Robert Brauneis Has Copyright Law in the United States Ever Aided Discrimination on the Basis of Race, Or Exploitation of Or

Has copyright law in the United States ever aided discrimination on the basis of race, or exploitation of or prejudice against racial minorities? Copyright legislation in the United States has never explicitly incorporated racial categories. No federal or state copyright statute has ever withheld copyright protection from authors of a particular race, or imposed different fees on authors or owners of different races, or otherwise explicitly used racial categories. Neither, as far as I know, has any judicial decision concerning copyright law ever explicitly discriminated against a racial minority. However, that does not exhaust the possibilities of copyright law aiding discrimination against or exploitation of minorities. Particular features or doctrines of copyright law may arguably facilitate racial discrimination or exploitation in a more focused manner. K.J. Greene and Candace Hines have argued that copyright law’s requirement of fixation, and particularly its former requirement of written notation for music, has denied protection to African American oral culture.1 They have also each argued that copyright law’s requirement of locating an individual author clashes with African American traditions of communal creation.2 In addition, Greene has argued

 Professor of Law and Co-Director of the Intellectual Property Law Program, The George Washington University Law School. I would like to thank the participants in the Second Annual Mosaic Conference, co-sponsored by the Intellectual Property Program, Marquette University Law School and the Institute for Intellectual Property and Social Justice, and particularly my co-panelists, Jasmine Abdel-Khalik, Cheryl Slay Carr, and Betsy Rosenblatt. I would also like to thank Peter Menell, Jennifer Urban, and all of the students in the Berkeley Law and Technology Scholarship Seminar. For their helpful comments, I would like to thank Charles Cronin, Jennifer Mbagwu, Kip Lornell, and Bhamati Viswanathan. I would like to dedicate this essay to my mother, Geraldine Ann Brauneis, 1930-2016, who believed deeply in and worked tirelessly to further social justice. 1 See K.J. Greene, Copyright, Culture & Black Music: A Legacy of Unequal Protection, 21 Hastings Comm./Ent. L.J. 339, 360, 378-379 (1999) (hereinafter “Copyright, Culture & Black Music”); K.J. Greene, “Copynorms,” Black Cultural Production, and the Debate over African-American Reparations, 25 Cardozo Arts & Ent. L. J. 1179, 1201 (2008) (hereinafter “Copynorms”); Candace G. Hines, Black Musical Traditions and Copyright Law: Historical Tensions, 10 Mich. J. Race & L. 463, 469-470 (2005). 2 See K.J. Greene, Copynorms, supra note 1, at 1201; Hines, Black Musical Traditions, supra note 1, at 469-470. For a contrasing perspective, see Olufunmilayo B. Arewa, Blues Lives: Promise and Perils of Musical Copyright, 27 Cardozo Arts & Ent. L.J. 573, 582 (2010) (“[T]he tendency to see blues music as a primitive form of

Electronic copy available at: https://ssrn.com/abstract=3591113 that previously required copyright formalities such as notice and registration particularly disadvantaged African American authors, who were less likely to have access to legal information,3 and that the idea/expression doctrine has denied musicians protection for their signature styles. 4 This essay follows in those footsteps, and investigates another set of copyright doctrines that operated to the disadvantage of minority creators. Those doctrines concern copyright protection of musical works and sound recordings, and operated to deny composers and musicians rights in what they added to songs during the process of recording versions of them. One goal of this essay is historical recovery. I will show that copyright doctrines that are facially race-neutral were formulated in a racially-charged context in which it was clear the doctrines were going to disparately impact African- American composers, musicians, and vocalists. And disparately impacted they were. Many record companies made dozens of commercially successful “mirror cover recordings” that featured white vocalists and musicians who copied without permission, compensation, or attribution music created by African-American musicians and vocalists. This recovery of this history also confronts us with the question of what we should do when adoption of a rule in copyright law facilities discrimination against or exploitation of racial minorities, or by extension, discrimination against or exploitation of other minorities or vulnerable groups. Should we leave matters of discrimination to general antidiscrimination law, or is there a place for considering disparate impacts when formulating facially neutral doctrines in copyright law? I will argue that there are good reasons for taking disparate impacts into consideration when formulating copyright policy. Part I provides an introduction to the historical and doctrinal context necessary to understand copyright, music and race in much of the twentieth century. Those details include the construction of the racial categories that were used in the marketing of sound recordings, and the operation of the mechanical compulsory license. Part II introduces the phenomenon of white “mirror covers” of songs originally recorded by African Americans – covers that were not just recordings of the same song, but also copied some or all of the musical arrangement and stylistic elements introduced on the original recording. Part III focuses directly on the litigation of and decision

collective folk production reflected widespread stereotypes about African Americans and was part of a conceptual framework of later borrowers that facilitated the free borrowing of such music, often without attribution, let alone compensation.”). 3 See K.J. Greene, Copyright, Culture & Black Music, supra note 1, at 353-354, K.J. Greene, Copynorms, supra note 1, at 1201-1202. 4 K.J. Greene, Copyright, Culture & Black Music, supra note 1, at 383; K.J. Greene, Copynorms, supra note 1, at 1201.

Electronic copy available at: https://ssrn.com/abstract=3591113 in Supreme Records, Inc. v. Decca Records, Inc,5 a landmark case concerning the protection of initial recorded versions of songs against “mirror covers” of those songs. It also describes the impact of the Supreme Records decision, both as judicial precedent and as a model for legislation adopted as part of the Sound Recordings Act of 1971 and the Copyright Act of 1976. Part IV describes the treatment of covers under current law. Part V discusses whether there are policy lessons to be drawn from this history. Part VI concludes.

I. THE HISTORICAL CONTEXT: SHEET MUSIC, RECORDINGS, THE MECHANICAL COMPULSORY LICENSE AND THE RACIAL SEGREGATION OF MUSIC MARKETS In the nineteenth century, music publishers made money from songs by selling sheet music. Recordings of songs were not widely commercially distributed until the early years of the twentieth century. Following decisions in several lower courts, the Supreme Court held in 1908 that the Copyright Act only regulated the making of copies of a song, and that a record was not a copy of a song, since it was just part of a machine that could render an audible performance of the song.6 Under that decision, the owner of copyright in a song had no control over the making of records of that song, and received no compensation from the sale of those records. Within a little over a year, Congress legislatively overruled that decision in part, and extended the exclusive rights of copyright owners in musical works to “mechanical” renderings of musical works. That category encompassed phonographs, player pianos, music boxes, and their exchangeable parts – including phonograph records. The rules regarding those mechanical renderings, which we would today call sound recordings, turned out to have several complexities, and those complexities have only multiplied over time. First, even though a mechanical rendering or sound recording could infringe copyright in a musical work, it was not itself protected by federal copyright law. Congress did not grant federal copyright protection to sound recordings until 1971.7 Second, the owner of copyright in a musical work had complete control over the making of the first sound recording of that work. It could refuse to license any sound recording; it could demand any royalty it wanted; and it could place any conditions it wanted on permission to record. Once the copyright owner licensed the first recording, however, the musical work became subject to a “statutory license,”8 which persists in modified form to this day.9 Anyone

5 90 F.Supp. 904 (S.D. Cal. 1950). 6 See White-Smith Music Pub. Co. v. Apollo Co., 209 U.S. 1 (1908). 7 See Sound Recording Act of 1971, Pub. L. No. 92-140, 85 Stat. 391 (1971). 8 See Act of Mar. 4, 1909, ch. 320, § 1(e), 35 Stat. 1080. 9 See 17 U.S.C. § 115.

Electronic copy available at: https://ssrn.com/abstract=3591113 else who wanted to could make another recording of that musical work, upon payment of a fee that was determined by the statute itself. That fee, 2 cents for every record sold, astoundingly remained unchanged from July 1, 1909 to January 1, 1978.10 Because the value of the dollar in 1978 was roughly only 15% of what it was in 1909, the compulsory license rate had by 1978 dwindled to about 1/3 of one cent in 1909 dollars.11 The standard explanation of why Congress created the mechanical compulsory license is that it was concerned about the lack of competition in the player piano roll market after one manufacturer, the Aeolian Company, had tried to corner that market by entering into exclusive licenses with most major music publishers.12 When it came to phonograph records, however, music publishers were already used to licensing many recording companies to create different recordings of a song. Consumers, who were purchasing a tinny, scratchy, “low-fi” novelty, were usually not that concerned about which recorded version of a song they purchased: [I]n the early decades of the twentieth century records still played a relatively minor role in popular music, and people thought of them very differently than we do today. . . . Some bands and artists had devoted fans, but most customers were apparently still shopping for songs rather than specific performances: they would hear [a performance of a song], go to a music store, and ask for a record of it. [Bandleader Paul] Whiteman’s name was regarded as a guarantee of quality, but if the Victors were out of stock, most people were happy to go home with an alternate version on

10 See Act of Mar. 4, 1909, supra note 8, § 1(e); Neela Kartha, Digital Sampling and Copyright Law in a Social Context: No More Colorblindness!, 14 U. Miami Ent. and Sports L. Rev. 218, 234 (1997) (noting that the compulsory mechanical license rate remained the same for decades, and arguing that that eroded copyright protection, particularly for African American composers). The Copyright Act now provides that the statutory license fee is to be set by an administrative tribunal, the Copyright Royalty Board (CRB). See 17 U.S.C. §§ 115(c)(3), 801(b)(1). As of this writing in early 2020, the CRB has set rates for physical copies and permanent digital downloads at the greater of 9.1 cents per copy/download or 1.75 cents per minute of playing time. See Library of Congress, Copyright Royalty Board, Determination of Royalty Rates and Terms for Making and Distributing Phonorecords (Phonorecords III), 84 F.R. 1918, 2035 (February 5, 2019). By now, however, a much larger percentage of recording industry revenue is derived from streaming, and streaming rates are set as more complicated percentages of streaming service net revenues. See id. at 2035-2036. 11 See Bureau of Labor Statistics, CPI Inflation Calculator, http://data.bls.gov/cgi- bin/cpicalc.pl (that calculator goes back only to 1913, so I used 1913 as a rough proxy for 1909). 12 See, e.g., Geoffrey P. Hull, Thomas William Hutchinson & Richard Strasser, The Music Business and Recording Industry: Delivering Music in the 21st Century 78 (3rd ed. 2011).

