What does it mean to remake the ear? In his early manuscripts, the young Karl Marx suggested that “the forming of the five senses is a labour of the entire history of the world.” Marx drew his example of the history of senses from music: “only music awakens in man the sense of music … the most beautiful music has no sense for the unmusical ear.”1 Thus the musical ear has a history: new modes of music awaken a new sense of music, a new musical ear. But this is not the history so often imagined by early musicology, an evolutionary progress from “primitive” to “civilized” musics. Rather, the history of the musical ear, its making and remaking—like the forming and reforming of the five senses in general—is the fundamental labor of cultural revolutions, those contradictory upheavals in the habits and manners, jokes and prayers, sounds and smells of daily life that accompany the struggle between modes of production, regimes of labor, technological grids, economies of sexuality, structures of domination and representation, and modes of emancipation and exploitation.2
Every mode of production has its sound, as F. Murray Schafer suggested in his visionary book, The Soundscape, its own divide between music and noise, and the conflict between modes of production—whether in the fast violence of war and political revolution or the slow violence of displacement and expropriation over decades or generations—is registered in the struggle between ways of hearing, between the sound of the dominant machines and the sound of the subaltern masses.
So what did this cultural revolution sound like? At first glance, this seems a simple empirical question: what did these musics sound like? Moreover, since many recordings survive, the sound of the music is accessible in a way unparalleled for earlier musics, whether learned or popular. Indeed, a new musicology is developing techniques of musical analysis, transcribing recorded performances and analyzing their musical structures and rhetoric. But it is a more complex issue for three reasons.
First, since recordings became the fundamental vehicle of the rise and spread of these musical vernaculars, it is worth stressing that the question “what did these musics sound like?” must be recast as “what did these recordings sound like?” The recording process was not transparent, and recorded musics were always an artifact of the recording process, their sound shaped by engineers and producers as well as musicians. Moreover, recording often changed musical cultures around the world, creating a gap between performed music and recorded music, a gap that can be often obscured if one assumes the transparency of recording.
For this reason, historians and ethnomusicologists interested in understanding the practices of musicking have long held a justified suspicion of commercial recordings: it is easy to misread musical practice by assuming that recorded music represents that practice accurately. In some cases, the technological limits and capabilities of recording create these misreadings: the limits of recording time gives a distorted sense of the length and structure of musical works, and the fact that different voices and instruments were more or less suited to recording meant that recorded performances often diverged radically from live performances by the same musicians. In other cases, decisions made by record producers created the gap: the demand for particular types and genres of music, both on the part of folklorists seeking “traditional” genres and on the part of commercial recorders seeking “saleable” genres, have long given a sense that musicians had more limited or restricted repertoires than was actually the case. Thus one must be careful in deducing musical practices from recordings; recordings are not transparent documents, mere evidence of an underlying musical culture.
Second, since twenty-first-century auditions of these records may be quite different to those of the 1920s and 1930s, the question “what did these musics sound like?” must be recast as “how were they heard?” It is difficult to hear these 78s as they were heard at the time, because our ears are the product of the very success of these recorded vernacular musics over the last hundred years. In some cases, the adoption of certain musics as national musics has contributed to the routinization of their sounds, making it difficult to hear the original “noise”; our boredom is, as Fredric Jameson has noted of aesthetic boredom generally, less an index of the music itself than a sign of our own resistance to its now conventional sounds.3 In other cases, timbres have been adopted and popularized by other idioms: as I noted earlier, the pervasiveness of the steel guitar in US country music has made it more difficult to hear the distinctiveness of the early Hawaiian recordings. As a result, the question “what did they sound like?” must further be rephrased as “what did people say they sounded like?” Thus Theodor Adorno’s very revulsion at the sound of the modern dance musics is a powerful sign of the remaking of the musical ear. Adorno recognized the timbres of the new musics—“the muted distortions of the horns, the chirping and vibrating tonal repetitions of the plucked instruments, the banjo and the ukulele, and even the harmonica”—but found them “unbearable” (he used the word in relation to the Wurlitzer organ).4 The “guitar, ukulele, and banjo, as well as the accordion” were, he insisted, “infantile instruments in comparison with the piano.”5” He hated the twang and the vibrato, the beat and the syncopation, of the dance musics. He saw—and deplored—their connection to military marches, but thought it was a connection to fascism rather than a legacy of colonial brass bands.6 Adorno’s hearing of the new sounds should be understood not simply as his individual “taste” but as a Geiger counter, registering the scale of the musical revolution taking place.
Third, the question “what did these musics sound like?” raises the question of whether it makes sense to group the sounds of these vernacular musics together. Even if these new musical vernaculars shared a common social and historical situation—their place in the working-class barrios of colonial ports and their relation to the migrations and movements, unsettlings and uprisings in the wake of the Great War—and even if the global recording industry remade world musical space around these musical vernaculars, can we conclude that they had a common sound? Did the records sound alike? Were they heard to sound alike? This, too, was contested in the polemics of the day. If it is a mistake to lump all of these musics together as “jazz,” “light” music, or “popular” music, it is equally misleading to separate them into entirely distinct and unconnected “national” or “ethnic” traditions. Rather, the sound of the records circulating through the archipelago of colonial ports was inflected by the sounds of three diasporas: a black Atlantic sound that emerged out of the slave trade and the post-slavery migrations of African-American communities from Brazil to the United States; a gypsy Mediterranean sound that emerged as the Roma diaspora inflected musics from flamenco to tzigane; and a Polynesian Pacific sound that resonated across the ports of the Pacific and Asia.
Moreover, I want to suggest that the sound of the vernacular phonograph records in each of the three arcs also depended on a new kind of ensemble, a “dance band”—the phrase in English dates from the 1920s7—a “bando regional” or “orquesta típica.” These vernacular recording ensembles were based on a musical division of labor between a front line of melodic instruments and a “rhythm section,” a term that also emerged at this time.
The sound of these “dance bands” was characterized by four distinctive and controversial elements that remade the musical ear: the “noisy” timbres of their instruments and voices; their “syncopated” rhythms; their “weird” tonalities; and their “recorded” improvisations. They were, as a journalist wrote at the time, “prophets of a noisy heaven and a syncopated earth.”8
As the repeated invocation of noise indicates, the recorded vernacular musics were first distinguished by the timbres of their characteristic ensembles, their instruments and voices. It is not surprising that these were heard as noise, because timbre is the product of the specific noise of an instrument: in acoustic terms, timbre consists both of the harsh sound of the attack, the first pre-pitched moment of the sound envelope, and of the peculiar mix of overtones or partials that color the fundamental frequency as the sound resonates and decays.9 Timbre long seemed an accidental and ephemeral aspect of music, escaping notation and evading the mathematics of harmony and rhythm; but recording gave new substance to timbre, as the sheer noise of voices and instruments reverberated even when the harmonies or rhythms seemed rudimentary.
The vernacular ensembles that recorded were, for the most part, neither the large hotel, theater, or concert orchestras of the city’s cultivated classes, nor the large collective singing and parading groups of the city’s workers and poor—the church and union choirs, the coros de clave and coros de son, and the carnival percussion paraders. Rather, they were mainly small bands of three to seven musicians—trios and quartets, “hot fives” and “hot sevens,” “sextetos” and “septetos”—that accompanied singers and performed instrumental dance numbers. Table 3 lists some of the characteristic lineups of the vernacular phonograph musics. Their timbres varied, but two aspects of these ensembles were particularly evident.
Table 3: Idioms, Ensembles, and Instruments
| Musical Idiom | Reeds, Flutes | Brass | Bowed Strings | Plucked Strings | Keyboards, Zithers | Free Reeds | Percussion | Number |
| son | trumpet | tres, guitar, bass | bongo, clave, maracas | trio, sextet, septet | ||||
| jazz | clarinet, saxophone | trumpet, cornet, trombone, tuba | banjo, guitar | piano | drum set | quintet, septet | ||
| ṭarab | nāy, flute | violin | ‘ūd | qānūn | riqq | quartet, quintet | ||
| fasil | kemence | ud | kanun | |||||
| rebetika | violin, lyra | guitar, mandola, lauto, bouzouki, baglamás | accordion | cembalo, santouri | ||||
| chaabi, rai | gasbas | violin | ‘ūd | qanun | tār, bendir, guellals | |||
| fado | guitarra, viola da França | duo, trio | ||||||
| South Asian theater music | violin | harmonium | tablā | |||||
| mariachi | trumpet | two violins | guitarrón, 2 guitars, 2 vihuelas | septet | ||||
| kroncong | flute | violin | kroncong, guitar, ‘ukulele | rebana | ||||
| tango | two violins, bass | piano | two bandoneons | sextet or guitar duo | ||||
| calypso | clarinet, flute | violin | cuatro, guitar, cello, banjo | vera, chac-chac | quartet | |||
| country | fiddle | guitar, banjo | autoharp | |||||
| flamenco | guitar | castenets | ||||||
| cai luong | bamboo flute | two-stringed fiddle | two-stringed lute | bell | ||||
| hula ku‘i | steel guitar, ‘ukulele, harp-guitar, guitar | ipu, uli‘uli, pahu | quartet | |||||
| tzigane | clarinet | violins, cello, bass | cimbalom | |||||
| palm-wine, highlife | kazoo | guitar | piano | concertina | castenets | |||
| taarab | violin | udi | qanun | |||||
| samba, choro | flute, clarinet | guitar, cavaquinho | surdo, tamborin, cuícas, pandeiro | |||||
| céilí | flute | fiddle | piano | |||||
| huangse yinyue | clarinet, saxophone | violin | piano | |||||
| beguine | clarinet, saxophone, bamboo flute | trombone | bass, violin | banjo | piano | chacha, tibwa, drum kit | ||
| jùjú | guitar, banjo | shekere, jùjú | ||||||
| kundiman | guitar | |||||||
| marabi | guitar | piano |
First, these ensembles usually combined instruments with distinct and often clashing timbres, and much of their musical interest lay in this timbral counterpoint. For example, the takht of modern ṭarab—associated with the early recordings of Umm Kulthūm—was, the musicologist Ali Racy argues, “a collection of khāmāt ṣawtiyyah, ‘sound timbres’ … Incorporating one of each type of instrument, for example one ‘ūd, one qānūn, one nāy, one violin, and one riqq, the takht amounts to a few layers of discernible timbral-acoustical lines.” Racy contrasts these multi-timbral ensembles with “‘unitimbral’ but register-separated combinations, such as Europe’s Renaissance recorder or viol consorts, or for that matter the classical string quartet.”10 Similarly, a recent jazz historian has argued that “the polyphonic basis of New Orleans jazz made a small number of melody instruments of differing timbres desirable.”11 Such collections of instruments with distinct timbres were common to the vernacular phonograph musics, spanning quite distinct musical idioms. Indeed, as early as 1924, the American philharmonic composer Virgil Thomson noted that “jazz”—by which he meant modern dance music with a syncopated melody over a foxtrot rhythm—“does not require balanced timbres,” being “contrapuntal rather than homophonic.”12
However, it was not simply this multi-timbral texture that generated the sounds so often heard as noise: almost all of the vernacular ensembles also juxtaposed mass-produced musical instruments, imported from European and American metalworking factories, with artisan-crafted indigenous instruments. A social as well as timbral counterpoint emerged, as imported industrial instruments—often with standardized tunings—joined the host of regional soundboxes which had arrived with rural migrants. The immense variety of these hand-crafted flutes, fiddles, scrapers, rattles, and drums led ethnomusicologists of the time, notably Curt Sachs and Erich Hornbostel, to develop the arcane vocabulary of modern “organology” (the science of classifying musical instruments): chordophones, aerophones, membranophones, and idiophones.
