Imagining Native America in Music

Reprinted from Imagining Native America in Music by Michael V. Pisani.
Copyright © by Yale University Press. Used by permission of the publisher.

  • READ an interview with author Michael V. Pisani.



    Style and Ideology in Indian Musical Portraiture
    (from Chapter 7, “In Search of the Authentic: Musical Tribal Portaits, 1890-1911,” pp.225-235.)

    The combined use of a program for both a character piece and a tribal melody, Kiowa or Dakota, implied that some composers in the early part of the twentieth century considered specific ethnic associations to be essential to the realism of their portrayal. This ideology was partly an extension of developments in poetry, art, and, most notably, photography, which, with the advent of periodicals such as National Geographic, seemed poised to eradicate all fanciful notions of savagery. “Each [of these] must be what it purports to be,” wrote Edward Curtis in 1907 to accompany his remarkable photographic display of Native Americans. “A Sioux must be a Sioux and an Apache an Apache; in fact, every picture must be an ethnographic record.” 19 Using stunning visual imagery, Curtis purported to illustrate what he saw, “not what one in the artist’s studio presumes might exist.” Thinking topographically, Curtis hoped his “ethnographic records” in vivid chiaroscuro would reflect the great national and cultural diversity of the many Indian nations.

    Photographic portraits, like published transcriptions of tribal Indian music, were slices of reality, answers to the romantic rendering of native America by many nineteenth-century painters. Yet photography, unlike an actual performance of Indian music (but like an instrumental character piece), is an artificial medium. It freezes images—sometimes very powerful and influential images—that are as much of the artist’s making as they are reflections of the subjects in it. Its ability to convey is of course limited by what the photographer frames or chooses to emphasize. Variety is the greatest strength of Curtis’s collections, particularly in his twenty-three-year publishing project entitled The North American Indian. Nevertheless, his determination to isolate authentic Indian cultures suggested a kind of imagined purity. Most efforts to define cultures as pure are deeply problematic from an anthropological perspective, even if they appear to succeed on a poetic level as art. Cultures are never entirely “authentic,” notes anthropologist James Clifford, but rather contingent and subject to local reappropriation. In other words, ethnographic identity must always be understood as mixed, relational, and inventive. The racial and cultural purity that Curtis sought in his Indian portraits was first of all selective. He would often ask his male subjects, for example, to wear “traditional” costumes rather than their coveralls for his photographs, and his images rarely show Indian cultures as complex and adaptive.20

    Original sheet music cover for Edward MacDowell’s New England Idyls (Boston: Arthur P. Schmidt, 1902).

    Carrying this analogy over into music, a transcribed Indian song notated by someone schooled in European musical practices differs from the actual performance of an Indian song in the same way that a photograph of an event differs from the event itself. Without knowledge about how to perform the song, its context, and allowances made for notational inaccuracies or compromises, the ink on paper fades to a pale substitution. This notated song is what we might call the first level of removal from a fully experiential cultural source.

    In 1899 Krehbiel, a supporter of both Dvorák and MacDowell, prepared a survey of the scholarly literature on Indian music transcription. His resulting New York Tribune essay demonstrated his awareness of the complex (and sometimes ineffectual) processes involved in collecting and notating such music. He also drew attention to the number of American composers who had begun to adopt written transcriptions of Indian music “as thematic germs” for their own compositions. By 1899 Indian music had lost much of its earlier stigma. In fact it had developed a cachet, especially as adapted by composers. I would refer to adaptations such as MacDowell’s character pieces as the second level of removal from an original musical source. Since the “folk melody” (or notated Indian song) is itself a mere lifeless “photo” once removed from the event, the composition is like a “portrait” of the musical photo. On the one hand, the composition reactivates the life of the notated melody (much as a documentary film that incorporates a photo might do), but, on the other, it adds a wholly imaginative dimension and crosses the threshold from life to art. Naturally, such musical portraiture fully engages the subjectivity of the artist, and each musical portrait, though it may use the same notated Indian themes as another work, results in a different interpretation of the subject it attempts to portray.

