Instrumental and Vocal Love Songs of the North American Indians

This thesis by Mary Frances Riemer provides an important link to understanding the traditional use of the flute in North American indigenous cultures. It was written in 1978 and surveyed some of the earliest recorded flute music.

I was able to obtain the original thesis from the Wesleyan University Library through inter-library loan. Her research is available in two open-access formats: its original format as a PDF file ([Riemer 1978] ) and this web page, which provides a re-formatted version of the content of the thesis.

If you have feedback on this material, please contact me at clint@goss.com.

Acknowledgment

I wish to thank my advisor, Professor David P. McAllester, for his encouragement and constructive criticism during the writing of this thesis.

Introduction

This study of instrumental and vocal love songs has evolved from and will attempt to continue the research begun in two articles, “Special Song Types in North American Indian Music” by George Herzog ([Herzog 1935a]) and “Musical Areas in Aboriginal North America” by Helen Roberts ([Roberts-HH 1936]), both written more than forty years ago. The Roberts paper, which in part delineated musical areas by means of the distribution of instruments, clearly pointed out the lack of any information concerning the music of melody-producing instruments in aboriginal North America.

While so much can be said of musical instruments, almost nothing is known of the music which might have been produced by those capable of more than one tone.
…
The flute and flageolet produced readily audible melodies; various writers who have heard this music have almost always remarked that it was plaintive and beautiful. Very few records of it have been made, however, and equally few examples notated. ([Roberts-HH 1936], page 27)

As far as my research indicates, no thorough discussion of the flageolet, which was used almost exclusively as a courting instrument, exists in the relevant literature, with the exception of Merriam's treatment of the flageolet among the Flathead Indians. Accordingly, the first chapter of this paper will consist of a study of the courting flute: its construction, distribution, the beliefs and behavior associated with its acquisition, and its social function will be discussed in detail.

The second and major section of this thesis will deal with the melodies played on the courting flute and with the closely related genre of love songs. Together they contribute a special class of North American Indian music. These songs and flute melodies have never been taken together as a group and studied for their stylistic unity. Very frequently found in the same geographical location, flute melodies and love songs had a fairly wide distribution north of Mexico, being most highly concentrated through the central U.S.A. in roughly the Plains–Plateau–Southwest areas. Since this music cuts across several musical areas, can it be shown to possess a homogeniety within itself and a consequent dissimilarity to the other music of its area? This was the fundamental question posed by Herzog in 1935. To attempt an answer, flute melodies and love songs from tribes located in four musical areas, the Western Great Lakes, Plains, Plateau, and Southwest will be analyzed, studies, and finally compared with the general stylistic traits of the music of each area.

Chapter 1 — The Courting Flute

1.1 — Terminology and Construction

Throughout the literature disccusing Native American wind instruments, the terms most frequently used are whistle, whistle flute, flageolet, open flute (vertical and transverse), and single and double reed pipe. Although the distinction between instruments using reeds as the vibrating mechanism and those without reeds is clear, the other terms are often confused. The instrument under discussion here is correctly called a flageolet and belongs to the group of instruments known as whistle flutes ([Apel 1970], page 925). By way of illustration, the European flageolet and recorder are also whistle flutes. These two instruments are distinguished from each other by the number of finger-holes, six and eight respectively, and by their positioning ([Bessaraboff 1941], pages 62–63). The functioning principle, however, is the same. In all whistle flutes the upper end of the instrument is stopped by a plug or fipple with only a narrow slit, called a flue, remaining. The breath is led through the flue toward the sharp edge of a small opening below the fipple. The Indian flageolet is a slight variation on this design in that the narrow air channel is formed outside the cylinder, thus requiring the addition of the characteristic wooden block, or “saddle.”

The materials and method of construction fo the flageolet were remarkably similar througout North America. Red cedar was the most commonly used wood, although other stright-grained woods such as box-elder, ash, sumac, elderberry, redwood, osage orange, and fir are also mentioned in the literature as being suitable materials. More recently flageolets have been made from metal gun barrels and nickel tubing. Features of these metal flutes, such as tone quality and pitch level, which differ markedly from that of a woode flute will be discussed in the following chapter, “Instrumental Love Songs.”

The instruments were generally about 1½ inches in diameter and 20 to 21 inches long; however, they could vary in length from 11 inches (Northern Ute / [Densmore 1922], page 23) to 24½ inches (Omaha / [Fletcher 1893], page 72). Densmore has stated that the length of the instrument wad determined by the stature of the player; for example, the distance from the inside of his elbow to the end of the middle finger ([Densmore 1926], page 95) or the length given by “spreads” of his hand ([Densmore 1929b] , page 167) were two rough measures used in calculating the dimensions of the flageolet.

To make a flageolet a straight section of wood was split lengthwise and the insides of each half were hollowed out to form a cylindrical bore. A block (A) — see figure 1 — was left inside the cylinder creating a solid unbroken partition between the upper and lower chambers. The chamber containing the mouth-end was proportionally shorter (1:3) than the body of the instrument. Small square holes (B and C) were cut into each chamber just above and below the partition. The surface around these holes was then made smooth and flat and a thin wooden or metal plate (D) laid over it. This plate had a rectangular hole cut into it which fit exactly over the two holes in the cylinder (Figure 2). Finally, a wooden block (E), flat on the underside and carved according to the maker's fancy or tradition on top, was tied or glued over this plate.

Air blown into the end of the shorter chamber flattens into a thin stream as it passes between the partition and the plate. At the entrance to the longer chamber the airstream impinges on the sharp edge or “lip” of the plate and sets the column of air in vibration. At this point the airstream divides. To allow surplus air to escape, the block is either positioned to leave the second hole partially uncovered or has a vertical groove carved into it.

Figure 1. Cross-section of external flue and wooden block of the Indian flageolet

Figure 2. Detail of flue (with block removed) to show position of thin wooden plate over the partition and two air-holes

The mouth-end was either blunt, tapered to an opening smaller than that of the tube, or shaped into a small tube which projected from the instrument to form a true mouth-piece. The flageolet had four to eight finger holes ([Baker 1882] , page 55) but six holes, in two groups of three, was the most common arrangement. These holes were burnt into the wood with a pointed hot iron. There is no indication that measurements were taken for the positioning of the holes. The guiding principle seems to have been the size of the player's hands, the holes being placed wherever the fingers comfortably rested ([Merriam-AP 1967], page 50; [Collaer 1968], page 100; [Densmore 1926], page 95). This rather arbitrary method of determining the position of finger-holes, as well as the variable length and diameter of the instrument, should theoretically make the tuning of every flute unique. In practice, however, the tone systems of most wooden instruments conform to a fairly standard pattern; the lowest tone of the flute lies most often between g' and b' with the other tones rising in intervals that are very close to Western diatonic steps of a major second and minor third. Semitones, when they occur, are always between the highest tone of the fingered scale and the overblow octave. For example:

Figure 3. Tone systm of Omaha flageolet melody (13)

(the tone system for every flageolet melody discussed in this paper is given at the bottom of each transcription, see pages 45–72 of the PDF version of this thesis).

In addition to the finger-holes that were stopped, the flageolet of the Flathead Indians had a seventh hole placed near the bottom of the instrument. Although it was never covered, the instrument was considered incomplete without it ([Merriam-AP 1967], page 50). The Omaha, Onondaga, and Chippwea flageolets are similar in this respect since they all feature four small holes arranged circularly near the bottom end. Baker's drawing ([Baker 1882] ) of a flageolet (tribe not identified) also shows these small holes arranged in a cluster. Finally, the flageolet was often bound in several places with coloured strings or leather thongs. These could be purely decorative but on occasion also functioned to hold the two halves together. The Chippewa, for example, sometimes used raw deer hide to bind the flute and, as the hide dried, it contracted to hold the two parts tightly together ([Densmore 1929b] , page 168).

Figure 4. Flageolets from the Sioux (top), Onondaga (middle), and Taos (bottom) Indians. From the private collection of Dr. David P. McAllester. Photograph by Susan McAllester.

Figure 5. Detail of Onondaga flageolet showing the elaborately carved block. Photograph by Susan McAllester.