Electronic copy available at: https://ssrn.com/abstract=3591113 Brunswick, Okeh, or Columbia.13 When record companies began to sell records to the public in the early 1900s, they “courted the total, undifferentiated American public.”14 Their catalogs arranged records by genre – “Dance Hits,” “Marches,” “Novelty,” “Classical,” and so on – but they assumed that any consumer might be interested in any or all of these genres. By the 1920s, however, record companies began to segment markets along racial and social lines.15 In particular, “race” records, contemplated as being made by and for African Americans, and “hillbilly” or “old-time” records, contemplated as being made by and for white, rural Southern Americans, were marketed separately from “popular” records, the main, residual category made by and for whites generally. Frank Walker, an executive with Columbia Records in the 1920s (and later with RCA Victor and MGM Records) may have put it most tellingly. The category of “race records,” he said, was created “in order to have a differentiation between that and normal phonograph records.”16 As Karl Hagstrom Miller writes, Separate catalogues suggested a correspondence between consumer identity and musical taste, one that was both holistic and exclusive. They implied that unique segments of the population were satisfied by particular kinds of records and yet uninterested in others. . . . The significant variety of musical sounds and styles that stuffed race record catalogues were held together and advertised as a cohesive whole by the fact that they were created by African American musicians. Companies, however, often refused to allow African American artists to record selections the musicians held dear – from pop songs and arias to hillbilly breakdowns – but did not fit within corporate conceptions of black music.17 William Howland Kenney adds: “From the time of Mamie Smith’s 1920 blues recording, white record entrepreneurs wanted African Americans to sing the blues. They insisted on that genre to the nearly total exclusion of the popular songs that appealed to large numbers of whites and at least some

13 Elijah Wald, How the Beatles Destroyed Rock ‘n’ Roll: An Alternative History of American Popular Music 85 (2011). 14 Karl Hagstrom Miller, Segregating Sound 188 (2010); see Arewa, Blues Lives, supra note 2, at 592-596: Promise and Perils of Musical Copyright, 27 Cardozo Arts & Ent. L.J. 573 (2010) (discussing the categorization of music along racial lines and the development of the genre of “race music”). 15 The use of the term “race” to denote a category of records originated with the label Okeh Records in 1923. See Jonathan Karp, Blacks, Jews, and the Business of Race Music, in Chosen Capital: The Jewish Encounter with American Capitalism 141, 144 (Rebecca Korbin, ed., 2012). 16 Quoted in William Howland Kenney, Recorded Music in American Life: The Phonograph and Popular Memory, 1890-1945 130 (1999). 17 Karl Hagstrom Miller, supra note 14, 188-189.

Electronic copy available at: https://ssrn.com/abstract=3591113 Blacks.”18 When Billboard magazine started publishing charts of best-selling records in the 1940s, it used the same categories, charting sales of “popular” records separately from “race” records and “folk” records. Many “race” records were produced by smaller, independent labels, particularly during and after World War II, when shellac shortages caused major labels to concentrate on their core white audiences.19 When major recording companies marketed records by and for African Americans, they usually did so through separate subsidiaries or labels, or using separately marketed series. Columbia used its subsidiary “Okeh”; Mercury used its subsidiary “Wing”; RCA used its subsidiaries “Groove” and “Bluebird”;20 MGM had its “Ebony Series”;21 Decca and Capitol released “Sepia” records.22 Although Billboard decided in June 1949 to change the name of the “race” category to “rhythm and blues” – “R&B” for short – everyone understood that the new term was still “a code word for black music,”23 music by and marketed to African Americans. As Brian Ward has put it, The marginalization and oppression of peoples of African descent in America has always been more than a purely legal, political, economic, and social phenomenon. It has involved an integrated system of thought, categorization and action which constitutes the fundamental grammar of American racism. As part of that system, the recording and broadcasting industries did not merely reflect the prevailing racial assumptions of the 1950s, they internalized them, functioned according to their dictates, and, in so doing, helped to perpetuate them. Racial conventions permeated the organization and structure of the music industry at every level. The very existence of separate “Race,” and from 17 June 1949, “Rhythm and Blues” charts for black popular music, symbolized the routine segregation of blacks in American society as much as the segregated schools and separate drinking fountains of the south, or the restrictive housing covenants and discriminatory hiring practices of the North.24

18 William Howland Kenney, Recorded Music in American Life, supra note 16, at 132. 19 See Reebee Garofalo, Crossing Over, supra note 120, at 116-117. 20 See Brian Ward, Just My Soul Responding: Rhythm and Blues, Black Consciousness, and Race Relations 27 (1998). 21 See, e.g., The Billboard, Oct. 23, 1948, at 26 (advertisement for MGM Records). 22 See id. at 29 (advertisement for Capitol Records); id. at 23 (advertisement for Decca Records). 23 Reebee Garofalo, Crossing Over, supra note 120, at 121. 24 Brian Ward, supra note 20, at 27-28. For further explorations of dividing musical genres by race, see Jo Haynes, Music, Difference and the Residue of Race (2013); Jack Hamilton, Just Around Midnight: Rock and Roll and the Racial Imagination (2016).

Electronic copy available at: https://ssrn.com/abstract=3591113 II. “MIRROR COVERS” AND RACE As noted above, from very early in the history of sound recordings, it was common for many record companies to record and release different versions of the same song, a phenomenon that was subsequently supported by Congress’s provision of a compulsory license in the 1909 Act. In racially divided markets, however, a very particular form of the phenomenon of multiple recordings developed. James M. Salem provides one description of this phenomenon: An original performance by a black R&B artist was released by a small, independent (sometimes black-owned) record company. Then it was covered by a white performer’s version of the song, released by a large, white-owned major record company with the capability for national distribution and promotion. The white version was not so much a cover as a copy, an attempted duplication of not only the melody of the song but the musical voicings and rhythmic quality of the arrangement, plus the singer’s distinctive vocal style as well in many cases.25 In some cases, the white version did make some changes to the original R&B release. It often aimed for a somewhat “smoother,” “less raw” musical sound, or it substituted somewhat more innocuous lyrics for language that was considered too racy.26 Even when a cover version made changes, however, it still copied many features that were not present in the original notated song but had been introduced by African-American composers, arrangers, and recording artists while preparing to record and recording the song. In discussion of what would eventually become the Sound Recording Act of 1971 and the Copyright Act of 1976, these recordings were referred to as “imitation” or “mirror records,”27 and so I will call them “mirror cover recordings” or “mirror covers” here. Why were white recording artists copying performances by African- American recording artists? Because it paid, for reasons that can be traced more or less directly back to race. As James M. Salem noted above, one reason it paid was that many of the record companies that released records by African American recording artists were small, local companies that did

25 James M. Salem, The Late Great Johnny Ace and the Transition from R&B to Rock n’ Roll 168 (1999). For previous discussions in legal academic literature of white cover recordings, see Candace G. Hines, supra note 1, at 484-486; Neela Kartha, supra note 10, at 232-234. 26 For example, Pat Boone, a white singer who achieved great success recording cover versions of earlier R&B releases, stated that when covering “Tutti Frutti,” earlier recorded by Little Richard, he changed “Boys you don’t know what she do to me” to “Pretty little Susie is the girl for me” in order to “make it more vanilla.” David O. Szatmary, Rockin’ in Time: A Social History of Rock-and-Roll 26 (8th ed. 2014). 27 See infra notes 100-101 and accompanying text.

Electronic copy available at: https://ssrn.com/abstract=3591113 not have developed national distribution channels. Major companies could rerecord a song, copying features of the small company’s recording, and distribute their versions nationwide. Yet that is an incomplete account, because it does not explain why the major companies decided that they needed to produce recordings featuring white artists. That was likely due to a confluence of many factors, all of which manifested racial prejudice in one way or another. Most directly, many members of the record-buying public would not buy recordings by African Americans. In fact, the prejudices ran so deep that record companies who marketed to whites would often decide not to produce any records featuring African Americans, or at least, as noted above, to brand those records differently. As Perry Bradford, an African American composer, musician and producer active beginning in the 1910s, put it, “‘Victor [Talking Machine Company] just couldn’t afford to lower their prestige’ by issuing records by Black artists.”28 Those prejudices might be further heightened when the recording artists were visible, as they might be on the covers of sheet music, or later, on television. In addition, the sale of records was promoted chiefly through play on radio stations, and many radio stations did not play recordings by African American artists. In part, that was due to the racial prejudices of their managers or employees, or due to their deference to audiences that they believed harbored those prejudices. In part, it was because the American Society of Composers, Authors, and Publishers (ASCAP), the first U.S. performing rights organization, put pressure on radio stations to play only songs that it licensed.29 Since ASCAP had excluded all but a few African American composers from membership, those songs were more likely to have been composed, and performed, by whites.30 Could white recording artists make cover versions of R&B recordings, copying features created for the first recordings of the songs, without the permission of the composers, arrangers, and musicians who created the first recordings, or the companies that produced them? That question was not settled when, in November of 1948, a company called Supreme Records sued Decca Records and Capitol Records over their cover versions of a song titled “A Little Bird Told Me,” previously recorded by Supreme. The litigation, brought in the Southern District of California, was quickly recognized as a test case. Decca, reported Billboard, “put its big legal guns”

28 William Howland Kenney, supra note 18, at 113 (quoting Perry Bradford . . . 29 See, e.g., Charlie Gillett, The Sound of the City: The Rise of Rock and Roll 21 (Da Capo 1996). 30 See Reebee Garofalo, Crossing Over: From Black Rhythm & Blues to White Rock ‘n’ Roll, in Rhythm and Business: The Political Economy of Black Music 112, 113 (2002) (“Of [ASCAP’s] 170 charter members, six were black . . . . [T]he vast majority of ‘untutored’ black artists were routinely excluded from the society and thereby systematically denied the full benefits of copyright protection.”).

