and meter also explains why the meter of a work can be changed without rendering the piece unrecognizable. The "Star-Spangled Banner" has been taken out of its original context of triple meter and transformed into duple meter as a march (Figure 21). The listener still recognizes the melody, although the durational value of many notes has been altered. The important factor is that the rhythmic and metrical accents retain the same relationships.

Example 20: "Joy" B Theme.

Example 21: Bagley, "National Emblem March."

Plaintiff's expert in Gaste v. Kaiserman employed the same process to play "Pour Toi" as a waltz, changing it from duple to triple meter.[82] He showed how meter alone might be relatively inconsequential to perception. But plaintiff's counsel in Gaste switched arguments during the trial. He objected to testimony on an example of prior art, because it had a different meter from the accused work. The objection was correctly overruled.[83] A relevant distinction exists between two works of differing meter, but the distinction does not render comparison of those two works meaningless.

The complexity of rhythm makes its analysis as an isolated parameter potentially useful.[84] An analysis that presents rhythm only as a succession of durational values, however, provides largely isolated phenomena that deserve little weight. Although some court decisions state that originality cannot exist in rhythm alone, that proposition is probably false.[85] Nevertheless, one would probably search popular idioms in vain for an example of protectable rhythm.

Variations of rhythm are neither a mere guise to hide copying[86] nor a total defense. A shared succession of pitches without regard to rhythm should not, in most cases, prove copying. Examples abound in which the same succession of pitches can be found in pieces that sound quite different.[87] The court should not overlook the power of rhythm, thus demonstrated.[88]

The analysis of rhythm as an isolated factor may provide highly relevant evidence. By the same token, omission of rhythm should not render analysis of other parameters irrelevant.

Harmony

Definitions of harmony most often contrast harmony with melody. The definitions posit harmony as representing the vertical aspects of music, its combination of simultaneous tones, whereas melody focuses on the succession of tones.[89] Such definitions only hint at the significance of harmony.[90]

A single combination of tones can be described with some objective certainty as a particular kind of sonority. Chords can be labeled. In this regard, harmony resembles pitch. Chords are merely combinations of pitches that are separable from what surrounds them. But the basis of harmony does not depend on vertical sonorities. Like most other musical parameters, harmony moves from one sonority to another, and it moves purposefully to an objective goal.[91] Progression, much more than content, distinguishes one work's harmony from another.

Traditional western music since c. 1700 and virtually all currently popular forms revolve around the phenomenon of tonality. The tonal system results from a hierarchy of tones, a continuum that stretches from the most stable (tonic) to the most dependent. All tones have a tendency to progress logically through this hierarchy until they reach the tonic. Although tonality strongly influences melodic progression, it is more easily understood and better isolated as a harmonic factor. In virtually all academic settings, the study of music theory begins with the study of tonal harmonic progressions.[92]

Comprehension of harmonic progression requires another reference to psychological expectations. Virtually all listeners can perceive tonic and sense the tendency of music to move toward tonic. In the same way that regular patterns set up a series of expectations concerning rhythm, the movement to tonic provides its own set of expectations. The listener compares what he hears to what he expects. These expectations are so strongly felt and so universally shared that analysis confidently distinguishes progression from retrogression. Isolated chords do not connote these expectations; consequently, the mere labeling of chords describes rather than analyzes the music.[93] Analysis examines harmony as it conforms to and deviates from shared, objective expectations.[94]

Tonic is often described as "home." The listener knows intuitively where home is and constantly plots his position in relation to home. He expects to return home following the logical path. Every chord that deviates from the expected progression tends to alter the listener's expectations of how he will return. The unexpected change may be transitory, and the listener will be relieved to find himself back on course. On other hand, the change of direction may be so purposeful that the listener is convinced to adopt a new home, what theorists refer to as modulation to a new tonic. If the harmonies wander aimlessly, the listener will become disoriented.

The analyst should be wary of positing that substantial similarities exist by virtue of most chords in a particular progression being the same.[95] A few differences may profoundly alter the listener's perception of what follows.[96] On the other hand, numerous differences may be nothing more than short detours within the same overall progression. Theorists employ analytical systems that either emphasize a chord's relation to tonic or its relation to its immediate resolution. Both relationships provide useful data, but the designations bear no resemblance. Harmonies may be multi-tiered, containing relationships within relationships, and two legitimate analyses of the same thing may look entirely different.

Ambiguity of harmony, like that illustrated in Figure 22, appears quite frequently.[97] A new key is established by a sufficient contradiction of the old one, something that convinces the listener to relinquish one expectation for another. Perception may be changed gradually by repetition of subtle differences or suddenly by a marked departure from the vocabulary of the old key. A transitory contradiction of the original key may create disagreement among analysts as to which relationships should predominate.

Example 22: Beethoven's Symphony No. 1.

Confusion arose in Gaste concerning the implications of key signature and whether modulation to a new key entailed a change of signature.[98] Lawyers can understand the key signature best as a presumption. If the writer does a competent job, he will use the key signature of the tonal center that predominates; but any key can be established regardless of signature if the writer is willing to rebut the presumptions with accidentals. Transitory changes of key do not require that the basic presumptions be changed. Although the signature usually remains constant within a work (or within a movement of a multi-movement work), composers commonly seek variety through a change of tonal center. Some of these key changes are extended and structurally significant. The decision to change the signature within a work usually depends on an assessment of how the music may be most easily read. Most important, key signatures are notational devices that make reading music easier; they are not heard. Sound determines the tonal center.

Harmonic progression deserves some special attention when analyzing alleged infringement. Court cases suggest that copyright law cannot protect harmony alone.[99] Again, this proposition standing alone seems wrong. One who re-harmonizes an identifiable copyrighted melody without permission will probably be liable. But the full implications of harmony are sufficiently variable to allow original expression. Harmonic progression tends to be a secondary factor in the popular forms of music most commonly litigated, but that does not justify denying it protection in other settings where it may predominate. The example of Figure 23 seems to present a protectable harmonic progression of a rather nondescript melody.[100]

Example 23: Dvorak's "New World" Symphony, 2nd movement.

Melody

Litigation often treats melody as an isolated factor, and experts have received little criticism for isolating melody from other parameters. Unexamined assumptions seem to produce this concentration on melody. The court presumes this parameter to be the predominant one,[101] although, as Figure 23 suggests, the presumption may need to be reassessed in certain cases. At any given point the music may not state a single, predominant melody. Melody might not predominate, and what predominates might not be melody.[102]

Definitions of melody put forward in court are as inadequate as those of other parameters. Expert witness Spielman posits melody as the combination of rhythm and pitch. This common definition finds support in the literature,[103] but it fails to include some of melody's most fundamental characteristics. This work has already shown that rhythm presents far more complex factors than the definitions commonly assume. Melody includes rhythm, but the definition must import rhythm in all its complexity. To say that melody's other factor is pitch suggests that a random assignment of pitches to a rhythmic figure will produce melody. Melody implies more. Just as the combination of pitches did not define the important aspects of harmony, a succession of pitches will not define melody. The pitches sounded by the melody continue to function within the tonal system; they possess the same tendency to move toward tonic. Melody must be explained with reference to the harmonic motion that it implies.[104] Melody, like harmony, adheres to the rules of tonal syntax.

One theorist defines melody as a succession of tones organized into recognizable patterns.[105] The definition, although insufficient, raises a critical question that the expert must answer: What makes the melody recognizable? If melody were merely a succession of pitch and durational values, it would be exceedingly difficult to learn and remember. But the average listener can retain a multitude of melodies in his memory for a lifetime. Indeed, were that not true, the law would have no reason to provide a remedy for infringement.

Listeners remember melodies by making two principal musical associations: the melody is heard in the context of meter and tonality. The listener may hear pitch, contour, interval, duration, and a host of other factors; but these aspects of music are too isolated to form an accurate basis for remembering an entire work. Remembering a melody is not significantly different from remembering a sentence. The succession of words matters less than how those words function within the sentence to give it meaning. As memory fades, one usually retains the meaning of the sentence and forgets the specific words.

Music analysis strives to explain how the listener hears music. So it would be far less valuable to examine a melody's isolated pitches, intervals, and durations than to relate melody to its rhythmic and tonal context.[106] Smaller, isolated features may provide good indicia of copying and may be relevant. The expert should not overlook this possibility and the court should admit evidence on the minutiae. But evidence of this type alone probably cannot prove copying, and one may seriously question whether an accumulation of minutiae proves copying. A copier must take some larger aspect of melody if the similarities are to be considered substantial. The copy should preserve the syntax, and the listener should make the same associations when he hears the two melodies before branding one substantially similar to the other.[107]

The fact that most listeners hear pitch relatively raises the question: relative to what? Theory students often begin by learning melodic intervals, one pitch relative to another pitch. Testimony often appears in which the expert points out similarities of melodic interval; for example, both "Feelings" and "Pour Toi" begin with the downward interval of a perfect fifth. Defendant introduced testimony that Stan Kenton's "Artistry in Rhythm" also began with the interval of a perfect fifth. But whereas "Feelings" and "Pour Toi" both stated the scale degrees 5 down to 1, Kenton's work stated 3 down to 6. The relationship between the two notes is the same; the relationship to tonic is quite different.