1.2 — Distribution

The flageolet had a fairly wide distribution north of Mexico but was most highly concentrated through the central U.S.A., in roughly the Plains–Plateau–Southwest area. An early distributional map by Helen Roberts ([Roberts-HH 1936], page 17) shows that the flageolet existed in an area extending from the Western Great Lakes to the lower Mississippi and west to the Colorado River. Two other locations are also maked: California, where only the Mojave knew the flageolet ([Kroeber 1965], page 22), and the Southeast. The Roberts map corresponds closely with an earlier listing by Curt Sachs, given below ([Sachs 1929], page 214) of the Indian groups who possessed the flageolet. (The figures in brackets indicate the number of finger-holes of the various flageolets).

¹ Also known as Chippewa.

The inclusion of Altmexiko (pre-Columbian Mexico) in this list should be noted since it has been suggested ([Galpin 1903], page 135; [Sachs 1929], page 214; [Roberts-HH 1936], pages 20 & 25) that the flageolet originated in Mexico and subsequently spread northward.²

² The flutes generally were known to the Indians of the Southwest in the prehistcric period has been verified by several excavations. Both a painted pottery plate from the Hohokam civilization ca. 8OO A.D. found at Snaketown, Arizona ([Collaer 1968], pages 48–49) and a pictograph ca. 700–900 A.D. in the Hagoe Canyon in northeast Arizona ([Brown 1967], page 83) portray stylized flute-players blowing long, end-blown flutes. Another archeological find in the Prayer Rock Valley, northeast Arizona has proven the existence of open, end-blown flutes in the area ca. 620–670 A.D. ([Bakkegard 1961], pages 184–186).

A search to augment these early distributional listings has revealed the existence of the flageolet among these additional tribes (see distributional map, Figure 6):

1. Iroquois — Eastern Woodlands ([Baker 1882] , pages 55–56; [Morgan 1904] , book 3, page 38; [Beauchamp 1905], page 177; [Miller-DC 1921], pages 511–512; [Speck 1945], page 77; [Kurath 1951], page 116; [Kurath 1956], page 4)
2. Menominee — Western Great Lakes ([Densmore 1932a], page 11)
3. Winnebago — Western Great Lakes ([Radin 1923], page 123; [Hofmann 1964])
4. Mesquaki — Plains/Western Great Lakes ([Busby 1886], page 82, [Kurath 1956], page 4)
5. Fox — Plains/Western Great Lakes ([Jones 1939], page 57, [Forsyth 1912], page 222)
6. Sauk — Plains/Western Great Lakes ([Forsyth 1912], page 222)
7. Mandan — Plains ([Densmore 1923] , pages 9–10, [Collaer 1968], page 100)
8. Hidatsa — Plains ([Densmore 1923] , pages 9–10)
9. Cree — Plains ([Wissler 1910], page 86)
10. Dakota Sioux — Plains ([Galpin 1903], page 134, [Collaer 1968], pages 100–101)
11. Teton Sioux — Plains ([Catlin 1876], page 243)
12. Pawnee ([Densmore 1929a], page 97)
13. Crow — Plains ([Lowie 1935], page 52, [Curtis-E 1909a] , volume 4, page 30)
14. Arapaho — Plains ([Hilger 1952], page 199)
15. Kutenai — Plateau ([Ray 1942] , page 187)
16. Thompson — Plateau ([Merriam-AP 1967], page 52)
17. Okanagon — Plateau ([Merriam-AP 1967], page 52)
18. Sanpoil — Plateau ([Ray 1942] , page 187)
19. Kalispel — Plateau ([Ray 1942] , page 187)
20. Flathead — Plateau ([Merriam-AP 1951], page 369)
21. Cour d'Alene — Plateau ([Teit 1930], page 165)
22. Nez Perce — Plateau ([Spinden 1908], page 231, [Curtis-E 1911c] , volume 8, page 157)
23. Tenino — Plateau ([Ray 1942] , page 187)
24. Umatilla — Plateau ([Ray 1942] , page 187)
25. Northern Shoshoni — Basin ([Steward 1941a] , page 251, [Steward 1943], page 336)
26. Bannock — Basin ([Steward 1943], page 336)
27. Northern Paiute — Southwest ([Stewart 1941], page 404)
28. Southern Paiute — Southwest ([Stewart 1942], page 295)
29. Taos — Southwest ([McAllester 1961])
30. Lipan — Southwest ([Gifford 1940], page 59)
31. Shawnee — Southeast/Plains ([Trowbridge 1939], page 39)
32. Alabama — Southeast ([Densmore 1937], page 278)
33. Creek — Southeast ([Swanton 1924], page 521)

Figure 6. Map showing Distribution of the Courting Flute

1.3 — Function

In the summer of the olden time there might often be heard at eventide the call of flutes. It was the youths upon the hill-side piping love-songs. Every one may know a love-song when he hears it, for the flute-tones are long and languorous, and are filled with a soft tremor. When a maiden heard the flute-music of her lover without, she always found it necessary to leave the tipi to draw water or to visit some neighbor.
 
In this song the maid asks leave of her mother to go to see her uncle, but the music tells that it is really her lover to whom she is going. The old people are not often deceived when the flute-music sounded. ([Densmore 1923] , page 261)

Although occasional reference is made to the use of the flageolet in ceremonies ([Radin 1923], page 123) and as a warning or war signal ([Densmore 1929b] , page 168, [Trowbridge 1939], page 39), its most common use seems to have been as a courting inbstrument. Consequently, its repertoire was almost exclusively restricted to the playing of love songs. Although both men and women could sing love songs, only men played the flageolet. It appears to have always been played as a solo instrument and never accompanied by the voice ([Driver 1969], page 195).

Because of its function, the flageolet was considered the personal property of the player and was rarely borrowed or loaned. Flageolet songs appear to have been individual creations and were, therefore, also personally owned. Something of the supernatural was often attached to songs played on the flageolet. In Menominee tradition it is said that a man who played the flageolet carried “love medicine” with him, an indication that songs possessed magical qualities ([Densmore 1932a], page 208). Among the Flathead Indians, Merriam discusses the relationship of a guardian spirit in association with flageolet melodies. The flageolet was usually made on the instruction of the spirit, who also gave the man the songs. When a song was played, it was not even necessary that the woman to whom the song was directed be able to hear it. “A woman always knows when someone is singing a love song to her. It is like a dream. The spirit that gave the man the song would be the one who caused her to know.” ([Merriam-AP 1967], page 60). As with the Flathead, intervention by the supernatural also occurred among the Crow Indians, where flageolets were often given to suitors in a vision. A supernatural being, usually in the form of an elk, would appear playing a flute and his music would cause all the female animals to run towards him. After such a vision the would-be suitor returned to camp and, taking his example from the elk, would make an exact copy of the kind of flute that had been revealed to him. With this flute he would be able to irresistably charm the woman he desired ([Lowie 1935], page 52)

A more elaborate version of the origin of the courting flute is still current in Plains Indian mythology. I Richard Erdoes' collection of Indian legends ([Erdoes 1976]), Henry Crow Dog, a Sioux from Rosebud, South Dakota, describes the creation of the first “Siyotanka.”

When the young hunter awoke, the sun was already high, and on a branch of the tree agains which he was leaning was a red-headed woodpecker. The bird flew away to another tree and then to another, but never very far, looking all the time over its shoulder at the young man as if to say “Come on!” Then, once more the hunter heard that wonderful song, and his heart yearned to find the singer. The bird flew toward the sound, leading the young man, its flaming red top flitting through the leaves, making it easy to follow. At last the bird alighted on a cedar tree and began tapping and hammering on a dead branch, making a noise like the fast beating of a small drum. Suddenly there was a gust of wind, and again the hunter heard that beautiful sound right close by and above him.
 
Then he discovered that the song came from the dead branch which the woodpecker was belaboring with its beak. He found, moreover, that it was the wind which made the sound as it whistled through the holes the bird had drilled into the branch. “Kola, friend,” said the hunter, “let me take this branch home. You can make yourself another one.” He took the branch, a hollow piece of wood about the length of his forearm, and full of holes. The young man walked back to his village. He had no meat to bring to his tribe, but he was happy all the same.
 
