REVIVING THE RHETORIC OF THE PUBLIC INTEREST: CHOIR DIRECTORS, COPY MACHINES, AND NEW ARRANGEMENTS OF PUBLIC DOMAIN MUSIC
PAUL J. HEALD ¤
Reprinted by permission
from Volume 46, No. 2 of the Duke Law Journal
46 DUKE L.J. 241 (1996)
Copyright © 1996 by Duke Law Journal
Professor Peter Jaszi recently issued a challenge to those concerned about growing threats to the public domain.1 According to Jaszi, advocates of a broadly accessible and easily usable public domain must develop a new rhetoric by exploring alternative ways of speaking about the values they espouse.2 It is time for such a challenge arch-protectionists are on the march, determined to recapture works from the public domain for their private benefit,3 and the old ways of advocating the public interest can no longer be counted on to carry the day. The public has demonstrated little understanding of the nature of the public domain and, unlike the publishing industry, has shown little ability to effectively protect its interest. As a remedy, Jaszi suggests that scholarly commentators bear the burden of explaining why the public domain is so important.4
Several perceptive explorations of the public domain have already appeared.5 Although these articles provide a theoretical framework for protecting the public domain, a focus on the particular rather than the general might better effect the rhetorical move proposed by Jaszi. Given the natural tendency of legal rights to express themselves in absolute terms to the exclusion of all else,6 perhaps the best way to defend the public domain is to demonstrate its particular importance to those who draw from it. To this end, we will enter the world of the choir director and conduct a case study in the public domain.
A confused choir director recently showed me a version of the William Billings anthem Jordan published in 1954,7 pointed to the bottom of the page, and asked me how the work of a composer who has been dead for almost 200 years could remain copyrighted.8 A glance at the sheet music revealed a copyright symbol. My snappy and not entirely helpful answer was that a sufficiently original arrangement or edition of a public domain composition is entitled to protection by copyright law.9 She then presented me with a copy of the original version of Jordan,10 long in the public domain, and asked me to identify the differences between the two works.11 Her critical follow-up question: could she photocopy the 1954 publication and distribute it to her choir? In other words, did the copyright symbol on the piece have any legal force? While her question in the case of Jordan was academic (since the public domain work was in her other hand), the legal status of recent versions of public domain music is, in general, extremely important to choir directors: original versions of the music they seek are usually very difficult to obtain.12
Although most published opinions discussing arrangements of public domain music suggest that the 1954 publication of Jordan is not sufficiently original to be protected by copyright, and therefore is probably amenable to the photocopy machine, the parameters of the applicable legal standard are hazy.13 A recent case finds that, to be protected by copyright, a new arrangement of pre-existing work must contain more than cocktail pianist variations . . . . [S]omething of substance [must be] added making the piece to some extent a new work with the old song embedded in it.14 Another, older case holds that inconsequential melodic and harmonic embellishments such as are frequently improvised by any competent musician are not sufficiently original to earn copyright protection for an arrangement of a public domain work.15 Another opinion suggests, however, that the addition of new fingering, dynamic marks, tempo indications, slurs and phrasing can be enough.16
None of the cases that discuss this issue provide an appendix that allows the reader to compare the public domain version of a composition with the allegedly original, derivative arrangement. The articulated tests, described as the mere mechanic test,17 the substantiality test,18 and the distinguishable variation test,19 are too vague in and of themselves to resolve disputes. The lack of visual comparisons and the limited precedential value of the relevant opinions20 complicates the choir director's determination of whether she can photocopy an arrangement of public domain music or must purchase multiple copies of the work.
The decision to photocopy or not to photocopy has significant consequences for the music consumer's pocketbook. Photocopies cost around three cents per page,21 while an original printed version of a choral work costs about thirty cents per page.22 The expense of buying rather than copying public domain sheet music is directly absorbed by the taxpayers who fund music education in public schools, the church congregations who must raise money for the church music budget, and the patrons of the fine arts who finance music ensembles with their admission fees or donations.23
To recognize the high cost of sheet music is not to assert that legitimately registered works should be pirated at the copy machine. The point is that the decision to purchase rather than to photocopy is a costly one, that as consumers, taxpayers, and churchgoers we are directly affected when the decision to purchase is made unnecessarily. Arrangements of public domain music not sufficiently original to be protected by copyright law are free to all,24 yet, much to our financial detriment, the law does not sufficiently identify exactly which of these arrangements are fair game.
Music publishers, taking advantage of this uncertainty, intimidate the public into buying what they already own by affixing copyright symbols to virtually all public domain music as well as trivially different arrangements of public domain music.25 Publishers
change titles without modifying lyrics or melody. They may claim full originality when they are really only finders of public domain songs. They register copyrights to such songs, claiming they are original works, and fail to set forth accurately the limited amount of any new material. They thereby falsely and unfairly obtain the benefit of the Copyright Act provision that places on an unauthorized user the burden to prove the invalidity of a certificate of copyright registration.26
This Article, in response to Professor Jaszi's challenge, uses the particular experience of a choir director and visual comparisons of public domain and copyrighted works to inquire into the standard that governs the copyrightability of musical arrangements and, at the same time, to strengthen the voice of the public domain.
I. THEORETICAL PRELIMINARIES AND INTRODUCTION TO THE RELEVANT LAW
The law governing arrangements of public domain music can perhaps be explained best in light of the policies that animate the Copyright Act.