Electronic copy available at: https://ssrn.com/abstract=3591113 to work on the case “because of its importance as a precedent-setting suit.”31 And indeed, the District Court’s decision in the case, in favor of defendant Decca (Capitol having settled), has had lasting effect, both as judicial precedent and as a model for a provision in the Sound Recording Act of 1971 and the Copyright Act of 1976. Thus, the facts and legal analysis in the case, and its impact, are worth a close look, and that is the task to which the next Part turns.

III. SORTING THROUGH THE SUPREME RECORDS DECISION: FACTS, LAW, AND IMPACT A. The Facts The dispute in Supreme Records begins with a song called “A Little Bird Told Me,” written by successful African-American composer and pianist Harvey Oliver Brooks. Brooks was born in 1899, and toured and recorded with blues singer Mamie Smith in the 1920s.32 He then settled in Los Angeles, and became the first African-American to write a complete score for a major motion picture, namely, “I’m No Angel,” the most popular movie of 1933, starring Mae West and Cary Grant.33 On October 27, 1947, Brooks registered a claim of copyright in “A Little Bird Told Me” as an unpublished work.34 The version of the song that was protected by federal copyright was therefore the version that Brooks had deposited with his registration application. We don’t have the deposit copy, but apparently it was the same as the manuscript that the court accepted as representing the song during the litigation.35 In 1947, Albert Patrick, an African American entrepreneur who was also a dentist, founded Supreme Records, a small, independent record label based in Los Angeles.36 He and Leroy Whyte, a bandleader, arranger, and trumpeter,37 recruited Paula Watson, a singer and pianist, to record for Supreme, and they began to search for suitable songs for her. In the early

31 Billboard, May 13, 1950, at 12. 32 See Leonard Feather & Ira Gitler, The Biographical Encyclopedia of Jazz 69 (2007). 33 See Wikipedia, “Harvey Oliver Brooks,” https://en.wikipedia.org/wiki/Harvey_Oliver_Brooks (last visited December 11, 2019). 34 See U.S. Copyright Registration No. EU101224 (Oct. 27, 1947), available at http://www.robertbrauneis.net/brooksregistration.pdf. 35 That manuscript is available at http://www.robertbrauneis.net/covers/brooksscore.pdf. 36 Deposition of Albert Patrick, February 3, 1950 and February 14, 1950, Supreme Records, Inc. v. Decca Records, Inc., No. 8929-Y, United States District Court, Southern District of California, Central Division, p.10, available at http://www.robertbrauneis.net/covers/patrickdeposition.pdf. 37 See Peter Vacher, Swingin’ on Central Avenue: African American Jazz in Los Angeles 175 (2015).

Electronic copy available at: https://ssrn.com/abstract=3591113 summer of 1948, Patrick, Whyte and Watson went to visit Harvey Brooks at his house.38 Brooks played “A Little Bird Told Me” for them, and Watson tried playing and singing it. The group decided that it would be a good song for Watson to record. Whyte, Patrick, and Watson spent some time at Patrick’s house working out how to arrange the song for the recording. The arrangement that Whyte ended up writing included new lyrics and melody for a male chorus to sing in response to the Brooks-composed lyrics that Watson would sing.39 The Whyte arrangement also added an introduction hummed by the male chorus; a distinctive repeating two-bar bass line on the first and third beats that alternated with hand claps on the second and fourth beats; and four added bars of music.40 Whyte wrote those elements on individual notated parts that he distributed to the musicians who recorded the song for Supreme.41 On the Supreme recording, which was made sometime in July of 1948, Paula Watson sang and played the piano. Her backing band included Tiny Webb on guitar, Jesse Sailes on drums, Chuck Hamilton on bass, and Maxwell Davis and Pete Peterson on saxophones.42 The instrumentalists also sang as the male chorus.43 Supreme Records obtained a license to record the song directly from Harvey O. Brooks44 and released its recording in August 1948, as the B-side of a 78-rpm record that included “Pretty Papa

38 Deposition of Albert Patrick, supra note 36 , at 7-9. 39 Reporter’s Partial Transcript of Proceedings, May 3, 1950, Supreme Records, Inc. v. Decca Records, Inc. No. 8929-Y, United States District Court, Southern District of California, Central Division, pp.11-14 (cross-examination testimony of Leroy Whyte) (hereinafter “Whyte Cross-Examination”), available at http://robertbrauneis.net/covers/whytecross.pdf. Whyte’s score, as it appears in the District Court record, is available at http://www.robertbrauneis.net/covers/whytescore.pdf. Table 1, available online, see http://www.robertbrauneis.net/covers/lyricstable.htm, shows on the left side the words that are in the Brooks manuscript, and on the right side the additional words added by Whyte that are sung in both the Supreme (Watson) recording and the Decca (Knight) recording. 40 Id. at 16. 41 Id. at 11-14. It is not clear whether everything that Whyte notated in the individual parts distributed to the musicians is contained in the score that is in the District Court record. 42 Whyte Cross-Examination, supra note 39, at 21. 43 Id. 44 I have not been able to locate the license. However, when Harvey O. Brooks assigns copyright in “A Little Bird Told Me” to music publisher Bourne Inc. on September 15, 1948, Brooks represents that there has been no previous publication or mechanical license granted, “with the exception of one license to Supreme Records Company. Assignment of Copyright from Harvey O. Brooks to Bourne, Inc. of September 15, 1948, as recorded in the U.S. Copyright Office, Vol. 683, page 4, September 22, 1948, available at http://www.robertbrauneis.net/covers/brooksbourneassignment.pdf.

Electronic copy available at: https://ssrn.com/abstract=3591113 Blues” on the A-side.45 The brief Billboard review of the release provides further evidence of the pervasiveness of racial stereotypes. It deems Watson’s “A Little Bird Told Me” to be “too pop-flavored for the race lists,” but comments in connection with “Pretty Papa Blues” that “Miss Watson has the race manner . . . and may be heard from.”46 Decca unsuccessfully tried to buy the rights to the Supreme recording.47 It then proceeded to make and release its own recording.48 That recording featured Evelyn Knight, a white singer who had previously recorded two top ten hits for Decca, as the principal vocalist. A vocal group called the Stardusters performed the backup vocals.49 Decca released its recording in October 1948, less than two months after the Supreme release. On November 19, 1948, Bourne Inc. published the first sheet music for the song, and registered a claim of copyright in that version, which, typical for its time, was ‘[f]or voice and piano, with chord symbols.”50 The cover of that sheet music featured, not Paula Watson, but Evelyn Knight, a move that was standard for marketing to white audiences.51 Also in November 1948, Capitol Records recorded a version of “A Little Bird Told Me,” featuring Blue Lu Barker, an African-American singer from New Orleans.52 By December 1948, at least three other recordings of the song had been released.53 The Watson and Knight recordings reached the Billboard retail

45 As of this writing, the Supreme recording is available on YouTube at https://www.youtube.com/watch?v=KKLIxpUNHp0. 46 See Billboard, October 16, 1948, page 112, available at http://www.robertbrauneis.net/covers/supremereview.htm. 47 See Supreme Records, 90 F.Supp. at 913 n.4. 48 Harvey O. Brooks had assigned copyright in “A Little Bird Told Me” to music publisher Bourne Inc. on September 15. Bourne Inc. then filed the required “Notice of Use on Mechanical Instruments” that announced the availability of the song for statutory licensing. See http://www.robertbrauneis.net/covers/bournenoticeofuse.htm. 49 The Decca recording session was held on October 12, 1948. See Michael Ruppli, The Decca Labels: the eastern sessions 1943-1958 218 (1996). It was released very quickly; Billboard’s review appeared less than two weeks later, on October 23rd. See Billboard, October 23, 1948, page 37, available at http://www.robertbrauneis.net/covers/deccareview.htm. As of this writing, the Decca/Evelyn Knight recording is available at https://www.youtube.com/watch?v=Oe9RwAFVn9U. 50 U.S. Copyright Registration No. EP31690 (November 19, 1948), available at http://www.robertrbrauneis.net/covers/bourneregistration.pdf. 51 An image of the sheet music cover is available at http://www.robertbrauneis.net/covers/littlebirdsheetmusiccover.htm. 52 As of this writing, the Blue Lu Barker recording is available at https://www.youtube.com/watch?v=-JfEiT8SFMc. 53 See Billboard, January 8, 1949, at 18. The three additional recordings noted on in Billboard were by Smokey Rogers on Capitol and Varsity Records, available at https://archive.org/details/78_a-little-bird-told-me_smokey-rogers-harvey-o.- brooks_gbia0007232a; by Jerry Wayne and Janette Davis on Columbia Records, available at https://www.youtube.com/watch?v=bFiD00WsHQM; and by Rose Murphy