Back in his tipi, he tried to make the dead branch sing for him. He blew on it, he waved it around — but no sound came. It made the young man sad. He wanted so much to hear that wonderful sound. He purified himself in the sweatlodge and climbed to the top of a lonely hill. There, naked, resting with his back against a large rock, he fasted for four days and four nights, crying for a dream, a vision to teach him how to make the branch sing. In the middle of the fourth night, Wagnuka, the bird with the flaming red spot on his head, appeared to him, saying, “Watch me.” And in his vision the young man watched — very carefully.
 
When he awoke he found a cedar tree. He broke off a branch, and working many hours hollowed it out delicately with a bow-string drill,³ just as he had seen wagnuka do it in his vision. He whittled the branch into a shape of a bird with a long neck and an open beak. He painted the top of the bird's head red with washasha, the sacred vermilion color. He prayed. He smoked the branch with incense of burning sage and sweet grass. He fingered the holes ad he had watched it done in his dream, all the while blowing softly into the end of his flute. Because this is what he had made — the first flute, the very first Siyotanka. And all at once there was the song, ghostlike and beautiful beyond words, and all the people were astounded and joyful. ([Erdoes 1976], pages 6–8)

³ This account of the method of construction is unique. As described earlier, all other sources indicate that the wood was split lengthwise and hollowed out to form a cylindrical bore. It is hard to imagine a bow-drill being able to make as large and as long a bore as is required.

Other origin myths have also been document. According to Mandan legend the flageolet was created by the Old Woman Who Never Dies by taking a long section of a large sunflower stalk, hollowing out its length, and cutting seven holes into its side. Each of the seven holes represented one month of winter and upon playing the instrument, snow would fall. ([Densmore 1923] , pages 80–84)

To conclude this chapter on the courting flute, an interesting and informative summary is provided by Belo Cozad, a Kiowa Indian who was recorded by Willard Rhodes in the late 1940's. On this recording, Cozad prefaces his performance of a Kiowa melody (14; see transcription) with a short personal biography and his version of the “story” of the Kiowa flute. The passage which follows is of particular interest for the first-hand information it contains about the source of his flute musich, its value and importance.

I'm a Kiowa tribe; my daddy he's the chief of the Apache Indian. He's hte first one who went to Washington city to see the Uncle Sam. A lot of Kiowas went with him, and they all die out. I'm seventy-seven years old now. I'm pretty old. And I like to give you some kind of new about this music — music I got, you know. If you'd like it I'll go and fetch it for your, sing for you and you can have that long as you live. And remember me and tell all your friends that you saw me right here at this Riverside Indian School. I like to play music for you and put some good songs that I know — I made it myself, good songs … for you and keep it as long as you live. I got this music from way back in Montana. One of a … poor boy, he's got no home and he went up on the mountain and stayed four nights there and he learned this music. He got it from some kind of spirit, he give it to him, show him to make it this way and to make it good music. And keep it as long as you live and you make it your good living because these trees … good trees, called cedar tree … It's a great tree, you know. And that's where he got this … From now on he's got this music and he's coming to well-off. He's got well-off womans and good home … raised children … I'm going to play it for you so I want you to hear good. (From Folk Music of the U.S. from the Archive of Folk Song, S I, b 10).

Clearly, the value and importance of music is foremost in the speaker's mind. He offers to play his music for his audience so that they may have it as long as they live. Power is attributed to good music and he makes a direct connection between it and economic success: “from now on he's got this music and he's coming to well-off.”

The story of his music's source substantiates the claim that, among Plains Indians, flute music is often received in a vision: “he went up on the mountain and stayed four nights there and he learned this music. He got it from some kind of spirit, he give it to him, show him to make it this way and to make it good music.”

Towards the end of his story, the speaker expresses his sense of a connection between the cedar tree, from which the flute is made, and a successful life: …you make it your good living because these trees … good trees, called cedar tree … This statement suggests that, to the Indian mind, the combination of the cedar tree, symbol of the powerful force of Nature, and music, which derives from supernatural sources and is also imbued with power, guarantees the success of any person who possesses flute music.

The speaker does not identify the song he plays nor does he describe the courtship function of flute music. However, this piece is very similar to the Kiowa flageolet melody (16) which appears on Side II of American Indian Soudchief Recording 248 under the title “Kiowa Indian Love Call.” [[Cozad-E 1964]]

Chapter 2 — Instrumental Love Songs

Among all recordings of American Indian music that have been published, only a small proportion is of flute songs. This is no to suggest, however, that the genre has been overlooked or neglected but rather that few flute melodies exist. As Nettl points out, “a typical tribal repertory may consist of several hundred vocal songs and a dozen flute melodies.” ([Nettl 1954], page 7) Although Nettl does not name the tribes upon whose music he bases this statement, it has been shown in the case of the Flathead Indians ([Merriam-AP 1967]) to be a fairly accurate estimate.

For this study, nineteen recorded examples of flute melodies from the Plains–Plateau–Southwest are are available and these have been transcribed to permit detailed comparison.¹ Most of these recordings have been issued by Ethnic Folkways and the Library of Congress.²

¹ In addition to this music, a Nez Perce flageolet melody published in Curtis, Edward S., The North American Indian, vol. 8, page 50; one additional Flathead melody from Merriam, Alan P., Ethnomusicology of the Flathead Indians, page 182; two Menominee melodies from Densmore, Frances, Menominee Music, pages 208–209; and three Omaha pieces, two from Fletcher, Alice C. A Study of Omaha Indian Music, page 151, and one from Fletcher, The Omaha Tribe, page 319, were also examined.
 
² A complete list of recording used appears on pages 154–155 of the PDF version of this thesis.

From the list which follows, it can be seen that flute music from the Central Plains has the best representation, while melodies from groups on the periphery of the Plains area, for example, the Chippewa and the Apache, are few. No recording of flute music from the extreme Northeast (Iroquois) or Southeast (Alabama, Yuchi, etc.) were found.

Although this list may reflect an impalance in available recordings, the predominance of Plains melodies is not unexpected since the practice of serenading with the flute appears to have been strongest in the Central Plains and references in the literature to this tradition are more numerous than for any other area.

In order to discuss regional differences, the flute melodies will be divided according to musical areas. It should be rememberd, however, that in dealing with music that is played on an instrument of fairly uniform design in all regions except the Southwest, it is to be expected that features such as range and tone system will conform to a standard pattern. Others such as performance style and ornamentation may be more variable but, nevertheless, will reflect an instrumental idiom.

2.1 — Western Great Lakes / Plains

The Western Great Lakes area, in which the Chippewa, Winnebago, Meskwaki, and Menominee are located, lies on the edges of both the Eastern Woodlands and the extreme northeastern Plains. Both Roberts ([Roberts-HH 1936], page 35) and Nettl ([Nettl 1954], pages 24–25) place this region within the Plains musical style.

Of the nine flute melodies which have been studied for this area, all but the Chippewa piece present a remarkably homogeneous pattern. The range of all the pieces lised between 12 and 14 semitones; i.e., a semitone below and above the full octave. Scales are most often pentatonic (anhemitonic) (3, 4, 7, D2) and hexatonic (5, 6, D1).³ One Menominee piece (2) is built on only four tones; while the Chippewa is based on seven tones without the octave repeitition. Both range and scale are sonsistent with the types of melodies that can be produced from the six-holed flageolet common to this area. The range, however, is smaller than would normally be found in the vocal music. According to Nettl's survey, “the range of most of the Chippewa and Menomini songs is very large; the average is the largest in North America. Of the Chippewa songs, only 9 per cent have a range smaller than an octave and 36 per cent greater than a perfect eleventh. The Menomini ranges are only slightly smaller.” ([Nettl 1954], page 25)

³ The melodies under discussion will be referred to by number. the two Densmore examples (Menominee) are designated D1 and D2. Transcriptions are given on pages 45–72 of the PDF version of this thesis.