In the Pacific ports, the early Hawaiian hula ku‘i ensembles combined guitars and ‘ukuleles with the ipu, the Hawaiian gourd drum;13 there are also fusions like Vietnam’s cai luong, which featured, as Jason Gibbs notes, two ensembles: one with Vietnamese instruments playing pentatonic melodies on plucked and bowed stringed instruments, and a second with Western military band instruments, playing “fanfare music.”14 In the Atlantic ports, one finds palm-wine music “created by coastal West African musicians who combined local stringed and percussion instruments (including the gombey frame-drums) with those of foreign sailors; that is, portable instruments used aboard ships such as the guitar, mandolin, banjo, harmonica, accordion, and concertina.”15 “Whereas old style music keeps to indigenous or traditional instruments,” J. H. Nketia wrote in one of the first musicological accounts of West African commercially recorded music, “new style music is more daring in its choice of medium. Bottles, cigarette tins, adaptations of the Western side drum, guitars, saxophones, clarinets, and other sound instruments are pounced on. Many bands attempt a ‘syncretism’: traditional drums are brought in company with guitars, castanets, and gongs or with tins, bottles, and so on.”16 To traditionalists everywhere, this was a sonic as well as ideological challenge.
One set of voices in this counterpoint were the imported industrial instruments, whose numbers grew exponentially in port cities in the early decades of the century. By the turn of the century, stocks of “melodeons, concertinas, mandolins, and guitars” were being exported to Africa;17 meanwhile, in China, “Cantonese folk musicians began to incorporate guitars, banjos, and saxophones into their regional music.”18 These brash mass-produced commodities of a growing musical-instrument industry were sold in standardized “models” in urban music shops and by mail-order catalog, together with cheap “method” books. They were usually portable chordal instruments, inexpensive, with little prestige. Plucked strings and free reeds resonated throughout these ensembles. The guitar was ubiquitous in the vernacular idioms, as were its near and distant relatives: the mandolin, son’s tres, samba’s cavaquinho, calypso’s cuatro, fado’s guitarra and viola da França, rebetika’s lauto and bouzouki, Hawaii’s ‘ukulele, kroncong’s kroncong, mariachi’s vihuela and guitarrón, as well as the four- and five-string banjos of New Orleans jazz and Piedmont old-time music, and the ‘ūds of ṭarab and chaabi. And there were a host of manufactured free reeds, including accordions, concertinas, harmonicas, harmoniums, bandoneóns, and melodeons.
As the guitar, a soft-spoken parlor instrument in the European and North America musics of the late nineteenth century, was adapted to factory-made steel strings in the 1880s and 1890s, inexpensive models proliferated and they became indigenous instruments throughout the Polynesian Pacific, the gypsy Mediterranean, and the black Atlantic. In Hawaii, King David Kalākaua’s 1880s patronage of the guitar—introduced to Hawaii by Mexican cattle herders—had made the instrument central to the new hula ku‘i, an emblem of Hawaiian cultural resistance. By the early twentieth century, Hawaiian guitarists had not only developed a variety of “slack key” tunings (in which strings were loosened to make patterns matching major or minor triads), but also invented a new playing style, sliding a steel bar over the strings of a lap-held guitar to create striking glissandos. This “steel guitar” (kika kila), usually attributed to Joseph Kekuku, a student at the Kamehameha School for Boys in the 1890s, became one of the major guitar timbres of the vernacular phonograph records of the 1920s and 1930s, not only in the recordings of Hawaiian virtuosos like Sol Ho‘opi‘i, Mike Hanapi, and David Napihi Burrows, but on recordings in Jakarta, Calcutta, Johannesburg, Athens, and Atlanta.19
In Andalusia, guitar playing—toque—emerged as an ever more central part of the flamenco tradition, taking its place alongside singing (cante), dance (baile), and handclaps (palmas), in large part because of the technologies of electrical recording that allowed the guitar to carry melodic leads as well as rhythmic accompaniments. “In 1925,” a flamenco scholar notes, “with the development of the microphone and the audio amplifier, the guitar came into its own. Amplification released it from its acoustic constraints, and the guiding hand of musical geniuses such as Ramón Montoya led it to new heights.”20
Meanwhile the guitar was also adopted across Africa, becoming an African instrument. “What sense does it make,” Kofi Agawu asks, “after a century and a half of regular, continuous, and imaginative use, to describe the guitar as a ‘foreign’ instrument in Africa?”21 “In the late 1920s and early 1930s,” musicologist Christopher Waterman writes of Lagos, “a ‘box’ guitar could be obtained for one pound, a price within the reach of many regularly employed wage-earning Africans.”22 “For those with cash to spare,” the collector Hugh Tracey noted of South Africa, “the guitar takes over the accompaniment with drums and rattles as before and a kind of Bantu calypso is the response to this now universal instrument.”23 It rarely earned respectability—Kwaa Mensah recalled that his uncle Kwame Asare “learned to play guitar against his father’s wishes, who thought only ruffians played guitar.”24 However, despite the fears of folklorists, it did not undermine vernacular music making. In some cases, the playing techniques of indigenous stringed instruments were transferred to the guitar: Michael Veal has pointed to the ways African guitarists used techniques of the likembe, nyatiti, and kora.25 In other cases, new techniques were developed to gain desired timbres. Kru seamen, originally from Liberia but living in “Kroo-town” settlements in West African ports, developed a two-finger playing style that spread up and down the coast: the Kumasi Trio’s guitarist Kwame Asare claimed “to have learned the basis of the two-finger style known as dagomba from a Kru sailor.”26 Not only playing techniques but even distinctive tuning systems traveled widely: “common Hawaiian slack-key tunings in C, G, F—for example—‘taro patch’ (5-1-5-1-3-5, ascending)—are identical to tunings used by Lagosian guitarists,” Christopher Waterman notes. “The international distribution of such tuning systems along trade routes has yet to be adequately investigated.”27
Equally important, and equally controversial, were the relatively new family of free reed instruments, manufactured boxes with flexible metal tongues—“reeds”—that vibrate as a result of air pressure or suction, usually mouth- or bellows-blown. These “squeezeboxes” emerged across Europe and North America in the decades of the first Industrial Revolution, influenced by the appearance of the Chinese mouth organ, the sheng, in Europe in 1777: first the accordion and harmonica in 1820s Vienna, followed, over the next three decades, by the English, or Wheatstone, concertina, the melodeon, the German Konzertina, later renamed the bandoneón, the cheaper diatonic Anglo-German concertina, and the French harmonium.28
Some had buttons, other keys; some were diatonic, with only the notes of a single key, others were chromatic. In general, their fixed tunings and “closed systems” that prevented the playing of dissonant chords made them relatively easy to master; in addition, they were loud and portable, enabling one-man dance bands.29 As a result, the historian Stuart Eydmann notes, “the 1870s and 1880s saw many working-class musicians turn to the more versatile and robust free-reed instruments which were being developed on the continent.”30 The free reeds followed labor migrants to settler colonies around the world: legend has it that the bandoneón was brought to Buenos Aires by German sailors.31 By the early twentieth century, the largest importers of accordions were the United States, Argentina, and Brazil.32 The “cheap ten-key Anglo-German concertina” was particularly popular in South Africa,33 and the 1927 Zonophone recordings of the West African George Williams Aingo included several concertina tracks.34 Meanwhile, in 1875, a Calcutta instrument maker added drone stops and a hand pump to the French harmonium: “By 1913, India had become the richest market in the world for harmoniums—for church services, for missionaries in the field, for accompaniment of urban musical dramas, and, increasingly, for playing in Indian classical music.”35
The free reeds were controversial everywhere. On the one hand, they were modern and urban; in Buenos Aires, the guitar was associated with the countryside, the bandoneón with the city.36 In India, the shift from the traditional bowed sarangi to the harmonium occurred not only because it was easier to play but also because it didn’t have the traditional association with courtesans.37 On the other hand, free reed instruments were shunned by the respectable, the cultivated, and the traditional.
“The harmonium has attracted more elite contempt than any other instrument in the history of Indian music,” the historian Matt Rahaim writes. Gandhi lamented the “execrable harmonium, concertina, and the accordion.” “All-India Radio banned the harmonium from its airwaves for over thirty years, and it has long been banished from the South Indian classical music stage.”38 In South Africa, Africans associated with mission education avoided the concertina, choosing piano and brass instruments.39 “The musical establishment throughout Europe was unified in the dismissal of the popular squeeze boxes,” historian Christoph Wagner notes.40 “The accordion,” Theodor Adorno wrote, “is a very primitive instrument” on which the “player strikes ready-made chords in a quasi-improvisatory manner.” Noting that it was known in Germany as a “sailor’s piano” and in the US as a “gypsy piano,” he argued that “its role may be roughly defined as a piano fit for camp life or collective life of any sort, involuntarily antagonistic to the private apartment.”41 Traditional musicians also found themselves threatened by the new technology: the Auvergne piper Antoine Bouscatel, the star of the turn-of-the-century Parisian bal musette, was challenged by the arrival of immigrant Italian accordionists: “The days of my bagpipes are numbered, and those of your hurdy-gurdy too! This character with his accordion carries with him our ruin! … The accordion is a miracle that falls from the sky. It is a revolution on the way. Did you hear? It is complete, it is hot, it is alive. And it is a whole orchestra, this instrument of the devil!”42
But it was also the sound, the timbre, the noise of these instruments that offended. The harmonium—argued its critics, English and Indian—could not glide between notes; its tuning was wrong, and it was un-Indian.43 Accordionists of the early decades of the twentieth century often used a “wet” tuning—“one reed tuned precisely to a note and supplemented by the other two reeds tuned sharp and flat around it”—which gave players, in the words of Michael Dregni, “the trademark plaintive sound with ‘vibration’ surrounding every note in a sentimental quavering like a street singer’s crooning.”44 It was this wavering intonation—characteristic of bent guitar strings as well—that Adorno heard in all the varieties of “jazz.”