A. The Copyright Bargain: Lessons from History and Some Economic Theory
The Copyright Clause of the Constitution27 embodies a hypothetical bargain struck between the polity and its artists: citizens give artists the right to exclude others from copying their creations, in return for free use of those creations after the expiration of a set time.28 England's experience under the Stationer's Company publishing monopoly,29 which led to copyright reform under the Statute of Anne,30 was not a lesson lost on the Framers of the Constitution.31 They knew that the grant of exclusive rights would be a cost borne directly by the public, and they demanded a direct public benefit in return.32 In theory, the possibility of obtaining a valuable monopoly provides economic incentives for the creation of new works which would eventually be dedicated to the public domain.33 For this reason, the first Congress promptly used its constitutional authorization and provided protection for maps, charts, and books in 1790.34
Works in the public domain at the time of the ratification of the Constitution (all of the works of J.S. Bach, for example), remained (and remain) free for all to use. Granting someone an exclusive right to a Bach Cantata would impose a cost on the public without any countervailing benefit. Since the ratification of the Constitution, Congress has with increasing aggressiveness established a system of rewards and incentives for artists.35 Nobody can prove that artists are more productive because of copyright law, but if they have been, the payoff has been spectacular.36 Consider the vast wealth of musical, literary, and other artistic works that now belong to the public-everything published before 1921,37 and even some created thereafter.38
Congress has always recognized the public's proprietary interest in the public domain. For example, nothing in an arrangement of a public domain music composition is protected unless it qualifies as a derivative work which consists of editorial revisions, annotations, elaborations, or other modifications which, as a whole, represent an original work of authorship.39 Even if a composer creates an original arrangement of a public domain work, the new copyright extends only to the material contributed by the author of such work, as distinguished from the preexisting materials.40 For example, only an original bass line added to a public domain composition would be protected by copyright. In L. Batlin & Son, Inc. v. Snyder,41 the Second Circuit provided a clear explanation why:
Absent a genuine difference between the underlying work of art and the copy of it for which protection is sought, the public interest in promoting progress in the arts-indeed, the constitutional demand could hardly be served. To extend copyrightability to minuscule variations would simply put a weapon for harassment in the hands of mischievous copiers intent on appropriating and monopolizing public domain work.42
In other words, the public should not have to pay a monopoly price for a work that is essentially the same as its public domain progenitor.
In addition, keeping the public domain free for exploitation makes economic sense. A musical work, like a lighthouse or a new word is a public good. That is to say, each successive use of it does not diminish its intrinsic value.43 Unlike a piece of land that is farmed or a car that is driven, no matter how many times a piece of music is performed, it does not wear out. No matter how many copies of a musical composition are made, still more can be made equally cheaply. In theory, once a public good is created, charging consumers to use it entails a deadweight loss.44 Limiting its use serves no purpose, since each successive use does not harm it. A composition is not like a lake full of fish whose exploitation must be limited to maintain its stock.
Nonetheless, under some circumstances it may be appropriate to charge consumers of public goods. It may be rational to grant their creators exclusive rights for fear that, absent a monetary incentive allowing for the recovery of sunk costs, some public goods will not be created in the first place.45 In theory, for example, some music may not be written in the absence of the monetary incentives provided by copyright law.46 This rationale, of course, has no application to works that have already been created, such as our vast supply of public domain compositions. We may need incentives for music to be written tomorrow, but not for music written seventy-five years ago.
The copyright bargain, whether one views it in economic terms or as the binding intent of the Framers of the Constitution, should and does extend to original derivative works. No one believes, for example, that Aaron Copland's Appalachian Spring should be uncopyrightable because it borrows one of its themes from the old Shaker hymn Simple Gifts, a piece long in the public domain.47 Similarly, the Electric Light Orchestra's Roll Over Beethoven is properly copyrighted in spite of the incorporation of bits of Beethoven's Fifth Symphony.48 Congress has provided incentives for composers to mine the public domain and create original new works. In theory, we would have fewer such works in the absence of protection. Given the importance of the public domain as an inspiration for new creations,49 Congress has rationally included original derivative works as a category of copyrightable expression.50
Congress has also just as rationally excluded unoriginal variations of public domain works from protection,51 thereby disarming those who may attempt to obtain a weapon for harassment by copyrighting works with only minor changes.52 In this regard, the Court has defined original in such a way as to preclude protection based merely on a work's intrinsic value or the amount of labor necessary to create it. In Feist Publications, Inc. v. Rural Telephone Service Co.,53 the Court denied copyright protection to the telephone white pages, emphasizing that the Constitutional requirement of authorship54 precluded protection of unoriginal works.55 The compilers of the white pages expended substantial effort in creating a very valuable new work, but that was not enough. Under the rationale of Feist, a court would be forced to conclude that the cost to the public of granting an arranger or an editor of public domain music a monopoly is not offset by any countervailing public benefit when the new work is essentially the same as the original. While new editions of old music are often useful and valuable, as a matter of constitutional law, Congress cannot grant copyright protection to unoriginal works no matter how much sweat has been expended in their creation.56 A new but unoriginal typesetting and printing of a Billings anthem may be useful and valuable, but it is not protected by copyright law.57
I have elsewhere argued that Feist's disavowal of copyright protection for sweat of the brow works may be inconsistent with the copyright bargain described above, noting that some valuable, but unoriginal, works may not be created in the absence of copyright incentives.58 If, as the economists suggest, copyright is about establishing wealth-maximizing incentives to bring about new creations, then Feist's affirmation of a Romantic notion of authorship may be misplaced. For example, Handel's Messiah is a public domain work, but Congress might rationally conclude that limited copyright protection for new editions of the Messiah is needed to ensure that the public's demand is satisfied. The Court in Feist prevents Congress from acting on such a conclusion. To this extent, the case strays from a purely economic justification for copyright, while expressly endorsing it.59 Although I seriously doubt that legislation is necessary to stimulate new editions of public domain music,60 Feist's rationale probably precludes it.