A consideration of melodic movement reveals another major difference between the instrumental and vocal music of the area In this eastern sub-section of the Plains, melodic movement is almost exclusively of the ‘terrace’ type ([Nettl 1954], page 25), to which none of the examples of flute music conforms. In contrast, all pieces (exclusing the Chippewa) show an initial leap of an octave to the highest tone and a gradual descent, occasionally only half-way, but most ogten through the full octave to the base tone.⁴ Instead of returning to the point midway in the melodic line, as would occur in the ‘terrace’ pattern, almost all subsequent phrases return to this same high note and repeat the descent. Again the reason for the difference between instrumental style and the typical vocal style may lie with the flute itself. Naturally if the instrument affords a range of only one octave, instead of the one and a half or some sometimes two of the vocal range, the descending line would be greatly limited by repetitions starting at consecutively lower points. The Chippewa melody, anomalous in other respects as well, has an undulating contour which is more typical of the Eastern Woodlands style.

⁴ In discussing melodic line and contour, some generalizations are made for the sake of clairty. For example, if a phrase is said to descend from highest to lowest tone it sometimes does not do this directly as but rather with some undulation, . The general movement, however, is strongly downward.

The intervals used in these melodies reflect an instrumental style. A typical beginning consists of an octave rise (the Chippewa piece begins with the interval of a perfect fifth) followed by large intervals of fourths and fifths. The latter parts of phrases have somewhat less movement and generally show a step-wise descent. (For example, see Figure 1)

Figure 1. Menominee flageolet melody (2)

The frequent use of wide intervallic leaps in combination with long-held tones at phrase-ends produces a quality of “spaciousness” ([Herzog 1935a], page 29), which is typical of both instrumental and vocal love songs. Again, the Chippewa melody is atypical in that the melody line is smooth, with step-wise movement predominating. Major seconds and minor thirds are common.

The rhythm of these flute melodies is free. Combined with irregular melodic lines often containing wide intervals, a free rhythm gives this music a rhapsodic quality. The tempo is relatively slow (M.M. = 72–88); however, in the case of the Meskwaki melody, the extensive use of ornamental trilling gives the impression of a faster tempo.

Figure 2. Meskwaki flageolet melody (7)

In contrast to these melodies with a free rhythm and lack of consistent meter, the Chippewa piece has a regular rhythm and underlying triple meter. It uses only two rhythmic figures, and , a feature which is not seen in the other pieces. In her work, Densmore has suggested that songs with a regular hrythm tend to be more modern and, to reinforce her statement, relates the following incident:

In recording a Chippewa song from an old Indian the writer found the rhythm peculiar, with frequent changes of measure lengths; later the same song was recorded by a young man, said to be an excellent singer. On comparing the phonographic records it was found that the younger singer had slightly changed the rhythm so as to avoid the irregularity in the measure lengths. The song had lost its native character and also its musical interest. ([Densmore 1918], page 59)

If one considers the Chippewa melody to be a more modern peice, then not only the rhythm but other dissimilarities can be explained. Either it is a new piece, strongly affected by European musical style, or it is an older melody whose ‘native character’ has been gradually lost until it now resembles a European folk melody.

In more typical melodies (2, 3, 4, 5, 6, 7), the ‘rubato’ tempo is in part aided by a wide distribution of durational values. All pieces except the Chippewa contain note-values from a sixteenth, or even thirty-second, to very long-held notes used at phrase-endings. Characteristic of neither Plains nor Eastern Woodlands musical style, this feature can be considered as idiomatic of the flute.

Flute melodies from the Western Great Lakes area exhibit a definite binary structure (2, 3, 4, 5, 6, 7). For example, the Meskwaki melody (7) which is fairly long and made up of many repeated phrases (iterative) has an exact repetition of material: ABBBA'B'B / ABBBA'B'B. Other forms are more progressive (i.e., contain new material in each phrase), as ABC / A'B'C (4) or ABC / DB'BC (6), but still show a division into two equal parts. Major sections of a piece are very clearly marked off by long-held notes, invariably the base tone or ‘tonic.’ These notes, which are one of the most characteristic features of flute music, function as short introductions, mark the ends of phrases and of major sections, and provide coda-like endings. In all instances this tone is the tonic or base tone. Other tones which are secondary to the base tone but still very prominent are the fourth, fifth, and octave above this note. An example of this weighting can be seen in the tone system of Winnebago melody (4).

Figure 3. Winnebago flageolet melody (4)

Finally, there are a number of other features which, together with the above, create an idiomatic flute style. The most characteristic of these is the intense vibrato with which the tonic of the melody is played. This apparently is essential to a good flute technique. According to Fletcher ([Fletcher 1911] , pages 371–372):

To be acceptable, a flute must give forth a full vibrating tone when blown with all six holes closed. It was interesting to watch men, old and young, take up a flute to test it; they would readjust the stop piece, bound to the top over the opening and usually carved, and if after several trials the instrument could not be made to give this vibratory tone the flute would be laid aside and no words would avail to make the man take it up and play a tune on it.

Although speaking here of the Omaha tribe, the same prominence of a full vibrato on the tonic is seen in the flute music of the Western Great Lakes area.

In this sampling, two of the pieces (4, 5) are played on flutes made form metal gun barrels. Both are able to produce a vibrato on their lowest tones, but in their higher range they are somewhat shriller. In addition, their base tones are slightly higher than those of the wooden flutes (i.e. b' and c'' instead of g' and a').

Because the base tone is played with such intensity, the octave above is often heard, either as an overtone or as a quick grace note. This appears to be a cultuvated effect, rather than accidental, since all pieces with the exception of the Chippewa contain many examples of it. Other grace notes within the melodic line are common (4, 5) as are downward glissandi, or falling releases (2, 4, 5, 6), rising releases (2, 3) and trilling (7). All of these ornamental devices occur to some extent in all pieces (again, except the Chippewa).

2.2 — Central Plains

Twelve melodies from the Sioux, Kiowa, and Omaha provide the material for studying the flute music of the Central Plains.⁵ As in the preceding group of melodies fromt he Western Great Lakes, the music is generally consistent within itself; i.e., many features are common to all pieces, with the exception of the two Sioux examples (8, 9). Through the following discussion it will become evident that the Sioux melodies should probably be considered, like the Chippewa melody, as newer pieces strongly influenced by European musical style.

⁵ Three of these melodies, designated F1, F2, and F3, are Omaha flageolet pieces taken from Fletcher, The Omaha Tribe, page 319, and Fletcher, A Study of Omaha Indian Music, page 151. See transcriptions, pages 65–67 of the PDF version of this thesis.

Whether traditional or modern, all the melodies are within the range of 13 to 15 semitones; i.e., an octave to a major ninth. Flutes on the Central Plains almost always had six finger-holes, although the Sioux also made instruments with only five stops. While capable of a slightly fuller scale, the majority of flute melodies are pentatonic (8, 10, 11, 12, 13, 14, 16). One Kiowa piece (15) is hexatonic, and the remaining Sioux melody (9) is tetratonic. THis predominance of the pentatonic scale conforms with Nettl's findings for vocal music in the central and southern Plains ([Nettl 1954], pages 27 & 29).

One peculiarity of scale was noted for a set of three Omaha melodies (10, 11, 12) which are all played by the same performer on the same instrument. In this music, the base tone is g', while the only ton consistently heard as its octave is g#''. This is the only occurence in all of the flute melodies of an augmented octave. Since, on these instruements, the upper octave is reached by over-blowing which has a natural tendency towards sharpening, the augmented octave is probably only a result of this. If, in fact, the g#'' were to have its own stop, this would be a very rare example of a tuning in semitones of the flageolet's lower range.

In common with flute music of the Western Great Lakes, that of the central Plains also fails to show a ‘terrace-type’ of melodic movement. This presents a significant departure from Plains vocal style, where melodic contours are almost entirely of this type. Three Plains flute melodies have the same kind of melodic pattern as seen in the Western Great Lakes area; i.e., an initial rise of an octave to the highest tone followed by a gradual descent to the tonic (12, 15, F1). Kiowa melody No. 15 exhibits a tightly constructed version of this general contour. It has some resemblance to the ‘terrace-type’ movement but does not adhere strictly to its pattern. After the initial rise to its highest tone, the melody descends through a perfect fifth, returns, and repeats this first phrase. After reaching the halfway point a second time, the melody continues its descent to the base tone. This half also is repeated, thereby completing the symmetry. Shown diagrammatically, the melodic contour would be:

Figure 4. Diagram of melodic contour of Kiowa flageolet melody (15)

The majority of the flute pieces, however have a simple arch-form (8, 10), a combination of arch-form and straight descent (13, 14, 16, F2, F3), or an undulating (9, 11) melodic contour. In the combined form, the melody begins fairly low on its scale and gradually rises until the highest tone is reached approximately midway in the piece. Once the highest point is reached, the melody gradually descends through the full octave to the base tone.