The “American accent of the saxophone”—as Cuban composer Emilio Grenet put it—was also a distinctive timbre of the vernacular phonograph records, and, in many cases, an emblem of “jazz.”45 “In Europe,” Adorno wrote, “the saxophone is considered representative of this sound [of jazz], the instrument against which the resistance has concentrated its forces.”46 To adopt the saxophone was to link oneself symbolically with jazz, as when the bandleader Pixinguinha brought the saxophone into Brazilian choro in 1923, when steel guitarist Sol Ho‘opi‘i recorded with Andy Iona in 1928 in one of the earliest hula sessions that incorporated saxophones, and when guitarist Ramón Montoya recorded with Fernando Vilches in 1933 on one of the first flamenco sessions with a saxophone.47
If the timbres of these industrial instruments reverberated from the vernacular phonograph discs, so, too, did the remarkable variety of vocal timbres. Unlike printed sheet music, which stripped the voice from words and melodies, the phonograph captured a semblance of the vocalist’s timbre, what Roland Barthes later called “the grain of the voice.”48 At the time, musicologists and folklorists noted that this was important not simply for accuracy but also because singers themselves privileged timbre. “Phonograms are immensely superior to notations of melodies taken down from direct hearing,” the German musicologist Erich Hornbostel argued, because “the singers themselves attach as much importance to the timbre of the voice and the mode of recitation as to anything else, and very often more.”49 Similarly, reviewing the early recordings of Portuguese fado, Rodney Gallop argued that “it is precisely in the manner of singing of fadistas such as Alberto Costa and Filipe Pinto, which no musical notation, but only the gramophone, can convey, that the chief character and charm of the fado lie.”50
The centrality of timbre in recorded music was also responsible for the then surprising shift in which the singer, rather than the song, became the center of the music industry. After all, ever since the explosion of the mass circulation of sheet music in the 1850s, the music industry had depended on the popularity and sale of songs. Publishers would try to attract a number of singers to put across the song in their live performances, and this model continued in the early days of recording. A striking example of the reversal from song to singer is noted by recording industry historian Alan Sutton: before 1926, Victor “would scout the South not for performers but for songs, with the idea of ‘refining’ the compositions, copyrighting them, and then handing them off to professional studio performers in New Jersey and New York.” However, the unexpected success of the southern Piedmont singer Ernest V. Stoneman and his family string band on the OKeh, Edison, and Gennett labels in 1925 and 1926 led Victor to sign Stoneman and begin seeking singers rather than songs; a few months later, they sent Ralph Peer, with Stoneman’s help, to find local singers in Bristol, Tennessee, resulting in the landmark recordings of Jimmie Rodgers and the Carter Family.51 One can also see this reversal in the writing of Abbe Niles: whereas his early essays focused on songs and their variants, his record reviews of 1928 called attention to the voice rather than the song. “Crave you Hawaiian singing,” he writes, “try ‘He Manao Healoha,’ or words to that effect by Kalama’s Quartet”; he also recommends Jimmie Rodgers’s “engaging, melodious and bloodthirsty ‘Blue Yodel’ … which started the whole epidemic of yodeling blues that now rages,” and Bessie Smith’s “immense voice” with its “tones of great bitterness and haunting sadness.”52
Many of the recordings of the late 1920s became famous for the vocal timbres they circulated: the yodeling of Jimmie Rodgers, the scat singing of jazz vocalists like Louis Armstrong, and the Hawaiian falsetto of singers like Mike Hanapi of Kalama’s Quartet.53 “In modern songs of considerable range one very often hears a peculiar break due to the slipping of the vocal cords which in a modified way resembles yodelling,” Helen Roberts wrote of Hawaiian singing in 1925. “This is so commonly encountered that it may be described as a feature of modern Hawaiian singing. Coupled with a habit, which is to be definitely traced to the ancient hula music, of gliding swiftly from a tone finished to the one to be attacked, slightly in advance of its normal appearance, it imparts a peculiar quality to the music which is quite foreign to our manner of singing.”54
In some cases sound recording circulated established vernacular singing styles; however, the adoption of the electric microphone also enabled a new range of intimate vocal timbres that came to be known as “crooning.” If the best-known crooner was the young Californian Bing Crosby, nonetheless versions of crooning emerged in a variety of vernacular idioms, perhaps most notably in the tremendous success of tango’s Carlos Gardel and samba’s Francisco “Chico” Alves.
The circulation of these distinct vocal timbres by recording led to their appropriation by the raciologies of the day, as folklorists and musicologists took timbre, like rhythm, to have a basis in the racialized body. “Timbre and recitation appear to be racial characteristics deeply rooted in physiological functions,” the German musicologist Erich Hornbostel argued in 1928, “and give therefore valuable evidence of anthropological relations and differences. Peoples and their music, then, are not so much distinguished by what they sing as by the way in which they sing.”55 The connection between race and vocal timbre ran throughout popular journalism. “The somewhat nasal singing of a native Don Juan or his female counterpart may not, at first hearing, find favor in ears trained to Western music,” E. V. Solomons wrote of kroncong records in 1933. “But, if one is prepared to listen to Krontjong music without prejudice, one will be forced to admit that the language of love is universal.”56 Abbe Niles argued that race records illustrated how “jazz draws on the timbres and accents of the Negro voice, weaving it into its fabric, contrasting its shrill or throaty melancholy with the voices of wood and brass, imitating, imitated by it, catching its tricks of expression, translating them, throwing them back again.” His praise for the recordings of Jimmie Rodgers was punctuated by the remark that Rodgers was a “white man singing black songs.”57
The association of vocal timbres and peoples—whether defined by race or nation, language or religion—was long-lasting, in part because of the materiality of sound recording’s schizophonia. In many cases the political and cultural meanings of the vernacular phonograph musics depended on the symbolic power of their vocal timbres. In some cases, singers turned the local idiom of a vernacular language into an emblem of a nation or a people. Umm Kulthūm’s resonance as the “voice of Egypt” lay not simply in her vocal power, both in volume and range, but, as Virginia Danielson has shown, on the clarity of her diction in Arabic—often remarked on by contemporaries—and on the color of her voice, its ghunna or nasality, and its bahha or hoarseness.58 Similarly, a recent critic has argued that Zhou Xuan’s diction in Beijing Mandarin was an important part of the making of Mandarin as the national language in China.59 In Hawaii, the recording of Hawaiian-language songs in vocal timbres drawn from traditions of chant were crucial to the survival of the language through decades of suppression; indeed, the ethnomusicologist Mantle Hood once noted the paradox that whereas, outside Hawaii, the timbre of the steel guitar became the emblem of Hawaii, inside Hawaii it was the vocal timbre of the sung words that was considered the most important aspect of the music.60
These recorded vernacular voices stood apart from the timbres of classically cultivated voices, and often validated an emerging vernacular aesthetic, constituting a “modern” voice. In her discussion of the “subaltern voice” in the Indian subcontinent, Lakshmi Subramanian notes the importance of the new timbres of Tamil vernacular singers like K. B. Sundarambal: though
what kept such voices outside the domain of classical music was in part a lack of training, and in part the way they interpreted music and its intentionality, … for some of the protagonists [in the debate over Tamil music], the preference lay with spontaneous full-throated singing unmediated by constructed notions of refinement and containment. A taste for this was developed by the circulation of commercial recordings and the growing popularity of early film songs, which became the logical site for Tamil music at the most immediate level. Singers like K. B. Sundarambal and Kittappa in particular embodied this register.61
Similarly, the fadista, Rodney Gallop wrote in 1933, “sings in the curiously rough, untrained voice and simple unpretentious manner which are dictated by tradition … The fado does not lend itself to bel canto, and the opera singer with his cultivated voice and professional manner would never be tolerated by the critical audience.” Rather, Gallop goes on, “against the strict common time of her accompaniment, the fadista maintains a rhythm as free and flexible as that of the jazz-singer, with whom she shares certain tricks of syncopation and suspension of the rhythmic beat which give the song a lilt as fascinating as it is difficult to reproduce.”62 Here we see the appearance of the second element that characterized the new ensembles of recorded phonograph music: their syncopated rhythms.
If the vocal and instrumental timbres—the music’s “noise”—marked these phonograph discs, so, too, did their “syncopated” rhythms. Virtually every commentator on recorded vernacular musics called attention to their rhythms, linking their syncopations to subaltern and colonized peoples, and seeing them as figures of social and sexual upheaval. “Down in the basement” of the “house of music,” a New Orleans journalist wrote in 1918, there is “a kind of servants’ hall of rhythm. It is there we hear the hum of the Indian dance, the throb of Oriental tambourines and kettledrums, the clatter of the clogs, the click of Slavic heels, the thumpty-tumpty of the negro banjo, and, in fact, the native dances of the world.”63 The vernacular musicians and ensembles often appropriated the term themselves: there were “Rhythm Kings” in Batavia, Buenos Aires, Johannesburg, Port of Spain, and New Orleans.
Others were more skeptical; the German critic Theodor Adorno discounted the originality of “those achievements of jazz in which people thought they perceived elements of a fresh beginning and spontaneous regeneration—its rhythms.” “The rhythmic achievements of jazz are mere ornaments above a metrically conventional, banal architecture, with no consequences for the structure, and removable at will,” he argued, “… the apparent variety of rhythmic constructs can be reduced to a minimum of stereotypical and standardized formulae.”64
Were these “rhythm kings” expressing a desire for freedom or simply playing standardized formulae? The intensity of this controversy over rhythm was as great as that over timbre. In part, it marked the emergence of a new ideology of rhythm deeply entwined with modern, colonial notions of race, one that continues to inform musical thought and practice. But it also grew out of the distinctive form and structure of the vernacular phonograph musics, figured by the centrality of their “rhythm sections.” Far from being “mere ornaments,” the rhythmic constructs were central to the architecture of the music.
The ideological link between rhythm and race that has inflected much musical discourse over the last century was forged during the modernist generation. In his essay, “The Invention of ‘African Rhythm,’” Kofi Agawu noted the “way rhythm as a separate dimension is singled out for special mention” in the influential essays on African music published by Erich Hornbostel and William Ward in the late 1920s.65 Both the German musicologist Hornbostel (working from phonograph recordings) and the British collector Ward (who worked with training college students in the Gold Coast) insisted on the radical difference between African and European conceptions of rhythm, and both maintained that African rhythm was more advanced: “African rhythm springs from the drummer’s motions and has far outstripped European rhythm, which does not depend on motion but on the ear,” Hornbostel wrote;66 while Ward argued that “in rhythm Africa is two centuries ahead of Europe … The sight of a native dancer keeping with different sets of muscles five different rhythms at once almost makes one believe that Africans have not merely cultivated their sense of rhythm far beyond ours, but must have started with a superior sense of rhythm from the beginning.”67
The invention of this idea of an “African rhythm”—spanning the peoples and cultures of the continent—was part of a modern racial discourse about rhythm that was historically quite new, as Ronald Radano has shown. In the nineteenth century, the sacred singing of the African-American jubilee groups had made vocal and melodic elements the “signature of race”; however, “within the space of thirty years … that perception would change dramatically. By the 1920s, when Americans imagined black music, they thought first of rhythmic practices, particularly in instrumental settings … the bodily affecting power of black rhythm consumed the attention of reporters and readers alike.”68 This rhetoric was often adopted by African-American critics discussing the new musics: in his pioneering essay on jazz in The New Negro, the journalist J. A. Rogers wrote of “that elusive something, for lack of a better name, I’ll call Negro rhythm. The average Negro, particularly of the lower classes, puts rhythm into whatever he does, whether it be shining shoes or carrying a basket on the head to market as the Jamaican women do.”69
However, this ideology of rhythm and race stretched beyond Africa and the African diaspora. The discourse about “gypsy rhythm” went back to Liszt’s assertion in 1859 that “the liberty and richness of its rhythms … more than anything else, tends to increase the admiration in which Gipsy music is held”: they were “distinguished both by a multiplicity and a flexibility nowhere else to be met with in the same degree.”70 And there was a parallel “invention of Polynesian rhythm,” as musicologists singled out rhythm in their accounts of the musics of the Pacific. “The hall-mark of Hawaiian music is rhythm,” Nathaniel Emerson wrote in his pioneering 1909 study of the hula, “for the Hawaiians belong to that class of people who cannot move hand or foot or perform any action except they do it rhythmically.”71 Similarly, the anthropologist Helen Roberts, who made field recordings of traditional chants in the mid-1920s, insisted that the proof that Hawaiians were “inherently musical” was “their excellent rhythm, about which even the earliest travelers to their shores remarked when witnessing their hulas and hearing the accompanying chants.”72 By the 1920s and 1930s, this association had become a commonplace, articulated in academic journals—the Musical Quarterly noted that “Drums were of utmost importance in all Hawaiian music, which was rhythmic rather than melodic”73—as well as popular journalism: the popular music writer Sigmund Spaeth wrote that missionaries to Hawaii had “found a heathen race musically addicted to rhythm and little else,” while Time reported that “contrary to popular belief, ‘Hawaiian music’ is not a pure racial product. Natives invented the rhythm, foreigners the melodies.”74
But the link between rhythm and race was not an eternal trope of racial thinking; nor was it a result of the encounter between traditional ceremonial musics of cultivated elites around the world. Rather the peculiarly modern notion of race and rhythm emerged alongside the eruption of the unsettling plebeian phonograph musics in the colonial ports; it offered a genealogy of these “mixed-race” musics. It is striking that both Erich Hornbostel and William Ward arrived at their sense of the radical difference between “European” and “African” rhythm through the mediation of the New World musics—blues, jazz, ragtime—which Ward dismissed as “a false and feeble imitation of the mixture of African and European music that developed among the American negroes.” For Ward, the popularity of jazz is “due solely to its rhythmic appeal; to the device, taken from West Africa to the Southern cotton-fields, and thence to Tin-Pan Alley, of superposing on the heavy regular dance beat a multitude of shifting, changing, shimmering subsidiary rhythms.”75 Similarly, both Emerson and Roberts arrived at their sense of Hawaiian rhythm through reflections on the contemporary vogue for commercial renditions of hula ku‘i.