B. The Inadequate Caselaw
Only one federal appellate decision in the last forty years discusses the standard of originality necessary for a new musical composition to constitute a copyrightable derivative work. In Woods v. Bourne Co.,61 the Second Circuit held that several versions of When the Red, Red, Robin Comes Bob, Bob, Bobbin' Along were insufficiently original to constitute copyrightable derivative works.62 Although the opinion does not reprint the original song or subsequent versions of it, the court's discussion is helpful. In 1926, Harry Woods brought a lead sheet, consisting of the melody and lyrics of the song, to Irving Berlin, Inc., whose in-house technicians then added harmony and made other changes.63 Finding the differences between Woods' lead sheet and the subsequent piano-vocal version that was eventually published to be insubstantial, the district court held the piano-vocal version to be unoriginal.64 It also found unoriginal a 1981 version of the song that substituted a moving bass line for the piano-vocal harmony, which had primarily consisted of quarter notes on the first and third beats of every measure.65
The Second Circuit affirmed both these holdings, approving the legal standard required by the district court:
[S]omething of substance [must be] added making the piece to some extent a new work with the old song embedded in it but from which the new has developed. It is not merely a stylized version of the original song where a major artist may take liberties with the lyrics or the tempo, the listener hearing basically the original tune. ... [C]ocktail pianist variations of the piece that are standard fare in the music trade by any competent musician [do not suffice].66
The Second Circuit emphasized that a trivial variation from the first work or the mere demonstration of `physical skill or `special training was not enough.67 Although the circuit court approved of the district court's factual inquiry into whether the piano-vocal arrangement was the result of the application of conventional rules of harmony,68 it did not endorse the district court's dictum that a derivative work must contain unusual vocal treatment, additional lyrics of consequence, unusual altered harmonies, [and] novel sequential uses of themes.69 Most importantly, the opinion recognizes the role played by a substantial originality standard in taking `a weapon for harassment [from] the hands of mischievous copiers intent on appropriating and monopolizing public domain work.70
Other decisions which discuss derivative musical arrangements are also quite sensitive to the danger that trivializing the originality requirement would present to the public domain. They all demand a palpable contribution by the arranger; whether they apply precisely the same standard of originality, however, is difficult to discern. Without the actual pieces of music before us, it is virtually impossible to conclude whether the reported decisions are really consistent. For example, Northern Music Corp. v. King Record Distributing Co.71 holds that neither rhythm nor harmony can in itself be the subject of copyright72 overtly conflicting with another decision from the same jurisdiction, Tempo Music, Inc. v. Famous Music Corp.,73 which finds a Duke Ellington harmony protectable.74 While these two opinions are hard to reconcile, even the Tempo court demonstrates an awareness of the danger of adopting a standard of originality that would demand only a trivial variation from the pre-existing work.75 Unfortunately, neither of these opinions are discussed by the Second Circuit in Woods II, a frustrating state of affairs given that both cases arise in the Southern District of New York.
A relatively small number of other cases are relevant. The court in McIntyre v. Double-A Music Corp.76 refused to find originality in an arrangement that added an introduction, some repetition, and inconsequential melodic and harmonic embellishments.77 In Wihtol v. Wells,78 the Seventh Circuit protected a hymn with new lyrics that borrowed from an old Latvian folk melody remembered from the composer's childhood. A very simple new bass line was found insufficiently original to be protected in Shapiro, Bernstein & Co. v. Miracle Record Co.79
With the exception of Consolidated Music Publishers, Inc. v. Ashley Publications, Inc.,80 which seems to hold several insignificant editorial changes copyrightable,81 the language and intent of the decisions are consistent with the copyright bargain discussed above. But while the cases do not conflict with the constitutional and statutory requirements of originality, they provide inadequate guidance to a consumer concerned with whether a particular arrangement is protected or not. The courts have not consciously protected insignificant arrangements or approved the weapon for harassment decried by the Second Circuit. However, the published decisions have also not thwarted attempts by music publishers to capture the public domain. These publishers, by affixing copyright symbols to public domain music and thereby intimidating music buyers, have managed to subvert copyright policy nonetheless.
I. SPURIOUS CLAIMS TO COPYRIGHT IN MUSICAL ARRANGEMENTS: THE THREAT TO PUBLIC DOMAIN MUSIC
I have elsewhere explored the causes of action which might be brought against a publisher who profits by improperly affixing a copyright symbol to a public domain work.82 I have not, however, previously detailed practices in the music publishing industry or provided guidance to those seeking to distinguish works in the public domain, which can be copied, from those that are protected, which can not. Appendices A1 to C2 (A1, A2 , B1, B2, C1, C2) to this Article provide examples of public domain works side-by-side with recently copyrighted versions, illustrating Krasilovsky and Shemel's warning that [m]uch of the music material in the public domain is tainted by vague and indefinite claims of copyright in minimal or obscure `new versions.83
A well-known, one-page Brahms chorale for women's voices, O bone Jesu, recently ordered from a national sheet music distributor for $1.60 plus $2.80 shipping and handling, appears as Appendix A2.84 At the bottom of the page, one finds a copyright symbol and the statement Copyright 1979 by Carus-Verlag Stuttgart. Undoubtedly, many choir directors and music educators would think twice about deliberately ignoring the registration symbol and would decide to purchase more copies of the one-page sheet music at $1.60, rather than make their own copies for three cents apiece. Inducing reliance on the part of music buyers and intimidating them into purchasing, rather than photocopying, music is a very profitable venture. In the case of O bone Jesu, reliance on the copyright claim by the music purchaser increases the buyer's cost more than fifty-fold.