    Musical portraiture similarly embodies constructs of native America that were based partly on source material (transcribed melodies) and partly on interpretation (the stylistic features of individual composers). Early twentieth-century composers, steeped in the notion that Indian cultures would soon vanish from the earth, felt it their right, if not their responsibility, to borrow what they saw as distinctive characteristics from Indian tribal musics. (Appendix 2 includes a selective list of such musical “tribal portraits,” works that attempt to depict some particular aspect of a specific native American culture. The bulk of the tribal-specific compositions date from about 1903 to 1924, after which interest waned. For purposes of comparison, the list also includes important works that are largely indexical and do not feature specific tribal music. While this list is selective, it includes works produced by major music publishers, and I kept the overall proportion of findings roughly the same so as not to skew the data.) Following the recognition of the rhetorical power of pentatonicism in the 1890s, three new techniques that developed during this concentrated period contributed to the ongoing syntax in music that reflected native America—all of them in some way derived from folk cultures, though not necessarily American Indian cultures per se. These techniques encompassed (1) melodic parallelisms (also associated with primitivism, alterity, and orientalism); (2) modality (associated with ancientness as well as the sacred); and, less commonly, (3) dissonance (associated with the “rawness” of the primitive experience). Let’s consider briefly the important roles of each of these with a few examples.


    It may be difficult to believe, but melodies stereotypically harmonized in parallel fourths or fifths, though ubiquitous in popular culture by the 1930s, were relatively uncommon at the beginning of the century. Westerners at this time would have had few opportunities to hear, for example, the sheng, a Chinese mouth organ that played melodies in parallel intervals such as fourths. Debussy and Ravel were familiar with it, of course, and imitated the effect in their later music. Puccini, though, was perhaps the first Western composer to use parallel fifths in La bohème (1896), where the fifths’ “openness” symbolized the bleakness of winter. Melodies harmonized in parallel fourths first turn up in America—indeed, anywhere in modern Western composition, to my knowledge—in an Indian setting. These appeared as part of the remarkable (and much underrated) publishing venture of Arthur Farwell, who included a significant number of Indian character pieces in the early years of his Wa-Wan Press (Newton Center, Massachusetts, 1901–11).21 Particularly striking among the Indian items in the Wa-Wan series were those of Harvey Worthington Loomis. Exactly when Loomis wrote these is unknown, but Farwell published them in two installments as Lyrics of the Red Man, Books 1 and 2 (1903 and 1904). Loomis featured many of these works in his concerts, among them (as already noted) his much-publicized 1905 lecture-recital on Indian music in New York City. Though each of the character pieces was supposed to reflect some aspect of contemporary tribal Indian life, Loomis, who was no stranger to reaching beyond the borders of his style, created a distinctive Indian idiom all his own. Like his contemporary exoticists Granville Bantock in England and Charles Koechlin in France, he was steeped in orientalism and deeply immersed in Arabic and Chinese musical styles (as well as American Indian). In 1899 Witmark published Loomis’s “Chinese Lullaby” (set to a verse by Edwin Starr Belknap), and as late as 1919 Oliver Ditson published “In Chinatown,” a musical postcard from San Francisco. Works such the recitation Sandalphon and the opera The Bey of Baba reveal his fascination with the Middle East. One of Loomis’s several dramatic pantomimes, The Garden of Punchinello, was even written for the Shah of Persia, Ahmad Shah Qujar, and in 1914 was performed in Tehran.22

    To make the connection between Indian folksong and composition perfectly clear, Loomis (or Farwell, as editor) preceded each miniature with a musical epigraph of the original source melody from Fletcher and LaFlesche’s A Study of Omaha Indian Music. Loomis’s titles, like “Prayer to Wakonda,” “Ripe Corn Dance,” or “The Chattering Squaw,” were clearly meant to stimulate performers’ and listeners’ imaginations. These were, at the very least, associations with specific aspects of modern Indian life, in contrast with, say, MacDowell’s poetic titles. (As part of a lecture, of course, their Indian context would have been further clarified.) From a technical standpoint, they are more pianistically demanding than many other Indian character pieces. They are also stylistically different from anything that came before—indeed, from all previous Western instrumental music.