For many, the distinctive nature of these new rhythms was captured in the notion of “syncopation,” a more or less technical term for the displacement of accents to the weak or off beat, that emerged, in these years, as a keyword in the popular discussion of the new rhythms. By 1928, at the height of the recording boom, the new edition of Grove’s Dictionary of Music and Musicians noted that “syncopation has become a general term for all that class of twentieth-century dance music which has sprung from the American adoption of rag-time.”76 “Syncopation is the reigning feature of both” the beguine and the rumba, the Martiniquan critic Andrée Nardal noted.77 But “syncopation” was also extended beyond its common connection to ragtime and jazz to the Polynesian musical arc and the gypsy musical arc. As early as 1909, Nathaniel Emerson wrote that Hawaiians had an “inclination to lapse from their own standard of rhythm into inexplicable syncopations,” the product of “an emotional susceptibility and a sympathy with environment.”78 Syncopation, the well-known “music appreciation” commentator Anne Shaw Faulkner wrote in 1921, “is found in its most intense forms among the folk of all of the Slavic countries, especially in certain districts of Poland and Russia, and also among the Hungarian gypsies.”79 And by 1933, the black South African critic Mark Radebe was defining “jazz” as “a perversion of some of the remarkable syncopating rhythms to be found in the Native music of many races.”80 As a result, syncopation came to have social and political, as well as musical, connotations. “Syncopation, this curious rhythmic accent on the short beat, is found in its most highly developed forms in the music of the folk who have been held for years in political subjection,” Anne Shaw Faulkner argued. “It is, therefore, an expression in music of the desire for that freedom which been denied to its interpreter.”81
The commonplace notion that the rhythms of these musics could be understood through the category of “syncopation” led many to hear them as simple ornaments and to dismiss their originality; “syncopation is new for popular music,” Adorno writes, “but by no means for art music.”82 However, as Carlos Sandroni has argued, the very term “syncopation” was less an adequate way of understanding the rhythmic character of the musics—he is writing about samba—than a metaphor that mediated between the vernacular musics and the dominant musical discourse. A concept drawn from Western musical theory, where it denoted a sense of rupture or dislocation and depended on the non-syncopated, it was an uneasy match for musics where the “irregular”—the variety of asymmetrical formulae—was the rule. Despite its inadequacy, the notion of syncopation was, he argues, adopted into the local knowledge of the music, giving it a musicological sanction and becoming a seal of authenticity: syncopation should be seen less as a scientific concept than a kind of “native-imported category like café or mango.”83
Thus the new rhythms of the vernacular phonograph records must be understood as more than simple syncopations, and less than the essence of a people or a continent. Rather, I would argue, they were the product of a new hybrid musical form: the “rhythm section.” For the distinctiveness of these vernacular ensembles derived not only from their vocal and instrumental timbres, but also from their characteristic divide between a front line of melody instruments and a “rhythm section” of bass and chordal instruments as well as drums and percussion instruments. The importance of these “rhythm sections” was noted at the time; the phrase first appears in English in 1924.84 There are analogous formations in other idioms: in the popular Cantonese opera of the 1920s, Su Zheng notes, the musical ensemble was called pai he/paak woh, “beat and accompaniment.”85 In Brazil, the sambistas who played percussion were known in the 1920s as “rhythmists.”86
The newness of the rhythm section involved four distinct if overlapping connotations: the appearance in the music industry of a range of traditional percussion instruments; the invention of the drum set; a structural divide among the instruments of the ensemble; and a formal principle of these vernacular idioms. First, the idea of the “rhythm section” marked the eruption into commercial music of the wide range of percussion instruments—ṭarab’s riqq, son’s bongos, hula’s ipu, samba’s pandeiro—that had been part of these new urban musical cultures, and that stood in sharp contrast to the mass-produced industrial instruments imported from Europe and North America as well as the aristocratic instruments fashioned by highly skilled artisans. “We still remember the marvelous stupor,” Alejo Carpentier wrote, “with which the people of our generation greeted, one fine day, the instruments that came from the eastern provinces, and that today are heard, poorly played, in all of the world’s cabarets … the bongo, on whose hide were heard the most sonorous glissandi with the palm of the hand; the creole timbales, secured between the knees, so nervous and mischievous, as they were struck with one or more fingers; the econes or cencerros (cowbells), little bells made of dull-sounding metal, played with a metal stick; the botijuela, a potbellied clay jar, with a narrow neck, from whose lips pour forth a sound analogous to the pizzicato of a bass.”87
These percussion instruments were often seen as uncouth noise-makers, sometimes banned from street festivals and excluded from the musical stage. In Cuba, son is generally taken to be the first music to “incorporate an Afro-Cuban drum (the bongo) performed with bare hands.”88 However, as late as the mid-1920s the bongo was banned from urban performances, though the Septeto Habanero was allowed to use it in private shows for Machado himself.89 “Thanks to the son,” Alejo Carpentier later wrote, “Afro-Cuban percussion, confined to the slave barracks and the dilapidated rooming houses of the slums, revealed its marvelous expressive resources, achieving universal status.”90
As a result of these exclusions, the “first” recordings of particular drums and percussion instruments have achieved legendary status, symbolic markers of the recognition of vernacular musics in many places. In Cuba, it is said that Eliseo Grenet’s productions of the late 1920s were “the first to adopt conga drums into stage entertainment.”91 In Brazil, the surdos, tamborins, and pandeiros of the samba schools first appeared in a recording studio, according to legend, in 1930 when the Bando de Tangarás, a group of young, middle-class whites, mainly amateur musicians, from Vila Isabel, fronted by the singer and radio host Almirante, recorded “Na Pavuna.”92 In Hawaii, the first commercial recordings of the ipu (a gourd drum), the uli‘uli (a gourd rattle), and the pahu (a wooden drum) were also between 1928 and 1930, most famously when Kalama’s Quartet decided to record “ancient chants” and fused their steel guitar hula ensemble with the traditional Hawaiian gourd drum.93
In another respect, the “rhythm section” marked the emergence of a new percussion apparatus, the “drum set” or “traps.” The development of pedals in the 1890s enabled drummers to combine a variety of percussion instruments into a mechanical drum machine that could be played by a single individual, thereby cutting labor costs in live entertainment industries. It seems to have first become a musical fixture in New Orleans: according to Royal Hartigan, the “snare and bass drums of the concert and marching bands in New Orleans provided a base to which, from 1900 to 1930, other accessories, or ‘trappings’—hence the name traps—were added.”94 “The trap-drummer of the early part of the century,” percussion historian James Blades notes, “… was connected with various forms of light entertainment, including the dance hall, the travelling show, the circus and the theatre pit. In the dance hall, he and the banjo constituted the rhythm section of the ragtime and early jazz bands.”95
By the early 1920s this mechanical combination of drums and cymbals had become an emblem of “jazz.” The “jazz-drum,” the 1922 Encyclopedia Britannica stated, “needs some explanation … The drummer uses a side drum, a big drum, and cymbals played with the feet, and various other instruments on which he beats a tattoo with his drum-sticks in alternation with the side drum. He is in fact a sort of one-man band in himself and adds considerably to the rhythm of the ensemble.”96 This “jazz-drum” turned up in vernacular dance ensembles around the world, regardless of their musical idiom: when it was adopted in the accordion-led musette bands of Parisian music halls, it was called “un jazz.”97 “Every jazz band in Asia, Africa, or Europe starts with the drum-and-trap accessories as a nucleus,” a New York Times reporter wrote in 1922. “This constitutes the jazz, the rest merely band.”98
But if the machine had its origins in New Orleans, the drum set was itself a hybrid of several vernacular traditions across the colonial ports. Chinese and Ottoman diasporas had brought percussion instruments of East and West Asia to the Caribbean and the Mediterranean, where they were assembled into the drum kit. Chinese percussion instruments were central to the early drum kit: “early drummers,” Royal Hartigan notes, “adopted the instruments they heard played by Chinese immigrants in urban areas in the late 19th and early 20th centuries, like the small Chinese cymbal (Bo), large gong (Da Luo), woodblock (Ban), temple blocks (Mu-Yu), and the first tom-tom (Bangu), usually a thick painted pigskin drum head tacked on to a red painted wooden shell.”99 Cymbals had entered European and American musics from the Turkish military bands of the Ottoman janissaries, and Istanbul—a key musical crossroads—remained a center of cymbal production, particularly by the Zildjian firm.100 Moreover, many of the drum sets that emerged included local percussion instruments as well as newly improvised sound-makers: “Nowhere else in the world, outside of New York, have the cymbals, bells, sirens, motor horns, cow bells, and all the clap-trap of the original ragtime bands been abandoned,” the New York Times journalist reported.101 “The trap-drummer was not only the metronome of the band,” drummer and historian James Blades concludes: “his purpose was to colour it with every sound possible from the instruments at his disposal, and to give the combination style with his ad lib syncopation.”102
Nonetheless, if the drum set turned up in “jazz” dance bands from Bombay and Shanghai to Rio and Cape Town, it was not adopted by all of the vernacular idioms. The drum set never became important in Hawaiian music where the rhythm section was dominated by a strummed ‘ukulele, often playing the olapa rhythm that had been associated with the ipu. Few of the gypsy-inflected idioms of the Mediterranean adopted the drum set; in tzigane, the hammered cimbalom was the key to the band’s rhythm section. Even in the Americas, one sees, as Matthew Karush notes, the “refusal of many tango bands to incorporate the drum set.”103 Noting that the drum set would later enter Cuban music via Cuban jazz bands, Ned Sublette points out that “the bass-snare-cymbal combination, so closely identified with African-American music, was never an important part of Cuban popular music which used timbales and African-derived percussion.” Thus it would be a mistake to conflate the spread of the drum set with the wider emergence of “rhythm sections.”104
For the rhythm section signified more than the eruption of local percussion instruments in the recording studio, or the spread of the drum kit. It was also the wider appearance of analogous forms of ensemble in modern dance bands—like the bando regional of Brazilian samba or the orquesta típica of Argentine tango—that were distinguished from both the street-parading percussion troupes and the violin-based “universal” orchestras of the hotel ballroom and concert hall. In Brazil, for example, recorded samba was based, as Carlos Sandroni has argued, on a “new instrumental synthesis of elements from Afro-Brazilian traditions and elements from the music played by middle-class urban groups,” which was called “‘regional,’ abbreviation of ‘regional orchestra,’ in the recording studios and the radios, to distinguish it from the ‘universal’ orchestra based on string and bow.” The “regional” ensembles combined what Sandroni calls a “chamber batucada,” in which the percussion, or batucada, of carnival parades was “reduced to a surdo, a pandeiro, one or two tamborins,” with “an instrumental ensemble of the type called ‘choro’ at the start of the century, that is, a harmonic base provided by guitars and cavaquinho joined by one or two soloists on the flute, clarinet, or mandolin.”105
The sambas recorded by these “regional” ensembles with their “chamber batucada”—one thinks of the recordings by Estácio’s Ismael Silva, or Noel Rosa’s 1930 recordings with the Bando Regional and Grupo Regional—differ in timbre and texture, as well as rhythm, from the orchestrated sambas of popular crooners like Francisco Alves and Mário Reis. It is a divide analogous to that between the “hot” jazz of Louis Armstrong’s Hot Fives and Hot Sevens and the “sweet” jazz of the hotel orchestras like that of Paul Whiteman, or between the steel guitar and ‘ukulele hulas of Kalama’s Quartet and the hapa haole film and hotel music of Harry Owens and His Royal Hawaiians. In each of these cases one sees a divide between the distinctive new ensembles of the vernacular phonograph musics, with their “rhythm sections,” and the incorporation of elements of the vernacular idioms into the “light music” mainstream of the commercial music industry, which remained dominated by the “universal” orchestra.