A comparison of the original 1865 publication85 with the recently ordered version86 reveals the entirely spurious nature of the latter's copyright claim. The 1979 publication of O bone Jesu is exactly the same as the public domain version. All notes and dynamic markings are identical. A more blatant abuse of the registration symbol, a symbol which carries with it the threat of significant civil and criminal penalties for copyright infringement,87 could scarcely be imagined. Music published abroad often presents this problem, given that copyright protection for new editions of music is more easily available there.88
Not all spurious claims are so blatant. Appendices B1 and B2 contain the original public domain version and a copyrighted 1954 version of William Billings' anthem Jordan.89 They are not quite identical. Although the notes are precisely the same in the two pieces, the 1954 version (Appendix B2) adds slurs90 whenever two notes are sung on the same syllable, and writes out the repeat of the refrain that Billings had economically marked instead with a repeat sign. The 1954 publication also includes a piano reduction for rehearsal only (the piece is meant to be performed a capella) which is no more than the notes the choir sings conveniently marked in one system.91 The reduction is not a new harmonization or a different tune to be played contrapuntally against the choir, but merely the same notes that the choir sings, written to make playing them on the piano easier while the choir sings those notes in rehearsal.
The copyright claim placed on the 1954 publication is almost certainly spurious. It flunks even the laxest originality standard proposed in the cases discussed above.92 Slurs are a convenience that add nothing original to the piece-it would sound the same without them. In the words of one authority, they are unnecessary in this sort of choral music.93 Writing out the repeat is merely realizing Billings' own scoring. And, as mentioned above, the piano reduction is just that a mechanical condensation of Billings' own work with nothing new added. We have no new words or harmonies or other new material of any kind. The 1954 publication, although not identical to the public domain work, is not an original work of authorship as required by the Constitution and the Copyright Act.94
A good illustration of a valuable, but probably unoriginal, publication the final chorus of Act III (Hallelujah, Amen) of Handel's oratorio Judas Maccabäus95 appears as Appendix C2. The 1933 publication allows a choir to perform the final chorus without having to buy the entire 200 page oratorio.96 The orchestral accompaniment, which a church choir director may not want, is also economically deleted. However, as noted in Part I above,97 the constitutionally and congressionally mandated test for copyrightability is not usefulness or value, but rather originality.
What is original about the 1933 publication? The credited editor, Norwood Hinkle, marks a repeat of the first twenty-two measures that is not indicated in the original.98 In a couple of places where Handel ties two quarter notes, Hinkle writes them out as a half note, while maintaining the original ties. Handel's work largely lacks dynamic markings. Hinkle marks mf on measure 5, f on measure 12, and ff on measure 14.99 He indicates the tempo at Allegro, a tempo giusto100 whereas Handel had indicated several pages earlier in the original score andante allegro.101 Most important, however, the melody, the harmony, and the rhythm printed in the two pieces are identical. The Hinkle edition is essentially the same as the original piece that Handel first directed in 1747.
These three comparisons reflect industry practice as gleaned from forays through my church's music collection and the University of Georgia's Music Library. Some spuriously copyrighted publications appear to be no different from the original (O bone Jesu); some show a few trivial differences (Jordan); and some reveal numerous trivial differences (Hallelujah, Amen). The effects of their copyright claims are identical, however-a great number of music purchasers choose to buy additional copies of sheet music instead of photocopying at less expense in deference to these works' spurious copyright symbols.
III. WHEN IS AN ARRANGEMENT OF PUBLIC DOMAIN MUSIC PROTECTED?
The problem created by spurious copyright claims is not insoluble. If the consciousness of music purchasers, and of the attorneys and professors who may advise them, is sufficiently raised (the objective of Parts I and II), and they are provided with a workable standard to judge the originality of the typical features of an arrangement of a public domain composition (the immediate objective), the costs of the current predicament can be significantly reduced.
A relatively limited number of musical techniques are used to add to, or delete from, a public domain work. The most common sorts of tinkering can be judged in light of copyright policy and the extant caselaw to define a fairly clear standard to govern whether an arrangement is original enough to be copyrighted.
Surprisingly, the Court's decision in a case not involving music, Feist Publications, Inc. v. Rural Telephone Service Co.,102 is perhaps most helpful. In finding the telephone white pages to be unoriginal, the Court found their arrangement to be mechanical, routine, practically inevitable, and an age-old practice, firmly rooted in tradition.103 The white pages were entirely typical, garden variety, and obvious.104 The Court indicated that conventionality and lack of significant choice are the keys to judging the originality of derivative works;105 this notion is very helpful in judging the originality of arrangements of public domain music.
The only appellate case decided after Feist that discusses originality in the context of musical arrangements echoes this sentiment. In Woods v. Bourne Co., the Second Circuit approved an inquiry that focused on whether the arranger exercised deliberate aesthetic choices or whether he was merely applying conventional rules of harmony.106 Conventional cocktail pianist variations are not sufficient. In fact, the ease with which inconsequential or conventional additions to music may be made may have resulted in a stricter standard than absolutely required by Feist itself.107
One finds a similar focus on convention in non-music decisions after Feist. For example, in finding a title insurance commitment unoriginal and unprotected by copyright law, the Seventh Circuit found that the selection of data to be included in the form was not a matter of discretion based on Mid America's personal judgment or taste, but instead it was a matter of convention.108 The contribution of the putative copyright holder was too rote and mechanical.109
The Mid America opinion denies copyright protection for much the same reasons that a patent is invalid if it represents only an insignificant improvement over a patent already issued-the improvement being obvious to someone experienced in the relevant art.110 Both the copyright law originality requirement and the patent law non-obviousness requirement focus on whether the derivative work is the result of conventions familiar to creators working in the relevant culture. In noting this similarity, a lawyer friend with a D.M.A. in piano performance suggested to me that an accurate articulation of the originality test might be whether a derivation is within the range of the musical culture.111 This formulation resonates with the patent law obviousness standard.
