FRANCES DENSMORE WAS A DISTINGUISHED MINNESOTA SCHOLAR who for over half a century worked to record and interpret the music of American Indians. At a time when most Americans were interested in eradicating Indian culture, she labored with a sense of urgency to document their customs and music. Essentially self-taught and with little institutional support, she tirelessly collected and transcribed native music, conducting field studies among tribes from the Pacific Northwest to the Florida Everglades. By the time of her death in 1957, she had made over 3,500 sound recordings, published more than 20 books and 200 articles, and lectured throughout the United States on Indian customs and music. Although her musical and analyses reflected a training in eighteenth- and nineteenth-century European music, her scholarly interest in the music of Native Americans made her an American pioneer in the field of ethnomusicology.

Frances Densmore was born on May 21, 1867, in the Mississippi River town of Red Wing. Her paternal grandparents, Orrin and Elizabeth Densmore, had migrated westward from New York state to Wisconsin in the 1840s. In 1857 they moved on to Minnesota and settled in Red Wing, where Orrin became the superintendent of a sawmill and a citizen of some prominence. He had strong interests in educational and scientific topics and recorded local weather data for many years, sending off monthly reports to the Smithsonian Institution.

Orrin and Elizabeth's oldest child Benjamin - Frances' father - was born in New York in 1831 and moved west with his family. Trained as a civil engineer, he served with the Sixth Minnesota Regiment during the Civil War and then returned to Red Wing, where he and his brother Daniel founded the Red Wing Iron Works. In 1866 Benjamin married Sarah Adalaide Greenland and the next year their first child, Frances, was born.²

During her early childhood Frances frequently saw Indians on the streets of Red Wing. As she recalled many years later, her curiosity was aroused and she became eager to learn more about these "strange people."³ She also remembered that her interest in their music was stimulated when she was very young: "our home commanded a view of the Mississippi River. Opposite the town, on an island, was a camp of Sioux Indians and at night, when they were dancing, we could hear the sound of the drum and see the flicker of their camp-fire. In the twilight I listened to these sounds, when I ought to have been going to sleep .... So I fell asleep with my mind full of fancies about the 'interesting people' across the Mississippi."⁴

Frances' formal musical education began early and was, she recalled, "Spartan in its severity. I was taught harmony on the keyboard while I still lisped, and took 'time out' if I played anything as frivolous as Pleyel's German Hymn with variations during a practice period." On the strength of this foundation she went on to undertake the most extensive and rigorous musical training then available in the United States. In 1884, at the age of seventeen, she entered Oberlin Conservatory in Ohio, where she studied piano, organ, and harmony. Returning to Minnesota after her graduation in 1887, she gave piano lessons at a St. Paul music store and served as a church organist. During the winter of 1888 she began further musical training in Boston, studying piano with Carl Baerrnann and counterpoint with the distinguished Harvard composer and professor, John Knowles Paine. After two winters of study in Boston, she settled again in St. Paul and rented a studio to teach piano.

In a brief autobiographical sketch Frances Densmore recalled that her scholarly interest in Indian music probably began while she was in Boston. About that time she had learned of the research being done by Alice Cunningham Fletcher on the music and customs of the Omaha Indians. Densmore acquired a copy of Fletcher's A Study of Omaha Music shortly after its publication in 1893, and it inspired her to begin her own lifelong study of North American Indian music and customs.⁵

Fletcher's book was an important early work in its field, as well as a model for Densmore's own studies. When Densmore started her work, the music of non-Western cultures, previously of interest largely as a novelty, was just beginning to develop as a subject for serious study. Of critical importance in this development was Thomas A. Edison's invention in 1877 of sound recording, a technology which provided a means of preserving unwritten music for study and analysis. Alice Fletcher's account of Omaha music was based on sound recordings made with a primitive but sturdy device suitable for prairie travel. Her book stimulated considerable interest in Indian music and firmly established the importance of sound recording for future work in the field.⁶

Fletcher's study included descriptions of Omaha customs and presented the Indian songs as transcribed by her collaborator, John Comfort Fillmore, director of the Milwaukee School of Music. Fillmore transcribed the songs with piano accompaniments, a procedure he explained in an analytical "Report on the Structural Peculiarities of the Music" which was appended to Fletcher's study. In the course of working with Fletcher's collection of songs, Fillmore had become convinced that Indians possessed an unconscious sense of harmony, by which they even conceived modulations from one key to another. He therefore presented Fletcher's Omaha songs with elaborate harmonizations not unlike those of nineteenth-century gospel hymns. Fillmore's theoretical work on native music was to have a considerable influence on Densmore's own ideas about the structure of Indian music.

In 1893 Densmore visited Fillmore in Milwaukee in order to become fully acquainted with his theory. She remembered that he explained his work in detail, "playing the harmonized songs on the piano and describing his experiences with the Omaha."⁷

Returning to Minnesota, she studied Fletcher's book and began to read systematically on the subject of Indians. "For the next ten years," she recalled, "I soaked my receptive mind in what army officers wrote about Indians, and what historians wrote about Indians, [along] with some of the publications of the Bureau of American Ethnology.... All this was preparation for my life work."'⁸

During the early years of her study of Indian music, Densmore continued her career as a professional, performing musician. She was active in St. Paul and Red Wing as a piano teacher, church organist, church choir director, and professional lecturer on various musical subjects to local clubs. In 1895, she added a novel lecture to her repertory. Now, in addition to her talks on Wagnerian operas and other conventional musical topics, she offered a lecture on Indian music based on the research of Alice Fletcher, with whom she was corresponding and who was offering her direction and encouragement.⁹

On December 4, 1895, Densmore delivered her first lecture on Indian music before the Schubert Club of St. Paul, which had assembled for the occasion at Conover Hall. The talk was illustrated with Omaha songs. Many years later Densmore recalled that she played the piano accompaniment while one of the Schubert Club members sang the songs. In later lectures Densmore sang the songs herself and provided her own accompaniment.¹⁰