Two general patterns for the use of intervals in Plains music emerge. Half of the melodies (8, 11, 13, 16, F2, F3) consistently use small intervals of seconds and thirds, a feature which coincides with the general trend of Plains vocal music ([Nettl 1954], page 29). The other half (9, 10, 12, 14, 15, F1) makes use of wide leaps of octaves, fourths, and fifths, and represents a more idiomatic flute style. Of these, Nos. 9, 12, 14, 15, and F1 show similarity to the typical Western Great Lakes melodic movement: i.e., large intervals occur in the early parts of phrases, while the latter parts have somewhat less movement and show a step-wise descent. In general, a descending melodic line will contain wider intervals of fourths and fifths, while an arch-form or undulating line will have a predominantly step-wise movement.

The rhythm of Plains flute melodies, in contrast to those of the Western Great Lakes, is more restrained and regular. Several rhythmic figures recur, giving these pieces a rhythmic unity which was not apparent in the Western Great Lakes music. For example, the figures and are prominent in three Omaha pieces (11, 13, F2), while or its variants and are seen in pieces from all three groups, the Omaha (12), Kiowa (15) and Sioux (9). The rhythms of nos. 10, 14, and 16 remain free and ‘rhapsodic’ and, in this respect, are similar to Western Great Lakes music. The two Sioux melodies (8, 9) with regular triple and duple meters respectively, are anomalous to this group of Plains melodies but are very similar to the Chippewa piece discussed earlier. Their extreme rhythmic regularity would suggest that, in this case as well, the melodies have been modified by European musical influence.

In summary, Plains flute music presents a wide range of rhythmic possibilities from very free (14, 16) to strictly metrical (8, 9). In his study, Nettl found the rhythm of Plains vocal music similarly complex ([Nettl 1954], page 29).

As expected, the distribution of durational values is wide in those melodies which are rhythmically free. Those with recurring rhythmic figures (9, 11, 12, 13, 15, F2) show less of a distribution, while the completely regular melodies (8, 9) are restricted to only two or three note values. The tempo of all these flute melodies is generally ‘andante.’ (M.M. = 80–92)

Like Western Great Lakes flute music, the majority of Plains melodies have a binary structure (8, 9, 11, 13, 15, 16) which is clearly demarcated by long-held tonic notes. Again, this pulsating base note also functions as an introduction, which can sometimes be rather prolonged as in the Kiowa melodies 14 and 16. It also signals the end of a piece in all cases except the two Sioux melodies (8, 9) where an unusual coda consisting of two overblown notes not found in the body of the piece is heard.

Figure 5. Sioux flageolet melody (9)

These codas can probably be considered as a characteristic manner of closing for Sioux love songs. Sung versions of this type of ending occur in several of the vocal love songs and will be discussed more fully in the following chapter.

Internally, the structure of Plains flute melodies is most often reverting (8, 10, 13, 14, 16) and only occasionally progressive (9, 10) or iterative (12, 15).⁶ This breakdown agrees with Nettl's findings for vocal music of the southern Plains where “reverting forms consisting of a number of short sections … account for almost half of the songs.” ([Nettl 1954], page 28). Elsewhere on the Plains, however, the most common form is an incomplete repetition which is not seen in these flute melodies.

⁶ Reverting: restatment of earlier material; progressive: no material repeated; iterative: repitition of material immediately preceding. (From Nettl, North American Indian Musical Styles, page 6).

As in Western Great Lakes flute music, the tonic of a piece is always the lowest tone and is also the most prominent by virtue of its position at the beginnings and endings of pieces and usually at phrase endings as well. The only exception to this is seen in Sioux melody no. 8. Here the fifth below the tonic is the most prominent. A review of the tone systems of these melodies, which give the relative weightings of each tone in a given scale, shows that the fourth, fifth, and octave above the base tone are also structurally important.

No additional ornamental devices are seen in Plains flute music that were not used in the Western Great Lakes area, although the style of playing in the Omaha melodies is relatively more staccato. Omaha melody no. 13 is also remarkable for the heavy ornamentation centering around the long-held tonic notes. In addition to the usual overblow octave grace notes, rather elaborate turns precede these notes in several places.

Figure 6. Omaha flageolet melody (13)

2.3 — Plateau

It has been hypothesized that the flageolet originated in Mexico and subsequently spread northward ([Galpin 1903], page 135; [Sachs 1929], page 214; [Roberts-HH 1936], pages 20 & 25). Roberts pictures the diffusion of Mexican influences into the continental U. S. A. as having fanned out somewhat in the shape of a mushroom — strongest in the central corridor of the southern and central Plains and less prominent in regions to the north, northwest, and northeast. Therefore, in moving away from the central Plains and into the Plateau area it is not unexpected to find few references to the courting flute and even fewer examples of its music. Only four examples of flageolet melodies are available for study: three from the Flathead and one from the Nez Perce Indians.⁷ Withing this small sampling, however, the melodies are similar in several respects.

⁷ Two Flathead melodies (17, 18) have been transcribed from Alan P. Merriam's recordings, Songs and Dances of the Flathead Indians; the third piece (M1) appears on page 182 of his Ethnomusicology of the Flathead Indians and C1 is a Nez Perce melody taken from Curtis, The North American Indian, volume 8, page 50.

The range of these melodies varies from 12 semitones (G1) to 15 semitones (17). ALl of the Flathead melodies are based on pentatonic scales (anhemitonic), while the Nez Perce melody is heptatonic (d', e', f#', g#', a', b', c#''). Curtis' description of the Nez Perce instrument as a seven-holed flute made from an elderberry stalk ([Curtis-E 1911c] , volume 8, page 157) is consistent with Merriam's; however, Merriam found that the seventh hole, placed at the lower end of the flageolet was never stopped ([Merriam-AP 1967], page 50). To produce the heptatonic scale of the Nez Perce melody, either the seventh hole of the instrument was placed closer to the usual bank of finger-holes and was covered, or the c#'' was produced with all the finger-holes open.

In melodic movement the four Plateau pieces are quite divergent. Only one (17) shows the strongly descending pattern typical of Plains and Western Great Lakes melodies. The remaining three have an undulating movement (M1, C1) or are arch-form (18). The intervals used in all four melodies are small. Step-wise movement predomintes, with some use of falling thirds (18, C1) and fourths (17, M1). These usually occur at phrase-endings.

The pieces are all played with a free rhythm and the distribution of durational values is fairly wide in all except No. 17. This Flathead piece presents a pattern of greater rhythmic stability than the other through its use of only three durational values, eighth, quarter, and half, and notes of the same duration often follow consecutively.

Figure 7. Flathead flageolet melody (17)

Phrase patterns tend to show incomplete repetition and are non-symmetrical (No. 18: ABA / BA' ; C1: AB / B ; and M1: ABB). Only no. 17 is binary and iterative (AA' / AA'). In these melodies, unlike the majority of Plains and Western Great Lakes pieces, there is no long-held base note to provide a distinct introduction, mid-section, or ending. No. 17 approaches this form somewhat with held notes at phrase endings but these are not on the tonic. In fact, in all pieces except M1 the tonic does not have greatest prominence, which is a major difference between flute music from the Plateau area and that of the Plains and Western Great Lakes.

The performance style of flageolet music in the Plateau area is generally rather subdued in comparison to that of the Plains. For example, neither long-held notes played with an intense vibrato nor overblown octave grace notes are heard. Some rising releases (18) and falling glissandi (M1) occur at phrase endings, and grace notes, typical of the instrumental style, are common (17, 18, M1). Two of the Flathead melodies are played on flageolets made of nickel tubing. These instruments produce a tone which is thin and light but without shrillness. Only the Nez Perce melody has the full tone and spacious quality typical of Plains flute music.