Regardless of the precise articulation of the test, the focus on conventionality is very helpful in evaluating the originality of the sort of activities that arrangers engage in.
A. Orchestral and Choral Reductions
A sufficiently original reduction of an orchestral work112 into a form playable on a keyboard is copyrightable. Although some decisions113 suggest a very stringent standard for musical reductions, in Wood v. Boosey,114 an English court found that a piano arrangement of an opera was original. The multiple lines of music played by various orchestral instruments in a complex operatic or symphonic work can be realized for piano in many different ways. In the absence of accepted musical conventions which would render reductions of an opera substantially the same,115 Wood v. Boosey would seem to be correctly decided.
Of course, a derivative reduction composed in the spirit of the original work which diverged significantly from the original in melody and harmony would also be copyrightable. Under Feist, the key to judging the originality of a reduction would seem to be the number of significantly different options that could be exercised by the arranger. One of the reasons why the telephone book in Feist was uncopyrightable was the lack of choices available to someone arranging phone numbers in a book-alphabetical order is the choice dictated by convention.116 A similar lack of choice in choosing and arranging the yellow pages subject headings drives the holding of unoriginality in Bellsouth Advertising & Publishing Corp. v. Donnelley Information Publishing, Inc.117 Under Feist, a mechanical reduction of an orchestral score that is dictated by musical convention, where the arranger exercises no significant choices, would not be original. Choices made outside the dictates of convention would be protected.
Most piano reductions of non-orchestral works, such as choral arrangements, are entirely dictated by the choral parts. Billings' Jordan in Appendix B2 provides a good example of this.118 Since the reduction is dictated by the vocal parts, it is unprotected. As already noted, most piano reductions of choral works are merely the sung notes reproduced for the pianist's convenience onto a system below the choral lines.119 No new musical material is added. On the other hand, an accompaniment that added original harmonizations or a new contrapuntal melody would be protected. The presence of a sufficiently original piano or organ part does not, however, absolutely prevent the use of the arrangement. Blocking out the protected material before photocopying is permissible.120 In addition, new music software makes it quite easy to print out only the public domain parts of a musical composition in an alternative and convenient format.121
B. New Harmonies
Although some older caselaw suggests that harmonic additions to public domain music are unprotectable as a matter of law,122 the opinions are ultimately unconvincing. The facts of Cooper v. James123 illustrate the problem. Many songs in the first published versions of the famous Sacred Harp hymnal were written in three-part harmony; they lacked an alto line. In other words, all women would sing the soprano line or the tenor line pitched up an octave. The copyright claimant in Cooper wrote alto lines for compositions in the hymnal and sued when they were copied by a competitor. The court held that the addition of an alto harmony could never be sufficiently original.124 When viewed from a late twentieth-century perspective, this holding seems hopelessly outdated both jurisprudentially and musically.
The court in Cooper applies the mere mechanic test of Jollie v. Jaques125 and reasons that originality is lacking because any musician of reasonable talent could compose an alto part. This may be true, but after Feist the question is not the talent level of the composer, but whether significant choices were exercised in the course of the harmonization. In other words, were these musicians of reasonable talent guided in writing similar alto lines by existing musical conventions? A line dictated by accepted rules of composition would not be original, but if the rules permitted numerous harmonic possibilities or if the new alto part broke significantly from convention, then it would be original. A musicologist familiar with the harmonic conventions of the era and a sense of the options available to a composer of an alto line written in the style of Southern Harmony could tell us whether the alto parts at issue in Cooper were original or not.126
The more recent decision in Tempo Music Inc. v. Famous Music Corp.127 would seem more in tune with both the principles of Feist and modern music theory. At issue were harmonies composed by jazz great Duke Ellington. The court rejected the defendant's claim that harmony results only from the formulaic application of centuries-old compositional rules.128 Recognizing the breakdown of conventions in modern Western music, the court noted that in the jazz music genre, musicians frequently move beyond traditionalrules to create a range of dissonant and innovative sounds.129 Emphasizing the choices available to (and made by) Ellington in his harmonization of Satin Doll, the court found the harmonies to be sufficiently original. While recognizing that in most instances, harmony is driven by melody,130 and may in those cases be unprotected, the court refused to deprive appropriate protection to composition which contains sufficient originality and creativity to warrant such protection.131
The inquiry into the originality of a harmony is similar to the inquiry into the originality of a reduction. What choices were available to the composer? Has the harmony been dictated by preexisting elements in the public domain work or is it a break from musical convention? Obviously, the originality of a harmony must be judged on an ad hoc basis and with the help of a musicologist who understands the nature of the real choices, if any, exercised by its composer.
C. Realizing Figured Basses, Rekeying, and Recleffing
Johann Sebastian Bach, like many of his contemporaries, did not write out in modern musical notation the harmony to his compositions, but instead indicated those lines numerically.132 This convenient device enabled a Baroque composer to save the time and space it would have taken to write out the composition completely. Although a talented keyboard player can play directly from the figuring,133 most modern arrangers of works containing figured basses write them out in standard notation. A realization of a figured bass line is almost certainly not protected because the realization evinces very little, if any, choice on the part of the arranger. Competent arrangers realizing a figured bass will inevitably write it out in much the same way, thereby failing to meet Feist's originality requirement.