From the beginning Densmore took her work as a lecturer seriously and paid meticulous attention to the manner of delivery. She took elocution lessons, and an undated lecture manuscript from about this time shows careful underscoring to indicate details of pronunciation and inflection. Between 1895 and 1904 she gave at least fifteen talks on Indian music in Minnesota, North Dakota, Illinois, and New York. Her appearances were promoted with printed circulars which included complimentary extracts from a variety of press notices and personal endorsements. In these early years she sang the Omaha songs from Fletcher's book, using the piano at times to "simulate a drum" as well as accompanying herself by striking two sticks together and clapping her hands. Then in 1903 she acquired a drum and four birch-bark medicine rattles used in Chippewa medicine rituals. Soon mastering the rudiments of performance on these devices, she expanded her repertory to include a talk on native musical instruments. She lectured in addition on such topics as "The Indian's Natural Sense of Harmony," "Primitive Rhythms," and "Indian Life Expressed in Music."¹¹

About this time Densmore also began to use two Omaha songs in an arrangement by St. Paul-born composer Arthur Farwell, whom she heard late in 1903 in a lecture-recital entitled "Myth and Music of the American Indians and its Relationship to the Development of American Musical Art." She subsequently obtained a copy of Farwell's "Dawn," a piano fantasy on two Omaha songs, and used it to demonstrate the possible application of Indian music to modern composition."¹²

In 1903 Densmore began her long and prolific career of writing on Indian music. In May of that year the Minneapolis Journal published her article, "The Song and the Silence of the Red Man." The piece was based on James Mooney's The Ghost-Dance Religion and the Sioux Outbreak of 1890 and was accompanied by photographs from the book. In this and other articles written about the same time, she treated the experience of Native Americans in a highly romanticized fashion, more as drama than as history. While her early writings are by today's standards sentimental and patronizing, they nevertheless display a genuine and sympathetic interest in Indians and their culture. This attitude continued to characterize her work as she moved into more rigorous studies, and perhaps it helped her to earn the confidence and cooperation of Indian acquaintances.¹³

In 1905 Frances Densmore made her first extensive acquaintance with Indians in their own cultural setting. Accompanied by her sister Margaret, who was to join her on many later field trips, she visited the Chippewa at the villages of Grand Marais and Grand Portage on Lake Superior. To reach them the Densmores boarded a small passenger boat which left Duluth on August 9 and sailed up the shore to Grand Marais. Upon arriving, Frances hired an Indian guide named Caribou to show her the sights of the town. With Caribou she also trekked through the woods to call on Shingibis, the Grand Medicine man, from whom she hoped to learn more about Chippewa music. "I asked Shingibis if he knew the hunting songs that have magic power over the animals," she wrote. "A quizzical smile crossed his kindly face as he replied with appropriate gestures that when he went hunting he did not sing, but kept very still, took good aim and shot the game. Slightly crestfallen I changed the subject." Later "after considerable circumlocution I returned cautiously to the subject of Indian music. 'When do you sing?' I asked most politely. Shingibis waved his hand in the direction of a church spire which pricked the distant green, and replied, 'White man goes to church over there. I don't go to church. Have Indian church at home. I sing when I have Indian church.'"¹⁴

Shingibis agreed to hold a religious ceremony that night, and after making arrangements about price and the length of the session, Densmore went back to Grand Marais. That evening she returned with four chaperones. She later described the ceremony: "Shingibis was watching for us, and ushered us into one of the wigwams, motioning us to seats on a cot which extended along one side, and introducing us to Blue Sky and one or two other Indians who sat on the opposite side. Around the fire in the center were fresh sprigs of arbor-vitae laid carefully on the ground. Shingibis seated himself upon some bright splint mats at the end of the wigwam. Beside him was a wooden box, upon which stood a nickel alarm clock and a small kerosene lamp. Before him stood his Grand Medicine tom-tom, with its top of soft brown deerskin.

"We sat in respectful silence until Shingibis was ready to begin.

"The songs were full of wild beauty, but I did not attempt to use a note book, - it would have seemed a sacrilege. Shingibis sang one after another, beating the tom-tom with the curved stick in his right hand and shaking steadily the Grand Medicine rattle in his left."¹⁵

This ceremony and its associated songs were of great interest to Densmore and marked the beginning of her research on the Midewiwin, or Grand Medicine Society of the Chippewa. When she visited the Grand Portage band, she gathered enough material for two short articles. She also made a number of lantern slides at the Chippewa villages which she later used to illustrate her lectures.¹⁶

In the summer of 1906 Densmore collected songs from two Sioux women at the Prairie Island village near Red Wing. Although she had written down Indian music from memory on earlier occasions, Densmore remembered this event as the first time an Indian singer actively cooperated by performing songs for transcription. One of the Prairie Island women sang a song of the Haethuska (Warrior) Society which was a variant of an Omaha song Densmore had often performed in her lectures. The following year she published the two Sioux songs with a brief account of the occasion and an explanation of the social significance of the songs. She embellished her transcriptions with piano accompaniments similar to those made by John Fillmore for Alice Fletcher's study.¹⁷

During the same summer of 1906, Densmore made an extended visit to Indian communities in northern and western Minnesota. At White Earth in the western part of the state, she attended on June 14 a Chippewa celebration marking the anniversary of the tribe's removal to that reservation. Accompanied by the Reverend Benjamin Brigham and Mrs. F. C. Wiswell, an Episcopal missionary who was teaching lace making to the Chippewa, Densmore witnessed the ceremonies. "The celebration was colorful and picturesque, with the Indians dancing in full regalia," she wrote. "Chief Wadena and Joe Critt were there and let me take their pictures. Chief Mejakigijig was among the dancers, with prominent chiefs and men from other reservations. There were war dances in which veterans of wars with the Sioux dramatized their victories, and 'squaw dances' in which an invitation to join was accompanied by a gift of beadwork or a few yards of calico. I heard for the first time the throb of the huge drum and the lusty songs of the men who sat around it. My new friends explained it all to me, described the participants and interpreted for me."¹⁸