However, the “rhythm section” was not just a form of ensemble but also an informing principle of the music, the key to the structure and organization of the vernacular phonograph musics. The different idioms of vernacular phonograph musics were usually distinguished by the patterns woven by their rhythm sections, composite rhythms that established the overall feel or groove. For their rhythmic character was not simply a matter of a single rhythm, a single voice. Rather, rhythm, like harmony, is constituted by the relation between multiple voices. Just as the idiomatic harmonies of these musics involved “voice leading” from one combination of voices (a chord) to another, as well as the underlying modes or keys (or, more abstractly, pitch classes) that the voices share, likewise their idiomatic rhythms were constituted by the interlocking of relatively fixed voices—the high-hat or the claves—and the relatively variable ones, all working within an underlying grid or frame that held them together.
Alejo Carpentier identified this dialectic between “rhythmic regularity” and “rhythmic variation” when he wrote one of the earliest analyses of the Afro-Cuban son. The passage is worth quoting at length, because it captured—far more than the idea of syncopation or the ideologies of race and rhythm—the musical revolution of these new rhythm sections:
The great revolution of ideas instigated by the son’s percussion was in giving us the sense of a polyrhythm subjected to a unity of time. Up until then, one spoke of the rhythm of the contradanza, the rhythm of the guaracha, the rhythms of the danzón (admitting to a plurality within that succession). The son, on the other hand, established new categories. Within a general tempo, each percussive element assumed an autonomous existence. If the function of the botijuela and the diente de arado was rhythmic regularity, that of the timbales was to enact rhythmic variation. If the marímbula worked on three or four notes, marking the harmonies with the insistence of a basso continuo, then the tres furnished a cadence. The bongo acted more freely, using a more direct percussiveness or a glissando technique. The other percussive instruments would manage themselves according to their tonal registers and possibilities, according to the performer’s imagination, as long as the singing—all of the musicians sang—was sustained by the percussion.106
This dialectic between “rhythmic regularity” and “rhythmic variation” that Carpentier pointed to in the son is echoed in many of the vernacular phonograph musics, whose composite rhythm depends on the interplay between a relatively fixed, often cyclic, ground or grid, kept by the “rhythm section,” and the more fluid and varied figures and inventions of the solo voices, whether singers or instrumentalists. And these composite rhythms become the basis for the distinct “rhythmic dialects” of the vernacular idioms. “Much of the expressive essence of Latin music,” musicologist Peter Manuel writes, “lies in the intricate composite rhythm created by the percussion, bass, and piano. … This composite rhythm is the product of a set of standardized accompanimental ostinatos.”107 The real difficulty in crossing idioms, a jazz musician once said to me about playing with a clave-based Afro-Cuban band or with a Brazilian band using samba rhythms, lay less in the melodic or harmonic material than in “feeling the time,” inhabiting the interplay of individual rhythms that constitutes a particular rhythmic dialect.108
The different rhythmic voices were often highlighted by contrasting timbres: “distinct timbres,” Thomas Brothers notes, “help the ear distinguish the layers.”109 This is a structural reason for the prominence I noted earlier of multi-timbral rather than uni-timbral ensembles in these musics; it also accounts for the curious centrality of instruments that were often taken to be simple noise-makers, “percussion instruments of sharp and penetrating timbre,” in Sandroni’s phrase. A striking example is the Afro-Cuban claves, the pair of wooden dowels that sound the distinctive rhythmic timeline—also called the clave in Cuban music—that organizes the idiom.110
In recent years, musicologists analyzing these idioms have developed several ways of expressing this dialectic of “rhythmic regularity” and “rhythmic variation,” distinguishing between the “fixed” and the “variable,” the “background” and the “foreground,” the “ground” and the “figure.” Thomas Brothers insisted on the “basic distinction of fixed and variable” in his account of the New Orleans jazz of King Oliver and Louis Armstrong, arguing that “the technique depends on a division of musical labor: one group of instruments or voices (or just a single part) maintains a fixed rhythmic pattern; the other plays in variable rhythms, with the intention of creating a pleasing mix of agreement and disagreement with the foundational pattern.”111 Kofi Agawu called attention to “the essential tension between a firm and stable background and a fluid foreground” in his discussion of West African dance musics.112 Jeff Pressing distinguished “a figure–ground relationship … with the rhythmic patterns the figure, and the ground defined either by pulse, meter, or timeline” in his account of “the rhythmic devices of African diasporic music [that] typically rely on the support of a firmly structured temporal matrix, typically called a ‘feel’ or ‘groove.’”113 Louise Meintjes also used the notion of figure and ground in her study of South African mbaqanga: “The figure is a motive that is subject to ornamentation, to variation in repetition, over a more or less steady (bass) line, termed the ‘ground.’” Linking social theory with musical theory, she argues that figuring is “a process of arguing musically, by means of repeated and varied motives, over ideas about social relations.” These analyses are not identical: some stress the interplay between the ground established by the rhythm section and the figuring of the soloists; others emphasize the interlocking of rhythmic patterns and underlying pulse within the rhythm section that establishes the music’s feel.114
Oddly, Adorno’s hostile account of the rhythms of “jazz” was itself based on an analogous recognition of this divide between the fixed and the variable: “rhythmically free, improvisational constructions complement each other in such a way that, taken together, they fit back into the unshaken schema after all. Hence, for example, to cite only the simplest and most frequent case in point, two measures in three-eight and a measure in two-eight are combined sequentially to make a four-four measure, as marked out by the drum.”115 Unlike recent musicologists, however, Adorno does not find in it a “pleasing mix” (Brothers), or an “essential tension” (Agawu). Rather, for Adorno, the fixed schema—“stereotypical and standardized formulae”—turn the variable figures—the “rhythmically free, improvisational constructions”—into “mere ornaments … with no consequences for the structure, and removable at will.”116 Adorno’s sense that the rhythmic figures were less essential to the music’s structure than the rhythmic ground, or that the ground was banal precisely because it was a fixed grid, seems strange to postmodern ears: a sign not only of his unwillingness to acknowledge the distinct structural principle embodied in the modern dance musics, but also of the aesthetic revolution taking place at the time.
Throughout the vernacular phonograph musics, there are common rhythmic grids or frames, which are often associated with a dance. Record labels on the 78 rpm discs of the 1920s prominently displayed the dance genre of the recording alongside the song title and musician’s name. A similar phenomenon is found on the flamenco discs of the 1920s, where the palo of the recording—the classification as seguiriya, bulería, soleá, among many other styles characterized by rhythm and mode—was usually more prominent than the “title”—often little more than the opening words of the song. Thus, one might think of the tango and the foxtrot, the hula and the samba, not only as specific dances, but as forms of musical time: tango-time, foxtrot-time, hula-time, and samba-time might be thought of the way one thinks of waltz-time or march-time.
In addition to the proliferation of names of dances and rhythmic styles, two different principles run through the dance-like grids played by the rhythm sections of the vernacular ensembles: timelines and chord cycles.117 Timelines are distinctive rhythmic figures or cells, repeated throughout a performance or composition. The term was developed by ethnomusicologists discussing the West African dance rhythms that are often played on a metal bell.118 A timeline, Kofi Agawu writes, “is a distinctly shaped and often memorable rhythmic figure of modest duration that is played as an ostinato throughout a given dance composition.”119 Well-known Ghanaian timelines include the “standard” timeline, the kpanlogo, and the Gahu.120 However, the term has been extended beyond West Africa by theorists of rhythm: the principle of the timeline can be seen in the rhythmic cells at the heart of the African-American vernacular musics of Latin America—the tresillo, the cinquillo, the son clave, the habanera or tango rhythm, and the bossa nova—as well as the compás of the flamenco of Andalusia, the talas of South Asia, and the Turkish aksak rhythms of the Balkans.121 Though the timeline was articulated in different ways—with metal bells, wooden sticks (claves), or accented hand claps—these periodic rhythmic figures are fundamental to each musical idiom. “The key to understanding the structure of a given topos” or timeline, Agawu argues, “is the dance or choreography upon which it is based … No one hears a topos without also hearing—in actuality or imaginatively—the movement of feet.”122
The principle of the timeline was adopted by many of the vernacular music ensembles not only because they inherited the traditional practices, but because, as Godfried Toussaint points out, “timelines act as an orienting device that facilitates musicians to stay together and helps soloists navigate the rhythmic landscape offered by the other instruments.”123 The timeline was usually articulated by an instrument in the rhythm section, like the claves in son or the matchbox in samba, or by an element of the drum set, like the snare or cymbal.
However, the ostinato of the timeline was usually combined with a second rhythmic principle that grew out of the new rhythm sections in which not only percussion instruments but bass and chordal instruments served as timekeepers: the chord cycle. The chord cycle was a commonplace series of chords played by the rhythm section to establish a periodic harmonic rhythm, a harmonic ostinato. Such chord cycles are found throughout the vernacular phonograph musics across each of the musical arcs of the colonial archipelago. Across the black Atlantic, one finds the palm-wine patterns of West Africa like the “Yaa Amponsah” cycle;124 the four-bar chord ostinato of South African marabi;125 the twelve-bar blues cycle that structured not only recorded blues but much of early jazz and country recording in the United States; the thirty-two-bar “rhythm changes,” the chord cycle of George Gershwin’s “I Got Rhythm,” which was adopted by jazz musicians in the early 1930s; and the beguine chord cycle that, Prieto argues, played “a role similar to the Blues or Rhythm changes … for the biguine of Stellio’s era.”126 Across the gypsy Mediterranean, they include the Andalusian cadence of flamenco,127 and the chord oscillation between relative major and minor chords in the rebetika of Piraeus.128 Across the Pacific, they include the ten-bar hula cycle played on the ‘ukulele,129 as well as the thirty-two-bar kroncong asli chord cycle that was, in the words of musicologist Philip Yampolsky, “an inescapable harmonic-melodic pattern” in Indonesian popular music: “Imagine if I Got Rhythm were the only set of changes available to jazz musicians.”130 Like the timeline, these chord cycles served both to mark time and to enable improvisation within the idiom.131
The distinctive new rhythms and “syncopations” of the idioms of the vernacular phonograph musics were thus a product of the interplay of timelines and chord cycles in the rhythm sections of these “dance bands” and “orquestas típicas.” Some musics were based mainly on time lines: in Cuban music, the timeline of the clave was so omnipresent that the composer Emilio Grenet wrote of its “rhythmical tyranny.” When a composition passes into “popular hands,” he argued, “the rhythm almost always recovers its predominance. In the language of the people this is called meter en los palos (to put into the sticks).” The sound of the wooden claves dominates the music with “relentless authority”; “the steps of our dancers … follow the claves as closely as the shadow follows the body.”132 Other musics were based more on chord cycles: as the gourd pattern of the ipu gave way to the strummed chords of the ‘ukulele, the ten-bar harmonic ostinato dominated the recorded hula ku‘i of the 1920s and 1930s. However, the circulation of phonograph records led to hybrids of the two principles. Musics based on timelines developed a repertoire of commonplace chord cycles when they were adapted to chordal instruments like the guitar, while musics based on chord cycles—like African-American blues and jazz, often distinguished by musicologists by the absence of timelines133—began to incorporate timelines: though Latin timelines rarely structured entire jazz performances, the tresillo, cinquillo, and clave became, as Chris Washburne has argued, ubiquitous in jazz.134 In the United States, as David Schiff puts it, “the clave pattern wore a variety of disguises.”135
Moreover, unprecedented syntheses, hybrids, and transformations of the timeline and the chord cycle appeared. Ghanaian highlife, for example, was rooted in dance forms based on timelines, but it developed a distinctive timeline with fewer attacks, one that, Kofi Agawu suggests, expressed a distinctive modern ethos.