Similarly, an arrangement is unprotected if it merely transposes the entirety of a public domain work into a new key to make it easier (or in the case of a bad or weird arrangement, harder) to sing or play. Transposition from key to key is entirely mechanical and free of choice. Any competent musician (or a computer program) will produce the same piece when asked to transpose Jordan from A major to C major. The same rationale that results in the protectability of the realization of a figured bass and many piano reductions dictates that the mere rekeying of a piece is unprotected.
The conversion of an older piece of music notated in an obsolete clef134 into modern bass or treble clef is also an unprotectable contribution. The recleffed work will sound exactly the same and the process of recleffing is entirely mechanical. No choice is exercised by the person (or computer) doing the work. The sole motivation of the arranger is the convenience of the performer who may not feel comfortable with an unfamiliar clef.
D. Typeface and Facsimile Editions
American law, unlike that of some foreign countries,135 does not extend copyright protection to typefaces.136 Therefore, a music publisher cannot obtain protection for a new publication of a public domain work simply by reprinting it in a different type. For similar reasons, a facsimile edition of a composer's original works (perhaps even of the initial handwritten drafts 137) is unprotectable. Although West Publishing has argued that its formatting of opinions and page placement is protected by copyright law,138 this assertion probably does not survive Feist.139
E. Critical Comments and Corrections
Significant editorial comments that accompany a public domain composition may be protected. A recent edition of the complete works of William Billings contains significant introductory, historical, and critical materials that are protected original expression, even though most of the compositions themselves can be freely copied.140 Sometimes such comments appear on the musical composition itself. As noted above, one can always avoiding infringing such material by blocking it out before photocopying.141 Copying the commentary itself may even be a fair use, if it is only copied to access the public domain material and is not itself exploited by the copier.142
Many editorial choices, however, may be too trivial to be protected at all. For example, the original version of Jordan,143 taken from a wonderful edition of his complete works edited by Karl Kroeger,144 is not exactly what Billings wrote. The collection, on occasion, makes corrections when Kroeger thinks that the initial publication was misprinted in some way. In Jordan, for example, he places the notation Forte over measure 25, where it sensibly belongs.145 This move puts it over So which clearly starts a new musical thought, rather than over to, which would be an awkward and unconventional place to begin an increase in volume. The notes to the Kroeger edition indicate that the first publication of Jordan placed the Forte erroneously over measure 26.146 We can be grateful to Kroeger, and to subsequent publishers who have moved Billings' Forte to the place where he probably intended it, but this contribution is unoriginal in the constitutional and statutory sense.
One commentator who argues for the increased protection of editorial changes reasons that the law should reward those editors who, through diligent effort and research, have preserved the integrity of the original.147 His rationale for protection reveals why most editorial changes cannot be constitutionally protected. First, according to the court in Feist, copyright rewards originality, not effort.148 The court expressly rejected the sweat of the brow rationale that was once generally accepted.149 Second, editorial changes that are designed to make a composition truer to the composer's intent would seem by definition to be unoriginal. Such efforts are by definition intended not to create a new work, but to preserve the essence of the old. Although one could imagine that a very speculative edition of an ancient musical fragment incorporating numerous creative choices might be protectable, the run-of-the-mill changes made to most commonly performed compositions are, for better or worse, usually unoriginal in the legal sense.
Some publications of public domain compositions include helpful notations indicating how the piece should be physically realized on a particular instrument. Compositions for string instruments in particular occasionally include fingering, bowing or articulation notations. These suggestions on how to play a series of notes are sometimes included for beginners learning how most easily to play a piece, or they may be suggestions for accomplished musicians experimenting with various ways of interpreting the piece.150 On some instruments, the choice of fingering or bowing methods can directly affect the sound made by the instrument and therefore the performer's interpretation of the piece.151
Notation that is designed to make a piece easier for the novice to play should be unprotected. Accepted methods of fingering or bowing that are dictated by ease of playing are not original. Significant interpretative choices made by performers of string music, however, can depend on various fingering or bowing options.152 The mere existence of options, however, is not dispositive. Although a musician may make one interpretive choice among several, the originality standard may still not be met within the relevant musical genre or culture.
An example may serve to illustrate the problem. Imagine a modern arrangement of the Bach Sonata in C Major for two violins and harpsichord153 which indicates that the strings should be fingered by a violinist wearing aluminum thimbles. Such an original and unconventional notation would arguably be protected. Imagine, on the other hand, the three alternative fingerings suggested by Ronald Smith for the first movement, measures 3-4, of the second violin part of the same Bach sonata.154 He explains with great insight the differences in sound produced by the fingerings that he offers, the aesthetic advantages of each, and the skill required to play each option. His discussion concludes that [b]ecause of the originality and intellectual labor required in designing them, fingerings are worthy of copyright protection.155 Without the teaching of Feist and Woods v. Bourne Co. at his disposal, he is not prompted to offer an opinion on how conventional his suggested fingerings are. Are these variations that any cocktail violinist156 would naturally offer? Are they obvious options well within the range of the musical culture? If they answer to these questions is yes, then it's difficult to see how any of the fingerings offered by Smith would be protected under current precedent.