The next year Frances, once more accompanied lay her sister Margaret, again attended the White Earth celebration of June 14. Afterward one of the Indians agreed to sing the songs into a recording machine. Thus arose Densmore's first opportunity to make recordings of Indian music. Borrowing the necessary equipment from a local music store, she set up a temporary recording studio in the store's backroom. The Indian Big Bear (Kitchimakwa) sat in front of the recording machine and filled twelve cylinder records with his songs.¹⁹

The summer's work was by no means finished. From White Earth, the Densmores went on to the Red Lake Reservation where they spent three weeks gathering further material about tribal life and customs. Frances then went alone to the Leech Lake Agency at Onigum. She arrived to find that Flat Mouth, hereditary chief of the Pillager band of Chippewa, was gravely ill. Medicine men were performing rituals in a desperate last effort to save his life. Frances Densmore was permitted to witness the extraordinary events: "Hour after hour I stood outside the circle of Indians, watching the medicine men and listening to their songs. As the end approached, Flat Mouth was carried into a teepee and a gun was fired when his spirit passed away. A funeral feast was held the next day and the Indians let me go into the lodge while it was in progress. They drew back the curtain that concealed the body of the dead chief and let me take a photograph. In all my experiences I have never felt so much alone!"²⁰

On her return to Red Wing, she wrote to the Bureau of American Ethnology at the Smithsonian Institution, reporting the events she had witnessed at Onigum and soliciting support for her efforts to record Indian customs before they disappeared forever. William H. Holmes, chief of the bureau, responded with a grant of $150 to support her work. With these funds Frances purchased an Edison recording machine. She returned to Onigum in September and persuaded the Chippewa to repeat the songs for her. "The Indians remembered my presence there at the time of Flat Mouth's death," she wrote, "and the medicine man who was in charge of the ceremony recorded many of his best songs. Several others recorded songs of the Grand Medicine and drew the pictures that represent the words of these songs. I tested the accuracy of this system of mnemonics by showing the pictures to members of the Grand Medicine Society at White Earth, a few weeks later, and they sang the same songs." In her account book for 1907 Densmore noted her travel expenses to Onigum and White Earth as "Trip No. 1 - September 25th - October 19th - average cost per day - $3.70." Thus began her professional field work and her association with the Bureau of American Ethnology.²¹

Early in 1908 the bureau summoned Densmore to Washington, D.C.., to report on her work. She presented her recordings and information on the Chippewa Grand Medicine ceremony, for which the bureau paid a fee of $300. This was the first investigation of native music supported by the bureau. Its annual report for 1907-08 described her work enthusiastically, noting that "the collection of phonographic records thus far obtained is extensive, and the investigation promises results of exceptional interest and scientific value."²²

Densmore's association with the bureau immediately brought her research to the attention of her professional peers. The Anthropological Society of Washington, taking advantage of her presence at the bureau, invited her to lecture. When she addressed the society on February 18, 1908, Alice Fletcher was in the audience. Thus Densmore finally met the woman who had been the greatest inspiration to her work.

The Bureau of Ethnology continued its financial assistance, and in 1909 she returned to Washington for a second report on her continuing research and to announce the completion of her manuscript on Chippewa music. In a second address to the Anthropological Society she described in detail the method she had devised for recording and transcribing, and the system she had developed for analyzing the songs. She also spoke of the sense of mission which inspired her work. "The purpose of the present work is the collection and classification of data with a view to determining the natural laws which govern musical expression.... It is as though one set out to chart an unknown land. Another has blazed the trail . . . It is not mine to prove that there is beauty and poetry in Indian music - that fact was established years ago by Miss Alice C. Fletcher. It is ours to follow.... In a few years the valuable Indian songs will cease to exist . . . We have taken from the Indian his land and his hunting ground but he is carrying his song with him, on his last long journey. Strange as it may seem the Indian its willing to give his songs... "²³

Densmore's Chippewa Music inaugurated a long series of scholarly studies based on her own field recordings and interviews and modeled after Alice Fletcher's A Study of Omaha Music. In her book, Fletcher had argued that music showed the influence of social life and thus she presented the Omaha songs according to their function in tribal life. With a similar anthropological orientation, Densmore described Indian music in the context of specific ceremonies and social customs, the particular circumstances in which she witnessed a ceremony, and information provided by her informants, illustrating her material with photographs of persons and artifacts.

She also provided analyses and transcriptions of 180 songs which she had collected in 1907, 1908, and 1909 at the White Earth, Leech Lake, and Red Lake reservations. These included songs of the Mide (Grand Medicine) and social songs - dream, war, love, and moccasin-game songs. Along with descriptive analyses of individual songs, she gave detailed information about their significance and tribal context. Tables summarized her analyses of the music in terms of six aspects of melody (tonality, tone material, beginnings of songs, endings of songs, first progressions, and accidentals), two aspects of rhythm (accent and metric unit), and structure-melodic or harmonic. These categories emphasized melody rather than rhythm, reflecting Densmore's fundamental orientation toward nineteenth-century Western music, in which the qualities of melody are substantially more developed than those of rhythm.

Although Densmore was later to vary the specific parameters of her tabulated analyses, she continued to use the basic methods and organization she had developed for Chippewa Music: a system of tables summarizing the occurrences of various musical characteristics of the songs, descriptive analyses of individual songs in relation to their function in the life of the tribe, and transcriptions of the songs in conventional notation.

In a second volume, Chippewa Music - II published in 1913, she began to express the results of her analysis in percentages as well as in tabulated figures.²⁴ In addition she expanded the number of parameters of analysis from nine to twenty-two, fourteen of them concerned with aspects of melody and eight with aspects of rhythm. She continued to use these twenty-two parameters until 1923, when she reduced them to eighteen in her publication on Mandan and Hidatsa Music. Thereafter she gradually simplified the tabulated analyses by decreasing the number of parameters until in her last book on Music of Acoma, Isleta, Cochiti and Zuñi Pueblos (1957) she included only eleven. Parameters of analysis were eliminated, she explained, when the "results of analysis were practically uniform" or when "tables did not seem of sufficient importance to be continued." In several books she eliminated tabulated analyses entirely, although these studies included descriptive analyses which were in fact based on unpublished tabulations.²⁵

In her monumental study, Teton Sioux Music, published in 1918, Densmore made an important addition to her analysis. Here she included cumulative tables which compared the characteristics of the music under sturdy - in this case, Sioux - with the cumulative characteristics of all the music she had studied to date. As a corollary, she also began to include a descriptive comparative analysis interpreting the information provided by the cumulative comparative tables, a practice she continued in her later books. Regrettably, however, she did not take the logical next step of transferring this information to maps which would have indicated the geographic distribution of specific characteristics in North American Indian music.