    In “The Chattering Squaw,” the work that caused such mirth among his New York audience, Loomis harmonized a pentatonic Cree melody in parallel fourths (table 6). While this feature might seem to us today to resemble some Tin Pan Alley Chinese stereotype, vaudeville songs about Asians at this time did not yet contain this particular feature. The parallel index entered the American popular song repertory sometime between 1903 and 1909, the same time that indexical Indian features began showing up in this venue.23

    A. Edward MacDowell, “Indian Idyl” from New England Idyls (1901), middle section
    B. Ernest Kroeger, “March of the Indian Phantoms” (1904), beginning
    C. Harvey Worthington Loomis, “Chattering Squaw” from Lyrics of the Red Man, op. 76, book 2 (1904), mm. 3–10

    The accompaniment for “The Chattering Squaw” in the piano left hand is specified as imitating a large and small drum. The underlying rhythm is a persistent [whole-note, half-note, half-rest]. The original melody is basically in pentatonic mode V (here Bb–G–F–Eb–C), but it also contains nonpentatonic inflections: a D neighbor tone above the high C and an Ab passing tone between G and Bb. Loomis’s parallelisms are actually quite sophisticated by comparison with the Tin Pan Alley parody of Chinese music, which is usually in strict parallel fourths. He keeps the Ab, for example, when the Cree melody reaches up to D, resulting in fourths that are sometimes perfect, sometimes tritone. The effect is odd but also shrill, which may have been intentional, given the nature of this droll song. In their use of melodic parallel fourths, Loomis’s Lyrics of the Red Man are the earliest works to draw upon Far Eastern allusions to portray American Indians. Even Debussy, who used parallel fifths in the bass and inner voices, did not do so in the melody (in print anyway) until 1903 (in “Pagodes” from Estampes) and more overtly in 1905 (La mer) and 1910 (Préludes, Book 1). Indeed, Loomis’s parallelisms in these Indian pieces seem to have been re-ceived as quite a novel effect. Critic Henry T. Finck’s response implies that it was a new phenomenon to him as well. Instead of perceiving a connection with Asian music or anything by Debussy, however, Finck wrote that “The Chattering Squaw” was a work “in which consecutive fourths and fifths pro-duce a cacophony reminding one of the medieval organum of Hucbald’s day.” We can presume that listeners in 1904 would have heard these melodic paral-lelisms as parody, as Far Eastern “oriental,” or even—depending on their familiarity with music history—as neomedieval. In any case, a new trope had been added to musical Indianism.24


    Modality, on the other hand, had served as a distinctive colorist device in nineteenth-century Europe as a marker for the folk (as in Chopin’s mazurkas) or for referencing older European church music (as in Liszt’s oratorios). Western European composers also used modality for other purposes. Gounod and Respighi, for example, read late nineteenth-century treatises on sacred music by Niedermeyer, Bragers, Arnold, and others on how to provide accompaniments for, and thereby modernize, Gregorian chant. These scholars tried to reconcile the medieval Roman church modes with the normative practices of nineteenth-century major-minor harmony. The ancient Greek modes, from which the church modes were derived, were said to embody traits of character —Dorian for bravery, Phrygian for warlike, and so on—and it is well known that Plato believed strongly in their effect on human behavior. After several centuries of neglect, interest in the modes began to surge dramatically in the 1870s and early 1880s; this was largely due to influential composers such as Musorgsky (who found his national voice partly in old Russian Orthodox chants) and Debussy, who, significantly, had adopted the practice from Musorgsky. But in the early twentieth century, staunch nationalists like Vaughan Williams came to modality through a profound interest in regional folk music, much of which still retained the ancient modes. Moreover, pentatonic melodies could be mapped onto modal scales, as Bartók discovered around 1905. Pentatonic mode V, for example, could be set to Dorian, Phrygian, or Aeolian scales.