The succession of intervallic relationships does not seem to provide a sufficient basis for hearing music as an intelligible whole. Even short passages present numerous intervals that must be tied together if melody is to be discerned. However, the relation of melodic pitch to tonic provides a constant reference that accounts not only for motion but for direction as well. It seems more appropriate, therefore, to analyze melody in terms of tonality (harmonic motion) than in isolated terms of pitch or interval.

An analyst can explain much about melody by breaking it down according to its rhythmic and harmonic motion. Of the two, harmonic motion may provide a more significant basis for comparison. This does not argue that the analysis of either in isolation is improper. This kind of reduction, like all others, should always be designed to facilitate the discovery of how music functions‑-in this case, how rhythm functions within the melody or how the melody conveys tonality.

Influence of Text and Stylistic Idioms

The analysis of a song normally cannot be divorced from the text. Many cases are tried in which the lyrics are not included in the infringement claim. Experts may ignore lyrics in those cases only to find themselves questioned on the subject during cross-examination. If the music was composed before a text was added and no alterations were made in the music to accommodate the text, then a case might be made that text should be ignored. But if the composer incorporates the text as he writes, the music probably reflects textual influences. Most composers find it easier to bend the music to the text than vice versa.[108] Text suggests its own rhythm.[109] Text may also determine to a significant degree the overall structure of the work,[110] the phrasing, the harmony, and treatment of the smallest details.[111] Slight rhythmic differences may appear in the music from one verse to the next. Thus, analysis of the text's form is vital. The expert cannot present a complete analysis without explaining these details unless the parties stipulate that textual influences are immaterial.

Defense expert Irwin Coster testified that the similarity in the first two notes of "Pour Toi" and "Feelings" had been dictated by the text of each.[112] But on cross-examination, he had to admit that he did not know whether the music or lyrics had been written first.[113] He had not examined the lyrics because no infringement of lyrics was claimed,[114] but he had opened the door to this line of questioning. Coster was forced to make a further admission regarding his analysis of the "Pour Toi" chorus. He had analyzed the melodic sequence of the chorus as centering on the first note of each grouping. His opponent claimed the second note of each grouping was more important. Both positions were defensible; Coster's analysis followed good musical logic. But the text of "Pour Toi" ("Tes yeux . . . ta bouche . . . ton corps . . .") placed emphasis on the second note of each grouping.[115] It made Coster's musical analysis seem wrong. To the extent text explains something about the music that is relevant, it should be included in the analysis.

Stylistic idioms may also dictate or influence certain musical parameters. Although idiomatic treatment may not be protectable expression in itself, that fact does not render the idioms irrelevant to the analysis. Pleading style is highly conclusory; it explains nothing.[116] If style forms a basis of what the expert seeks to prove, then he should identify which elements define the style and explain how they influence other parameters. Stylistic formulae do not preclude originality; they merely stipulate a certain set of commonplace devices and an emphasis on certain parameters. The composer must add original expression in a parameter that remains relatively free of the stylistic dictates, and the expert must separate the original from the commonplace. For this reason, an expert whose knowledge is limited to a particular style may be able to provide only half the equation. He can explain only what is stylistic and what is not; he cannot compare one work to another.

Conclusion

The isolation of parameters provides a useful method of examining the function of various aspects of music. Knowledge gained from this kind of analysis enables the expert to make more relevant comparisons. Objections to this kind of reduction should be overruled; the knowledge obtained is relevant to discovering the significant features of the work examined. However, reductions made purely for the purpose of comparison to a second work should be disallowed. The distinction, though not apparent in some cases, remains an important one. The expert should be able to demonstrate the relevance of the reduction to the individual work before using it to compare that work to another.

There is no virtue in reduction as such; only in the kind of reduction that makes something intelligible to you that wasn't otherwise.[117]

Whereas the validity of segmentations can be tested by requiring all segments to be catalogued, isolation of every parameter does not reveal trustworthiness. Parameters overlap. The sum is greater than its parts.[118] The validity of this kind of reduction must be tested by how well it explains some internal function of the work analyzed.

Hierarchical Reduction

A third type of reduction involves the simplification of music, a stripping away of elaborations to reveal the more functional aspects lying beneath the surface. This technique presents a more complex problem. It requires, perhaps, the greatest musical judgment and confounds the musical layman more than segmentation and isolation of parameter. Compounding the problem, it relies on criteria less likely to lend themselves to visually apparent explanations, such as rests separating phrases, or to roughly definable terminology, as in the isolation of parameters. Partly for these reasons, it does not appear as a regular analytical device in courtroom testimony.

Yet "an analysis that does not simplify the music is a complete waste of time."[119] Analysis must reveal function, and hierarchical reductions enable the analyst to present functional aspects of music unadorned. Proper segmentation cannot be achieved without an understanding of function, and so the analyst will perform certain hierarchical reductions in his own mind before attempting to make temporal divisions. To do otherwise would be unmusical and pointless.

Notation itself, the printed page on which most music is preserved and sold, represents a gross reduction and requires much re-assimilation on the part of the performer.[120] Notational simplifications are accepted in the courtroom largely without question. But a line exists that seems to bar the witness from going further. An unspoken rule suggests that all simplifications done in order to reach the printed score introduced as an exhibit at trial are to be ignored as though they had never happened. Yet no reductions beyond that of the score can be tolerated.

The terminology commonly used by witnesses and counsel implicitly supports this fiction; their discussions often revolve around "notes." Notes, as explained above,[121] are mere symbols of musical events, the result of serious generalizations, classifications, and simplifications. The notated page may represent choices made in the most casual fashion regarding the extent to which the music will be reduced; these decisions may have been made by the composer, publisher, arranger or other transcriber. In general, the fewer features stipulated by the writing, the more freedom the performer will enjoy in reconstructing the work. But in all cases, the music must be reconstructed, because all of it cannot be captured in print.[122]

The writer usually has the performer in mind. But a good performer reduces the page further in his own mind in order to understand the music's function and unity, and he adds much to his performance that did not appear on the page. Even if the transcriber were to be credited with making the most scrupulously precise decisions regarding the extent of notation, those decisions would have been made for the benefit of the performer, not the analyst. Like the performer, but in a kind of mirror image, the analyst uses the notated page merely as a starting point‑-a point with virtually no relevance to his purposes. He adds innumerable factors when he looks at the page and hears the music in his imagination, and he reduces and classifies the music in order to reach the most complete understanding of how it works. Analysis is an aural process, not a visual one. Because the analyst's reductions have a purpose different from the transcriber's, there is no reason for the court to attach significance to the latter. The analyst must reject what is written merely for the performer's benefit and employ the reductions and simplifications that make analytical sense.

If, as has been demonstrated, the notated page represents a somewhat arbitrary stopping point on a long continuum of reductions and simplifications, then it should not have the control that it does. The analyst should be free to simplify the written music further. Of course, the music's fixation in a tangible medium of expression determines what is protected.[123] But this is not the inquiry here. Although the extent of fixation determines the scope of protection, it does not reveal the nature of the thing protected. Analysis involves explaining and comparing the two pieces of music in a search for similarities and differences. Whether those similarities are found in protectable elements goes to the separate question of misappropriation. Further, appeals to the criterion of fixation would arise only if the analyst were adding musical elements that previously had not been captured in tangible form. Reductions confront the court with the opposite situation: there should be no objection to further reductions of the fixed score, because the more abstract materials revealed through further reductions have been fixed no less securely than the surface materials printed on the original score.

In circuits following the "total concept and feel" test, counsel might object that reductions are irrelevant. But in the Ninth Circuit, the total concept and feel standard applies only to the extrinsic test for similarities of expression. Experts may testify regarding the intrinsic test for similarities of idea, and, during trial, the inquiries into similarity of expression and similarity of idea will usually be before the court simultaneously. Therefore, unless the inquiry before the court concerns the extrinsic test alone, an objection to analytical reductions as being irrelevant should fail. The only other possibility seems to be for the objecting party to assert that the reductions are designed to uncover similarities of expression rather than idea. But one of the many flaws in the Ninth Circuit intrinsic-extrinsic bifurcation is the inability to define the distinction between idea and expression. The two are particularly interdependent in music. The court would be wrong to sustain an objection to hierarchical reductions on the grounds of relevance, because, as will be explained below, this very type of reduction is the surest route to uncovering the abstract "idea" in music.[124]

The Necessity of Background Analysis

One cannot hope to understand something by examining only its facade. Appearances are deceiving in music just as in any other subject; therefore, a competent analysis must pierce the surface of the music and explain its inner workings. Heinrich Schenker introduced his analytical methodology by stating that the art of music is much simpler than it appears, but its simplicity does not lie on the surface. "Every surface, seen for itself alone, is of necessity confusing and always complex."[125] The analyst must find reasons for the surface manifestations.