G. Rhythm: Meter, Ties, and Bar Lines
Although an early case suggests that rhythm is not protectable,157 one can imagine how extensive rhythmic changes might make a public domain work virtually unrecognizable. While an original rhythmic contribution to a public domain work is conceivable (for example, imagine a rap version of the fourth movement of Beethoven's Ninth Symphony), many arrangers tinker with rhythmic notation merely to make performance of the piece easier. For instance, although the director of our church choir remetered several measures of a piece we recently sang, her motivation was not to create an original work of authorship but to facilitate her choir's reading of the piece. She felt that the shifting meter of the original which enabled the maintenance of even note values was distracting, and rewrote it in a single meter with shifting note values to make the piece easier to sing.
My director's practice-one that is not uncommon-raises two issues. First, one must judge the originality of her choice to change the metering by comparing it to musical conventions governing the transposition of meter.158 Second, one must ask the question whether a change in the written expression of music which does not change the sound of the piece when performed could ever be the subject of copyright. At least some opinions suggest that the originality of new version of an existing musical work turns on whether it sounds sufficiently different than the existing one.159 Under this test, my director's changes are surely unoriginal-they effect no difference at all. Since the reason for reducing music to tangible written expression is to stimulate a particular sound when it is performed, this focus on the lack of aural difference may be consistent with the rationale of the originality requirement. My director's changes had the same effect as the work of the skilled artisan who carefully cleans the mold and grime off of the fresco of the Last Supper. They both make the underlying work of art more accessible, but their contribution is not original in the copyright sense.160
Raising a similar issue are many works written centuries ago, such as Gregorian chants, that contain no bar lines or indication of meter at all. One must remember that just because a piece lacks rhythmic markings does not mean that an exact rhythm is not generally understood to be implied. Markings may be lacking, not because the composer did not care how a piece was sung, but rather because he knew that given existing conventions, no real performance options existed. The musical culture's equivalent of the Uniform Commercial Code notion of usage in the trade fills many gaps that we may perceive in an old piece of music. An arranger who does nothing more than realize in modern notation what was, and may still be, understood as the conventional way to perform the piece has not made an original contribution.
An arranger, however, may go further and change the intended feel of a piece by shifting the overall meter and rebarring it, for example, taking a waltz and playing it as a two-step in cut time, without changing melody or harmony.161 Would such a rhythmically different arrangement of a public domain work be protected by copyright law? Protecting a cut time version of a public domain waltz would serve little or no purpose. It would reward the labor and industriousness of the arranger for making the mechanical adjustments to the original work, but would not really reward an original musical contribution. The arranger may have been the first to conceive of arranging the work in cut time, but copyright law does not offer protection to ideas.162 The tangible expression of an idea, which may be protected, is in this case a mechanical operation that involves no significant choice to be exercised by the arranger.
Obviously, some rhythmic changes are significant and could be protected. Walter Murphy published disco versions of several public domain works on his 1972 A Fifth of Beethoven album. Such unconventional rhythmic gerrymandering may well be original; the fact that much tinkering with rhythm is insignificant does not mean that an arranger cannot, in fact, make an original contribution by making rhythmic changes.
H. Dynamic Markings and Editing
New arrangements of public domain compositions frequently alter musical dynamics by adding or deleting slur marks and repeats, indications of crescendo and diminuendo, or volume markings. Such changes affect the mood, flow, or feel of the piece. Many public domain works contain sparse dynamic markings or lack them altogether.163 An arranger of such a work may add dynamics to realize the sound she believes the composer originally intended for the work. As noted above, many older works lacked rhythmic or dynamic markings altogether because generally accepted musical conventions dictated how the piece would be performed.164 Alternatively, the arranger may be trying to adapt a public domain work to her own vision. For example, a section to be sung marcato165 may be changed to legato166 or a section may be changed to be sung piano instead of forte. In fact, choir directors spontaneously make these sorts of changes to a piece during rehearsal in an attempt to produce precisely the desired musical effect.
The originality of changes in dynamic markings turns on the overall effect they have on the sound of piece. As the Second Circuit stated in Woods v. Bourne Co., the arrangement must be to some extent a new work with the old song embedded in it but from which the new has developed.167 Even though they may not affect the melody, harmony, or rhythm of a musical composition, a significant audible difference may be perceptible since markings are directions to performers telling them: This is how you should sing this composition. The critical inquiry is whether the changes in dynamics are obvious-that is, suggested by relevant conventions-or not.
The issue of extensive, but obvious, textual editing arose in Grove Press, Inc. v. Collectors Publishers, Inc.168 In that case, the editor of a public domain biography made over 40,000 changes to grammar, syntax, and word choice. This extensive editing was found not to be protected by copyright law. The court found that mere editing of the sort any high school English student could do was not sufficiently original.169 The new edition undoubtedly had a different feel than the public domain book, but the court required more of an original contribution to its essence. No one denied the value of the editorial contributions to the new publication, but the test under Feist, again, is not value or usefulness but originality measured as departure from convention.170
Viewed in light of the copyright bargain discussed above, denying protection to insubstantial changes in dynamics makes sense. Massaging an existing work is not the same as creating an original derivative work. If one could obtain copyright protection merely by adding a forte to a measure of a public domain composition, a dangerous weapon for harassment would be available for anyone to exploit.171 The originality standard would be no standard at all if a trivial variation would suffice for copyright. Would extensive and radical changes in dynamic markings be original? Substantial additions that worked a significant change to the essence of the piece could suffice. Additions that are typical, routine, obvious, or practically inevitable, to use the Feist phraseology, should be insufficient.