    Modality was exceptional in nineteenth-century American orchestral and vocal music, showing up only in the “Indian” works of Arthur Foote and MacDowell. Yet by 1915, when Frederick Converse fashioned a sophisticated modal accompaniment as a solution for a migrating Cheyenne melody in The Peace Pipe (written for the Chautauqua Festival), modal settings of Indian music had become fairly common through tribal portraits and larger works. One obvious marker of modality, the lowered leading tone, “inevitably suggests antiquity,” wrote the Iowa theorist and composer Horace Alden Miller in New Harmonic Devices (1930). Another particularly “Indian” device, according to Miller (who wrote quite a few Indian works himself), was the use of minor dominant chords (which of course incorporate the lowered leading tone). In a later study, Modal Trends in Modern Music (1941), Miller demonstrated how the harmonization of a Chippewa (Ojibwa) song could be accomplished without any dominant sevenths, relying solely on minor subdominants and other substitution chords. Miller even included examples of “Indian cadences” in his book. The lowered seventh figures prominently in many of these. One that Miller isolates as a particularly “interesting cadence” (B minor seventh to G major seventh) does have the uncommon effect of resolving to a dissonant chord. More important for our subject, however, the penultimate chord in this cadence (as well as the “minor dominant sevenths” that Miller highlights) is identical with the sonority of overlapping fifths in Dvorák’s “New World” Scherzo (beginning in m. 5: E–B–G–D), which, as discussed in chapter 4, established a kind of primeval sound through modal suggestion.25

    Natalie Curtis noted in 1913 that visiting European musicians such as Felix Mottl, Vasily Safonov, and Ferruccio Busoni, all of whom took “keen interest in our native music,” found that Indian melodies needed a wider range of modal possibilities than simply major or minor accompaniments. The use of modes and other harmonic devices to avoid traditional harmonic associations has also been noted in the compositions of Amy Beach, not least in several works based on Indian themes, especially those of the Alaskan Inuit. Adrienne Fried Block has pointed out the expressive possibilities that Beach found in the use of folk music, particularly in her larger and more complicated forms. “Indian music,” wrote Block, “provided her with themes of the utmost sim-plicity, almost without harmonic implications. In setting the theme Beach had to both simplify her harmonic style and find compositional means that al-lowed for development while not overwhelming the primitive themes.”26

    While composers such as Beach searched well into the 1920s for more effective ways to incorporate Indian themes (such as in the string quartet), modal harmony continued to offer an alternative to pure major-minor diatonicism, not only in Indianist music but also in works based on African American and Anglo-American spirituals and folk song, both of which retained many of the old modes. (We might ask why chromatic harmony, certainly another viable musical language in this post-Wagnerian age, was rarely used to invoke native America during this period. Given its fundamental connection to European musical modernism and expressionism, it may have been seen as too neurosis-laden or relating specifically to states of mind and therefore in-compatible with the tropes of Indianism, which were either of the body and nature-based or of the soul and hence spirit-based.) Modality in Indian musical portraiture in the early twentieth century may have helped to infuse the works of MacDowell, Miller, Beach, Herbert, and even Sousa with a sense of authenticity (because of their ancientness). Perhaps because of its innate centuries-old characteristics, modality also served to deepen listeners’ spiritual connection with the subject matter, regardless of how they may have felt about the contradictions between the nature of Indian music and the essentially Western medium of expression.


    As editor of the Wa-Wan Press, Farwell wrote short essays to accompany each volume of this Americanist enterprise. In his introduction to volume 2, no. 12 (1903), he professed that he had arrived at nine “Articles of Faith” with which composers could turn to American Indian music, thereby finding the spiritual depth lacking in much contemporary classical music. At the same time, however, he had discovered “new motives and rhythms” in Indian music that were synchronous with modern sensibilities. He began with Fletcher’s Omaha collection, of course, claiming that he had studied the songs as mono-phonic works, (i.e., without Fillmore’s harmonizations). But when he set some of them himself in American Indian Melodies—his first character pieces—he rarely diverged from Fillmore’s original harmonies. This may have been due largely to the fact that when he began setting Indian songs around 1900, there were no other models to follow, and his compositional style was still, as a young man, relatively undeveloped.27

    The first of Farwell’s American Indian Melodies, “The Approach of the Thunder God,” could be compared with Fillmore’s original setting to show that, with the exception of the added MacDowell-like “Indian” progression (iv6–ii65–i) in m. 9 and in the final cadence, Farwell makes only a few minor alterations in the accompaniment: some repeated notes and occasional smoother voice leading. Farwell’s settings thereafter quickly grew more elaborate. Works such as Dawn, Ichibuzzhi, The Domain of Hurakan (all 1902), the first Navajo War Dance (1904 [later titled "No. 2" by John Kirkpatrick upon its belated publication in 1947]), and especially “Pawnee Horses” from the suite From Mesa and Plain (1905) are not just illustrations of how to harmonize an Indian melody; they are fully realized compositions that convey a solid compositional technique, an emotional depth, and a keen ability to convey the impression of place and experience in a quasi-narrative setting.28