"Analyzing surface features is pointless unless you do so in terms of the background structure."[126] This statement concedes more than it should. In fact, examining surface features alone cannot rise above mere description to the level of analysis. The expert who refuses to reduce the music to its fundamental structure in order to explain why the surface appears as it does provides no help to the trier of fact. His purported analysis simply points out what is obvious and applies a label to it. Such an expert testifies merely as a seasoned observer of music, not as an analyst.

Cook argues that analysis must find a background structure in relation to which the musical surface can be viewed as a kind of elaboration.[127]

[T]he musical surface acquires it significance through the elaboration of underlying patterns‑-through function, in other words. If one were to use such techniques for analyzing musical function as the basis for comparisons between different pieces of music‑-whether from one culture or several‑-then the conclusions might have a degree of musical significance which can hardly be attained merely by evaluations of surface features, however comprehensive and painstaking.[128]

The most relevant objection to surface descriptions of the music lies in the fact that they say little of significance about the compositional process. Surface similarities are always subject to the defense of coincidence, and that defense often provides an adequate and credible explanation. The process can easily degenerate into the rather useless exercise of counting notes and arguing over how many is too many.[129] The expert should offer more than conclusory characterizations of the similarities. He should point out not only where the works are similar but how they are similar, and he should be able to explain in understandable language why those similarities are or are not substantial. These goals can be accomplished only by moving beyond the surface to the true substance of the music and by examining how form and function control the compositional process. Surface analysis alone too often produces facile and naive results.

Because the court has no history of admitting this kind of reduction, witnesses understandably hesitate to be the first to introduce it. This common shortcoming of infringement suits tends to keep both plaintiff's and defendant's analyses closely tied to surface features. As a general proposition then, the party whose case is best argued by similarities or differences at the surface enjoys a decided advantage; the party whose case requires the least analysis benefits. In effect, the party who can present the two works as written and say that the pieces speak for themselves‑-they are similar where they are similar‑-presents what appears to be unimpeachable truth.

Few counter-measures effectively offset the impression that the witness has engaged in deliberate obfuscation by removing notes. He may cite the writings of prominent theorists to bolster the necessity of these reductions. The judge may empathize more than the jury; the judge has been educated in a field that requires analysis of abstract principles. But the judge may also know from experience that an expert can be found to testify on most any point, no matter how tenuous its grounding in truth. Consequently, the judge might improperly exclude this evidence, or, if the judge admits it, the trier of fact might hold it suspect. The solution may have to await judicial recognition of the true nature of music analysis and the regular issuance of instructions to the jury on what analysis entails. The first judicial formulation will likely survive as common law for a long time; if the court doesn't get it exactly right, what now is merely difficult may become impossible.

The author suggests the following as a model jury instruction and as a guideline for bench trials:

The law calls for dissection and analysis on the issue of copying. Dissection involves dividing music into its smaller component parts and reducing music to reveal its more functional aspects. This dissection may be accomplished by (1) temporal segmentation, such as focusing on a theme within the whole piece, a phrase within the theme, a motive within the phrase, etc.; (2) isolation of a parameter or group of parameters, such as the omission of harmony or rhythm, in order to focus on other parameters; (3) hierarchical reduction, such as the elimination of ornamental features and the gradual stripping away of small figurations to reveal overall function, purpose, and design; or (4) the likely combination of any or all of the above.

Hierarchies in Music

The discussion above regarding harmonic progressions and melody demonstrates that music is not two-dimensional. Critical relationships are heard in numerous parameters, in both transitory features and in music's larger aspects. The analyst must account for these different layers of musical context, because associations and meanings at every layer determine the response of all listeners, average and professional.

If one starts with the concept of embellishments in music, hierarchical reduction (the elimination or illumination of a particular layer) becomes easier to comprehend. The Baxter case focused on the eighth-note figure F E D E that was virtually identical between the two works at issue. The plaintiff characterized the figure as essential; the defense called it a "turn"‑-an embellishment common in the baroque and classical periods. There also appeared in the relevant portion of plaintiff's "Joy" two grace notes, which appear in Figure 24 in small type. Plaintiff argued that these grace notes were inconsequential; defense expert Barlow disagreed. As the name suggests, grace notes ornament the music‑-they embellish and refine it.[130] Both plaintiff and defendant argued that one type of embellishment mattered and that the other did not. The debate was entirely result-oriented.

Example 24: "Joy" B Theme.

The history of musical ornamentation is far too complex to guide the discussion here. In the baroque period some ornaments were notated, some were not. The composer expected the performer to add certain ornaments, and performers sometimes added more than was anticipated. Seemingly contradicting the term, some ornaments were functional. For example, plucked instruments, such as harpsichord and lute, cannot sustain tones like wind instruments, bowed instruments, and voice. Ornamentation provided a means of conveying sustained tones on these instruments.[131] Ornaments are correctly understood as "diminutions"‑-something made smaller or, more particularly, the breaking up of the notes of a melody into quick figures.[132] Diminution refers to the process by which intervals formed by long notes are expressed by notes of shorter value.[133]

The English-language equivalent of the term diminutions is divisions. From an English treatise dating from 1659, we see that the tradition of systematic and explicit teaching of diminution technique is indeed a long and honorable one. The process with which Simpson's treatise is concerned is "breaking the ground," which he describes as follows: "Breaking the ground (a melodic bass figure) is the dividing its notes into more diminute Notes. As for instance, a Semibreve may be broken into two Minims, foure Crotchets, eight Quavers, sixteen Semiquavers, etc."[134]

Just as embellishments are diminutions of longer notes, those longer notes are themselves diminutions of more basic structures. Music, therefore, reveals numerous hierarchies as one moves from surface diminutions to more abstract background features:

The tradition of embellishment manuals, stretching from Ganassi (Opera intitulate Fontegara, 1535) to Virgiliano (Il dolcimelo, c. 1600) and then on to the 17th-century vocal and instrumental tutors, was primarily concerned with teaching graces and passaggi to performers. This was done by means of tables of ornaments, extended practical examples and formulated rules. In these manuals is established the fundamental concept of "diminution." This concept has two aspects to it: (1) the subdivision of a few long note values into many shorter values; and (2) the application to an "essential" melodic line of a layer of less essential linear material. In both aspects, a hierarchy is created, and in both the possibility exists of the hierarchy becoming multi-layered as an already embellished line is subjected to further embellishment. On the face of it this was the purely transient affair of the virtuoso performer. In reality much 16th-century music contained elements of embellishment as it was written down; and the modern style of 17th-century seconda prattica subsumed ornamentation within its notated exterior. The compositional notion of invention (or adopting) a basic structure and then elaborating it, which goes back at least to the 9th century and was developed as contrapunctus diminutus by 14th-century theorists, was crystallized in this instructional tradition and was absorbed deep into European musical consciousness. . . . This tradition was to be of incalculable importance to the theories of Heinrich Schenker at the beginning of the 20th century.[135]

Schenker posited that music composition involves a series of diminutions, a process of adding detail that mirrors the analytical process of stripping it away.

Schenkerian analysis is in fact a kind of metaphor according to which a composition is seen as the large-scale embellishment of a simple underlying harmonic progression, or even as a massively-expanded cadence; a metaphor according to which the same analytical principles that apply to cadences in strict counterpoint can be applied, mutatis mutandis, to the large-scale harmonic structures of complete pieces.[136]

Figure 25 shows the process of hierarchical reduction that moves from foreground through middleground to background. Brahms wrote variations on this theme by Handel, all admittedly derived from the theme. The music analyst will want to know the many different ways Brahms took material from the theme to construct the variations. In order to discover Brahms' compositional process, the analyst cannot rely on surface features alone. Looking for the same succession of notes in the melody yields little useful information. Yet the listener need not be a skilled analyst in order to hear all twenty-five of Brahms' variations as a coherent work; the mind comprehends the similarities.

Example 25: Hierarchical reductions of Brahms' Variations on a Theme by Handel (Theme).

Example 26: Hierarchical reductions of Brahms' Variations on a Theme by Handel, Var. III.

Variation III appears quite dissimilar on the surface. Surface differences between Handel's theme and Brahms' Variation III (Figure 26) are striking, but the variation expresses a functional structure that is almost identical. Schenkerian analysis reveals those similarities that the layman may perceive and be unable to articulate. It explains why some pieces sound alike even though they appear quite different. By the same token, two works may contain numerous surface similarities but yield a different functional structure. Schenkerian techniques will explain why such works may yet sound significantly similar.