IV. ADMINISTRATIVE OR STATUTORY SOLUTIONS
As the preceding discussion suggests, users of public domain music trying to determine whether a registered work is in fact original find little concrete guidance in legislation or in the existing caselaw. In this respect, future courts will hopefully be able to provide some assistance with the help of expert witnesses. A court sensitive to copyright policy should be capable of determining which sorts of arrangements of public domain music are original. Furthermore, the decision in Feist should be enough to establish the importance of the originality requirement. Nevertheless, although the judiciary and secondary sources can help by clarifying the legal parameters of the originality requirement, only Congress or the Copyright Office can help provide the relevant facts needed by those making decisions about whether to photocopy or to buy arrangements of public domain music: what changes did the arranger actually make to the public domain piece? No informed decision as to the originality of the arrangement can be made unless the additions and deletions made to the public domain work have been ascertained. In spite of the fact that I live in a city with a decent music library, I have had trouble collecting the public domain versions of various well-known works. Imagine the plight of the choir director or school administrator who has neither a major university music library in town nor the time to do significant sleuthing.
In rare instances, a consumer can obtain the relevant information from the Copyright Office. The Copyright Office requires the applicant for a copyright in a derivative work, like a musical arrangement, to disclose both the preexisting material on which the arrangement is based and also to make a brief, general statement of the additional new material covered by the copyright claim for which registration is sought.172 For a moderate fee, the Copyright Office will conduct a search and supply the application record.173 The information provided, however, will only be useful if the applicant was both honest and detailed in his or her application. Unfortunately, the directions on the application allow the arranger to respond very generally.174 More importantly, in many cases the requested information cannot be obtained in a timely fashion, or is too costly.
In economic terms, we have a problem with prohibitively high search costs. A music buyer is unlikely to spend significant time in the library or contact the Copyright Office in order to ascertain whether a particular arrangement is original unless a significant amount of money is at stake. As noted above,175 although a huge sum is at stake in the aggregate, each individual purchasing decision does not generally implicate sufficient savings in photocopying to justify incurring substantial search costs.
Several possible solutions suggest themselves. First, the Copyright Office, taking its cue from Patent Office practice, should require that prior art (the pre-existing composition) be submitted along with the derivative work sought to be registered. This would greatly facilitate the identification of precisely what an arranger has added to, or deleted from, an existing work. It would also clarify the nature of the claims made under the registration, because as a matter of law, material contained both in the pre-existing work and in the derivative work would not be protected. In addition, a composer should have to list all the elements of the new arrangement that she feels are original.176 An application for a derivative musical work would thus be like a patent application, where the inventor must specify precisely what claims he is making.
The Copyright Office or Congress could also employ the Internet to drastically reduce consumer search costs and improve the quality of administrative decision-making. The Copyright Office could, for example, establish a web site accessible on the Internet that would contain the information provided by applicants for copyrights in derivative musical arrangements. Consumers of public domain arrangements would then have instant computer access to the elements claimed as original in the arrangement. They could then make an informed choice whether to buy or to photocopy. Such a web site would likely be inexpensive to establish and maintain given current volume. The Office might even list information from past applications already on file. These administrative suggestions could eventually reduce the millions of dollars needlessly spent by public institutions, churches, and other consumers on unoriginal arrangements of public domain music.
Consumers of sheet music may also be able to force reform through litigation. I have previously outlined four causes of action that could successfully be brought against those who make spurious claims of copyright to public domain music.177 Consumer class actions based on breach of warranty, unjust enrichment, fraud, and consumer protection or false advertising statutes offer a promising means of dissuading those who profit from false claims. The implication of a private cause of action in the criminal prohibition of falsely affixing a copyright symbol is also a possibility.178 This statutory provision may provide yet another avenue of deterrence. Of course, private litigation would be unnecessary if the Department of Justice played a more active role in policing violations. A number of visible and substantial fines might provide the greatest disincentive of all against the plundering of the public domain. Sheet music without a copyright registration is hard to find now, but an aggressive litigation strategy may make the absence of the symbol on public domain works more common.
Although the Constitution, the Copyright Act, and judicial opinion clearly require arrangements of public domain music compositions to be original to receive protection from copying, music publishers' persistent claims to public domain works have created a climate where most consumers instinctively pay for new sheet music rather than copy it. High search costs and a dearth of appellate precedent make it difficult even for sophisticated consumers to investigate and challenge music publishers' spurious and very profitable claims to copyright. This Article has demonstrated, however, that caselaw, informed by copyright policy as explicated in Feist, can provide sufficient guidance-most changes that arrangers make to public domain compositions can be adequately judged by the musically-informed layperson to be original or not. The most persistent problem lies in the determination of exactly what changes have been made. Only the Copyright Office or Congress can provide sufficiently cheap access to this information.
When publishing companies attempt to extend the protective reach of copyright laws to their advantage, they rely on the moving metaphor of the Aspiring Artist.179 Even though the Supreme Court explains that copyright law makes reward to the owner a secondary consideration,180 it also recognizes that public welfare is advanced through the talents of authors.181 The Court sees the entire picture: the interests of both the public and the author are at stake in the debate over the parameters of copyright. The metaphor of the Aspiring Artist, then, only captures half of the story.
Other stories need to be told. The public domain will not continue to be robust if it remains a conceptual abstraction-direct and indirect consumers of music have little voice when the public domain is balanced against the more tangible interests of the Aspiring Artist. For this reason, the public domain needs its own avatar in the public debate. As this Article has demonstrated, the Choir Director and Musical Arranger are viable candidates. The Movie Director, borrowing liberally from public domain scènes à faire, plot twists, and stock characters, may be another. The Digital Future Coalition, a loose association of computer programmers and others with information age interests, nominates the Internet Surfer.182
Convincing economic argument in favor of protecting the public domain is one tool of persuasion. The music publishing industry, however, is best confronted by another-specific examples of the way in which the public domain benefits individuals. The experience of the Choir Director is a good place to start.