Beginning with Chippewa Music, and in all her subsequent publications, Densmore transcribed Indian songs in conventional Western notation as unaccompanied melodies. In this important respect she departed from Fletcher's model and from her own earlier work, in which she had embellished the renderings of Indian melodies with realizations of what she heard as implicit harmonies. These early transcriptions were in fact more arrangements than transcriptions. But in Chippewa Music Densmore notated only the melody of the songs.²⁶

Throughout her career Densmore was aware of the problems involved in using traditional Western notation to transcribe Indian music. In listening to Indian singers she had noted that they often sang certain "irregular" intervals, deviations from the tones of the tempered scale, some of which were smaller than a quarter tone. In Chippewa Music she introduced special signs to indicate these slight variations of pitch and note duration which defied conventional notation. As she explained in the preface: "It is acknowledged that ordinary musical notation does not, in all instances, represent accurately the tones sung. . . . If a new and complete notation were used in recording fractional tones it should be used in connection with delicately adjusted instruments which would determine those fractional tones with mathematical accuracy. The present study is not an analysis of fractional tones, but of melodic trend and general musical character; therefore the ordinary musical notation is used, with the addition of a few signs in special cases." Here Densmore assumed that her background in Western music would enable her to distinguish "general musical character" and did not consider the possibility that the fractional tones might themselves be the essential characteristic. Moreover, she overlooked the fact that while a given system of notation may successfully capture the character of music produced by one culture, such a system may be relatively ineffective when applied to that of another culture.²⁷

Despite her conclusion that the fractional tones were not an essential feature of Indian music, Densmore remained curious about singers' deviations from the tempered scale, and in 1915 she went to Carl E. Seashore of the psychological laboratory at Iowa State University for help in devising a way of testing the accuracy of pitch discrimination among Indians. With Seashore's guidance, Densmore developed a test using a set of tuning forks which produced tones of only slightly different frequencies. She gave the test to a number of Mandan, Hidatsa, Chippewa, and Sioux, hoping to learn whether they intentionally produced the minute intervals when they performed their music. Seashore's analysis of the test data verified her suspicion that Indians were capable of singing the tiny intervals deliberately. This information did not, however, persuade her to develop a more sensitive notation. Rather, throughout her career she continued to defend her transcriptions in conventional notation as generalized representations of the Indian songs."'²⁸

Early in her career Densmore experimented with the potential of graph notation for transcription. She never intended, however, that her graphic representations of Indian melodies should replace the use of conventional notation but rather that the two types of transcription be used concurrently. She first developed a system for making graph transcriptions in conjunction with her study of Teton Sioux music. This work included graphic representations introduced, as she explained, "for the purpose of making the trend of Sioux melodies more apparent to the eye than in musical transcription.

By plotting only the tones which her analysis had identified as accented, Densmore hoped to reveal the general course or trend of the melodies. Her examination of the graphic transcriptions of 240 Sioux songs revealed five primary melodic types. She used this information to explore a possible correlation between melodic type and songs having similar uses or titles, a correlation she had already tentatively suggested in Chippewa Music - II. She found that two of the melodic types occurred throughout the series of songs studied but that comparison of the plots to the titles of the three remaining types revealed that songs having similar uses or titles resemble each other. Densmore's use of graphs is one of the earliest experiments with an alternative approach to transcription in the history of ethnomusicology.²⁹

Although Densmore was primarily interested in studying the melodic trends and general character of Indian music she showed considerable interest in various electronic instruments which could provide more exact transcriptions of sound than were possible by ear. These instruments were of particular interest to her as a means of testing the validity of her own transcriptions. In 1918 she contacted Dayton C. Miller, chairman of the department of physics at Case School of Applied Science in Cleveland. Miller had recently invented an instrument known as the phonodeik which was capable of making a film record of sound. The phonodeik operated by means of a conical horn which focused sound on a diaphragm; the resulting movement of the diaphragm was then photographed and projected on a screen. Densmore hoped that the phonodeik could be helpful in verifying her observation of a strange incongruity between the rhythm of the singer and that of the instrumental accompaniment in much Indian music. In her study of Chippewa music she had first observed that the musicians seldom struck the drum and sang a tone simultaneously. She wished, however, to confirm this incongruity of song and accompaniment which was foreign to Western music.³⁰

Her analysis of ninety-one Chippewa songs with drum accompaniment had revealed sixty-three songs in which the voice and drum had independent metric units. Although she had attempted to transcribe this phenomenon by ear, she hoped that the phonodeik would provide a means of determining the exact relation of the drum beats to the voice. Miller made a film record of part of a cylinder record in which the voice and drum seemed to have independent rhythms using about twenty-three seconds of the recording to produce about thirty-eight feet of film. His reading of the film record verified Densmore's discovery that there were indeed two independent rhythms, but he felt on balance that the results of the experiment hardly justified the great amount of work involved.³¹

In 1934 Densmore again visited Carl Seashore's laboratory to discuss the latest developments in phonophotography. As a demonstration of his apparatus, Seashore produced a tone-photograph of one of Densmore's cylinder records. A graphic analysis was then made from the tone-phonograph and Densmore was reassured to learn that the graph gave essentially the same results as her transcription by ear.³² Although Seashore's equipment could produce more detailed and accurate transcriptions than those made by ear, Densmore persistently clung to her position that the object of transcription was to show the musical essentials, not minute details which might obscure the larger picture. Thus she continued to make transcriptions in conventional notation by ear.