Although each hierarchical level dispenses with non-essential notes and designates some notes as more important than others, Schenkerian techniques offer fewer opportunities for tailoring an analysis to suit the argument than the layman might suspect. Established principles have been thoroughly documented. Most importantly, all expressions ultimately derive from the organic unit of the tonic triad. This organic unit of tonality accounts for what the court calls limited permutations of notes that are pleasing to the ear.[137] The factors that make those permutations pleasing have a basis in music, a background progression, that Schenker labels its fundamental structure (Ursatz).[138] The analyst must adhere to sound musical principles if his analysis is finally to yield this fundamental structure. Arbitrary choices lead to dead ends. Further, the same principles of harmony and counterpoint that apply to the foreground in traditional analysis also apply to every level of the reduction. Choices that reflect poor analysis remain subject to cross-examination to the same extent as those entailed in other methodologies.

Hierarchical reductions can be tested more easily than segmentations. The same rules of completion should be applied. The analyst should account for at least three hierarchical layers and trace the work to its fundamental structure.[139] This procedure will yield a most reliable indicator of trustworthiness. If the essential elements are removed and the unessential retained, the analysis will become incoherent as it moves toward the background. Figure 27 shows the reductions culminating in the deep background (Ursatz) on the top line.

Example 27: Schenker's analysis of "Ich bin's, ich sollte büssen by J.S. Bach.

Some music theorists have criticized Schenkerian analysis as being irrelevant to non-tonal 20th-century music. The corollary to this flaw is that Schenker's techniques are exceptionally well-suited to the analysis of tonal music. A Schenkerian analysis provides a clearer picture of the tonal implications of melody than other methodologies. It shows how the small- and large-scale relationships combine to form an overall impression of the work.

Variations on Schenker's methodology accomplish the same thing. Leonard Meyer developed similar techniques for analyzing hierarchical structures in music.[140] Benjamin Boretz expanded Schenkerian analysis in order to apply its techniques to non-tonal music.[141] Much of the scholarship in music theory has for the past twenty years explored the possibilities of Schenkerian and neo-Schenkerian analysis. Schenker currently holds the predominant position in this kind of music analysis, but experts may want to employ similar techniques that they find more understandable to the lay trier of fact. Whether or not he uses Schenkerian techniques, the analyst must present something that accomplishes the same thing. Even Schenker's critics do not argue that he set out to do something improper, only that this essential aspect of analysis can be accomplished through better means.[142]

Earl Spielman included what he calls "macro analysis" in his pre-trial report for the Overstreet case.[143] Spielman, acting as expert for the plaintiff, reduced the melodies of numerous examples of prior art that had been submitted by defendant. He introduced the process as follows:

Macro Analysis of predominant pitches avoids undue attention to insignificant details, passing tones, upbeats, etc., and allows the observer to concentrate on just the primary tonal considerations. The underlying premise of Macro Analysis is that similarities in the predominant pitches of two works is a more accurate reflection of the overall similarity of the works, even if additional aspects of pitch concurrence occur in more detail elsewhere.[144]

Spielman's reductions reflect the purpose and general techniques of a Schenkerian analysis; his analysis reveals the middleground of the prior art. Spielman has not introduced this type of evidence in his courtroom testimony, because the choices made in the reductions appear subjective. However, had Spielman represented one more level between the foreground and his middleground and carried the process through to deep background (something he necessarily thought but did not write), the basis of his choices would be identified. He could defend those choices according to musical logic and withstand cross-examination.

A small effort was made in this direction in Baxter when Spielman introduced a comparative chart showing how the examples of prior art revolved around a C major triad.[145] This rather tentative approach not only pointed out the relatively obvious similarities between "E.T." and "Star Wars" but the relevant and less obvious similarities in the "Dallas" theme.

Conclusion

The importance of this analytical approach to the discovery of infringement lies in the purposeful revelation of musical functions at all levels. Evidence relevant to copying will naturally appear near the surface; the question of its substantiality requires reference to the context in which the similarities appear‑-some short distance into the background.[146] Formal segmentation provides too blunt an instrument to accomplish the subtle refinements of analysis that are likely to reveal the best evidence concerning copying. Formal segmentation tends to explain similarities of device, and such similarities by themselves cannot demonstrate copying.

Schenker's revelations are better understood not as reductions but as elaborations.[147] For this reason, the various layers tend to reveal the composition process itself. No better evidence can be found to show what the composer did and how he did it. Whether the composer actually followed this process consciously does not matter. The technique reveals the music in both its structural and psychological aspects and provides the lay trier of fact with an indispensable guide to the context in which similarities appear.

---

[1] Nicholas Cook, A Guide to Musical Analysis (New York: George Braziller, 1987), 16.

[2] Heinrich Schenker, Free Composition, trans. and ed. Ernst Oster (New York: Longman, 1979), xxiii.

[3] Plaintiff's expert Winter developed his "six criteria" for comparison based on the similarities between the two works, whereas defense expert Spielman's five criteria were constants. Baxter v. MCA, Inc., 812 F.2d 421 (9th Cir.), cert. denied, 484 U.S. 954 (1987), on remand, Case No. 88-6660 (C.D. Cal. 1988), aff'd, 907 F.2d 154 (9th Cir. 1990) [unpublished opinion], Record at 182 and 881-83.

[4] "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." Baxter, Record at 186 (Winter testifying).

[5] See, e.g., Baxter, Record at 881-83 (Spielman testifying). However, Winter used criteria dictated by the similarities between the two works. Record at 182.

[6] Defense expert Coster testified that he does exactly the same kind of analysis in all cases. He makes a chart of both songs and goes bar for bar, beat for beat. "That is the accepted way of analyzing all over the world." Gaste v. Kaiserman, Case No. 86 Civ. 5671 (S.D.N.Y. 1986), aff'd, 863 F.2d 1061 (2d Cir. 1988), Record at 680-81. Barlow testified that his method of analysis is "generally accepted." Ensign Music Corp. v. Avis, Inc., Case No. 80 Civ. 7346 (KTD) (S.D.N.Y. 1980), Record at 18. Parsons testified that, although he had never done a comparative analysis before, the process would not differ for popular and classical music. Selle v. Gibb, 567 F. Supp. 1173 (N.D. Ill. 1983), aff'd, 741 F.2d 896 (7th Cir. 1984), Record at 198-99.

[7] Spielman pointed out that there are certain standard procedures but no standard methodology in music analysis. Tree Publishing Co., Inc. v. Overstreet, Case. No. 3:87-0032 (M.D. Tenn. 1987), Defendant's Deposition of Spielman, at 13.

[8] Form and style influence the course of analysis, suggesting the relative utility of certain methodologies. One analytical method is not enough. If your only tool is a hammer, every problem becomes a nail. Michael R. Rogers, Teaching Approaches in Music Theory (Carbondale: Southern Illinois University Press, 1984), 28 (citing Andrew Maslow).

[9] "The ear moves beyond surface similarities to hear more subtle distinctions. It is in this realm of subtle distinctions that composers work." Keyt, An Improved Framework for Music Plagiarism Litigation, 76 Calif. L. Rev. 421, 428 (1988).

[10] See, e.g., Arnstein v. Porter, 154 F.2d 464, 476 (2d Cir. 1946); Dorchester Music Corp. v. National Broadcasting Co., 171 F. Supp. 580, 586 (S.D. Cal. 1959).

[11] Selle v. Gibb, 567 F. Supp. 1173 (N.D. Ill. 1983), aff'd, 741 F.2d 896 (7th Cir. 1984), Record at 77.

[12] Different parameters of music may function ambiguously or even at cross purposes. See discussion, infra 293.

[13] Games Magazine, January, 1984.

[14] See Grosvenor W. Cooper and Leonard B. Meyer, The Rhythmic Structure of Music (Chicago: University of Chicago Press, 1960), 6.

[15] Of course, bar lines may coincide with the proper divisions.

[16] Tree Publishing, Inc. v. Overstreet, No. 3:87-0032 (M.D. Tenn. 1987), Deposition of Spielman, at 197-99.

[17] Gaste v. Kaiserman, Case No. 86 Civ. 5671 (S.D.N.Y. 1986), aff'd, 863 F.2d 1061 (2d Cir. 1988), Record at 490 (Question by plaintiff's counsel Bart to plaintiff's expert Ricigliano).

[18] See Harold R. Laycock and Quentin R. Nordgren, First-Year Music Theory (New York: Appleton-Century-Crofts, 1962), 17; Ralph Turek, The Elements of Music: Concepts and Applications (New York: Knopf, 1988), 105.