    Farwell was probably the first creative artist since George Catlin to draw existentially on American Indian subjects. He strove to escape the nineteenth-century’s “imaginary Indian” and, like Edward Curtis, moved toward more realistic interpretations of living Indians and toward the representation of raw experience. His (actual) second Navajo War Dance for piano (1905) is particularly remarkable in its radical shift away from the more staid rhythmic and harmonic techniques of the previous generation of American composers. Dark and musically disjointed, its melodic inflections and rhythmic vigor reflect Farwell’s encounters in the Southwest with Indian ritual music. He did not forget the thrill of excitement he first experienced from hearing two young Indian men sing Isleta Pueblo songs at the home of ethnologist and entrepre-neur Charles Lummis in Los Angeles, nor the Navajo chanting and drumming he transcribed from Lummis’s collection of cylinders. In 1909 he recalled the inception of the Navajo War Dance:

    It was at this time that I made my first really savage composition on Indian themes. . . . I had earlier inclined to the more pastoral songs and peace chorals, and folks reasoned naively that these could not represent the Indian, since the latter was a savage. Evidently I must reform and do something really Indian. The theme of the Navajo War Dance was something to make your blood curdle and your hair to stand on end.29

    What was “blood curdling” about the work, however, was not any particular “savageness,” but rather Farwell’s experimentation with modernist techniques in harmony and rhythm in combination with the Navajo song. He peppered his accompaniment with nonfunctional dissonances and varied the length of successive phrases to avoid a sense of predictability. The first presentation of the theme, for example, cadences on the seventh measure, with an extra measure of repetition added to round off the phrase to eight. The next phrase cadences after only five measures, but Farwell adds a measure of interruption that seems imagistically evocative of a dancer’s leap. After yet another extra measure (which sounds like imitations of vocal “yips”), a third phrase begins, though first with a one-measure false start. Played “with savage aban-don,” this phrase then proceeds to cadence unexpectedly on the sixth measure. The cadences are all on stark open fifths, but the harmonies that accompany the melodic phrases are quite pungent. Farwell also arrived at the solution—just about the same year as Bartók did—of setting a pentatonic melody to a nontraditional triad, one derived from the pentatonic mode of the source tune (in this case, C–D–G). He later adapted this piano work for the Westminster Choir in Princeton (Four Songs on Indian Themes, 1937). The eight-part a cappella choir sings triplets with Indian vocables: “weh-eh ah / weh-eh ah / weh-ah ha / hi.” In this revised version, the composer added even more dissonances and some new rhythmic features drawn from his experience of Navajo performers. The lower basses and altos hold firmly to a throbbing two (the drum) while the rest the chorus continues the triplets (the voices). In Navajo music, singers use freer rhythms, often in threes (which represents the spirit) against the regular patterns of the drumming (which represents the physical).



    19Edward S. Curtis, “North American Indians,” National Geographic 18, no. 7 (July 1907): 483.

    20Clifford, Predicament of Culture, 10. On Curtis’s methods of composing his photographs, see Lyman, Vanishing Race and Other Illusions.

    21Puccini used melodic parallelisms in Edgar (1889), but not fourths or fifths. Act 2 of La bohème opens in a festive mood with parallel triad chords. It can be argued that the parallel fifths at the opening of act 3 shadow the earlier parallels and symbolize the fact that the warmth of love has turned to ice. The last “Indian” set in the Wa-Wan Press series appeared in 1906; two short piano miniatures by Carlos Troyer appeared in separate issues the following year and were its final Indianist works.

    22Henry T. Finck wrote that Loomis’s reputation rested on his songs; see Finck, My Adventures in the Golden Age of Music (New York: Funk and Wagnalls, 1926), 275. A short assessment of the first half of Loomis’s career can be found in The Musiclover’s Calendar 2 (1906): 73-75. Several of Loomis’s dramatic pantomimes were produced by the Academy of Dramatic Arts in New York. The majority of his works can be found in the Loomis Collection in the Music Division of the Library of Congress. The manuscript score of The Garden of Punchinello, with the original scenario by Kendall Banning, is at Dartmouth College Library.