Electronic copy available at: https://ssrn.com/abstract=3591113 sales charts, 54 and those recordings and the Barker recording reached the Billboard radio play charts.55 The Supreme/Watson recording ended up selling well, about 250,000 copies,56 but the Decca/Knight recording sold much better, about two million copies.57 Supreme’s founder Albert Patrick believed, correctly, that Decca and Capitol had copied just about every feature that the Supreme recording had added to the Brooks composition – “an introduction, overlaps and handclapping, choral responses, a certain verbal deviation from the wording of the song, and the introduction of [several] bars of music at about the middle of the song.”58 By the end of November 1948, Supreme sued Decca and Capitol. Capitol eventually settled the suit, paying Supreme $1000, but with a provision requiring Supreme to refund that amount if Decca eventually prevailed.59 Decca did not settle, and the case went to trial. B. The Law Judge Leon Yankwich presided over the trial, and wrote the opinion deciding the case. Yankwich had already gained prominence in copyright law through his invention of the doctrine of “scenes á faire” in a 1942 opinion,60 and he would go on to write many articles in copyright and other legal fields.61

on Victor Records, available at https://www.youtube.com/watch?v=b_oHgWYvjlM. Those recordings vary in their similarity to the Supreme recording. The Rose Murphy recording, a jazz cover, is the least similar, with a different introduction, no backup singers, no handclapping, and different additional lyrics. 54 The Billboard “Retail Record Sales” and “Race Records” charts for February 12, 1949, on which the Knight and Watson recordings appear, respectively, are available online at http://www.robertbrauneis.net/covers/charts.htm. 55 See Billboard, January 29, 1949, p. 25 (“Records Most Played by Disk Jockeys” chart lists the Knight recording in first place, the Barker recording in seventh place, and the Watson recording in eighth place). 56 The estimate of 250,000 was produced by adding the month-by-month sales figures that the President of Supreme Records, Albert Patrick, provided in his deposition. See Deposition of Albert Patrick, supra note 36, at 18-23. Other estimates have varied widely. The court in Supreme Records provides the figure of 100,000, see Supreme Records v. Decca Records, 90 F. Supp. 904, 905 (S.D. Cal. 1950); Albin Zak states that the record eventually sold over a million copies. See Albin J. Zak III, I Don’t Sound Like Nobody: Remaking Music in 1950s America 143 (2010). 57 See “Evelyn Knight, 89; versatile singer had two no. 1 hits in 1949,” Los Angeles Times, October 31, 2007, available at http://articles.latimes.com/2007/oct/31/local/me- passings31.s1. 58 Supreme Records, 90 F. Supp. at 911. 59 See Affidavit [of Albert Patrick] in Support of Motion for Substitution of Attorneys, January 24, 1950, pp. 2-3, Supreme Records v. Decca Records, No 8929-Y (S.D. Cal.), available at http:/www.robertbrauneis.net/covers/patrickaffidavit.pdf. 60 See Cain v. Universal Pictures Co., 47 F. Supp. 1013 (1942). 61 See, e.g., Leon R. Yankwich, What is Fair Use?, 22 U. Chi. L Rev. 203 (1954); Leon R. Yankwich, Parody and Burlesque in the Law of Copyright, 33 Can. Bar Rev.

Electronic copy available at: https://ssrn.com/abstract=3591113 Well before trial, Supreme Records had agreed to a stipulation that its claim for relief was based solely on a theory of unfair competition, and not on either statutory (federal) copyright or the common law of literary property.62 Supreme most likely made that concession because its counsel viewed its claim as arising solely out of the performance of Paula Watson and her backup band as fixed in Supreme’s sound recording. As previously mentioned, before the passage of the Sound Recording Act of 1971, sound recordings did not enjoy federal copyright protection.63 As for common-law protection, there was a split of authority on whether public distribution of a sound recording ended that protection.64 Supreme’s recording had of course been publicly distributed, and Supreme may have believed that the Southern District of California would rule that that public distribution had terminated any common-law protection. As already mentioned, Supreme did not merely create a sound recording of “A Little Bird Told Me.” Leroy Whyte, presumably while in Supreme’s employment, created a notated arrangement of the song, and Decca’s recording used many elements added by that notated arrangement. Why didn’t Supreme sue for infringement of that arrangement? The notated arrangement did not enjoy federal copyright protection, because it had neither been registered as an unpublished work,65 nor published in notated form.66 What about common-law protection? As noted in the previous paragraph, there was a split of authority on whether public distribution of sound recordings that embodied a musical work, including a musical arrangement, was considered to be “publication” of that musical work that

1130 (1955); Leon R. Yankwich, Originality in the Law of Intellectual Property (Its Meaning From a Legal and Literary Standpoint), 11 F.R.D. 457 (1952). In his opinion in Supreme Records, he also delights in exhibiting his knowledge of the history of music and drama, filling footnotes with discussions of how “call and response” dates back to ancient Greece, medieval England, and Germany. Supreme Records, 90 F. Supp. at 911 n.2, 911 n.3. 62 See Supreme Records, 90 F. Supp. at 906. 63 See supra note 7 and accompanying text. 64 Compare Waring v. WDAS Broadcasting Station, Inc., 184 A. 631, 635-636 (Sup. Ct. Pa. 1937) (common-law rights in a performance fixed in a sound recording were not destroyed by public distribution of that sound recording) with RCA Mfg. Co. v, Whiteman, 114 F. 2d 86, 89 (2d Cr. 1940) (common-law rights in a performance were destroyed by public distribution of the sound recording in which it was fixed), cert. denied, 311 U.S. 712 (1940). 65 See Act of Mar. 4, 1909, ch. 320, § 11, 35 Stat. 1075, 1078 (providing for federal copyright protection for certain unpublished works, including musical compositions). 66 If the distribution of Supreme’s records had been ruled to publish Whyte’s arrangement, Supreme would have forfeited federal copyright in that arrangement, because the records did not display proper copyright notice. For images of the labels on the Supreme and Decca releases of “A Little Bird Told Me,” see http://www.robertbrauneis.net/covers/alittlebirdtoldmelables.htm; for the requirement of publication with notice, see Act of Mar. 4, 1909, ch. 320, § 9, 35 Stat. 1075, 1077.

Electronic copy available at: https://ssrn.com/abstract=3591113 would destroy common-law protection. In addition, although we don’t have the exact language of the license that Harvey O. Brooks granted to Supreme, it is quite possible that it referred to rights of mechanical reproduction, and would have been construed not to include the right to create a notated arrangement. At this point you may be asking yourself, if the Supreme Records case was not about copyright, why am I discussing it as if it were? The answer to that is at least twofold. First, although Judge Yankwich acknowledges that Supreme was making only an unfair competition claim, he discusses unfair competition not just as a matter of likelihood of confusion, but as a matter of protecting creative efforts assessed using copyright concepts like standards of originality. Thus, the opinion veers into quasi-copyright. Second, subsequent discussion of the case viewed it as having articulated a rule that should be adopted as federal copyright legislation, and it was eventually so adopted. Thus, the opinion was seen both by its author and by subsequent observers as determining the proper bounds of copyright-like protection. Pursuing first the likelihood of confusion inquiry, the opinion asks whether the Decca recording is so similar to the Supreme recording that “the average person who would listen to the two recordings . . . would confound [one] with the other.”67 There are some faults to find with Judge Yankwich’s analysis of this issue. Part of Yankwich’s discussion seems simply to ignore important similarities between the two recordings. He states that “even the most untrained ear would conclude that the only similarity discernible is that which comes from the identity of the melody.”68 Surely, however, if the “untrained ear” in question understood that the Supreme recording had added the backup singers who repeatedly respond to the lead singer’s lines, it would not fail to recognize that the Decca recording also had backup singers, and that they sang identical lines. Judge Yankwich is also unafraid to present his musical tastes as objective judgments, and to use them as a basis for declaring that the two recordings could not be confused. Those tastes decidedly weigh against the Supreme recording and in favor of the Decca recording. The Paula Watson recording, Yankwich declares, is “thick, mechanical, lacking inspiration, containing just the usual accompaniments and the usual intonations which one would find in any common recording.”69 By contrast, he asserts, the Decca recording “is rich, against a musically colorful background. It sounds full, meaty, polished. The difference derives from the different quality of the voices of the artists, the more precise, complex and better organized orchestral background, the fuller harmonization of the responses, [and] the

67 Supreme Records, 90 F. Supp. at 912. 68 Id. 69 Id.

Electronic copy available at: https://ssrn.com/abstract=3591113 clearer intonation and expression.”70 In light of the subsequent history of popular music these aesthetic judgments seem antiquated, and they also seem affected by what Brian Ward calls the “system of thought [and] categorization . . .which constitutes the fundamental grammar of American racism.”71 During the trial, counsel for Supreme objected to Decca’s attorney’s questions about the identities of and instruments played by each musician on the Supreme recording. Decca’s attorney replied: If your Honor please, one of our defenses is going to be, to show that one of these records was a record aimed for what is known in the record business as the race market, and the other record, our record, was aimed for popular consumption, and the instrumentation on these records was different because they were aimed for different markets.72 Judge Yankwich overruled the objection and let the questioning proceed. In his mind, apparently, there was a sufficient relationship between the identities of the musicians and racially divided markets in which one race was content to listen to arrangements that were “thick, mechanical, lacking inspiration” while the other was attracted by “rich . . . full, meaty, polished” music. While Yankwich’s aesthetic judgments and categorical pronouncements are deeply questionable, it is not clear that his conclusion as to likelihood of confusion is wrong. Even a listener who was more appreciative of the Supreme recording, and less appreciative of the Decca recording, might well not confuse them. Thus, if the unfair competition analysis was limited to confusion, Supreme could well have lost no matter who the finder of fact was. However, Yankwich’s opinion reaches out to consider the question of whether the elements that the Decca recording copied from the Supreme recording should be accorded some other form of protection. And although Supreme had given up any claims to common-law or statutory copyright, Yankwich’s analysis uses concepts and precedents derived from copyright, and appears to declare Supreme’s creative additions as unqualified for copyright protection. As we will see, that is how the opinion is received in the recording industry – as granting license to freely copy anything added to a song in a recording – and that is how it is eventually enshrined in legislation. In one passage, Yankwich appears categorically to reject any further protection, concluding that “a mere recording of an arrangement of a musical composition by one who is not the author of the composition is

70 Id. 71 Brian Ward, supra note 20, at 27-28. 72 Whyte Cross-Examination, supra note 39, at 10-11.

Electronic copy available at: https://ssrn.com/abstract=3591113 [not] a property right which should be given recognition in equity.”73 That rejection is based largely on analogizing the creative elements added in the Supreme recording to acting performances. Yankwich is correct that under leading precedent, then as now, “mere portrayal of a character by an actor in a play, which is the creation of another, is not, of itself, an independent creation.”74 Elements such as “gestures or motions of actors,”75 their facial expressions, the timbre of their voices, their pronunciation and rhythm when delivering lines, are not considered to create an independently copyrightable work.76 It is also true that some of what Watson and the other musicians added to the Brooks composition could similarly be considered to be “performance elements.” Those would include elements such as the timbre of Watson’s voice, and her timing and pronunciation when singing. However, those “performance elements” hardly exhaust what the Supreme recording added to the Brooks composition. As noted above, Leroy Whyte produced a notated version of “A Little Bird Told Me” that included a new male choral part with new melodies and lyrics; new music for the instruments in the band, including a new bass line; and four new bars of music in the middle of the song. None of those are mere uncopyrightable “performance elements,” but are instead components of a potentially copyrightable musical work – a derivative work with respect to Brooks’s original composition. Perhaps in tacit recognition that Supreme Records’s contributions to its recorded version of “A Little Bird Told Me” were not limited to performance elements, Judge Yankwich also articulates a standard of creativity that must be met for protection, and explains why the Supreme Records arrangement does not meet that standard. The standard of creativity is an extremely high one that most readers will be familiar with from patent law. Yankwich begins by declaring that the elements added to “A Little Bird Told Me” on the Supreme recording “are well known in the art,”77 patent language for expressing a lack of novelty. That is certainly true at the high level of abstraction that Yankwich describes them – “handclapping” and “choral responses” as such. However, he then suggests that, since each element separately lacks novelty, the arrangement can be protected only if the “combination of old elements . . . achieve[s] a better or different result.”78 Without much further analysis, he declares that the combination “did not produce any distinct result in this case.”79 He also states that the elements introduced on the Supreme recording are of “a type

73 Id. at 908. 74 Id. 75 Id. 76 See, e.g., Garcia v. Google, Inc., 786 F.3d 733, 741-743 (9th Cir. 2015). 77 Id. at 911. 78 Id. at 913 n.4. 79 Id.