[19] Motive is "a term used in a variety of senses in thematic and phrase-structure analysis. It usually denotes a short utterance that retains its identity as a musical idea as it often `appears in a characteristic and impressive manner at the beginning of a piece' (Schoenberg). It is usually thought of in melodic terms, and it is this conception that is sometimes rendered as `figure.' Another term often encountered in the context of the melodic motif is `cell,' which connotes a still smaller unit of melody, as in the opening theme of Beethoven's Piano Sonata in E Major, op. 109." Ian Bent, Analysis (New York: Norton, 1987), 122.

[20] Turek, supra note 18, at 105. See also New Harvard Dictionary of Music (1986), s.v. "Motive."

[21] Anheuser-Busch, Inc. v. Elsmere Music, Inc., 633 F. Supp. 487 (S.D.N.Y. 1986), seems to provide an example. The case involved the Budweiser jingles. In the original version, Steve Karmen's familiar music accompanied "For all you do, this Bud's for you." Although Budweiser had made that figure significant by repeatedly broadcasting it, the final descending figure (the melodic progression 6‑4‑2‑1) had not generated nor been derived from other material of the complaining work. Familiarity with the original Budweiser theme resulted from its use as a musical trademark, not as a motive. Karmen alleged similarities in a subsequent advertising campaign in which Tom Anthony's music accompanied the slogan "You make America work, and this Bud's for you." The alleged similarities were confined to the last four notes of the of accused work and relied on its similar placement as a tag ending. Record at 198-99. See Pre-Trial Analysis of Spielman, Appendix G, at . The defense depended in part on limiting the inquiry to the last four notes. The prominence of the musical trademark raised interesting questions regarding whether such a short figure might be protectable and whether the figure's prominence might in itself make any similarities substantial. Karmen's expert, Abba Bogin, suggested a cumulative effect: if 75% of everything is the same, then one begins to hear all as similar. Record at 199; Appendix G, at 1248. Earl Spielman, testifying for Anthony, analyzed both works in their entirety in order to show that Karmen's ending stated new material (a mere figure) while Anthony's was based on repeated material (motives which had generated a quite different work). Record at 668-71. The court determined that the works did not exhibit substantial similarities.

This result can be contrasted to that of Bright Tunes Music Corp. v. Harrisongs Music, Ltd., 420 F. Supp. 177 (S.D.N.Y. 1976), aff'd, ABKCO Music, Inc. v Harrisongs Music, Ltd., 722 F.2d 988 (2d Cir. 1983), in which George Harrison's "My Sweet Lord" was held to have infringed the Chiffons' "He's So Fine." Each song states two key motives that repeat with slight variations and embellishments. 420 F. Supp. at 178. Virtually no other material appears. The paucity of material hurt Harrison's case. Once similarities were established between the motives, they had to be considered substantial. See Pre-Trial Analysis of Greitzer, Appendix E, at 1213.

[22] Rudolph Reti, The Thematic Process in Music (New York: Macmillan, 1951).

[23] See Laycock and Nordgren, supra note 18, at 18.

[24] Turek, supra note 18, at 107.

[25] See Laycock and Nordgren, supra note 18, at 18.

[26] A typical musical rhyme involves parallel beginnings of phrases that progress to different cadences.

[27] A period consists of two related phrases, the second coming to a stronger close than the first.

[28] A sequence is the reiteration of a figure at a different pitch level.

[29] The chord in measure 4 is not B minor as the piano arrangement indicates, but a G major 7. Measure 3 of the theme should be understood as a secondary dominant (D7 or V7 in G) implying a transitory key change to G major. Further, G is the root of the chord in measure 4, a note not accounted for in the B minor designation.

[30] Sequence is such a strong device that its contrapuntal aspects tend to obscure everything except the fact of repetition and the motion outlined by the beginning point of each repetition.

[31] Record at 116-17. The judge instructed the jury: "The idea or thought involved in selecting the use of repetition or sequence to design a song is not protected by copyright. There can be no copyright on an idea." Record at 1078. The judge incorrectly determined that sequence must be idea rather than expression. This error points out the tendency to treat any aspect of music that can be defined as non-expressive. See also, Pre-trial report of Barlow, Selle v. Gibb, § 123 at 47: "Whatever the creative status of the Motif of "Coda," the underlined numbers (notes) in the Sequential Repetition are not the product of additional creative effort."

[32] Mazer v. Stein, 347 U.S. 201 (1954) (Copyright protects originality, rather than novelty or invention, and confers only the sole right of multiplying copies, and in absence of copying, there can be no infringement of copyright). "`Original' in reference to a copyright work means that the particular work `owes its origin' to the `author.' No large measure of novelty is necessary. . . . No matter how poor artistically the `author's' addition, it is enough if it be his own." Alfred Bell & Co. v. Catalda Fine Arts, Inc., 74 F. Supp. 973, 102-03 (S.D.N.Y. 1947), aff'd, 191 F.2d 99 (2d Cir. 1951).

[33] Judge Learned Hand observed, "True, it is the themes which catch the popular fancy, but their invention is not where musical genius lies, as is apparent in the work of all the great masters." Arnstein v. Edward B. Marks Music Corp., 82 F.2d 275, 277 (2d Cir. 1936).

[34] Fed. R. Evid. 403.

[35] This phenomenon is somewhat like the mechanic who, upon reassembling the machine, has parts left over.

[36] See, e.g., Baxter, Spielman Pre-Trial Report, Chart 1, (Appendix C, at 889).

[37] Examples include such common devices as syncopation (a contradiction between the underlying pulse and surface rhythm) and hemiola (groupings that contradict the established metrical grouping, such as groupings of two in triple meter).

[38] Subtlety of meter does not facilitate getting masses of people to march in step or dance to the beat.

[39] "If the Guidelines have accomplished anything thus far, they should have instilled a settled habit of regarding music first as a process of growth, then attempting to understand this growth by an analysis that fully reflects the character of musical flow. The fluidity of music must always be perceived as a stream (whether a spring, brook, or river) with tributaries of many sizes from many directions, never as a scattering of ponds and lakes, or worse still, as buckets and boxes into which a composer pours his thoughts." Jan LaRue, Guidelines for Style Analysis (New York: Norton, 1970), 115-16.

[40] Leonard B. Meyer, Emotion and Meaning in Music (Chicago: University of Chicago Press, 1956), 14. Meyer defined three stages of perception: the "hypothetical meaning" before a sound-pattern has been heard, the "evident meaning" when the sound-pattern becomes a concrete event, and "determinate meaning" that arises out of the total experience. Bent, supra note 19, at 59-60. Eugene Narmour developed this notion into an "implication-realization" model. See Eugene Narmour, Beyond Schenkerism (Chicago: University of Chicago Press, 1977), 136 ff. "Musical formations carry certain implications as to how they will continue. Those implications are manifold; they relate to the way things have gone in the work so far (`idiostructure'), but also to the stylistic context of the work (`style structure'). . . . Each parameter of a musical event carries its own separate implications. The implications in play at any one moment are bound to be to some extent mutually contradictory, hence no more than part-realization is possible. That which is unrealized remains potential. Realization yields `closure,' non-realization `non-closure.' Bent, supra note 19, at 69.

[41] Webster's New Collegiate Dictionary (1980), s.v. "Parameter."

[42] See New Harvard Dictionary of Music (1986), s.v. "Parameter."

[43] The law hears arguments on many complex subjects and is quite accustomed to seeking out accurate definitions. It is unclear why the district judge in Selle v. Gibb failed to consult an authoritative source before including erroneous definitions in his instructions. See infra note 65.

[44] Charles A. Culver, Musical Acoustics, 4th ed. (New York: McGraw-Hill, 1956), 83-85. Relative intensity of the overtones determines timbre; it makes a clarinet sound different from an oboe.

[45] A pitch class is, for example, the set of all Gs regardless of octave.

[46] Hindemith noted this vast terrain between pitch classes and the ear's attempts to generalize, giving the example of five different types of the interval of a third. Common notation recognizes only two: the major third (e.g., C ‑ E), and the minor third (e.g., C ‑ E‑flat). Hindemith's five types are: major (frequency ratio 4:5), minor (5:6), "too small" (6:7), "too large" (7:9), and one in between (9:11). All of these ratios lies within the first sixteen partials of the overtone series, and sixteen partials are required to define the basic content of the tonal system. The interval of a half step is the ratio 15:16. Fritz Winckel, Music, Sound, and Sensation: A Modern Exposition (Max Hesses Verlag, 1960; repr., New York: Dover, 1967), 124. "[T]he ear usually hears what it wants to hear, even if that does not correspond to the acoustically given interval." Id. at 128.