Densmore intended her transcriptions to provide a generalized picture of Indian music, using intervals, or steps of the scale, borrowed from the vocabulary of Western music. She gradually became aware, however, of the limitations of a theory of Indian music which explained musical phenomena wholly in terms of Western music. Early in her career she began to raise questions about Fletcher's and Fillmore's theories based on the Indians' unconscious sense of harmony.

Fillmore had looked to acoustical principles to support his theory, noting that the "essential thing in all music is the relation of tones to a tonic or key-note; and the tones most nearly related acoustically to any given key-note are the tones of its triad."³³ Here Fillmore was referring to the acoustical fact that a vibrating or sounding body produces not only the fundamental tone but also a series of overtones which are less distinctly audible. The first four overtones above the principal C, for example, are the tones of the triad C-E-G. While Fillmore was correct in noting that the triad is generated acoustically, he was incorrect in assuming that this phenomenon provided a comprehensive, rational basis for the relationships among tones as they occur in all music. Acoustics do not in fact even provide a satisfactory explanation for the relationship of tones and chords in eighteenth- and nineteenth-century Western music. Fillmore, however, went one step further and argued that the tonal relationships exemplified by Western music with its key structure and harmony were universal and thus he justified his practice of harmonizing the Indian songs with tonal chordal accompaniments. The resulting arrangements sounded strangely like gospel hymns.

Densmore first expressed her uneasiness about interpretations of Indian music based on European musical theory in a short article published in 1906. While she praised Fillmore's work enthusiastically, she also observed that harmonizations made by others less skilled than he were often unconvincing. She hoped that someday an Indian musician would "take the melodies which we are collecting, and . . . translate for us their hidden harmonic thought, expressing it in terms of the white man's culture. He will do by intuition that which our most conscientious work cannot accomplish."³⁴

A year later, when she published the two songs sung for her by the Prairie Island women, she mentioned the enthusiasm with which the women had greeted her harmonized versions of their music. Fillmore, observing similar reactions, had interpreted them as proof of his harmonic theory. He noted that whenever the songs of the Omaha "were played for them on a piano or organ, they were not satisfied without the addition of chords to the melodies. . . . It seems proper to draw the conclusion that the sense of harmony is an innate endowment of human nature, that it is the same for the trained musician and for the untrained primitive man, the difference being purely one of development.''³⁵

Densmore, however, drew a rather different conclusion from the favorable response to her harmonizations. She suspected that it provided more information about the Sioux singers' desire to please her than about their unconscious sense of harmony.

When she published her study of Chippewa music in 1910, she made several important modifications in Fillmore's theory. He had argued that melody in native music was shaped by the Indians' "natural perception of the harmonic relations of tones." Densmore took a less dogmatic stand, observing: "The sequence of tones in Chippewa songs show that certain songs are harmonic and others melodic in structure.... Songs are classified as harmonic if their accented tones follow the intervals of diatonic chords, and as melodic if their contiguous accented tones have no apparent chord relationship." By melodic Densmore referred to melody completely independent of any implied harmony. Of the 180 songs analyzed in Chippewa Music, she identified 139 as melodic in their structure.³⁶

In her second study of Chippewa music, she retreated more overtly from Fillmore's position, observing that the "sequence of tones in many of the songs is such that the 'harmonic relation' is extremely complicated, if indeed (in some instances), it can be said to exist." By 1943, when she published a book on Choctaw music, she eliminated the category called "structure - melodic or harmonic" as a basis for analysis. Although she gave no reason for this change, it is apparent that she simply no longer thought of Indian music in terms of harmony.³⁷

In an article that appeared in 1917, Densmore's deviations from the ideas of her mentor are most clear. Although these deviations are interesting as theoretical modifications of Fillmore's concept of the keynote or tonal center, they are more important as they reveal Densmore's understanding of Indian music in terms increasingly independent of Western music. Her discussion was still heavily indebted to Fillmore, but she carefully avoided any specific reference to the keynote and tonality in the sense in which he had understood them. Instead she vaguely defined the keynote as the tone which was apparently most important and which functioned in some way as the focus of the melody. She avoided any reference to a tonal relationship between the keynote and the other tones of the Indian songs, stating simply that the keynote could be identified by the "test of the ear" thus leaving the matter to intuition. She suggested a variety of standards for determining the keynote, such as frequency of appearance, rhythmic stress, placement at the beginning or final tone, and function as the lowest tone. Instead of using criteria based on the theory of Western tonal music, she expanded her perspective and considered ways of organizing musical sounds other than those with which she was familiar.³⁸ Throughout her career, Densmore continued to generalize about Indian music. Unlike Fillmore, however, she based her generalizations on actual counts of various musical phenomena. On the basis of these data, she grew increasingly doubtful of the existence of a universal tonal system leased on acoustics.

As she examined the music of various Indian tribes, Densmore observed several other patterns of melodic organization in addition to the triad structure seen in much Chippewa music. Indeed, in Chippewa Music she had identified a few songs in which the interval of the fourth was particularly prominent. In songs collected during the 1920s among the Nootka and Quileute of the Pacific Northwest, she noticed the frequent occurrence of the interval of the fourth and of phrases comprised of the tetrachord pattern (as in C-D-E-F or E-F-G-A) as prominent structural elements. (She later observed this same characteristic in Menominee, Winnebago, Cheyenne, and Arapaho songs.) When she published Nootha and Quileute Music in 1939, she classified the Indian songs based on the tetrachord as a group analogous to those based on the triad. She wrote: "Two types of melodic structure occur in songs of the American Indians, the tetrachord (complete and incomplete), and the triad formation. Combinations of the two are also found, as well as freely melodic songs that have no framework and songs that are said to have been composed in imitation of the sounds produced by a wooden flute.''³⁹ Her analyses of Nootka and Quileute music finally put to rest Fillmore's theory of the harmonic structure of Indian music.