    23My assessment is based on the songs published between 1900 and 1920 on Asian subjects in the online MftN and LEVY databases. Angelo Read, the Canadian composer and conductor, illustrated melodic similarities between “Indian music” and “Asiatic music in “The North American Indian and Music” (Musical America, 13 July 1907). Such ideas, expressed elsewhere during this period, were aligned with those of Francis E. Leupp, Theodore Roosevelt’s commissioner of Indian Affairs, and other “race fusionists,” as they were called. It is not possible to determine exactly when American sheet music publishers began to issue popular songs that depicted Asian subjects. For more on this early twentieth-century repertory, see Tsou, “Gendering Race,” and Lancefield, Hearing Orientality. Other American composers of this time who took an interest in Asian subjects were Edgar Stillman Kelley (“Aladdin” Suite, 1892), Charles Tomlinson Griffes (Sho-Jo, 1917), and Henry Eichheim (Oriental Impressions, 1928).

    24Finck’s comment on Hucbald appears in “What is American Music?” 12. Other composers of serious instrumental character pieces after Loomis adopted this dichotomous “medieval”/”oriental” technique for their Indian music. Notable among them, perhaps, was Blair Fairchild in Some Indian Songs and Dances (1927). Fairchild’s adaptations of Indian themes in an orientalist light clearly related to his general interests; many of his other compositions were inspired by the Middle and Far East. I have been unable to find an example of Loomis’s technique in any music from the preceding three decades, as opposed to other parallel harmonic progressions—such as those involving full triads—that can readily be found elsewhere in nineteenth-century music.

    25John Vincent (The Diatonic Modes of Modern Music, 71) has found the III7-i cadence (but without the added seventh in the III chord) in the slow movement of Dvorák’s “New World” Symphony, the second act of Rusalka, Sibelius’s First Symphony, and Ravel’s L’heure espagnole, among others, where he notes that “the blandness of the III to i close offers grateful relief from the directness of the classical V-I cadence which in many cases [i.e., illustrated by the examples provided] would be too severe or too brusque.”

    The role of modality in American music has an interesting back story, In her famous essay “The Roots of American Culture” (188-89), Constance Rourke pointed out that New England in the early nineteenth century (which at the time represented a microcosm of American sophistication) turned its back on the emerging, and essentially rural, shape-note singing, which used modal harmonies. Its approach was considered unsophisticated and primitive. Because it was used in Englang long before the sixteenth century, “it belonged to an order of feeling antedating the complexly intellectual religious thought belonging to the Puritans; and this may be the true reason why, even with the lead of Billings, even with the inheritance of ballads and glees and catches cast in these older forms, shape-note religious song, which for many years used mainly the pentatonic or hexatonic scale, did not take firm root in New England.”

    26The Curtis quotation is from “Perpetuating of Indian Art,” 631. The Block quotation is from “Amy Beach’s Music on Native American Themes,” 162. See also Block, “Amy Beach’s Quartet on Inuit Themes.”

    27Farwell, who had prepared his 1900 lectures for Cornell University largely from August Wilhelm Ambros’s Geschichte der Musik, was no doubt well aware of Ambrosos earlier “harmonizations” for some non-Western monodic melodies in his book, for example, the Chinese songs.

    28All of these pieces are reproduced from original editions in Vera Brodsky Lawrence, ed., The Wa-Wan Press, 1901-1911, vols. 1, 2, and 4 (New York: Arno Press, 1970).

    29Farwell, “Second Trip West,” in Wanderjahre, 123. This Navaho War Dance, from From Mesa and Plain: Indian, Cowboy, and Negro Sketches, five works for piano (Wa-Wan Press 4, no. 28, 1905) was actually the second such titled work Farwell wrote. An earlier unpublished Navaho War Dance (1904) was edited by John Kirkpatrick in 1940 and later published (somewhat confusingly) as Navaho War Dance No. 2 (Music Press, 1947).