2.4 — SouthWest

In the Southwest, the Apache appears to be the only group to use the courting flute, and this area was found to have very little flute music. The only example of Apache flute music available for this study is taken from the recording, Music of the Pueblos, Apache and Navaho, made by McAllester and Brown in 1961. At that time the collectors found that “among the Apaches almost nobody plays the flute today.” ([McAllester 1960] , page 472).

The instrument made by the Apache is a whistle flute of river cane and has only three finger-holes. The range of this melody is limited to eight semitones (a perfect fifth) and its scale is tetratonic (d",f",g",a"). The melodic movement of each short phrase is in arch-shaped contours and gives the piece an overall undulating effect similar to that seen in the Flathead melodies. This feature conforms to Nettl's finding for vocal music of the Apache in which the melodic movement also tends to be in arc-shaped contours ([Nettl 1954], page 22). Intervals are small and step-wise movement predominates. There is some use of falling thirds; again, a similarity with Flathead melodies.

The rhythm of this melody is free and the tempo fairly slow (M.M. = 69). Only three durational values are used consistently: , a feature in common with the vocal music of the Apache which uses few (usually only two) durational values ([Nettl 1954], page 22). Structurally the piece is very simple: a repetition of three short phrases giving an AA'A'A' form. Although there is no base note introduction, each major section is ended with a long-held tonic note. The third and fifth above the tonic are featured prominently, a trait which is typical of the vocal music as well. The degree of ornamentation in this piece is remarkable, and each melodic repetition is varied in this manner. Because the Apache flute is made from river cane it does not afford the fuller, more vibrant tone of the wooden flageolet. There is, however a light vibrato on the base note but typically the instrument is played “with a breathy quavering technique.” ([McAllester 1961], page 11).

2.5 — Summary of Characteristics of Instrumental Love Songs

In summarizing the preceding discussion of twenty-six flageolet melodies, certain characteristics recur that can be considered features of an instrumental style while others show a similarity to the typical vocal style of a given area.

The range of these melodies is, of course, dependent upon the instrument to a certain degree. Twelve to fifteen semitones is the average range; somewhat less than vocal music of the Western Great Lakes but about the same as for the Plains. The great difference is seen in the flute music of the Southwest where the Apache flute produces melodies of only half an ocatve's range while the vocal music of the area often covers one and a half octaves.

Scales are most often pentatonic. This feature coincides with the general trend of vocal music in all areas discussed. Tetratonic and hexatonic scales occur less frequently and heptatonic scales are rare.

In Western Great Lakes flageolet music, a melodic pattern of repeated descent from the highest note of the piece emerges as the most important type. Plains instrumental music modifies this feature somewhat by alternating straight descent with arch-form phrases. Arch-form and undulating contours are also frequent in flute music of the Plains, Plateau, and Southwest areas. Only vocal music of the Southwest has a majority of songs in arch-form. In contrast to these various melodic contours, the melodic line of vocal music from the Western Great Lakes and Plains is almost always of the ‘terrace-type’.

An instrumental style is evident in the use of intervals and the many wide leaps of octaves, fourths and fifths that are idiomatic to flageolet melodies. A typical intervallic pattern is seen in Western Great Lakes and Plains music. Phrases begin with numerous large intervals and then towards phrase-endings become more restricted in movement and show a step-wise descent. Half of the Plains and all of the Plateau and Southwest melodies use small intervals of seconds and thirds, a feature which is typical of vocal music of the Plains and Plateau areas.

A very free rhythm characterizes flageolet music. Combined with a wide distribution of durational values, this unmetered rhythm creates a spacious and rhapsodic quality typical of instrumental love songs. In the few melodies (Plains) where the distribution of note values is not as wide, rhythmic figures recur to give unity to the music. A slow to andante tempo (M.M. = 72–92) is common to all flageolet pieces and is also a characteristic feature of the instrumental style.

Despite the improvisatory and rhapsodic impression created by a very free rhythm, large intervallic leaps, and a wide distribution of durational values, flageolet pieces always have a tightly constructed form. Music of the Western Great Lakes and Plains is most often binary in overall structure, with reverting forms most common internally. The iterative form is less frequent and progressive structures, least common. The reverting and iterative forms tend to compensate for the free rhythms by creating a structural unity through repetition. A distinctive feature of flageolet music is its use of the long-held tonic note as a structural divider. In Western Great Lakes and Plains music this tone almost always functions as an introduction and ending, and quite often clearly demarcates mid-sections and section endings. This, however, is not a feature of Plateau and Southwest flageolet music. Tones which are a fourth, fifth and octave above the base tone are also structurally important.

Finally, there are a number of ornamental features which together create an idiomatic flute style. The most characteristic of these is the intense vibrato with which the tonic of the melody is played. Grace notes, an octave above the tonic and created by overblowing, are a typical feature and appear to be a cultivated effect. Grace notes within the melodic line, turns, mordents, and trills are commonly used, as are downward glissandi and rising releases at phrase endings.

Transcriptions

Signs used in the Transcriptions

Transcribing Indian melodies in ordinary musical notation is somewhat like forcing a square peg into a round hole. it can be accomplished by dint of sufficient exertion, but the original form will have suffered. The vital part of these melodies can be expressed in our notation, but many a delicate nuance of wild and wayward beauty will have disappeared. (Henry F. Gilbert, “Note on the Indian Music”, in Edward S. Curtis, The North American Indian, Volume 6 (Cambridge: University Press, 1911), p. 166.)

Signs used in transcriptions

Chapter 3 — Vocal Love Songs

Before proceeding to a discussion of their musical aspects, it is necessary to try to define love songs in terms of American Indian culture for it is insufficient and, as will be shown, incorrect to assume that an Indian love song carried sentiments similar to those of a love song in European culture.

There is conflicting evidence as to whether love songs associated with courtship even existed in American Indian society before European contact. On the one hand, Fletcher ([Fletcher 1911] , page 319) maintains that the Omaha did have love songs prior to this period and that they were considered acceptable forms of expression. In contrast, Densmore quotes Indian informants who claimed that loves songs “were not sung in the old days” ([Densmore 1926], page 85). At that time songs concerning marriage were most likely to be closely associatged with war, as the following two texts from the Sioux illustrate:

You may go on the warpath. When you return I will marry you.
 
As the young men go by I was looking for him. It surprises me anew that he has gone. (It is) something to which I cannot ebe reconciled. ([Densmore 1918], pages 371–372).

Others, which are commonly termed ‘love songs,’ express loneliness for a close family member or sentiments of sadness and mourning for a child, wife, or husband who had died. Some of these are extremely touching in their simplicity. The following is the text of a song from the Tlingit in which an old man who is dying addresses his young wife:

Shake hands. I want to hold your hand before I die. I'm going to be sorry about you when I die. ([Laguna 1972], page 1295).

Closer to the notion of a love song in Western society are the “songs of affection” which might be sung by persons who had been married for many years. Among the Pawnee, for example, these were considered expressive of “honourable” sentiments and a clear distinction was made between this type of song and the “modern” love song which later developed in Indian society as a result of European influences. The Pawnee associated the singing of courting songs wich a lower class of people who lived near towns, worked for Europeans, and drank whisky. ([Densmore 1929a], page 96). As one Pawnee writer has rather strongly stated: “… charms, songs, etc., to lure women were furnished by sexual perverts who lived somewhat apart and were in social disrepute.” ([Murie 1914], page 640). It is quite probable that love songs were also viewed negatively by older people because they underscored the erosion of traditional parental authority in matters of marriage.

Despite this rather negative status of love songs, Densmore acknowledges that they were very prevalent on reservations in the early decades of this century ([Densmore 1926], page 87) and that among the Chippewa they were a favourite form of musical expression ([Densmore 1931], page 16). A brief glance at the number of love songs that are included in her volumes on Chippewa music confirms this.

Even in those songs which can be termed modern courting songs, texts did not often refer directly to another person expressing sentiments of affection. Rather, they tended to be songs of sadness, loneliness, or disappointment that were sung by oneself. Mention of weeping only occurs in love songs and is often associated with intoxication ([Densmore 1932a], page 210). The following song texts from the Chippewa are illustrative of the general tone of love songs:

To a very distant land he is going, my lover, soon he will come again ([Densmore 1913], page 301).
 
and still I have lost my sweetheart ([Densmore 1913], page 280).
 