Electronic copy available at: https://ssrn.com/abstract=3591113 which would occur to any arranger, a kind of musique a faire”80 – a formulation that invokes the patent standard of nonobviousness, and that interestingly extends Yankwich’s own “scenes a faire” doctrine to music. Given the low “modicum of creativity” standard that we now generally associate with copyright81 and supporting precedent for that standard that dates back to well before Supreme Records,82 Judge Yankwich’s application of a patent standard may seem abnormal. In the context of musical arrangements and performances, it is not quite as aberrational as it appears. As Joseph Fishman has detailed,83 there is a small line of cases running from 1851 up through the twentieth century that applies a heightened standard of originality to musical arrangements.84 There was precedent for applying such a standard to performances as well.85 Although Judge Yankwich does not cite any of these cases, he may have known of them. Even those cases, however, are limited to music – to the addition or variation of notes and chords to be played or sung – or to various “performance elements.” By contrast, Yankwich seems to apply the high standard even to the new text that Whyte wrote, noting that it consists in part of words that were supposedly commonplace at the time, namely, “certain expressions which have been referred to as jive expressions, such as yeah, dog, solid and the like.”86 Thus, Yankwich ends up applying a uniquely high standard of originality to deny protection to all of the new creation embodied in the Supreme recording.

80 Id. 81 Feist Publications v. Rural Telephone Co., 499 U.S. 340 (1991). 82 See, e.g., Bleistein v. Donaldson Lithographic Co., 188 U.S. 239, 250 (1903) (“[A] very modest grade of art has in it something irreducible, which is one man’s alone”). 83 See Joseph Fishman, Originality’s Other Path (SSRN or published cite?) 84 See, Jollie v. Jaques, 13 F. Cas. 910 (C.C.S.D.N.Y. 1850) (No. 7,437) (piano arrangement of piece originally composed for clarinet); Cooper v. James, 213 F. 871, 872 (N.D. Ga. 1914) (fourth vocal line added to pieces originally written or arranged for three voices); Norden v. Oliver Ditson, 13 F. Supp. 415 (D. Mass. 1936) (changes in music to accommodate English translation of original Russian lyrics); Woods v. Bourne Co., 841 F. Supp. 118 (S.D.N.Y. 1994), rev’d in part on other grounds, 60 F.3d 978 (2d Cir. 1995) (piano arrangement of original melody). 85 In Waring v. WDAS Broadcasting Station, Inc., 184 A. 631 (Sup. Ct. Pa. 1937), the Pennsylvania Supreme Court applies a similarly heightened standard to performances captured in sound recordings. The Waring court comments that “the ordinary musician” would not necessarily gain a property right in his or her performance of a musical work, since such a musician “does nothing more than render articulate the silent composition of the author.” Id. at 634-635. However, the court finds that the plaintiff in the case, Fred Waring, had gained such a right, because he had “contribute[d] by his interpretation something of novel intellectual or artistic value.” Id. at 635. 86 Id. at 912.

Electronic copy available at: https://ssrn.com/abstract=3591113 C. The Impact Leon Yankwich’s opinion in Supreme Records is not a model of either factual or legal clarity, and it could conceivably have remained an obscure, one-off decision. In fact, however, the decision had an impact that was both immediate and lasting. The immediate impact on Supreme Records was to render it insolvent, and cause it to cease operations. Supreme had already sold Paula Watson’s recording contract to Decca in late 1949.87 Watson, who had recorded a total of eight songs for Supreme in 1948, went on to record six songs for Decca in late 1949 and early 1950, and four more songs for MGM in 1953.88 She then moved to Europe, where she remained the rest of her life. Evelyn Knight, who was ten years older than Watson, had her last recorded hit in 1951, and retired from performing by 1954.89 The impact of the Supreme Records decision on the greater record industry was, as Larry Starr and Christopher Waterman put it, to “ope[n] the floodgates for cover versions during the 50s.”90 The case was commonly understood in the music industry to hold that anything added to a song by arrangers, musicians and vocalists while making a recording was not protected against copying. A Billboard magazine article published days after the release of the Yankwich opinion summarized its principal holding in these two sentences: “Judge [Yankwich] ruled that only a copyright owner holds [property] rights for original music, but that other arrangements are not considered as property. Moreover, after hearing both platters in court, Judge Yankwich ruled that there was nothing original in Supreme’s arrangement which could be made the subject of property ownership.”91

87 See Chip Deffaa, Blue Rhythms: Six Lives in Rhythm and Blues 152 (1999). 88 See Blues & Rhythm Series Classics: Paula Watson 1948-1953, Classics No. 5104 (booklet essay by Dave Penny) (2004). 89 See Matt Schludel, Evelyn Knight; Torch Songstress Began Career in D.C. Clubs, Washington Post, October 28, 2007, available at https://www.washingtonpost.com/wp- dyn/content/article/2007/10/27/AR2007102701264.html. 90 Larry Starr & Christopher Waterman, American Popular Music from Minstrelsy to MTV 196-197 (2003). For other acknowledgement of the impact of the Supreme Records decision, see Russell Sanjek, American Popular Music and Its Business: The First Four Hundred Years, Volume III From 1900 to 1984 325 (1988) (“Any legal barrier to the practice [of producing covers that were “almost exact musical duplications”] was removed by the courts . . . with a decision in the ‘Little Bird Told Me’ case. . . . [The] verdict . . . declared that musical arrangements were not copyrighted property and therefore not subject to the law’s protection.”); Brian Lukasavitz, Blues Law: A Little Bird Told Me about Cover Versions, American Blues Scene Magazine (2012), available at http://www.americanbluesscene.com/2013/01/blues-law-a-little-bird-told-me-cover- versions/ (“Finding for Decca records, the [Supreme Records] Court paves the way for what would become an endless number of cover version recordings of blues and R&B tunes for the pop music consumers.”). 91 Billboard, May 13, 1950, at 12.

Electronic copy available at: https://ssrn.com/abstract=3591113 From the current perspective of well into the twenty-first century, it may be difficult to imagine a musical culture like that of the 1950s and early 1960s, when white “mirror covers” of African American recordings became ubiquitous. Some readers may recall that Elvis Presley, Bill Haley, the Beatles, the Rolling Stones, and Eric Clapton all made hits by covering R&B recordings. However, other recording artists who are less known now but were extremely popular in their heyday, including Pat Boone, Georgia Gibbs, the McGuire Sisters, and the Crew Cuts,92 built their careers on such covers. At least one entire record company, Dot Records, focused much of its energies on making and selling white covers of R&B hits.93 In 1955, African American singer LaVern Baker became particularly angry when Georgia Gibbs made a recording of a song called “Tweedle Dee” that copied the arrangement in Baker’s earlier recording “note for note,” and that became much more successful than Baker’s recording.94 When Baker wrote a letter to Representative Charles Diggs of Michigan calling for an amendment to the Copyright Act to protect singers from “modern-day pirates,” she cited the Supreme Records case as evidence that current law would not protect them: To illustrate the practice [of duplication of successful disks], the letter mentioned the case of Paula A. Watson, who recorded “A Little Bird Told Me” on Supreme in 1948 – a disk which was “copied in its entirety by Evelyn Knight on Decca.” In the ensuing test case, the court held that the Copyright Act protected a song – but not an arrangement. . . . “Since there is no court to uphold my right [,” Miss Baker writes, “] maybe my plea for protection will merit some attention from you.”95 Thus, the Supreme Records case, along with several other features of copyright law, left African American composers and musicians vulnerable. Well into the 1960s, there was a major market for covers by white recording artists that copied many features of the original recordings by

92 For one chart of original black R&B recordings and white covers, see Steve Chapple & Reebee Garofalo, Rock’n’Roll is Here to Pay: The History and Politics of the Music Industry 239 (1977). For a book-length account of the phenomenon, see Bob Leszczak, Who Did it First?: Great Rhythm and Blues Cover Songs and Their Original Artists (2013). 93 See Deena Weinstein, Rock’n America: A Social and Cultural History 59 (2015) (“Dot Records became a major label by releasing so many covers.”). 94 Gibbs made covers of a number of LaVern Baker’s recordings, as well as recordings by African American singer Etta James. In a bit of critical humor, Baker once took out a life insurance policy on herself with Gibbs as beneficiary and sent it to Gibbs with a note stating “Georgia, you need this more than I do because if anything happens to me, you're out of business.” The Penguin Encyclopedia of Popular Music 62 (Donald Clarke, ed.) (2d ed. 1998). 95 The Billboard, March 5, 1955, at 13.

Electronic copy available at: https://ssrn.com/abstract=3591113 African Americans.96 If the original African American recording artists were not also the composers of the songs, they would get nothing. If they were the composers, they would get only the statutory license rate, the real value of which had been diminished greatly by inflation. Although the rules were facially neutral – in theory, they allowed African Americans to similarly record mirror covers of white recording artist originals – there was little or no market in that direction, and so the predictable effect was decidedly one-sided.