The Highlife ethos is essentially different from that of the traditional dances (like Agbadza or Atsiagbeko) … it is more committed to exploring—with a mixture of optimism and irony—the recent urban or modern turn in African history. Highlife is inflected by European ways of musical doing; it enshrines syncreticism as a creative force. This bundle of attributes may be read into the Highlife timeline, which occupies a lighter register in the hierarchy of African expressive forms.”136
But if highlife had a distinctive timeline, it also developed a repertoire of chord cycles drawn from early recordings like the Kumasi Trio’s 1928 “Yaa Amponsah.” In one of the earliest analyses of the vernacular phonograph musics, Robert Sprigge showed how the melody of “Yaa Amponsah” incorporated both the rhythm of the clave and the chordal patterns produced by Jacob Sam’s two-finger guitar style.137 In this way, the timbral and rhythmic syncretism of “Yaa Amponsah” leads to the third element of these new musics, their “weird” harmonies.
The pervasiveness of these deceptively “simple” chord cycles led to a curious paradox in the way the harmonies of the vernacular phonograph idioms were discussed. On the one hand, their harmonies were routinely discounted as rudimentary, hardly worth mentioning; on the other hand, they were often heard as strange or, in English-language commentary, as “weird.” It is not difficult to collect accounts dismissing the harmonies. The blues were, Abbe Niles wrote in 1926, “unconsciously so constructed that if the singer wished to accompany himself, he need know only three chords.”138 The sense that these idioms were harmonically simple was, as I have suggested, often based on a misunderstanding of the rhythmic centrality of the chord cycles: they are less “chord progressions,” leading voices to modulate and resolve, than “changes,” as the jazz tradition called them, marking time, and serving, as we will see, as a framework for improvisation. “If repetition of a harmonic progression seventy-five times can keep listeners and dancers interested,” Kofi Agawu argues, “then there is a power to repetition that suggests not mindlessness or a false sense of security (as some critics have proposed) but a fascination with grounded musical adventures.”139
The paradox is that these apparently simple chords often sounded weird, producing strange blends that fit neither the framework of common practice harmony nor that of the traditional modes. As a result, they offended the ears both of those trained in common practice and of those trained in traditional modes, maqam, and ragas. Perhaps the most celebrated early example of the resulting clash was in gypsy music: “the civilized musician,” Franz Liszt wrote in 1859, “is at first so astounded by the strangeness of the intervals employed … that he can find no other way of settling the matter in his mind than that of concluding the dissonances to be accidental; that they are mere inexactitudes; or, to be quite frank, faults of execution.” Nonetheless, “the harmony acquires a strangely dazzling character,” and Liszt went on to theorize a “gypsy” scale.140 The sense that different musical idioms could be defined by specific notes and scales was fundamental to the musicology of the period: it ranged from Carl Engel’s coining of the term “pentatonic” to characterize folk music scales in the 1860s141 to the early twentieth-century debates over the differences between European and Indian scales, as in A. H. Fox Strangways’s influential 1908 essay on “The Hindu Scale.” “The measurement of scales,” musicologist Matt Rahaim observes, “had become a sort of sonic phrenology.”142
In the decade before the recording boom, another celebrated example of these strange intervals emerged: the “blue note.” The blue note was apparently first named during the blues “craze” of the 1910s: an early use cited by the Oxford English Dictionary—“The lank trombone … unburdens its pent-up soul and sobs to high heaven the unspeakable agony of the famous ‘blue note’”—is particularly interesting, because it indicates that the term was both already famous and yet still unusual enough to put in quotation marks.143 The blue note was immediately heard as “weird”: “the weirdest blue notes ever heard,” a writer in the Dramatic Mirror noted in 1919, “began to flow … from the clarionet”; and W. C. Handy, the self-proclaimed “father of the blues,” called the blues guitar sound he encountered in Tutwiler, Mississippi, in 1903 “the weirdest music I had ever heard.”144
However, the “gypsy” scale and the blue note were less marks of distinctive “folk” scales, than the product of the clash between modal traditions and the common practice harmony embodied in printed popular sheet music. So-called gypsy notes were not part of the traditional musics of the Roma peoples of the European subcontinent, as Bálint Sárosi showed. Rather, they emerged in the “verbunkos idiom,” the popular urban Hungarian music, disseminated by sheet music with parlor piano accompaniments, but usually played by gypsy musicians in small orchestras built around a violin lead with a rhythm section anchored by the cimbalom, a hammered dulcimer. This fusion was heard as “gypsy music” from the time of Liszt to the era of Bartók; it was a major idiom of the 1920s recording boom, recorded and marketed as “tzigane.”145 The resulting clash between the common practice harmonies of the Hungarian popular sheet music and the performance practices of the gypsy violinists and cimbalomists—“the emphasis is never on what but on how they perform,” Sárosi writes—generated the characteristic “gypsy” notes.146
Similarly, the “blue note” was less a specific note in African-American folk singing than, as Abbe Niles noted in his introduction to W. C. Handy’s 1926 collection of sheet music “blues,” Handy’s influential solution to the problem of representing the blues singing he had heard in the terms of commercial sheet music. “Writing down his tunes with the memory of how the Negroes had sung, Handy was first met by the problem of perpetuating the typical treatment of the tonic third—the slur of the voices, whatever the song. This aberration he chose to represent by the frequent introduction of the minor third into melodies which … exhibited a prevailing major.” Handy’s “interpolated minor third”—which “acquired a name of its own: ‘the blue note’”—figured the clash between the traditional bending and worrying of African-American modes of singing and the parlor piano harmonies of Tin Pan Alley sheet music.147
However, this clash between common practice harmony and forms of singing shaped by intonational languages and modal musical practices was not only the product of the friction with the notational conventions of the sheet music industry. It was also a fundamental part of the colonization of the ear, as part-singing and solfège (the use of the do, re, mi syllables as training in common practice melody and harmony) were instituted in colonial choirs and choruses around the world, part of what Kofi Agawu has called the “violence of imposing harmony.”148
As a result, in all three arcs of the colonial archipelago, the new vernacular idioms were shaped by a clash between two practices of pitch. Early New Orleans jazz embodied, Thomas Brothers argued, the musicians’ response to the contradiction between blues training and solfège training: “blues finds expressive depth through subtle shading and dramatic bending of pitch, while solfège is designed with just the opposite goal in mind—to internalize the distance between pitches with such precision that there is absolutely no deviation from the measured scale.” Though “Baptist hymns, Sousa marches, and Ragtime songs were rigidly shaped … with the Eurocentric harmonic system in mind,” Brothers suggests, “it must have been perfectly fine to sing the hymns and popular songs with complete disregard for their harmonic implications.”149 Thus the “blue note” might be seen not only as the result of the attempt to notate the blues for commercial sheet music, but also as a sign of the tension between a blues ear and a well-tempered ear, a tension that marked the two worlds of New Orleans musical culture. For the Creole musicians as for Handy, Brothers argues, “Eurocentric harmony is a primary system of pitch syntax that must not be violated”; for Louis Armstrong and the emerging jazz tradition, on the other hand, “Eurocentric harmonic syntax is secondary and subservient.”150
In an analogous situation, many African musicians were caught between traditional African pentatonic sound fields and melodic shapes (beginning at a high point and coming down) and European cadences (with their stepwise melodic ascents and descents); composing under a tonal regime meant the grafting of one tonal system onto another or trying to inflect the tonal system.151 “Highlife melody,” Agawu writes, “is in a perpetual state of tension, seeking to balance the pull of an intonational language-based pattern against that of an invariant, harmonically driven, precompositionally determined one.”152
Hawaii was likewise the scene of a clash between the vocal glissandos of Hawaiian mele and hula that were imitated by the steel guitar, and the common practice harmonies of the Sankey gospel hymns that had been translated into Hawaiian. “In a century of acquaintance with harmonized European music the Hawaiians have learned to sing in parts,” the anthropologist Helen Roberts noted in 1925, “but unless they have been well schooled in choral singing, or are singing from notes, one frequently hears harmonies which are unorthodox according to classical rules but which, though novel, are seldom displeasing and always refreshing to ears on which certain combinations of sounds have fallen so long that they are unconsciously anticipated.”153
The clash between local melodic modes and common practice harmony was also the result of the widespread adoption of imported chordal instruments. As early as 1935, the song collector Rodney Gallop noted the way the adoption of the guitar had shaped the music of the fado. “The vocal line of the Fado is clearly conditioned by the harmonic accompaniment, chords of the tonic and dominant repeated in strictly symmetrical rotation, from which it is never divorced and without which indeed it would be difficult or even impossible to sing.” This guitar-influenced vocal style was, he suggested, the outcome of the tension between traditional modes and tonal harmony: “For centuries the Portuguese guitar which … was ‘well-tempered’ long before the days of Bach, has been used to accompany peasant song. Its harmonic system is based not on the famous Andaluzian ‘Phrygian’ scale of Southern Spain, but on the modern major and minor.” Thus “the instrumentalist … a semi-professional” had a “natural tendency to modernise old tunes, to tidy up what to him appear loose ends of rhythm and tonality to harmonise them according to modern ideas.”154
In many cases this produced what musicologist Peter Manuel has called a “vernacular tonality,”155 distinct from common practice harmony and modal practice alike. Ethnomusicologists have analyzed non-Western modal systems, Manuel writes, but they have neglected “the syncretic musical systems that have arisen … as products of the confluence of modal traditions with Western chordal harmony.” Each of these vernacular tonalities took shape when chordal instruments were adopted to accompany the vocal phrases of modal melodies; the chords—major and minor triads—were used not for their traditional functions in common practice harmony, but for “color … sonority, or … simple oscillation.”156
Manuel developed this hypothesis through his exploration of a “Mediterranean tonality,” in which local variants of the Arabic Hijaz mode or maqam, with its characteristic augmented second interval, were harmonized with European triads in urban musics from Seville and Cádiz to Istanbul, Athens, and Smyrna. In these musics, he argues, “chordal accompaniments consist either of static non-directional oscillations between a tonic and a secondary chord, or else they function as enhancements of a melody which remains predominantly modal.” For example, in the Phrygian tonality of the gypsy-derived cantes of flamenco, “the guitar accompaniment, where present, invariably consists of an ornamented oscillation between two chords—usually, the tonic and the flat supertonic—with occasional forays in the minor fourth chord, thence descending to the tonic via the familiar iv-III-II-I pattern [the “Andalusian cadence” of, for example, Am-G-F-E].” This Mediterranean tonality was often circulated by musical diasporas, including Jewish klezmorim, professional secular musicians who transmitted “stylistic features and musical genres … across international borders,” and Gypsy musicians who “in Eastern Europe as in Andalusia … tended to preserve and perpetuate older … musical practices, synthesizing them with modern, generally more Western styles, and thereby playing important roles in the evolution of a system of modal harmony which, in many respects, is common to the entire region.”157
In a series of essays, Manuel extended this analysis of the “Mediterranean tonality” to other forms of “vernacular tonality,” syncretic regional harmonic practices that are “qualitatively distinct” from common practice harmony, including what he calls “dual tonicity” in Latin American musics like the guajira, the country-style Cuban music of the early twentieth century (well known through the song “Guantanamera”), and the urban son, one of the fundamental musics of the recording boom of the late 1920s.158 Though “most of the harmonies of son,” Manuel writes, “… are fairly straightforward, based on I-IV-V pillars,” many son tracks recorded during the first half of the twentieth century “illustrate what would seem to be a curious indifference to finality, concluding, for example, on the subdominant, or … on what might otherwise be heard as the dominant.” This yields a “distinctive tonicity … even more resistant to Western musicological analysis than is the Andalusian tonality of flamenco.” Though Manuel acknowledges that these harmonic progressions can be (and have been) analyzed in common practice terms, he argues that these “simple and familiar-looking chordal vocabularies operate in a form of tonality quite distinct from common practice, with its relatively unambiguous sense of tonicity.” They are better understood as “a set of vernacular, guitar-derived conventions, consisting of simple chordal ostinatos, that emerged in the formative period of tonality and subsequently followed relatively independent trajectories,” enjoying “a prolonged life in various Latin American folk genres” as well as becoming “a harmonic feature of mid-twentieth-century Cuban popular music.” Manuel’s hypothesis of modal harmonies in the “vernacular margins of the Western mainstream” is a powerful way of understanding analogous harmonic practices in the recorded vernacular musics of the Polynesian Pacific and the black Atlantic.159
In Hawaii, a syncretic harmony was enabled by the central chordal instrument, the ‘ukulele, a Hawaiian adaptation of the machete brought by contract laborers from the colonial Atlantic island archipelago of Madeira in 1879. “The ukulele is now the common companion of every group of Hawaiian youths,” Helen Roberts commented in 1925, “who generally achieve with it only the necessary chords with which to accompany their songs, and these, like the vocal harmonies, are not always orthodox but often delightfully unexpected.”160 In West Africa, such a syncretic harmony developed in highlife: “highlife’s tonal types juggle the pressures of a closed I-V-I progression with various open modal structures.”161 Local modal sound fields and traditional playing techniques also changed the way chordal instruments were played: for example, in palm-wine guitar music, the two-note accompaniments of the traditional seprewa harp lute fused with the chordal forms of the guitar. Despite the differences between the two instruments, musicologist Kwadwo Adum-Attah writes, “the rural musician did not find it difficult … to play the old melodies on the guitar.” “The tunes of the traditional Asante ‘blues’ known as Odonson were transferred from the seprewa to the guitar.” However, “as a result of being exposed to Western hymn singing and its four-part chordal structure, the rural musician started to add more chords based on Western harmonic practices to his guitar accompaniment.”162
These modal harmonies or vernacular tonalities may have emerged as the result of colonial music industries and colonial musical training, involving the spread of commercial sheet music, the training in part-singing by missionaries, the travels of semi-professional dance musicians, and the circulation of mass-produced chordal instruments. However, the musical consequence was a soundscape defined by these weird if rudimentary harmonic rhythms, chord ostinatos, and oscillations, that, far from impoverishing the music, gave rise to a fourth aspect regularly noted of the recorded vernacular musics: their new forms of virtuosic improvisation.