I go around weeping for my love (free translation) ([Densmore 1913], pages 220–221).
 
Although he said it, still I am filled with longing when I think of him ([Densmore 1910], page 154).

Other songs are more light-hearted and have a touch of humour. Of the following song, Densmore ([Densmore 1910], page 151) has written that “in the old times an Indian maid would lie face down on the prairie for hours at a time singing this song.”

Why should I, even I be jealous because of that bad boy?

From Menominee comes the following “taunting” song:

You had better go home, your mother loves you so much. ([Densmore 1932a], page 210).

To conclude this discussion of song texts, one final example of a love song, exhibiting a finely developed sense of poetic expression, is quoted:

A loon I thought it was,
But it was my love's splashing oar.
To Sault Ste. Marie he has departed,
My love has gone on before me,
Never again can I see him.
([Densmore 1910], pages 150–151).

Vocal love songs, like Ghost Dance and Peyote songs, have been considered a separate genre by a number of writers.¹ As a special song type, love songs have characteristics which are unique to them and are not seen in other vocal music of a given area. How this song type may have evolved and a description of its features is the topic of this chapter.

¹ Herzog includes vocal love songs as a distinct category in his “Special Song Types in North American Indian Music” ([Herzog 1935a], pages 23–33). Densmore repeatedly places love songs under a separate heading in Chippewa Music I and II and in Menominee Music, and discusses features that are unique to them. In Ethnomusicology of the Flathead Indian, Merriam takes up Herzog's theory of special song types in connection with his own study of Flathead Indian music ([Merriam-AP 1967], pages 316–321).

Vocal love songs are generally found in the same geographical location as flageolet melodies and, as Herzog points out, this suggests a close connection between the instrument and the song type ([Herzog 1935a], page 27). In all probability vocal love songs derived from flageolet melodies. Quoting a native informant, Densmore ([Densmore 1932a], page 208) writes: “Long ago there was a kind of singing which had no words and was an imitation of the flute. This was intended as a love song and it was different from any other kind of singing.” It has been found that the Winnebago ([Herzog 1935a], page 27; [Densmore 1930], page 658), Dakota ([Herzog 1935a], page 28), and Pawnee ([Densmore 1930], page 658) also believed that love songs originated from flageolet melodies. If love songs are a result of a transfer from flageolet to voice, certain features of the instrumental melodies can be expected to recur in the vocal songs. Love songs will be discussed in this context and an attempt will be made to determine whether vocal love songs have greater similarity to their instrumental counterparts of to the typical vocal song style of a given musical area. For consistency and, in it hoped, clarity of organization, vocal music will be discussed according to the same geographical groupings as were employed in the previous chapter on flageolet melodies.

3.1 — Western Great Lakes/Plains

For the study of Hestern Great Lakes love songs, a sample of thirty-one melodies is available. All but one of these have been collected and transcribed by Frances Densmore and are published in Chippewa Music I and II ([Densmore 1910] & [Densmore 1913]) and in Menominee Music ([Densmore 1932a]); Bulletins 45, 53, and 102 from the Bureau of American Ethnology.² The Chippewa music was recorded by Densmore in 1908 and 1910 at White Earth, Red Lake, and Waba'cing in Minnesota and on the Lac du Flambeau Reservation in Wisconsin. In 1951 a long-playing recording, Songs of the Chippewa (L22), containing seven of these love songs was issued by the Library of Congress. These have been re-transcribed and are appended to this chapter as songs No. 20–26, pr. 112–121.³ Similarly, an album entitled Songs of the Menominee, Mandan and Hidatsa (L33, 1953) contains one Menominee love song and this has also been re-transcribed (No. 28, p. 123). Given a large body of material, no one piece will be discussed in detail, but since the purpose of this analysis is to distill the major characteristics of Western Great Lakes love songs, a large sampling should produce a fairly accurate picture.

² For a complete list of the Densmore songs that have been analyzed, see pages 106–107 of the PDF version of this thesis.
 
³ An eighth Chippewa melody from an album collected by Charles Hofmann is designated no. 27, see transcription page 122 of the PDF version of this thesis.

The range of most Chippewa and Menominee love songs is 13 to 15 semi tones (18 examples); hmvever, another eight songs (or almost 26%) have very large ranges of one and a half to two octaves. Referring back to flageolet melodies, it was noted that the range of all the Western Great Lakes pieces lay between 12 and 14 semitones, while the average range for vocal music of this area was much larger. It would seem here that, in terms of range, vocal love songs have been influenced by both their instrumental counterparts and by the typical musical style of the area, but that the influence of flageolet melodies predominates.

Scales are primarily pentatonic (12 examples) and hexatonic (7 examples). This corresponds to the previous finding for flageolet melodies; whereas typical vocal music of the area is usually based on pentatonic or tetratonic scales ([Nettl 1954], page 25). Some scales of seven tones and four tones occur (six and five examples, respectively) and one love song using only three tones was found.⁴ In general, songs with pentatonic scales have an average range of one octave while those based on hexatonic and heptatonic scales have the largest ranges.

⁴ The tritonic melody is anomalous to this group of love songs in many ways. Densmore ([Densmore 1913], page 300) relates that the singer had learnt this song as a young girl more than sixty years previous and it is quite possible that the repetitious text and bare triadic melody are a result of being imperfectly remembered.

The melodic movement of vocal love songs in the Western Great Lakes area exhibits a fairly consistent pattern. In most songs, individual phrases have an undulating line but the tendency over the entire piece is one of gradual descent from highest to lowest tone. Songs 20, and 22–25 are good examples of this type of melodic pattern. A few love songs have an undulating line which starts or ends on or near the same tone (27, MM page 210 top) and three are simple arch-forms (21, 28, MM page 211 bottom). The type of descending line seen in flageolet melodies of this area is rare among their vocal equivalents. It appears that the melodic movement of vocal love songs corresponds neither to the typical vocal song of the Western Great Lakes, which is almost exclusively of the terrace type ([Nettl 1954], page 25), nor to the descending line of flageolet melodies, but ratehr is closer to the undulating contour typical of Eastern Woodlands style ([Nettl 1954], page 34).

Unlike flageolet melodies which make frequent use of larger intervals of fourths, fifths, and octaves, the melodic lines of vocal love songs are smoother and contain smaller intervals. Most often love songs of this area have a step-wise movement but with numerous intervals of thirds and fourths interspersed (23, 25, 28). Occasionally intervals of thirds and fifths (27) or fourths and fifths (24) combine with a step-wise movement. Approximately 40% of all the vocal love songs employ small intervals of a major second and minor third (22) but even in these the interval of a fourth is still fairly prominent. Only two pieces (21, 27) exhibit the initial octave leap typical of flageolet melodies; however, the internal structure of thier melodic lines is smooth, with step-wise movement and a few thirds predominating. In general, the melodic intervals of Western Great Lakes love songs correspond more closely to the typical vocal style of the area; i.e., with thier use of smaller intervals of a second and third ([Nettl 1954], page 25), rather than to flageolet melodies.

The free rhythms common to most flageolet melodies of the Western Great Lakes area are lacking in vocal love songs.⁵ The majority of songs reveal fairly regular rhythms and have underlying duple meters (22, 23, 25, 26, 27, 28) or a triple meter (21). Only two pieces are rhythmically free (20, 24). In this sub-area, Nettl points out the frequent occurance of the isorhythmic principle in vocal music ([Nettl 1954], page 26). Although this feature was not observed in the flageolet melodies, seven of the vocal songs (20, 22, 23, 24, 26, 27, 28) contian isorhythmic phrases; i.e. phrases in which the rhythmic values of the notes remain the same even through pitches change. For example, the notes of the first, second, third, and final phrases of Chippewa songs 24 have the following rhythms (or very slight alterations thereof:

⁵ In this discussion of rhythm, Densmore's transcriptions have not been enumerated since, in re-transcribing the eight songs that were available on recordings, her rhythmic values were found to be unreliable and numerous changes were necessary.