IV. MIRROR COVERS UNDER CURRENT LAW Copyright law has changed a great deal since the Supreme Records case was decided in 1950. How would the case be decided under current law and practice? In 1971, Congress passed a law granting federal copyright protection to sound recordings fixed on or after February 15, 1972.97 Federal law preempts state-law claims like common-law copyright and misappropriation,98 and subsequent developments have greatly limited the possibility of suing for likelihood of confusion as to the authorship of two works under copyright.99 Hence, a modern-day Supreme Records case would more likely be decided under federal copyright law. However, that does not mean that it would be more likely decided in favor of Supreme Records. The 1971 Act provided sound recordings with narrower protection that other works of authorship, and a principal limitation was derived directly from a common understanding of the holding in Supreme Records. In 1957, the Copyright Office cited Supreme Records in “The Unauthorized Duplication of Sound Recordings,” one of the studies it prepared in preparation for a comprehensive recodification copyright law, as principal support for the position that imitation of “the style of performance and manner of interpretation” captured in a sound recording is not actionable.100 Discussants in a succeeding series of roundtables referred to Supreme Records as the “mirror record” case or the case on “mirror copying,” and likewise saw it as precedent for the position that performances captured on a sound recording could be freely imitated.101 Drafts of what would become

96 There are some examples that stretch into the 1970s, like James Taylor’s 1975 cover of Marvin Gaye’s “How Sweet It Is (To Be Loved By You),” and there are undoubtedly other examples of which I am unaware. 97 See Sound Recording Act of 1971, Pub. L. No. 92-140, 85 Stat. 391 (1971). 98 See 17 U.S.C. § 301. 99 See, e.g., Dastar Corp. v. Twentieth Century Fox Film Corp., 539 U.S. 23 (2003). 100 Barbara A. Ringer, The Unauthorized Duplication of Sound Recordings, Copyright Law Revision Study No. 26, at 10 (1957). 101 See Copyright Law Revision Part III, page 196 (comments of Sydney A. Diamond, London Records) (referring to “the way in which Judge Yankwich did decide

Electronic copy available at: https://ssrn.com/abstract=3591113 the 1971 Sound Recording Act and the 1976 Copyright Act included language implementing that position, and it survives in the current version of §114(b) of the Copyright Act: The exclusive rights of the owner of copyright in a sound recording under clauses (1) and (2) of section 106 do not extend to the making or duplication of another sound recording that consists entirely of an independent fixation of other sounds, even though such sounds imitate or simulate those in the copyrighted sound recording.102 Thus, under current law, as long as Decca “independently fixed” its sound recording, which it did, it could imitate any feature in the Supreme recording without violating the copyright in that recording. That is only part of the story, however, because it does not consider copyright in musical works, which is not subject to the same limitation as copyright in sound recordings. If, during the course of making an “independent fixation” of sounds, someone is deemed to have imitated a musical work and not just a sound recording, that could be copyright infringement. Thus, if a sound recording contains a copyright-protected musical work, the latitude to imitate it through independent fixation is limited by what features of it count as parts of the musical work. If the musical work contained in the sound recording is thin or nonexistent, the latitude to imitate is broad; if it is thick, the latitiude is narrower. For our purposes, then, the relevant question is, under current law, how much of what is added to a previously notated song while making a sound recording of it counts as a copyright-protected musical work? It turns out that there are three parts to the answer to that question. The first two parts of the answer, which concern standards of fixation and originality, give grounds for optimism. Those standards have moved in the right direction, and could if necessary be given a further nudge. The last part, however, which concerns copyright owner control over derivative works, is more complicated and less tractable. First, how do musical works need to be fixed, and what kinds of features can be components of a musical work? Until the passage of the Copyright Act of 1976, musical works had to be fixed in visible notation.103 That means that any musical features that appeared on a sound recording but were not notated were not protected as part of a musical work – even if they were “notes” – series of pitches with defined durations – that could have

in the so-called ‘mirror record’ case”); see also id. at 75 (comments of Leon Kellman, American Guild of Authors and Composers) (referring to “mirror copying”). 102 17 U.S.C. §114(b). 103 See White-Smith Music Pub. Co. v. Apollo Co., 209 U.S. 1 (1908). Although the Copyright Act of 1909 added rights of mechanical rendering, it did not overrule this part of the White-Smith ruling.

Electronic copy available at: https://ssrn.com/abstract=3591113 been notated. Thus, for example, on the Paula Watson recording of “A Little Bird Told Me,” the male chorus hums an introduction in harmony, and later also sings the line “She believes that it’s true” in harmony. In the Leroy Whyte notated score that we have available, only a single melody is notated for both of those male chorus parts. Under the requirement of fixation in visible notation, only that melodic line, which is sung by one of the male chorus members. would count as part of a musical work. The pitches that the other chorus members sung would only count as part of the sound recording, and could therefore be copied by means of independent fixation. Only if they had been notated would they count as part of the musical work. That all changed when the 1976 Act become effective. Under that Act, musical works need not be fixed in visible notation, but can be fixed in “phonorecords,” that is, copies of sound recordings.104 That means that much more of the “arrangement” made while recording a song would be subject to copyright protection as a musical work. It is still unclear exactly how much would. It seems certain that the pitches played and sung without ever having been notated would be protected. Whether other elements, such as a singer’s timing that anticipates or lags the beat, or the timbre of the singer’s voice, or the choice of instrumentation, would also count as part of the musical work, is still open to question.105 Some courts have held that the language now in § 114 of the Copyright Act, and its legislative history, preserve a domain of “style of performance and manner of interpretation”106 that can be freely imitated through independent fixation.107 Second, what standard of originality should be applied to derivative musical works, also known as musical arrangements?108 Recall that Judge

104 See 17 U.S.C. §§ 102(a), 101 (“fixed”). 105 See Robert Brauneis, Musical Work Copyright for the Era of Digital Sound Technology, 17 Tulane J. Tech. & Intell. Prop. 1, 35-43 (2014) (reviewing four different tests for separating musical works from sound recordings in phonorecords, and concluding that none of them work well). 106 Barbara A. Ringer, The Unauthorized Duplication of Sound Recordings, supra note 100, at 10. 107 See, e.g., Newton v. Diamond, 388 F.3d 1189 (9th Cir. 2003) 108 “Musical arrangement” is a term that appears three times in the Copyright Act. See 17 U.S.C. § 101 (“derivative work”) (defining “derivative work” as including “musical arrangements”); 17 U.S.C. § 101 (“work made for hire”) (providing that a “musical arrangement” is a “supplementary work” that can quality as a specially commissioned work made for hire); 17 U.S.C. § 115(a)(2) (providing that a statutory mechanical license “includes the privilege of making a musical arrangement,” within specified limitations). I would like to suggest that the term has some connotations of modesty and lower status of the task – that an “arrangement” is just figuring out how to adapt music to different instruments, or a different rhythm or style, and might not cover all musical derivative works. It is not clear, for example, that we would naturally refer to Sergei Rachmaninov’s “Rhapsody on a Theme by Paganini” as an “arrangement” of Niccolo Paganini’s “Caprices for Solo Violin,” although it would count as a derivative musical work, as it incorporates a significant portion of Paganini’s 24th Caprice. That

Electronic copy available at: https://ssrn.com/abstract=3591113 Yankwich articulated a very high standard of originality, patent-like in formulation, which resulted in the denial of protection to everything added to “A Little Bird Told Me” in the Supreme recording.109 As noted above, there are isolated subsequent decisions that have applied a similar standard.110 Yet that patent-like formulation is wildly out of step with the “modicum of creativity” standard applied to works of authorship in general,111 and even with the slightly higher standard of “substantial as opposed to a trivial variation” often applied to derivative works.112 Under a lower standard, most musical arrangements, and certainly the arrangement in the Supreme Records case, should be copyrightable. A weakness in the Supreme Records opinion provides a principal reason that a lower standard is desirable: to avoid the subjective and perhaps prejudiced musical assessments of a judge. To Leon Yankwich’s ears, which may have been trained on “popular records” rather than “race records,” Paula Watson’s recording was “thick, mechanical, lacking inspiration, containing just the usual accompaniments and the usual intonations which one would find in any common recording.”113 Others might well hear more nuance and inventiveness. To paraphrase Justice Holmes, “It would be a dangerous undertaking for persons trained only to the law to constitute themselves final judges of the worth of [musical arrangements], outside of the narrowest and most obvious limits.”114 So far, so good. However, the third issue is the most enduringly difficult: the matter of protection of derivative works generally. Both the Copyright Act of 1909115 and the Copyright Act of 1976116 have provisions denying copyright to a derivative work that uses an underlying copyrighted work without permission of the owner. The 1976 Act provision states that “protection for a work employing preexisting material in which copyright subsists does not extend to any part of the work in which such material has also means that to call something a “musical arrangement” is to subtly suggest that it is of a lesser degree of creativity, of lower status, deserving less respect, just as some might use the term “interior decorator” to denote a task and a role of lesser status than “architect.” 109 See supra text accompanying notes 77-80. 110 See Woods v. Bourne Co., 841 F. Supp. 118 (S.D.N.Y. 1994), rev’d in part on other grounds, 60 F.3d 978 (2d Cir. 1995) (piano arrangement of original melody). 111 See Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340. 345 (1991). 112 L. Batlin & Son Inc. v. Snyder, 586 F.2d 486, 490 (2d Cir. 1976) (en banc). 113 Id. 114 Bleistein v. Donaldson Lithographing Co., 188 U.S. 239, 251 (1903) (“pictorial illustrations” in the original). 115 Act of March 4, 1909, § 6 (“[C]ompilations or abridgements, dramatizations, translations, or other versions of . . . copyrighted works when produced with the consent of the proprietor of the copyright in such work, . . . shall be regarded as new works subject to copyright under the provisions of this Act.”) (emphasis added). 116 17 U.S.C. § 103(a).