One of the remarkable paradoxes of the vernacular phonograph musics was that the unprecedented inscription of these un-notated musics in shellac led to a new concentration on—indeed a revaluation of—the place of musical improvisation. The elements of music that had seemed most fleeting and unrepeatable—extemporized variations and ornaments—were now fixed on disc and subject to mechanical repetition. The recorded improvisation was a new form of reification, as a momentary relation between people became a material object; but it also initiated a new dialectic between recording and musicking, as improvisations were learned and copied from records, inspiring a spiral of improvisations on improvisations.
Improvisation in music had of course long existed, but, in nineteenth-century Europe and the European settler societies of the Americas, it lost recognition and value as printed composition came to dominate both art music and popular song.163 Improvisation was often dismissed as inessential, at best a virtuoso ornamentation, at worst a traducing of the composer’s intentions.164 Moreover, as improvisation declined in prestige, it became increasingly associated with subaltern castes of musical performers: “Extemporisation or improvisation,” the 1927 edition of the authoritative Grove’s Dictionary of Music and Musicians stated, is “the primitive act of music-making, existing from the moment that the untutored individual obeys the impulse to relieve his feelings by bursting into song. Accordingly, therefore, amongst all primitive peoples musical composition consists of extemporization subsequently memorized.”165 Gypsy musicians were particularly associated with improvisation in the late nineteenth and early twentieth centuries.166 Gypsy music, Liszt had argued, “more than any other belongs to the domain of improvisation, without which it does not exist.” The gypsy master “most to be admired is he who enriches his theme with such a profusion of traits (appoggiaturas, tremolos, scales, arpeggios, and diatonic or chromatic passages) that under this luxuriant embroidery the primitive thought appears no more prominently than the fabric of his garment appears upon his sleeve.”167 A half-century later, the improvising violinist of the tzigane orchestra was described as follows by an Irish observer: “In Hungary, the gipsy ‘Primás’ develops his rhapsody in the following way: first of all the slow, sad lassu in which the solo violin improvises arabesques and embroiders the Magyar tune possessing so many memories for his audience; meanwhile the accompanying fiddles and the cymbalum support his improvisation and when his inspiration wearies and sinks they enable him to soar again.”168
By the 1920s, this association of improvisation with the tzigane orchestras of eastern Europe made them a point of comparison with the new jazz bands. “A semblance of this lost, and rediscovered, art [of improvisation] is contained in the music of the Russian and Hungarian gypsies,” the musicologist Carl Engel noted in 1922. “Just as that music is a riotous improvisation, throbbing with a communicative beat, so is jazz. Just as the gypsy players are held together by an identical, inexplicable rhythmic spell, following the leader’s fiddle in its harmonic meanderings, each instrument walking in a bypath of its own, so is the ideal jazz band constituted—that is, the jazz band made up of serious jazz artists.”169 When the French composer Darius Milhaud first heard African-American jazz musicians in New York in 1922, his account was an uncanny echo of Liszt’s encounter with gypsy musicians a half century earlier:
In the jazz of the whites everything has been worked out to perfection and studied in the most thorough way. Among the Negroes there is far more improvisation. But what tremendous musical gifts and what power of performance are necessary to bring improvisation to such a pitch of perfection! In their technique they possess great freedom and facility. Each instrument follows its natural melodic line and improvises even while it adheres to the harmonic framework which underlies and supports the piece as a whole. We find this music perpetually employing a rich and confusing interweaving of elements. It uses major and minor chords together with quarter tones, which are produced by a combination of glissando and vibrato techniques—an exaggeration of the trombone tones, as well as vigorous vibration of the trumpet pistons and strange uses of the fingers on the violin strings.170
Throughout this discourse, improvisation was understood primarily as a performance practice, a supplement to an original song or dance. Despite Liszt’s praise, gypsy musicians were more often reproached for “tasteless ‘ornamentation.’” 171 Bartók’s 1931 argument that “even the much vaunted performance of the gypsies is lacking in uniform character … [and] authenticity” is echoed in Adorno’s 1933 argument that “even the much-invoked improvisations [of jazz], the hot passages and breaks, are merely ornamental in their significance, and never part of the overall construction or determinant of the form.”172 Though there is, as we will see, a material circumstance underlying this perception, it must be insisted that the improvisations of the vernacular phonograph musics such as tzigane and jazz, like those of other cultures of musical improvisation, were not fundamentally a matter of performance practice, of idiosyncratic ornamentations and elaborations. Rather, they were forms of what Derek Bailey called “idiomatic improvisation,” rooted in a sense of belonging to a musical speech community.173
The ability to speak freely in a musical vernacular is based on a knowledge of the musical idiom, its figures, its gestures, and its grooves. “Improvisation occurs always within a set of rules, or rather, conventions,” the pioneering musicologist Christopher Small has argued, and it depends “on the existence of a commonly agreed language.”174 The musical idioms of the colonial ports were such commonly agreed languages, embedded by ear in lips and hips, fingers and vocal cords. Their rules and conventions, vocabularies and rhetorics, were learned by playing in the small ensembles—the trios and sextets—that proliferated in the ports’ streets and parks, in the back corners of taverns and the bandstands of dance halls. Some idioms accented vocal improvisation, others dance improvisation, still others instrumental improvisation. A 1909 account of the improvised Hawaiian mele claimed that many Hawaiians “possessed the gift of improvisation in a remarkable degree.”175 The improvised music of Cairo’s ṭarab was based on the melodic designs embodied in the various modes, the maqamat. As a result, ṭarab improvisations were, as Ali Racy writes, “devoid of verse-like repetitions; solo-oriented; and nonmetric, or ‘rhythmically free.’”176
At the other end of the Mediterranean, the flamenco guitarists of Seville and Cádiz were known for their improvisations. “Take a characteristic gipsy guitarist in a tavern,” the critic and folklorist Walter Starkie wrote in 1935, “and you will find that he divides his accompaniment into three parts. First of all we have the rasgueado or general improvised prelude leading into the second part which is called the paseo or promenade. Then comes the theme which leads to the third part called falsetas or variations.”177 Across the black Atlantic, the emerging dance musics were characterized by forms of improvisation: in Johannesburg, marabi was, in the words of the Drum journalist Todd Matshikiza, “largely the illiterate improvisations of the musicians of the day”;178 in Buenos Aires and Montevideo, the early tango and milonga were “played by ear and the music was largely improvised.”179 In Rio de Janeiro, choro was an improvised music played by groups known as “pau e corda” (wood and strings): “The flute played the ornamented melodies, while the guitar and cavaquinho provided the improvised harmonic and rhythmic accompaniment as well as the melodic counterpoint.”180 In Havana, son “was the first [music] to prominently feature musical and vocal improvisation”; its montuno section, musicologist Peter Manuel writes, “usually commences with the trumpet player improvising two-bar phrases in alternation with the choral refrains (coros); typically, after four of these exchanges, the lead singer improvises his own soneos in alternation with the coro, which may continue indefinitely, perhaps interrupted by a solo on the bongo, or, less often, the tres.”181 “The swing idea of free improvisation by the players was,” Louis Armstrong said in 1936, “at the core of jazz when it started back there in New Orleans.”182
Out of these ensembles came two distinct forms of improvisation, determined by who set the “time.” In some cases, the time was established by a rhythm section, and solo voices conversed in the framework of a cyclic pattern, whether a periodic harmonic rhythm (a repeated “chord progression”) or a timeline stated by clave or bell. In other cases, the time was established by the call of a lead voice—expanding and contracting the time—to which other voices and instrumentalists responded: one hears this in the ṭarab recordings of Umm Kulthūm as well as in the flamenco recordings of Pastora Pavón.
Paradoxically, the word improvisation was often avoided by improvising musicians. “Idiomatic improvisers,” Derek Bailey points out, “in describing what they do, use the name of the idiom. They ‘play flamenco’ or ‘play jazz’; some refer to what they do as just ‘playing.’” “In most of the world’s musical traditions,” Christopher Small concludes, “the word ‘improvisation’ has little significance, since what we have been calling improvisation is just the normal way of musicking; they call it, quite simply, playing, and the idiom in which they work as, equally simply, ‘the way we play.’” 183
If most vernacular musicking is, in this sense, improvisation, the real question posed by the moment of the vernacular phonograph musics is not “why were they improvised?” but “why did the idea of improvisation become so prominent?” Why did many of these musics develop a rhetoric, even ideology, of improvisation? I would suggest that the improvisations that flourished in the vernacular phonograph musics were not simply the appearance on record of long-standing practices of musicking. Rather, these “recorded improvisations” mark the emergence of a new, specifically modern, culture of musical improvisation that depended on a feedback loop between three elements: urban vernacular idioms that prized competitive virtuosity, embodied in public contests; an industry of printed sheet music that inspired the practice of “faking”; and the emergence of electrically recorded discs that captured and circulated the improvisations of star musicians.