In studies of the Tule Indians of Panama and later among the Yuman, Seminole, and certain Pueblo tribes, Densmore found an elaborate patterned type of melodic design which she referred to as "period structure." By this term she meant a definite form consisting of several periods or discrete musical phrases which were repeated in regular order. In Yuman and Yaqui Music, she explained the complex period structure of many of the Yuman songs: "In describing the form of Yuman songs the interpreter said, 'There is always a chorus near the end of a song that goes up higher.' He said it is the custom that 'the song shall be sung four times and the chorus twice,' also that 'if the chorus is sung a third time the ending is on a high note.' The term 'chorus' is derived from a knowledge of the white man's songs and indicates a pleasing part of the song but not a change in the number of singers. It usually contains about eight measures and is not repeated. The other portions of the song are sung from 2 or 3 to 11 times and are accurately repeated."⁴⁰ The periodic structure was thus defined in terms of internal repetitions of phrases without reference to any particular intervals such as Densmore had noted, for example, in the Chippewa melodies based on the triad and in the Nootka and Quileute melodies based on the tetrachord.

When Music of the Indians of British Columbia appeared in 1943, Densmore reported that in the more than 2,500 Indian melodies analyzed up to that time, she had found four basic patterns of melodic structure: "(1) a formation on the simplest overtones of a fundamental, generally called a triad formation, (2) a formation based on the interval of a fourth, (3) a typical folk-song structure . . . and (4) a period formation."⁴¹ Giving only passing attention to the patterns already described in earlier studies, she concentrated her attention on the folk-song structure, which she introduced here for the first time.

She used the term "folk-song structure" to describe a large group of Indian songs in which the whole-tone interval was particularly prominent. It is significant, however, that she did not refer simply to the "whole-tone structure" in Indian music but instead borrowed the phrase "folk-song structure" from an apparently alien music. She was not unaware of the implications of this terminology as she intended to suggest a relationship between Indian music and English folk song. As she explained, "The Indians of British Columbia have been in contact with people from Scotland for many generations.... it appears that the songs recorded in British Columbia bear interesting resemblances to the songs recorded at Neah Bay and on the Mexican border, as well as resemblances to Scotch songs and to the accepted basis of English folk song. The foregoing observations are offered as an aid to further study, not as presenting any hypothesis or theory.... These observations suggest influences from the east, across Canada, and also from the south, along the coast of the United States."⁴²

In an effort to provide background for her argument, Densmore quoted a passage from the English folk-song scholar, A. H. Fox Strangways: "'the folk singer has not only no harmony . . . but no feeling for it . . . one note of the tune has an affinity for (or an antipathy to) some others; connections are thus formed, and structure is made possible. Shortly, the folk-songster is satisfied with affinity between notes.... The nucleus of his scale is three notes a tone apart (F. G. A, for instance) which have this affinity; above and below this are two outliers, C, D, which also have it; beyond those five he takes notes tentatively.'"⁴³

This extract notwithstanding, Densmore made only limited use of Fox-Strangways' conception in her subsequent discussion of folk-song structure and completely ignored his reference to tones within the context of a pentatonic scale (C, D, F, G, A). Indeed, she applied the concept of folk-song structure as she understood it to Indian melodies with a total compass of only three tones. Her application of Fox-Strangways' definition constituted a redefinition of the term. Although her analysis would have been stronger and more in keeping with her usual approach if she had simply noted the frequency of whole-tone progressions in some Indian music and then postulated the influence of English and Scotch folk song, she apparently felt that Fox-Strangways' explicit denial of a universal feeling for harmony would reinforce her own ideas as they had developed over the years.

Through the process of recording, transcribing and analyzing thousands of Indian songs, Densmore gradually evolved a theory of Indian music based on the objective data. While she continued to use notation and terminology originally developed for the interpretation of Western music, she developed an increasingly sensitive appreciation of the limitations they imposed. Perhaps most important, she moved from her initial conception of Indian song as melody with implied harmony to an understanding of Indian song based on only the actual linear melody heard. This freed her to discover the principles which actually functioned in Indian music and indeed to appreciate it on its own terms.

Densmore's efforts yielded major published studies on the music of the Chippewa, Teton Sioux, Northern Ute, Mandan and Hidatsa, Tule Indians of Panama, Papago, Pawnee, Menominee, Yuman and Yaqui, Cheyenne and Arapaho, Santo Domingo Pueblo, Nootka and Quileute, Choctaw, the Indians of British Columbia, Seminole and Acoma, Isleta, Cochiti, and Zuni pueblos. With the exception of her work with the Cheyenne and Arapaho and the Indians of Santo Domingo Pueblo, which were conducted for the Southwest Museum of Los Angeles, all her research was done under the auspices of the Bureau of American Ethnology.

She was employed by the bureau from 1907 until 1933, when over-all government economy measures necessitated severe cut-backs and it was no longer possible for the bureau to continue its assistance. In 1939, the bureau resumed its support and Densmore, although seventy-two years old, took up her field activities with the same intense enthusiasm which characterized all her earlier work.

In July, 1940, Densmore began work for the James J. Hill Reference Library in St. Paul on a brochure to accompany the Hill collection of Seth Eastman's watercolors of American Indians. She arranged the paintings so as to reconstruct as nearly as possible the sequence of Indian history during the period of Eastman's work and wrote an introduction to the collection. More than half the forty-six paintings were scenes near Fort Snelling, where Eastman had served as an army officer from 1841 to 1848. Densmore's introduction included frequent quotations from Mary H. Eastman's books, especially Dahcotah; or, Life and Legends of the Sioux Around Fort Snelling. As Densmore explained, "It is a privilege to present her work beside that of her husband after so many years."⁴⁴