[47] Culver, supra note 44, at 88-89. Extreme refinements may occur intentionally, such as a piano tuner leaving small deviations among unison strings to give the sound character. Other refinements, such as the regular compression and expansion of certain melodic intervals by performers, may be the result of psychological forces. The ear hears these deviations in frequency but does not consider them significant enough to reclassify pitch. Winckel, supra note 46, at 127-29.

[48] In fact, western notation entails more subtle distinctions than the pitch class designations admit. One must account for sixteen partials (divisions of a vibrating string) of the overtone series in order to define the intervallic content of common notation. There are fifteen major keys (three more than pitch class names) and seventeen pitch names that do not resort to double sharps and double flats. Enharmonic equivalents (e.g., d-sharp equal to e-flat) exist only in even-tempered tuning. Distinctions in tuning systems that require the most refined discriminations of frequency were a common subject of the theoretical writings of ancient Greece.

[49] "Blue notes," for example.

[50] Phenomenologists hold a similar view of pitch. "[P]itch is `transparentized' in a musical context, which is to say that we experience music through the pitch, rather than the pitch itself. More simply, we hear the musical activity of the pitch: it is receding, projecting, emerging, interrupting or being interrupted; changing in tone quality or intensity, glaring, glowing, echoing, etc. The most mysterious thing about a pitch is that it simultaneously presents the experiences of duration and change: the `same' sound constantly renews itself and passes away, even while enduring. In this sense, the phenomenal aspect of pitch is some kind of activity, however minimal, and is therefore involved in the presentation of some kind of time and space, although it may be of a highly undifferentiated kind. We can think of pitch in music as analogous to cadmium red in painting; again, pitch is to music what Paul Scofield is to King Lear, and in general, what a performer is to a performance. Pitch is not even to be confused as the reason for, or cause of, the musical event: it is simply the medium." Thomas Clifton, Music as Heard: A Study in Applied Phenomenology (New Haven: Yale University Press, 1983), 20.

[51] Alfred M. Shafter, Musical Copyright, 2d ed. (Chicago: Callaghan, 1939), 195.

[52] According to Shafter, "[p]ractically every original idea the composer can think of has appeared somewhere before; it is a matter of probabilities, and every day the number of new possibilities grows less." Shafter, supra note 51, at 196. Keyt correctly notes the diametrically opposite view held by most musicians and philosophers: "a seemingly infinite array of possible choices involved in the compositional process." Keyt, supra note 9, at 431 n.50 (citing Cox, Are Musical Works Discovered? 43 J. Aesthetics & Art Criticism 367, 370 (1985).

[53] See, e.g., Arnstein v. Edward B. Marks Music Corp., 82 F.2d 275, 277 (2d Cir. 1936) (The seven notes available do not admit of so many agreeable permutations that we need be amazed at the re-appearance of old themes); Schultz v. Holmes, 264 F.2d 942 (9th Cir. 1959) (The common utilization by different compositions of a few notes such as herein found to exist occurs frequently in the field of popular music, particularly because of the limited number of pleasing tonal combinations); Gaste v. Kaiserman, 863 F.2d 1061, 1068 (2d Cir. 1988) (We are mindful of the limited number of notes and chords available to composers and the resulting fact that common themes frequently reappear in various compositions, especially in popular music); Darrell v. Joe Morris Music Co., 113 F.2d 80, 80 (2d Cir. 1940) (While there are an enormous number of possible permutations of the musical notes of the scale, only a few are pleasing, and much fewer still suit the infantile demands of the popular ear). See also, May, So Long as Time Is Music: When Musical Compositions Are Substantially Similar, 60 S. Cal. L. Rev. 785, 790 (1987); Sherman, Musical Copyright Infringement: The Requirement of Substantial Similarity, 22 Copyright L. Symp. (ASCAP) 81, 124 (1977); Orth, The Use of Expert Witnesses in Musical Infringement Cases, 16 U. Pitt. L. Rev. 232, 234 (1955). In Herald Square Music Co. v. Living Music, Inc., 205 U.S.P.Q. 1241 (S.D.N.Y. 1978), the judge began the trial by asking if the defense wanted to submit a trial memo rebutting that of plaintiff's. Plaintiff's memo, he noted, "equates the plagiarism of musical works with those of written works although in the usual written work there are a thousand or tens of thousands of words, and in the usual musical work there are a couple of dozen notes. If that is an unfair comparison I would think you would want to demonstrate that to me." Record at 2.

[54] The three spatial dimensions do not limit the expressive range of sculpture. Primary colors do not inhibit the painter. Unlike language, music is free of the literal connotations. Music above all the arts can reveal the artist's idea "unobscured by adventitious literal meanings. That is presumably what Walter Pater meant by his much-debated dictum, `All art aspires to the condition of music.'" Susanne K. Langer, Philosophy in a New Key (Cambridge: Harvard University Press, 1942), 257; quoting Walter Pater, The Renaissance: Studies in Art and Poetry (1908; 1st ed. 1873), 140.

[55] "Music is unlike language in crucial respects. For one thing, music has many more parameters." Narmour, supra note 40, at 21. The linguist Roman Jakobson stressed the fact that the rules of music, as of poetry, do not concern etymology or vocabulary. Music analysis, he argued, should focus on phonology (the science of distinguishing between elements in a stream of vocal linguistic sound and the apprehension of the rules by which the sounds are linked together). Bent, supra note 19, at 59.

[56] "What is crucial is relational richness, and such richness (or complexity) is in no way incompatible with simplicity of musical vocabulary and grammar." Leonard B. Meyer, "Grammatical Simplicity and Relational Richness: The Trio of Mozart's G Minor Symphony," Critical Inquiry 2 (1975-1976): 694.

[57] The misperception may derive from Marks v. Leo Feist, Inc., 290 F. 959 (2d Cir. 1923). "Musical signs available for combinations are about 13 in number. They are tones produced by striking in succession the white and black keys as they are found on the keyboard of the piano. It is called the chromatic scale. In a popular song, the composer must write a composition arranging combinations of these tones limited by the range of the ordinary voice and by the skill of the ordinary player. To be successful, it must be a combination of tones that can be played as well as sung by almost anyone. Necessarily, within these limits, there will be found some similarity of tone succession." Id. at 960. This statement does not imply any limitation of music's expressive capacity.

[58] Listeners cannot even identify timbre in isolation. In one experiment a sustained tone played on various instruments was recorded. The "attack," the instruments' initiation of the tone, was edited out of the recording. Observers mistook a tuning fork for a flute, an oboe for a clarinet, a cello for a bassoon, a cornet for a violin, and a French horn for a flute. Winckel, supra note 46, at 34.

[59] Often explained as acute pitch memory, perfect pitch usually refers to the ability to identify the pitch of a given tone without reference to any other tone. It is the ability to remember frequency. One with relative pitch might be able to discern all of the relationships in music without relating them to any absolute value of frequency‑-like understanding an algebraic formula without knowing the value of "x."

[60] See Boretz, "Musical Cosmology," Perspectives in New Music 15 (1977): 128.

[61] See Marvin Thostenson, Fundamentals, Harmony, and Musicianship (Dubuque: Brown, 1963), 5. But the duration designated is merely some fraction (half, quarter, eighth, etc.) of an unknown constant.

[62] Baxter, Record at 836-37.

[63] Selle, Record at 106 (Plaintiff's counsel Engerman questioning Selle).

[64] The same error occurs in Dorchester Music Corp. v. National Broadcasting Co., 171 F. Supp. 580, 585 (S.D. Cal. 1959).

[65] Selle, Record at 1079-80 (jury instructions). The judge may have drawn from Northern Music Corp. v. King Record Distributing Co., 105 F. Supp. 393, 400 (S.D.N.Y. 1952): "Technically analyzed, a musical composition is made up of rhythm, harmony and melody. Originality, if it exists, must be found in one of these. Rhythm is simply the tempo in which the composition is written. It is the background for the melody. There is only a limited amount of tempos; these appear to have been long since exhausted; originality of rhythm is a rarity, if not an impossibility. Harmony is the blending of tones; this is achieved according to rules which have been known for many years. Being in the public domain for so long neither rhythm nor harmony can in itself be the subject of copyright."

[66] "Tempo is the speed of the beat." Turek, supra note 18, at 18.

[67] Pulse marks even divisions of time and the basic durational unit in a piece of music. Turek, supra note 18, at 18.

[68] Gaste, Record at 468 (Plaintiff's expert Ricigliano testifying).

[69] Ensign Music Corp. v. Avis, Inc., Case No. 80 Civ. 7346 (KTD) (S.D.N.Y. 1980), Record at 15-16.

[70] Rhythm is "[i]n its primary sense, the whole feeling of movement in music, with a strong implication of both regularity and differentiation." Harvard Dictionary of Music 729. An admittedly simplified definition: "Music occupies space in time. It consists of a series of impressions made up of combinations of sound and silence following one after the other. The relation of these successive impressions to each other and to the passage of time is a phase of rhythm." Thostenson, supra note 61, at 5.