The first element in this feedback loop was a product of the urban musical economy. In the local venues of live entertainment that exploded in the late nineteenth and early twentieth centuries—from vaudeville theaters and dance halls to sports stadiums—forms of virtuosic performance and athleticism flourished, creating stars and becoming a kind of entertainment currency. In the vernacular musics, improvisation became not just the normal practice of variation and extemporization within the conventions of an oral tradition, but part of a culture of improvisatory display in a continual round of contests among musicians. Some were informal, like the rodas of Rio’s choro or the “jam sessions” of jazz (the phrase emerges in the late 1920s).184
But many were public competitions and contests—concursos in Spanish—sponsored by dance halls, radio stations, and record companies. In Batavia, there were regular kroncong competitions—concours kroncong—from 1915 into the 1920s; some matched kroncong groups against one another, others were song duels between lead singers.185 The 1922 Concurso de Cante Jondo in Granada not only marked the resurgence of traditional flamenco but led to Odeon recordings of the guest artist, Manuel Torre, and the prizewinner, El Tenazas.186 The first son concurso in Havana in 1926 was won by the Sexteto Habanero for their performance of “Tres Lindas Cubanas”; they recorded it for Victor a few months later.187 From 1924 to 1930, the record label Discos Nacional sponsored tango contests in Buenos Aires: “the finest composers and poets presented their work, which was then judged by the members of the audience. The prize-winning compositions were made into records and the entire contest proceedings were broadcast” on radio.188 In Atlanta, the Georgia Old-Time Fiddlers’ Convention took place annually from 1913 to 1935, and success in the fiddling contest often led to recording contracts for local fiddlers like Fiddlin’ John Carson and string bands like Gid Tanner and his Skillet Lickers.189 And in Trinidad, the world of calypso depended on regular competitions during carnival season. “Competitions were a major source of income,” the historian Gordon Rohlehr writes: the calypsonian “had to be where the competitions were … in Princes Town or Sangre Grande, Port of Spain, Tunapuna or San Fernando.”190 These formal and informal contests put a new emphasis on improvisatory prowess; as Thomas Brothers notes of early New Orleans jazz, “quickness, extended-range playing, speed of execution, fresh ideas, harmonic experimentation—all were sources of reward and prestige in the manly musical world of New Orleans … The twin pressures of commercial reward and the climb toward urban sophistication helped stimulate improvisation.”191
The second element of this feedback loop was the place of notated music in these vernacular idioms. The notion of these musics as “improvised” depended on their relation to the commercial hegemony of the popular song, circulated as sheet music. In the face of popular printed music, modern “improvising” emerged in tension with “reading,” as can be seen in the curious appearance of the notion of “faking” music. In the US in the 1910s, the terms “fake,” “improvise,” and “interpolate” were used interchangeably; by 1924, a critic in the music trade paper, the Musical Courier, joked that “improvise” was “the polite word for ‘fake.’”192 The two contrary, if complementary, senses of faking captured the clash between “ear” cultures and “reading” cultures that ran through many of the musics of the colonial ports. On the one hand, vernacular musicians who played by ear and did not read European notation were confronted not only with the ubiquitous sheet music of popular song, but with the “charts” of commercial dance bands, “stock arrangements” of popular hits,193 and the “method books” of instrumental instruction. Flamenco method books dated from Rafael Marín’s Método de Guitarra pro Música y Cifra in 1902,194 while the Kamiki Hawaiian Guitar Method Book, which appeared in a number of editions from 1916, was only the best known of many Hawaiian steel guitar manuals. By the mid-1920s, reams of such printed materials accompanied the recording boom: for example in Batavia, the Odeon record dealer Tio Tek Hong, who issued kroncong disks, also published a series of kroncong songbooks for violin and mandolin from 1924 on.195
In the face of this print music culture, “faking” was the ability—at once respected and disrespected—to improvise a song (or a part in an arrangement) without reading the notation. On the other hand, those vernacular musicians who could “read”—having received formal training on their instruments through religious schools, military bands, or traditional musical apprenticeships—often found that this training did not prepare them for the practices of the urban dance musics. The figures and gestures of the music rarely appeared in the notated sheet music. So, for reading musicians, “faking” was the ability to improvise beyond the bare bones of the sheet music (bootlegged collections of lead sheets came to be known as “fake books”196). By the mid-1920s, publications appeared that instructed reading musicians in the art of faking and improvising. Reading musicians could pick up riffs from a new genre of sheet music that printed solos, like Kaai’s Method and Solos for Ukulele or Tiple (1926; Ernest Kaai was the leading Hawaiian ‘ukulele player of the era) and Louis Armstrong’s 125 Jazz Breaks for Cornet (1927, notated two-bar turnarounds in a variety of keys) and 50 Hot Choruses for Cornet (1927, more than a quarter of which were improvisations on tunes by Jelly Roll Morton).197 When the composer Milhaud visited the US in 1924, he found the method books of the Winn School of Popular Music “extraordinarily valuable … as regards improvisation,” showing how “such devices as arpeggios, trills, runs, broken chords, omissions, dissonances, embellishments, ornaments, variations, and cadenzas … are introduced ad libitum at the end of the parts of the various instruments, but in such a way that the rhythmic regularity of the whole does not suffer.”198
The third element in this feedback loop was the emergence of recorded improvisations, a trend that accelerated with electrical recording. The earliest acoustic recordings of these vernacular musics tended to ignore improvisation, in large part because the early recording industry borrowed its commercial model from the sheet music industry: it was selling songs rather than performances. Moreover, the technical difficulties of acoustic recording meant that there was little room for spontaneity in early recording; relatively anonymous studio musicians trained to play for acoustic horns were usually more successful at recording than were musicians known for their live performances.
The microphone of electrical recording changed this dramatically: it not only created new kinds of “location” or “field” recording, but also what one might call a new “moment” recording. The spontaneous performance in the studio became more important than the carefully rehearsed rendition of a song. At first, these improvisations remained restricted to relatively short passages: “hot riffs” within the three-minute limits of the 78 disc and the standardized structure of the commercial popular song. But as performances rather than songs became the currency of the recording industry, virtuosic improvisers became celebrities, and their records circulated widely. For example, in Rio de Janeiro, the recordings of the flutist Alfredo da Rocha Viana, Jr., better known as Pixinguinha, marked a distinct shift; he was “one of the first choro musicians to improvise over more than a few bars, a practice which soon became characteristic of the genre.”199 In Cairo, Umm Kulthūm was celebrated for performances and recordings in which she could “spontaneously produce multiple versions of a single line.” Using “familiar and well-loved modes: rast, bayati, nahawand, and huzam … Kulthūm’s varied repetitions were,” Virginia Danielson writes, “… intended to bring the listener closer to the mood of the line with each successive iteration. The tools she brought to bear were tone color, rhythm, and articulation rather than melodic excursions that might obscure the words.”200 In the US, electrical recording triggered an explosion of jazz improvisation: Victor’s electrical recordings of Jelly Roll Morton’s Red Hot Peppers in the fall of 1926 and the spring of 1927 were followed by OKeh’s first electric session with Louis Armstrong’s expanded Hot Seven in May 1927, the session that produced the famous stop-time trumpet solo on “Potato Head Blues,”* a landmark in “playing the changes.”201 In Cuba, Armstrong’s equivalent for the son was Félix Chappottín, the trumpeter who first recorded with the Septeto Habanero in 1928, inaugurating, on tracks like “Coralia,”* the distinctive improvisational style that was to make him “the most important trumpet player in Afro-Cuban music history.”202
The new prominence of improvisation was also evident as instrumental recordings appeared in idioms that had been dominated by singers, like Hawaiian hula and Andalusian flamenco. In hula ku‘i, steel guitarists adopted a handful of standard tunes for displays of instrumental virtuosity. “Hilo March,”* composed in the late nineteenth century by Joseph Kapaeau Ae‘a for a wind band, became, in the 1920s, a popular vehicle for a generation of young Hawaiian steel guitarists, including Sol Ho‘opi‘i, Mike Hanapi of Kalama’s Quartet, and M. K. Moke of Johnny Noble’s band.203 They also recorded a number of instrumentals featuring the popular “chimes” effect—using the guitar’s string harmonics—like Ho‘opi‘i’s “Chimes” (a version of the popular “Maui NoKa ‘Oi,” often known as “Maui Chimes”) and Moke’s “Moana Chimes.” In flamenco, two guitarists who had accompanied singers on dozens of discs in the 1920s, playing brief improvised falsetas—the gitano veteran Ramón Montoya, a regular at Gramófono sessions, and the younger virtuoso from Seville, Niño Ricardo, Montoya’s equivalent at Regal—made the first recordings of solo flamenco guitar in 1928.204 Montoya—“whose virtuosity is quite astonishing,” a record reviewer wrote in 1930—went on to pioneer concerts of flamenco guitar, with a celebrated 1936 appearance in Paris that was linked to the release of an album of instrumentals, Arte clásico flamenco.205 The younger Ricardo also became well known for improvising lengthy falsetas, and his recorded improvisations later became flamenco standards.206
Paradoxically, it was the repetitive nature of phonograph culture—record listeners playing the same improvised performance over and over again—that produced a distinctly modern conception of musical improvisation. As a self-conscious ideology, the notion of “improvisation” was most elaborated in North American jazz, and many of those who pioneered forms of recorded improvisation—including Sol Ho‘opi‘i, Pixinguinha, and Django Reinhardt—were often characterized as “jazz” players. The absence of this ideology of improvisation has been noted in some idioms: “there was little free or ‘jazz’ solo improvisation in marabi,” David Coplan writes, and Eric Prieto argues that “the art of improvisation was simply not a priority of the biguine style popularized by Stellio and his contemporaries … nowhere do we find a hot chorus of the kind that a New Orleans musician like Louis Armstrong or Sidney Bechet would have used to put his mark on a tune.”207 Nonetheless, marabi and beguine were cultures of idiomatic improvisation, not of notation. Moreover, the combination of urban musical contests, printed method books, and phonograph discs that produced modern improvisation was found across the archipelago of colonial ports, and so modern improvisation is better understood as the reverberations of a variety of musical idioms and practices.208 The recordings of the flute solos of Pixinguinha, the trumpet solos of Armstrong and Chappottín, and the guitar solos of Ho‘opi‘i and Montoya pulled their respective musical idioms away from the time and space of local dancers and singers and into the time and space of circulating phonograph records. Tempos became faster, harmonies denser, and instrumental prowess ever more evident, as many of these idioms developed increasingly elaborate modes of improvisation.
These new forms of improvised virtuosity—together with the timbral fusions of local percussion with industrial instruments in an explosion of rhythm sections creating a host of distinct, and sometimes mutually incomprehensible, rhythmic dialects and vernacular tonalities—were thus not so much a new sound—“jazz”—nor a new “world music,” as they were a profound revolution in musical practices, a remaking of the musical ear across each of the musical arcs of the colonial archipelago: the black Atlantic, the gypsy Mediterranean, and the Polynesian Pacific. Embedded in the grooves of shellac discs, this noise uprising was not only circulated and amplified at the time; it also set the stage for a series of remasterings, as the sonic legacy of this brief recording boom shaped the century that followed.