Electronic copy available at: https://ssrn.com/abstract=3591113 been used unlawfully.”117 Under that rule, as long as the song “A Little Bird Told Me” is under copyright, an arrangement of the song would be denied copyright protection if the authors of the arrangement did not get permission to use the song. Moreover, permission to use the song can be conditioned on conveyance of copyright in the arrangement to the owner of copyright in the song. Under current law, that conveyance could even take the form of a work-made-for-hire agreement, leaving no power to terminate the transfer in the future, because “musical arrangements” are mentioned in the Copyright Act as an allowable category of specially commissioned works made for hire.118 Thus, if Supreme Records were making the first recording of “A Little Bird Told Me” today, Harvey O. Brooks could demand that Supreme and all of the people that were involved with recording the song sign agreements that would result in Brooks being considered the author and initial owner of any arrangement created during the process of recording the song. Brooks could also condition permission to use the song on dedication of the arrangement to the public domain, which under the 1976 Act is the default status of arrangements made when recording a cover version of a song under the § 115 mechanical license.119 There are two reasons why that makes protection for what is added during recording difficult. The first is that protection and ownership all depend on the relative bargaining power and sophistication of the parties. Musicians and recording companies have to have leverage with composers and music publishers, and have to understand that they have that leverage. As for the latter, diminished access to resources and knowledge is another result of general racial subordination and isolation. As Reebee Garofalo has noted, because African American musicians “seldom had access to good advice about record contracts, royalty payments, marketing, promotion, or career development . . . , they were routinely swindled out of their publishing rights and underpaid for record sales.”120 There are countless examples of such uncompensated or undercompensated appropriation.121

117 Id. 118 See 17 U.S.C. § 101 (“works made for hire”) (listing “musical arrangement” as a type of “supplementary work” eligible for a to be a specially commissioned work made for hire). 119 See 17 U.S.C. § 115(a)(2) (providing that a musical arrangement made while making a recording under the § 115 mechanical license “shall not be subject to protection as a derivative work under this title, except with the express consent of the copyright owner.”). 120 Reebee Garofalo, Crossing Over: From Black Rhythm & Blues to White Rock ‘n’ Roll, in Rhythm and Business: The Political Economy of Black Music 112, 123 (2002). 121 See, e.g., K.J. Greene, Copyright, Culture & Black Music, supra note 1, at 372- 373, 376-377 (discussing the phenomenon generally and the cases of Frankie Lymon, Little Richard, and Chuck Berry in particular); K.J. Greene, Intellectual Property at the Intersection of Race and Gender: Lady Sings the Blues, 16 Am. U. J. Gender Soc. Pol’y

Electronic copy available at: https://ssrn.com/abstract=3591113 Second, splitting ownership of a song and an arrangement is a special case of the more general issue of dispersed ownership of layers of copyright, which raises transaction costs for using works, and which Congress and courts have often tried to avoid. (To be fair to Judge Yankwich, he also raises this concern in his opinion in Supreme Records.122) If a song and an arrangement of that song are separately owned, anyone who wants to make a recording or performance that would count as use of both the song and the arrangement would have to obtain permission from two different parties. Those parties then need to price their permission knowing that another permission fee will need to be paid, a very difficult task. And it is not clear that the existing statutory mechanical license or performing rights organization distribution schemes have any provision for dividing payments to owners of multiple layered works. If the owner of copyright in a song gave permission to a recording company to protect a recorded arrangement as a separate derivative work, then presumably that recorded arrangement would also become subject to the statutory mechanical license, and anyone who wanted to cover that arrangement would have to pay two mechanical license fees. As a result, what happens somewhat more often is that the recording artist (or record producer) who makes the first recording of a song negotiates for co-writing credit for the song itself. In the 1920s, Al Jolson, the most famous singer of that era, reportedly regularly bargained for songwriter credit while he was negotiating to record a song.123 He was often successful: the ASCAP repertory database lists him as co-writer on 43 songs.124 At that time, when most revenue from songs still came from the sale of sheet music, we can view Jolson’s authorship claim as fictional. He likely did not write any of the notated songs sold as sheet music, but only promoted them by singing them on records and in live performances. By the 1950s, however, songs were generating most of their revenue from recording royalties (including both mechanical license income and public performance income) rather than sheet music sales, and the initial recorded versions often became the canonical ones. Under those circumstances, it was far less fictional to claim that the recording artist was a co-author of “the song,” even if he or she became involved after the melody and words were notated. Thus, for example, Elvis Presley’s co-writing credit for “Heartbreak Hotel,” shared with composers Tommy Durden and Mae Boren

& Law 365, 370 n. 29 (citing sources discussing cases from Scott Joplin to Huddie Ledbetter to Jelly Roll Morton). 122 See Supreme Records, 90 F.Supp. at 909 (noting the problems created when copyright in a work is “segmentized”). 123 See Geoffrey P. Hull, Termination Rights and the Real Songwriters, 7 J. Entertainment Law & Practice 301, 301 (2005). 124 See id.

Electronic copy available at: https://ssrn.com/abstract=3591113 Axten,125 might best be understood that way. Presley may well have contributed compositional as well as performance elements to its canonical recorded version (although he of course also contributed his audience goodwill, which is not authorship). Similarly, many bands from the late 1960s onwards, including the Doors, Black Sabbath, Genesis, the Sex Pistols, U2, REM, and Coldplay, have shared songwriting credit equally across all band members.126 That can also be seen in part as recognition that whoever came up with the melody and lyrics, each member of the group contributed musical elements that are fixed in the sound recording. More recently, in some genres of music, the creation of a song does not start with notating a melody or words, or even with recording a performance of a melody and words. Rather, it starts with creating a rhythm track, which may be programmed as much as performed, and then layering melody, words and other instrumentation on top of that track.127 When that “track- and-hook” method of creation is used, songwriting credit may indeed be allocated among everyone who had a hand in creating the recorded song, including those we would have previously thought of as producers (a title that now has gained an alternative meaning, as the person who creates the rhythm track), arrangers, and performers. That is a large part of the reason why hit songs have had an increasing number of average songwriting credits over the past 40 years. In the 1960s, 1970s, and 1980s, the average number of songwriters credited for each top 10 song was just under two. In the 1990s, that figure rose to 3.13; in the 2000s, to 3.50, and in the 2010s, to 4.07.128 To the extent that songs continue to be written in notated form and then arranged in preparation for recording, it could be also possible to come to some industry-wide agreement about default rules for dividing licensing revenue between composer and arranger. Rules for dividing revenue are no strangers to the Copyright Act. For example, § 114(g) of the Act provides that 50% of the revenues from digital streaming of sound recordings will be paid to the owner of copyright in the sound recording; 45% will be paid to the featured artist or artists; 2½% will be paid to the nonfeatured musicians; and 2½% will be paid to the nonfeatured vocalists.129 More recently, Title III of the Music Modernization Act, known separately as the Allocation for

125 See id. 126 See Richard Osborne, Doing the Splits: The Creative Accounting of Songwriting Shares, available at https://www.academia.edu/28445319/Doing_the_Splits_The_Creative_Accounting_of_ Songwriting_Shares. 127 See generally John Seabrook, The Song Machine: Inside the Hit Factory (2015). 128 See Daniel Sanchez, The Average Hit Song Has 4+ Writers and Six Different Publishers, Digital Music News, August 2, 2017, available at https://www.digitalmusicnews.com/2017/08/02/songwriters-hit-song/. 129 See 17 U.S.C. § 114(g)(2).

Electronic copy available at: https://ssrn.com/abstract=3591113 Music Producers Act, sets up some complicated rules for providing some portion of those streaming revenues to sound recording producers, suggesting a realization of the need to broaden recognition of creative contributions to sound recordings.130 Analogously, one can imagine default rules about an allocation of some percentage of mechanical and public performance revenues to arrangers and performers associated with the initial recording of a song.

IV. POLICY IMPLICATIONS At a relatively modest level, the policy implications of the history of white mirror covers recordings could be spelled out reasonably simply. The inability of those made the first recording of “A Little Bird Told Me” to receive any compensation or attribution for the use of their creative work in subsequent mirror cover recordings resulted from the interplay of three copyright doctrines, and their unsympathetic application in the Supreme Records case. For those who think that the African American creators of musical elements used in white mirror covers should have received some compensation and attribution, two of those doctrines have moved in the right direction, and the third has been subject to increased circumvention through bargaining. The policy recommendations could be just to ensure that doctrine and bargaining practices continue to continue to move in that direction. Musical works can now be fixed in sound alone. The recommendation would be that what counts as a musical work within a sound recording should not be defined too narrowly, to ensure that it does not exclude important elements contributed in the recording process. The general threshold standard of creativity for works of authorship is now very low. The recommendation would be that it should be kept low when assessing copyright protection for musical arrangements. It is still the case that derivative works will only gain copyright protection with the permission of the underlying work owner. While that doctrine is unlikely to be changed, there has been increased bargaining about songwriting credit as the sound recording, rather than the notated song, has become the paradigmatic unit of music creation. The recommendation would be that that bargaining be acknowledged and facilitated, and that all participants in the process of creating a sound recording be educated about the possibility of bargaining for songwriting credit or for a share of musical work revenues. Those modest policy recommendations aside, there should be a broader conversation about whether and how copyright law should address social justice concerns, and how the history of white mirror cover recordings could help us understand what the best approach is. I have no more than tentative

130 See Allocation for Music Producers Act, Title III of the Music Modernization Act, Pub. L. 115-264, 132 Stat. 3676, 3737 (2018).