[71] White defines rhythm as the organization of a group of musical sounds according to the time interval between them, according to accents, and according to their duration. John D. White, Understanding and Enjoying Music (New York: Dodd, Mead, 1968), 17.

[72] "Explanations of Rhythm commonly include only immediate durational aspects and the concept of meter. For purposes of style analysis both of these approaches must be considerably expanded and a further category added to account for rhythmic influences contributed by other elements." LaRue, supra note 39, at 90.

[73] See Turek, supra note 18, at 18.

[74] "Meter is the basic pattern of organization and accent used in each measure." Thostenson, supra note 61, at 6.

[75] See supra page 254.

[76] "Meter is the grouping of pulses into patterns of two or more beats by means of accent." Turek, supra note 18, at 19.

[77] Emphasis may be real or only perceived. See New Harvard Dictionary of Music (1986), s.v. "Accent."

[78] "The center of this [rhythmic] hierarchy is the individual pulse and the grouping of pulses into a consistent meter that we feel as a rhythmic basis underlying many styles, even when neither pulse nor meter is consistently expressed in any single line. The persistence of this core of continuous expectation can be neatly demonstrated by the phenomenon of syncopation, which continues to be felt as a rhythmic dissonance even when no part reaffirms the continuum against it. The continuum may also establish larger expectations, such as recurrent phrasing of 4 + 4 + 4 bars, and in the other direction, smaller assumptions such as consistently duple or triple fractures of the pulse into a subhierarchy of microcomponents." LaRue, supra note 39, at 90.

[79] The device was described by Riemann as an "annexed motif" (Anschlussmotiv): "a subsidiary phrase unit place immediately after the strong beat of the main phrase unit. It serves to generate a second strong beat where a weak beat would normally occur." Bent, supra note 19, at 92 and 109-10.

[80] Baxter, Record at 698-99.

[81] Record at 1102-03.

[82] Record at 499.

[83] Record at 694-95.

[84] "[I]t is tempting to identify Rhythm as the arch-ambiguity, the single most mysterious and problematical of musical elements.  Aware as we are now, however, of the nearly total interconnectedness of the elements . . . we may at most conclude that the puzzling aspects of rhythm illustrate with particular sharpness a general, fundamental ambiguity that is characteristic of music." LaRue, supra note 39, at 88.

[85] "It has been said that originality in rhythm is an impossibility, and this view is probably correct." Sherman, supra note 53, at 126 (citing Northern Music Corp. v. King Record Distributing Co., 105 F. Supp. 393 (S.D.N.Y. 1952)).

[86] Cf. Shafter, supra note 51, at 199. "Time and accent, as elements of rhythm, have been used to disguise plagiarized melodies." The author suggests that this occurs very infrequently. Shafter continues: "Some of the classic works have been `jazzed up' so that their composers would never recognize them. The rate of performance ipso facto determines the general nature of the piece. Presumably, that majestic work of Chopin's, popularly known as the `Funeral March,' could be transformed into a `hot' dance melody with jungle rhythm, if the arranger were skillful and resourceful enough to accomplish that miracle." Id. at 199-200. See also, Orth, supra note 53. Keyt counters, "These self-styled music critics seem to envision a world full of composers endeavouring mightily to conceal acts of plagiarism, even going so far as to change all the notes." Keyt, supra note 9, at 424 n.17.

[87] See, e.g., Sherman, supra note 53, at 126-29. Examples in the cases presented in Chapter Four include "Dragnet" imbedded in "Feelings" (Gaste, Record at 656) and "Rudolph, the Red-Nose Reindeer" as equivalent to "Rock of Ages" (Baxter, Joint Appellee's Brief (No. 88-6660), at 33 n.37, 907 F.2d 154 (9th Cir. 1990)) (Appendix C, at 1061). One who enjoys this might also compare Beethoven's Sonata No. 4, op. 7 (3d movement) and the "Star-Spangled Banner," but the exercise seems rather irrelevant to the discovery of infringement.

[88] In many instances, a work's rhythm performed without pitch may be more recognizable than its pitch without rhythm.

[89] See, e.g., Harvard Dictionary of Music, 2d ed. (1969), s.v. "Harmony." "The vertical structure of a composition in contrast to the horizontal or melodic structure."

[90] Vertical structures may define chords, although not every vertical structure does, but chords do not define harmony.

[91] "For Schoenberg, structural function deals always with progressions, not with chords, and those progressions are considered in terms of their roots; function gives a sense of direction to a progression and determines its role with the scheme of regions which make up monotonality." Bent, supra note 19, at 116; citing Arnold Schoenberg, Structural Functions of Harmony, rev. ed. (London: 1969).

[92] Rameau provided both the first systematic description of tonality and the influence that focused analysis on harmony at the expense of melody. "To perceive that all the harmonies and their different properties originate in the triad and the chord of the seventh is not enough; beyond this it is necessary to note that all the properties of these two primary harmonies depend absolutely on the harmonic center and its progression. The intervals of which they are composed are intervals only in relation to this center, which makes use of the same intervals to form its progression, on which progression the order and progression of the two primary harmonies alone depend. These intervals are all comprised in the third, fifth, and seventh. . . . This reduction of the intervals has yet another exact relation to that of the harmonies: the inverted intervals form inverted harmonies, the extended intervals form harmonies `by supposition,' the altered intervals from harmonies `by borrowing,' the whole arising from our three primary intervals, from which the fundamental harmonies are formed, and related solely to our harmonic center." Rameau, Traité de l'harmonie (Paris, 1722) ed. Oliver Strunk in Source Readings in Music History (New York: Norton, 1950), 569.

[93] LaRue compares the harmonic continuum to that of language:

chords --> progressions --> tonality

words --> grammar --> syntax

LaRue, supra note 39, at 40.

[94] Noam Chomsky posited that all humans possess an inborn and universal grasp of the formal principles of musical syntax. Bent, supra note 19, at 71-73; citing Noam Chomsky, Aspects of the Theory of Syntax (Cambridge, Mass., 1965).

[95] Nizer's account of Baron v. Leo Feist, Inc., 78 F. Supp. 686 (S.D.N.Y. 1948), aff'd, 173 F.2d 288 (2d Cir. 1949), states that "L'Année Passée" had sixteen chords and that "Rum and Coca-Cola" (Appendix H, at 1278) had fourteen "identical" chords "in the same places and in the same consecutive order." Louis Nizer, My Life in Court (Garden City, N.Y.: Doubleday, 1961), 249. If like segments were being compared (implicit in Nizer's account), this result would seem to be impossible.

[96] Keyt recommends that courts examine "semantic similarity" in addition to literal similarity. "The trumpets are loud" has three literal similarities of words with "The ties are loud." However, the change of one word changes the meaning of another. Similarly, "The trumpets are loud" has a different significance when spoken at a band concert or in a library. Keyt, supra note 9, at 430 n.44.

[97] "[Wallace] Berry's treatment of tonality recognizes the ambivalence of harmonic events: a single harmony may perform different functions at different levels of structure. Tonality's capacity to project multiple levels gives to music its sense of depth of significance." Bent, supra note 19, at 71.

[98] See Record at 752-56.

[99] See Northern Music Corp. v. King Record Distributing Co., 105 F. Supp. 393 (S.D.N.Y. 1952).

[100] See also, Henry Purcell's "Fantasia Upon One Note."

[101] See Keyt, supra note 9, at 431-43 & n.46.

[102] The parties in Baxter argued over which voice should be considered the melody in measures 72 and 73 of "Joy." The argument concerned the significance of the thematic statement. Contrary to some of the testimony, loudness does not equate with significance. In MCA, Inc. v. Wilson, 425 F. Supp. 443 (S.D.N.Y. 1976), aff'd and modified, 677 F.2d 180 (2d Cir. 1981), defendant's expert admitted on cross-examination that it was hard to tell which voice should be considered the melody, so he chose the loudest. He made his transcription from a recording of a stage production, and the singers were dancing and moving in front of a stationary microphone. As different singers came close to the microphone, they may have caused the expert to change his impression of which voice was singing the melody.

[103] "Melody is succession of pitches in rhythm." Turek, supra note 18, at 100.

[104] Ernest Ansermet, postulating a "melodic path," classified intervals as "active extrovert," "active introvert," "passive extrovert," and "passive introvert." Bent, supra note 19, at 63; citing Ernest Ansermet, Les fondements de la musique dans la conscience humaine (Neuchâtel, 1961). Ansermet's melodic path can be traced to Rameau, whose theories solidified the concept of tonality. "At first sight it would seem that harmony arises from melody, inasmuch as the melodies which the single voices produce become harmony when they are combined; it has, however, been necessary to determine in advance a path for each of these voices in order that they may agree together. No matter, then, what order of melody we may observe in the individual parts, taken together they will scarcely form a tolerable harmony (not to say that it is impossible that they should do so) unless this order has been dictated to them by the rules of harmony." Rameau, supra note 92, at 570.