In the 1940s a generous private gift to the federal government made it possible to preserve Densmore's collection of more than 2,400 wax cylinders in more permanent form. In November, 1941, she was appointed to work with the collection and to prepare a handbook, a task she completed in 1943. In 1948 the enormous task of transferring the recordings from cylinders to discs was begun. With characteristic tenacity Densmore, now eighty-one years old, supervised this project from her home in Red Wing. Her correspondence for this period is proof of her full participation. In a letter written on July 20, 1951, for example, she apologized for the delay caused by what she described as a "slight bronchial pneumonia." She continued, "I must maintain my reputation for complete work." In addition, she reviewed the entire collection to select representative songs for transfer to long-playing records which were made available to the public. Between 1951 and 1953 the Library of Congress released seven of these records with the accompanying booklets she had prepared.⁴⁵

In 1954, Densmore, now eighty-seven, gave a series of seminars on Indian music at the University of Florida in Gainesville. She also took advantage of the opportunity to carry on further field studies among the Florida Seminoles. On March 24, she wrote back to her friend Mary Biederman in Red Wing: "I am getting to the end of my first month at the University of Florida and the month has really been more successful than I had dared to expect, in the interest that has been aroused in Indian music. I have given six short lectures . . . I will probably go to the town of Dania at the end of the week to begin more observations on the Seminoles." In April, after a month of field work, she wrote again with indefatigable enthusiasm, "Here I am - back in civilization . . . The trip yielded valuable results - some entirely new data."⁴⁶

Three years later, on June 5, 1957, Frances Densmore died at Red Wing not long after her ninetieth birthday. During her lifetime she was a tireless recorder of Native American music. Her extensive and original work contributed significantly to the study of non-Western music. It was important not only as one of the earliest efforts to study music in its cultural context but also as one of the first efforts to recognize and understand Indian music as profoundly different from Western music. Unfortunately, however, within the discipline of ethnomusicology some contemporary scholars, noting her use of the vocabulary associated with the theory of Western music, have tended to overlook the significance of her work. They have assumed that because she retained some traditional terminology she was not aware of the fundamental differences between Western and Indian music. Nothing could be further from the truth.

The current scholarly interest in American Indian cultures should bring new attention to her work. Restudy of her writings and recordings would not only bring long-overdue recognition to Densmore's pioneering scholarly accomplishments but also stimulate further study of her beloved Indian music. Such an enterprise would be in the spirit of Frances Densmore.

NOTES

1. Stuhler and Kreuter, ed. Women of Minnesota (St. Paul: Minnesota Historical Society Press, 1977)

2. For the history of the Densmore family, see C. A. Rasmussen, A History of the City of Red Wing, Minnesota, 284 (n.p., 1933), and Franklyn Curtiss - Wedge, ed., History of Goodhue County, Minnesota, 757 (Chicago, 1909).

3. "Chronology of the study and presentation of Indian music by Frances Densmore from 1893 to 1944," 1, Archive of Folk Song, Densmore Papers, Library of Congress. Hereafter cited as "Chronology." Throughout her life Frances Densmore kept meticulous professional and personal records. In addition to her publications, these include diaries, complete in 46 volumes for the years 1907-33, scrapbooks, correspondence, several retrospective accounts of her experiences, and various unpublished manuscripts, including the texts of several early lectures.

4. For information here and in the paragraph below, see Charles Hofmann, comp., ed., Frances Densmore and American Indian Music, 1 (Museum of the American Indian Heye Foundation, Contributions, vol. 23 - New York, 1968).

5. Densmore, ''[Auto]biography,'' n.d. but written not earlier than 1952. Densmore Personal Papers, National Anthropological Archives, Smithsonian Institution. Fletcher's study, originally published as Archaeological and Ethnological Papers of the Peabody Museum, vol. 1, no. 5 (Cambridge, Mass., 1893) was reprinted in 1967.

6. In 1952 Densmore noted in her scrapbook: ''when a child, I was taken to an exhibit, or demonstration, of a recording phonograph that had recently been invented. This was probably in 1878 or 1879." Archive of Folk Song, Densmore Papers, Scrapbook 8, 1949-54, Library of Congress. Densmore's library contained a copy of Emile Berliner, The Gramophone by its Inventor: Paper Read before The Franklin Institute, May 16,1888 (Washington, D.C.. 1894). Berliner had invented a recording instrument in 1877, the same year Edison invented the phonograph. Densmore noted in her copy of Berliner's book that he had given it to her about 1913, six years after she had made her own first sound recordings.

7. "Chronology," 2.

8. Hofmann, ed., Densmore, 2.

9. "Chronology," 3.

10. Schubert Club (St. Paul), Programs, 1899-1924, December 4, 1895, in Minnesota Historical Society (MHS) Library.

11. See a note by Densmore in 1945 attached to "Manuscript of a Lecture on Indian Music," dated before 1899, in Archive of Folk Song, Densmore Papers, Library of Congress. The manuscript is essentially the same as a lecture delivered in Chicago in 1899; the complete text is in Hofmann, ed., Densmore, 2-12. No dated manuscripts of lectures given before 1895 have survived, but a "Partial list of lectures, addresses and papers on Indian music and customs," may be found in Densmore Personal Papers, National Anthropological Archives, 4250 Box 2, Item 7, Smithsonian Institution. See also "Chronology," 3, 4. Topics are listed in printed programs for her lectures.

12. Farwell, who had a strong interest in Indian Music, in 1901 founded the Wawan Press at Newton Center, Mass., to promote the publication and performance of works by American composers and compositions based on American Indian melodies.

13. Minneapolis Journal, May 23, 1903, p. 11. See also Hofmann, ed., Densmore, 11.

14. Hofmann, ed., Densmore, 25; Densmore, "Three Indian Types," in Indian School Journal, November, 1906, p. 35.

15. Densmore, in Indian School Journal, p. 36.

16. See Densmore, "Prelude to the Study of Indian Music," in Minnesota Archaeologist, 11:28 (April, 1945), reprinted in Hofmann, ed., Densmore, 21-24. The two articles by Densmore are "The Song of Minagunz, the Ojibwa," in Indian School Journal, pp. 23-25; "An Ojibwa Prayer Ceremony," in American Anthropologist, new series, 9:443 (April-June, 1907).