Electronic copy available at: https://ssrn.com/abstract=3591113 thoughts about that conversation, which may be best presented as responses to a series of reservations to having copyright policy be involved at all. So here are the reservations, followed by my responses. 1. The entire phenomenon of white mirror cover recordings is in the end traceable to unregulated private consumption decisions – the preferences of prejudiced audiences for recordings by whites – and those should be addressed by attempts at changing attitudes, not law. That suggestion runs counter to a basic premise of American antidiscrimination law. Although that law does not regulate private consumption decisions directly, it also generally does not let businesses justify discrimination on the ground that they are merely responding to consumer preference. If the law can intervene when an airline hires only female flight attendants because their passengers prefer them,131 it should be able to intervene when a recording company produces a white mirror cover version because there is an audience that prefers to see white faces on recording artists. It is important to note that copyright law itself has historically followed the same pattern. It does not impose liability for receiving or possessing unauthorized copies of works, or for viewing unauthorized public performances. Thus, at least before the development of home copying and peer-to-peer distribution technologies, copyright law also left private consumption decisions unregulated, and operated only as a constraint on business operations. 2. Discrimination should be addressed by antidiscrimination law, not by intellectual property law. A principal problem here is that antidiscrimination law is unlikely to work. In the case of white mirror cover recordings, the decision that antidiscrimination law would regulate is the employment or contracting decision of the record company that produces the cover recordings. In the Supreme Records case, that would be Decca Records’s decision to employ or contract with Evelyn Knight, rather than with an African American singer, to make its recording of “A Little Bird Told Me.” Yet every vocalist has a distinctive voice, and it would usually be difficult to rebut a record company’s assertions that it chose the white vocalist because of his or her voice, and not because of his or her race. Moreover, in the era of music streaming platforms as the dominant source of recorded music revenue, a white mirror cover recording could be made and distributed without any employment or contracting decision ever having

131 See Diaz v. Pan Am. World Airways, Inc. 442 F.2d 385 (5th Cir. 1971), cert. denied, 404 U.S. 950 (1971) (holding that the preferences of airline passengers for female flight attendants did not make sex a bona fide occupational qualification for flight attendants).

Electronic copy available at: https://ssrn.com/abstract=3591113 been made. A white recording artist self-produces a mirror cover and makes it available on a streaming platform. It becomes vastly more popular than the African-American-produced original. That difference is reflected in the royalties paid to the owners of the two recordings. There is no employment or contracting decision left to regulate. More generally, the institution of property, including intellectual property, can provide and has provided important protection for minorities beyond that afforded by antidiscrimination law. As N.D.B. Connolly put it, “at the end of the nineteenth century . . . black people’s civil rights and voting rights stood on shifting sand. . . . By contrast, when it came to black people’s legal claims on property, courts and other adjudicating bodies generally proved more responsive, for there it was the integrity of capitalism, not white power, that was at issue.”132 While civil rights have come a long way, it is not at all clear that civil rights are now so perfectly protected that there is no work left for property to do. 3. Any sympathy for Paula Watson, Albert Patrick, and the others who worked on the first recorded version of “A Little Bird Told Me” is ultimately grounded on a theory of desert, and desert theories have no place in discussions of copyright or intellectual property. It is true that some of my own sympathy for Paula Watson and her collaborators seems to be connected to an implicit judgment that they created something for which they should have received some compensation when it was successfully exploited by others. I could test the emotional power of at least one kind of desert by asking myself whether my sympathies be any different if it was another African American recording artist who had made the mirror cover that had eclipsed the popularity of the original recording. And I think the answer is that they would be somewhat different – some of my sympathies are triggered by the practice of using white recording artists to make the covers – but they would not be gone. It is also true that the dominant theory of copyright in academic circles is based on incentives to create, and rejects any appeal to desert. There are some carefully argued reasons why, but incentivist academics probably also need to come to grips with the fact that rejection of desert sets them apart from the intuitions of the vast majority of nonacademics, and likely from their own intuitions when their vigilance lapses. Another aspect of the response to this reservation, however, is that it is not clear that we need desert to support policies that promote protection for

132 N.D.B. Connolly, A World More Concrete: Real Estate and the Remaking of Jim Crow South Florida 29 (2014).

Electronic copy available at: https://ssrn.com/abstract=3591113 minority follow-on authors. It is interesting to explore how the white mirror cover recording phenomenon might appear when viewed through the lenses of various incentive-based theories. Even from an unreconstructed efficiency perspective, under which we would have to take patterns of consumer demand completely for granted, there are empirical conditions that would support protection. Perhaps somewhat shockingly, the problem might have to be that too few good white mirror covers are being produced, because African American musicians didn’t have sufficient incentives to produce the original recordings. As with many incentivist narratives, one problem is that there were many African American original recordings made even under the shadow of potentially uncompensated copying. Some would argue that that means that the prospective revenue from freely imitable recordings provides sufficient incentive to make them, and no further protection is needed. The counterargument is that Paula Watson, Albert Patrick, and other African Americans of the era may have made even more recordings had they been compensated for the white mirror covers. That is very difficult to know, which is a characteristic vulnerability of incentivist arguments. Not all incentivist theories have to be quite so blind to distributional consequences. If we thought that there should be incentives for a diverse range of creators to create, across racial, ethnic, gender, and other lines, we might be concerned if we saw that the decision not to protect a particular kind of creativity was having a disparate impact on a particular group of creators. 4. The injuries of white mirror cover recordings stemmed from lack of attribution as well as lack of compensation and permission, but U.S. copyright law has always been weak on attribution. Both of those are true. Part of the injury of allowing Decca and Evelyn Knight to freely appropriate all of the elements added to “A Little Bird Told Me” is that the contributions of African Americans like Paula Watson, Albert Patrick, Leroy Whyte, Tiny Webb, Jesse Sailes, Chuck Hamilton, Maxwell Davis and Pete Peterson were rendered invisible. U.S. copyright law does not incorporate a right of attribution extending to musical works or sound recordings. There is an industry practice of crediting that to some extent follows recognition of authorship by copyright law – Harvey O. Brooks is credited on the Decca recording even though U.S. copyright law does not directly require that credit. Thus, if separate authorship and ownership of an arrangement were recognized, or if the arranger bargained for co-authorship of a song, credit might follow. Nonetheless, attribution does remain a weakness of both law and practice. But for the Supreme Records litigation,

Electronic copy available at: https://ssrn.com/abstract=3591113 the participation of Leroy Whyte and the musicians in creating “A Little Bird Told Me” might be completely lost to history, because they were not credited on the original Supreme release. And even with a fully implemented right of attribution, the race of the authors could be rendered invisible. A very high percentage of purchasers of the sheet music for “A Little Bird Told Me,” featuring a photo of Evelyn Knight on the cover, probably did not know that composer Harvey O. Brooks was African American, even though his name appears prominently on the document. 5. In general, the prejudiced taste of audiences is too transitory a matter to trigger a copyright policy response. In particular, the phenomenon of white mirror cover recordings waned decades ago, and so it seems odd to consider possible policy reactions to it. A lot of copyright policy is based on patterns of consumption that may change over time. For example, the ban on rentals of musical sound recordings and software,133 but not on rental of movies, books, or spoken word recordings,134 is based on experiences of consumer behavior patterns that might well change over time. It is true that the particular phenomenon of white mirror cover recordings waned decades ago, though there still may be isolated instances. However, an advantage of considering a policy decision that was made some time ago is that we can see how the decision played out over the years after it was made. It seems unrealistic to deny that many consumers of works of authorship still make consumption decisions based on race, and that to deny that some in the copyright industries act to feed those preferences. More recent discussions of acting suggest that the effects of the preferences of some consumers for white actors are still very much present in the theater and motion picture industries.135 6. Every copyright doctrine that could possibly be adjusted to protect African American musicians from uncompensated exploitation is very broad, and those adjustments would have consequences far beyond that

133 See 17 U.S.C. § 109(b). 134 See Brilliance Audio v. Haights Cross Communications, Inc., 474 F.3d 365 (6th Cir. 2007) (holding that the rental exception in § 109(b) extends only to sound recordings of music). 135 See, e.g., Christina Shu Jien Chong, Where Are the Asians in Hollywood: Can Sec. 1981, Title VII, Colorblind Pitches, and Understanding Biases Break the Bamboo Ceiling, 21 UCLA Asian Pac. Am. L.J. 29 (2016); Unmasking Tonto: Can Title VII “Make It” in Hollywood, 37 Am. Indian L. Rev. 549 (2016); Angela Onwuachi-Willig, There's Just One Hitch, Will Smith: Examining Title VII, Race, and Casting Discrimination on the Fortieth Anniversary of Loving v. Virginia, 2007 Wis. L. Rev. 319 (2007).

Electronic copy available at: https://ssrn.com/abstract=3591113 particular problem. That is likely true, but it is true for many copyright doctrines and many policy concerns. All the policy concerns have to be expressed, and in some cases there has to be some difficult balancing in formulating doctrines. However, in this circumstance, it is not clear that there are strong policy concerns weighing against recognizing that some range of features fixed only on phonorecords count as part of a musical work and not just a sound recording. Nor do there seem to be strong policy concerns against treating musical arrangements like other derivative works, rather then subjecting them to uniquely high creativity standards. The doctrine that copyright in derivative works hinges on permission from underlying work owners is probably supported by enough policy concerns to resist modification for purposes of supporting minority creators, but one can imagine other policies that could be subject to adjustment, such as enforcement of oral agreements to share songwriting credit when there is evidence that both parties contributed creative content to a sound recording that embodies the song, and hence that the agreement does not involve any transfer of copyright interest. 7. Expanding copyright protection will sometimes burden minority creators rather than support them. That is undoubtedly also true. For example, expansion of copyright protection of sound recordings in a way that requires licensing for even the tiniest of samples136 may burden creation of music in genres like rap and hip hop that are disproportionately created by minorities. That just means that copyright policy is complicated.

VI. CONCLUSION Although U.S. copyright law has always been facially race- and gender- neutral, copyright policy has been made in contexts in which the disparate impact of particular rules has been apparent. This essay has attempted to recover the history of one of those instances, in which rules about the scope of protection for sound recordings and the standard of creativity applicable to musical arrangements were formulated in a context in which it was clear that African American musical creators had a particular stake in the outcome. The disregard for and injury suffered by those creators can never be fully remedied. However, we can commit ourselves to acquiring sensitivity to such injuries, and to serious decision of the ways copyright law can and cannot address them.

136 Compare Bridgeport Music, Inc. v. Dimension Films, 410 F.3d 792 (6th Cir. 2005) (holding that there is no de minimis exception to copyright infringement for copying of sound recordings) with VMG SALSOUL, LLC v. Ciccone, 824 F. 3d 871 (9th Cir. 2016) (holding that there is such an exception).

Electronic copy available at: https://ssrn.com/abstract=3591113