[105] Laycock and Nordgren, supra note 18, at 10.

[106] "Originality is better viewed as a function of the interaction and conjunction of [rhythm, harmony, and melody] than of any element alone; a change in one element necessarily affects our perception of all others." Keyt, supra note 9, at 432.

[107] "Intervals and motivic patterns are the words and phrases of Melody. While musicological comment on Melody often includes tables of intervals, this rather clinical approach reminds one of a list of words without any explanation of meanings. We do not gain much insight from the comment "The piece contains many augmented fourths and major sevenths," a mere enumeration of components without real illumination of function." LaRue, supra note 39, at 83.

[108] This fact further contradicts notions that musical expression is more limited than literary discourse.

[109] "The most obvious influence of text rhythm can be seen in the patterns of individual words, and careful composers have usually tried to match notes to the stress and length of syllables as naturally as possible. . . . Sometimes a text phrase will suggest a special rhythmic atmosphere for a section. . . . In a larger dimension, word rhythms produce consistent meters that often influence the musical meter of a whole piece." LaRue, supra note 39, at 150-51.

[110] LaRue, supra note 39, at 148-52.

[111] Recall Nizer's common error of the chromatic chord implying the pathos of the word "streetwalker." See supra note 95, and supra Chapter 4, note 18.

[112] Gaste, Record at 657.

[113] Record at 729.

[114] Record at 726-63.

[115] Record at 775-76.

[116] The defense in MCA, Inc. v. Wilson, 425 F. Supp. 443 (S.D.N.Y. 1976), aff'd and modified, 677 F.2d 180 (2d Cir. 1981)., was based on the theory that style dictated similarities. Defense counsel argued that the accused work reminded listeners of the complaining work not so much because of similar melodic content or harmonic structure but because of the arranger's style. Record at 339.

[117] Cook, supra note 1, at 24.

[118] "Music, like language, is an articulate form. Its parts not only fuse together to yield a greater entity, but in so doing they maintain some degree of separate existence, and the sensuous character of each element is affected by its function in the complex whole. This means that the greater entity we call a composition is not merely produced by mixture, like a new color made by mixing paints, but is articulated, i.e. its internal structure is given to our perception." Susanne K. Langer, Feeling and Form (New York: Scribner, 1953), 31.

[119] Cook, supra note 1, at 24.

[120] See id. at 225.

[121] See supra page 280.

[122] "As we find it today, our conventional notation is still a mixed symbolic-linear music-writing in which the symbolic element is the more highly organized and therefore dominates. It is practically entirely prescriptive in character. Emphasis is upon structures‑-principally of pitch and meter. It does not tell us much about the connection of the structures. It does not tell us as much about how music sounds as how to make it sound. Yet no one can make it sound as the writer of the notation intended unless in addition to a knowledge of the tradition of writing he has also a knowledge of the oral (or, better, aural) tradition associated with it‑-i.e., a tradition learned by the ear of the student, partly from his elders in general but especially from the precepts of his teachers. For to this aural tradition is customarily left most of the knowledge of `what happens between the notes'‑-i.e., between the links in the chain and the comparatively stable levels in the stream." Charles Seeger, "Prescriptive and Descriptive Music-Writing," The Musical Quarterly 44 (1958): 186.

[123] See 15 U.S.C. § 102.

[124] See Keyt, supra note 9, at 437 (equating hierarchical analysis with Learned Hand's abstractions test).

[125] Schenker, supra note 2, at xxiii.

[126] Cook, supra note 1, at 205 (explaining Schenker and Blacking).

[127] Id. at 208.

[128] Id. at 214.

[129] See, e.g., Northern Music Corp. v. King Record Distributing Co., 105 F. Supp. 393, 398 (S.D.N.Y. 1952). "Examination of the musical score of both songs reveals substantial melodic similarity. Exhibit 18 shows the first eight bars‑-or the "A" strain of both songs. Tonight He Sailed Again has 25 notes in these eight bars; I Love You, Yes I Do has 28 notes in the same bars. Comparison shows that 16 of these notes are common to both songs. This coincidence of notation is significant because it occurs at the same point in the melodies. Similarity in the melodies of both songs is impressed upon us when the first eight bars of both songs are played simultaneously. So played, without harmony, we, paying special attention, detected with untrained ear but three discords."

[130] The durational value of grace notes is not taken into account in common notation; their value must be subtracted from one of the surrounding notes.

[131] "More than any other instrument, the lute called for sustaining ornaments in all voices. While in the lute music of the renaissance the florid sections of the melody were carefully written out in the tablature, they were indicated in baroque tablatures by means of newly invented symbols. The use of ornamental symbols was again derived from the virginalists. Instrumental ornaments were as essential a factor in the chordally conceived lute music as vocal ornaments were in the monody. In spite of the full accounts of lute ornamentation in Mersenne and Mace, its variety and subtlety is hard to describe." Manfred F. Bukofzer, Music in the Baroque Era (New York: Norton, 1947), 166.

[132] See New Harvard Dictionary of Music (1986), s.v. "Diminution."

[133] Allen Forte and Steven E. Gilbert, Introduction to Schenkerian Analysis (New York: Norton, 1982), 1.

[134] Id. at 8; citing Christopher Simpson, The Division-Viol or, The Art of Playing Extempore upon a Ground, 2d ed. (London: 1665; published in facsimile by J. Curwen & Sons, London, 1955).

[135] Bent, supra note 19, at 9.

[136] Cook, supra note 1, at 36.

[137] Darrell v. Joe Morris Music Co., 113 F.2d 80, 80 (2d Cir. 1940) (While there are an enormous number of possible permutations of the musical notes of the scale, only a few are pleasing, and much fewer still suit the infantile demands of the popular ear).

[138] "Schenker's unique view of a musical composition is that works that are tonal and exhibit mastery are `projections' in time of a single element: the tonic triad. The projection of this triad comprises two processes, its transformation into a two-part `fundamental structure' called the Ursatz, and the `composing-out' (Auskomponierung), i.e., the elaboration of the structure by one technique or more of prolongation. The Ursatz is made up of a linear descent to the root of the triad‑-the `fundamental line'

(Urlinie)‑-accompanied by an `arpeggiation' in the bass (Bassbrechung), from the tonic to the dominant and back to the tonic." Bent, supra note 19, at 81.

[139] "Schenker's levels are determined primarily by the Ursatz; they are displayed in his analyses on three basic layers, called `foreground,' `middleground,' and `background,' corresponding to lower, middle and higher levels. But these three strata are only general demarcations, and within any one of them many more hierarchical orderings are found, indicated by note size, type of stem, verbal abbreviations, parentheses, and so forth. How a piece evolves or is prolonged can be seen by employing the concept of the Ursatz on each of the three basic levels. Analytically, these are represented from foreground to background in voice-leading graphs. The analysis results in a hierarchical `tree.' Looking up the tree, one can see how the overall structure is `achieved'; looking down, one can see how it is `generated'; look across, we have a kind of `flow chart' of harmonic process." Narmour, supra note 40, at 5.

[140] Leonard B. Meyer, Explaining Music, (Berkeley: University of California Press, 1973), 109.

[141] See Cook, supra note 1, at 121. Bent points to Schenkerian analyses in Music Forum of medieval and Renaissance music by Salzer, Bergquist, and Schachter, of late Romantic music by Bergquist and Mitchell, of contemporary music by Travis, and of non-Western music by Loeb. Bent, supra note 19, at 68.

[142] See, e.g., Narmour, supra note 40, at 2.

[143] Tree Publishing Co., Inc. v. Overstreet, Case. No. 3:87-0032 (M.D. Tenn. 1987). The case alleged infringement of Nat Stuckey's "Sun Comin' Up" by the Randy Travis recording of "Diggin' Up Bones." The parties settled before trial.

[144] Overstreet, Musicological Exhibits (September 8, 1987), submitted by Spielman. The relevant passage appears in Appendix D, at 1169.

[145] Exhibit 492; see Baxter, Record at 954-58. Appendix C, at 895.

[146] "The thing created is best viewed as a structure of relationships. Because the sounds in a composition are dependent on one another for their meaning, musical meaning is solely a function of context. Thus, preservation of context must be a crucial element of copying. It is not enough to compare only strings of acoustical events. The comparison must include the structures that the sounds articulate." Keyt, supra note 9, at 437.

[147] The New Grove Dictionary of Music and Musicians (1980), s.v. "Analysis," by Ian Bent, 356.