17. Densmore, "Two Dakota Songs," in Indian School Journal, April, 1907, pp. 32-34; "Chronology," 1.

18. Densmore, in Minnesota Archaeologist, 11:30.

19. A number of the recordings made at the White Earth Reservation in 1907 were deposited with the Bureau of American Ethnology. They are now in the Archive of Folk Song, Densmore Papers, Library of Congress. For a list of these recordings, see 'Catalogue of Phonograph Records of Indian Music in the Archives of the Bureau of American Ethnology," 3-6, in National Anthropological Archives, Smithsonian Institution.

20. Densmore, in Minnesota Archaeologist, 11:31.

21. Densmore, "The Study of Indian Music," in Smithsonian Institution, Annual Report, 1941, 529 (Washington, D.C., 1942), reprinted in Hofmann, ed., Densmore, 101-114. The series of events associated with the death of Flat Mouth are recounted in Densmore's first major publication, Chippewa Music, 51-55 (Bureau of American Ethnology, Bulletin 45 - Washington, D.C., 1910). This volume and other Densmore books cited in this chapter were reprinted by Da Capo Press in 1972.

22. Bureau of American Ethnology, Twenty-ninth Annual Report, 1907-08, 19 (Washington, D.C., 1916).

23. Bureau of American Ethnology, Thirtieth Annual Report, 1908-09, 21 (Washington, D.C., 1915); Densmore, "Lecture on the Music of the American Indians before the Anthropological Society of Washington, April 26, 1909," p. 1, Densmore Personal Papers, National Anthropological Archives, Smithsonian Institution.

24. Densmore, Chippewa Music-II, 35 (Bureau of American Ethnology, Bulletin 53 - Washington, D.C., 1913).

25. These studies appeared as Bureau of American Ethnology, Bulletin 80 and 165 (Washington, D.C., 1923, 1957). On the elimination of tabulated analyses, see, for example, Densmore, Cheyenne and Arapaho Music (Southwest Museum, Papers, no. 10 - Los Angeles, May, 1936). For the study mentioned in the paragraph below, see Densmore, Teton Sioux Music (Bureau of American Ethnology, Bulletin 61 - Washington, D.C., 1918).

26. See, for example, Densmore, in Indian School Journal, April, 1907, pp. 32-34.

27. Densmore, Chippewa Music, 3. For the special signs, see p. xix.

28. On pitch discrimination, see Densmore, "American Indian Music,'' in Journal of the Washington Academy of Sciences, 18:398 (August 19, 1928); Densmore, "The Use of the Term 'tetrachord' in Musicology," in Journal of Musicology, 1:16 (March, 1940); and "Regional Peculiarities of Indian Songs,'' 1, typescript of a paper read by Densmore before the American Association for the Advancement of Science (AAAS), Minneapolis, June 28, 1935, in Archive of Folk Song, Densmore Papers, Library of Congress.

29. Densmore, Teton Sioux Music, 51; Densmore, Chippewa Music-II, 50-58. For another example of Densmore's graph transcriptions, see Mandan and Hidatsa Music, 34. Later scholars have used graph notation to provide more detailed transcriptions than are possible with conventional notation.

30. Densmore, Chippewa Music, 5.

31. "Department of Physics, Case School of Applied Science,'' a report by Dayton C. Miller, in Archive of Folk Song, Densmore Papers, Library of Congress; Densmore, in Smithsonian Institution, Annual Report, 1941, 543.

32. Densmore, "Methods of Recording Indian Songs," 9, typescript of a paper read before the AAAS, St. Louis, January, 1936; Densmore, in Smithsonian Institution, Annual Report, 1941, 543.

33. John C. Fillmore, "The Harmonic Structure of Indian Music," in American Anthropologist, new series, l:308 (April, 1899).

34. Densmore, "A Plea for the Indian Harmonization of All Indian Songs," in Indian School Journal, February, 1906, p. 14. The works of contemporary Indian composers such as Louis Ballard, who uses traditional Indian melodies in writing for ensembles of Western and traditional Indian instruments, are in a sense fulfillments of Densmore's vision. Even more prophetic was her view that the Native

American by virtue of his ethnic inheritance is himself the best interpreter of Indian culture.

35. Densmore, in Indian School Journal, April, 1907, p. 34; Fillmore, in Fletcher, Study of Omaha Music, 61.

36. Fillmore, in American Anthropologist, 1:315; Densmore, Chippewa Music, 8.

37. Densmore, Chippewa Music-II, 3; Densmore, Choctaw Music (Bureau of American Ethnology, Bulletin 136, Anthropological Paper No. 28 - Washington, D.C., 1943).

38. Densmore, "Recent Developments in the Study of Indian Music," in Proceedings of the Nineteenth International Congress of Americanists, 298 (Washington, D.C., 1917), reprinted in Scientific American (supplement), 85:253 (April, 1918) and Etude, 38:267 (October, 1920).

39. Densmore, Nootka and Quileute Music, 44 (Bureau of American Ethnology, Bulletin 124 - Washington, D.C., 1939).

40. Densmore, Yuman and Yaqui Music, 18 (Bureau of American Ethnology, Bulletin 110 - Washington, D.C., 1932).

41. Densmore, Music of the Indians of British Columbia, 95 (Bureau of American Ethnology, Bulletin 136, Anthropological Paper No. 27 - Washington, D.C., 1943).

42. Densmore, Music of the Indians of British Columbia, 98.

43. Densmore, Music of the Indians of British Columbia, 97.

44. Mary Eastman's book was published in New York in 1849. Densmore, The Collection of Water-Color Drawings of the North American Indian by Seth Eastman in the James Jerome Hill Reference Library, Saint Paul, 11 (St. Paul, 1954).

45. "Handbook of the Smithsonian-Densmore Collection of American Indian Sound Recordings in the National Archives," unpublished manuscript dated 1943, listed in Hofmann, ed., Densmore, 127, now located in Archive of Folk Song, Densmore Papers, Library of Congress. In the same archive is Densmore's letter to Duncan Emrich. July 20, 1951.

46. Densmore to Biederman, March 24, April 26, 1954, in Densmore Papers, Goodhue County Historical Society, Red Wing.