Similarly, the only available example of a Menominee love song is asymmetrically isorhythmic, being composed of two phrases which are then repeated:

The only melody of this group (27) which is sung with a drum accompaniment is strictly isorhythmic:

Vocal love songs of the Western Great Lakes do not have the same wide distribution of durational values found in their instrumental counterparts. Most of the songs are dominated by smaller note values, , and there is a notable absence of the very long-held tones that were prominent at the beginnings and ends of flageolet melodies. Several pieces make extensive use of dotted rhythms, (27, 28) or (25), a feature in common with typical vocal music of the area ([Nettl 1954], page 26). Four of the Densmore transcriptions (CM II page 219, 225, 282; MM page 210 top) indicate a wider range of note values but since these cannot be checked, no definite statement concerning their occurrence can be made. There is, however, general agreement on the tempo of vocal love songs. Referring to Chippewa and Menominee songs, Densmore states that they are slower than other classes of songs ([Densmore 1951], page 16^**; ([Densmore 1932a], page 212) and Herzog claims that “love songs are on the whole … among the slowest Indian songs” ([Herzog 1935a], page 29). My own metronome readings are a consistently slow, M.M. = ca. 62–92. The slowness of tempo of both instrumental and vocal love songs strongly suggests a connection between them.

^** Editor's note: It is not clear what resource was intended by Mary Frances Riemer for the citation [Densmore 1951]. It may have been [Densmore 1910] or [Densmore 1913].

In terms of form, all songs exhibit a clearly defined structure. This finding, however, is in disagreement with Herzog who states that instrumental melodies and their vocal counterparts “often enough, consist of a similarly free and loose cumulation of phrases” ([Herzog 1935a], page 29). The majority of love songs from this area (60%) have an iterative structure (26, 27, 28). Progressive and through-composed forms account for another 28% (20, 21, 22, 23), but reverting forms are few (24, 25). This distribution of structural types does not correspond with Nettl's findings for typical vocal music of the area which is predominantly progressive ([Nettl 1954], page 26). Unlike the instrumental melodies, vocal love songs are not clearly divided into sections by long-held base tones and only one vocal song (21) has the repeated-note introduction which is typical of flageolet melodies.

Even without repeated statements of the base note, the tonality of almost all vocal love songs is clear. The tonic is most often the lowest tone, towards which the whole melody gravitates, and it is usually the most prominent tone as well (22, 23, 24, 25, 26). Both flageolet melodies and typical vocal music of the area follow this pattern ([Nettl 1954], page 25). Occasionally a vocal melody will descend to a fourth below the tonic (27, 28) but without weakening the tonality. In vocal songs, fourths, fifths, and octaves above the base tone are not as consistently prominent as in flageolet melodies but are, nevertheless, important tones of the scale (22, 23, 24, 25, 28).

Although a discussion of the preceding characteristics has revealed similarities between vocal love songs and their instrumental counterparts, two significant features, the manner of performance and vocal quality, clearly indicate their connection. These, however, are the most intangible features, impossible to notate adequately and very difficult to describe in words.⁶ The most readily apparent and striking characteristic of vocal technique is the nasal, drawling tone with which love songs are rendered (20–25, 28). According to Densmore, this nasal tone was used by the Chippewa and Menominee to imitate the sound quality of the flute ([Densmore 1932a], page 208). To further enhance this imitation, a singer would sometime wave his hand slowly before his mouth to interrupt the flow of breath and produce pulsations ([Herzog 1935a], page 28; [Fletcher 1893], page 11). The occasional single sharp call at the end of the piece (20, 23) is reminiscent of the typical high grace note ending of flageolet melodies. The unique manner of voicing of love songs almost demonstrates their similarity to flageolet melodies. For example, the rising flissando attack at phrase beginnings (21, 22), glissandi between wider intervals (20–25, 28), and rising (20, 23–24) and falling (22) releases are all features of instrumental love songs. Song no. 20 is perhaps the best example of the unique manner of performance of vocal love songs since it contains virtually all of the characteristics mentioned above. In addition, this song has one peculiarity of performance not heard in the other love songs. At the beginning of phrase B', the singer performs a sudden dimunendo and sings two ‘portamento’ notes on the vocable or syllable “mu-um.” Attempts at swelling or diminishing a tone are “sometimes noticable in love songs” ([Fletcher 1893], page 11) and, in this case, it sounds very like an imitation of a flute call. This manner of performance is remarkable since attempts at interpretative singing are almost unknown in American Indian music.

⁶ Densmore's transcriptions have not been included in this discussion as her notation gives very little indication of features beyond pitch and rhythmic not evalues. She does, however, repeatedly refer to the distinctive vocal style of love songs.

3.2 — Central Plains

Vocal love songs from the Plains are are represented by ten melodies from the Sioux, Omaha, and Kiowa. ⁷ As a group, these songs are quite homogenous and have many features in common with flageolet melodies from the Western Great Lakes and Plains areas.

⁷ Three of these transcriptions from Fletcher, Alice C., The Omaha Tribe, page 320 (two Omaha pieces designated F4 and F5) and from Curtis, Edwards S., The North American Indian, volume 3, page 150 (one Sioux piece designated C2), for which no recordings are available. See transcriptions, pages 133–135.

The range of all the Sioux and Kiowa love songs is a uniform thirteen semitones; i.e. a full octave. In this respect, they conform exactly to their instrumental counterparts; while the two Omaha melodies, with very large ranges of 23 and 27 semitones (just under and over two octaves), resemble some of the Chippewa love songs just discussed. The most common scale is tetratonic (30, 31, 32, 33, 34, C2), followed by hexatonic (35, F4). Only one melody is pentatonic (29) and one, heptatonic (F5). This finding is somewhat unexpected since pentatonic scales were seen in the majority of Plains flageolet melodies and they are also the most unusual scales for vocal music of the central and southern Plains ([Nettl 1954], pages 27 & 29).

The melodic movement of vocal love songs of the Plains area exhibits two main patterns. The more prominent iss an undulating line (29, 30, 32, 33, 34, 34) which, in several instances (30, 32, 33, 34, 35) starts with a leap of a fourth, fifth, or octave to the song's highest tone before beginning an undulating descent. It will be recalled that this initial leap and gradual descent characteristic of flageolet Melodies from the Western Great Lakes and Plains area. The second type of melodic movement is ‘terracing’, a pattern very typical of Plains music but which has been noticeably absent from instrumental and vocal love songs. Both Omaha pieces and the Sioux melody from Curtis exhibit a terrace type of descent.

The kinds of intervals used in vocal love songs of the Plains do not present a uniform picture. Four of the melodies (30, 31, 32, 33) contian wide leaps of fourths, fifths, and octaves and these usually occur at beginnings of sections, with step-wise movement predominant in the latter halves. This use of intervals is very similar to that seen in flageolet melodies of the Western Great Lakes and Plains areas. Compare:

Figure 1. Opening phrases of Menominee flageolet melody (2) -top- and Sioux vocal love song (33) -bottom

Other melodies employ numerous fourths and fifths (34, 35, C2), while only three songs (F4, F5, 29) consistently use small intervals of seconds and thirds, a feature which coincides with the general trend of Plains vocal music ([Nettl 1954], page 29).

Like Plains flageolet melodies, vocal love songs from this area exhibit fairly regular rhythmic patterns. Two songs (29 and 30) are the vocal equivalents of flageolet meodies 8 and 9 which, it was noted, had very regular meters and had probably been modified by European musical influences.⁸ Excluding these, four songs remain that have some regularity of rhythm: no 33 and C2 with an underlying duple meter and no. 31 and 32 with an underlying triple meter. Rhythmic stability in this group of Plains love songs is created in two ways; by a narrow distribution of durational values and through the use of isorhythmic material. For example, songs 31 and 33 are restricted to smaller note values which are used consistently throughout the pieces. Isorhythmic patterns, seen often in vocal love songs of the Western Great lakes area, occur in the Sioux melody (C2) and in one Omaha piece (F5):

Figure 2. Isorhythmic patterns in Sioux love song (C2) -top- and in Omaha love song (F5) -bottom.

⁸ Those melodies having instrumental and vocal equivalents will be discussed in the following chapter.

Only three songs are rhythmically free (34, 35, F4). In general, vocal love songs fron the Plains are rhythmically similar not only to Plains flageolet melodies but also to vocal love songs of the Western Great Lakes.

Most often vocal love songs are sung without accompaniment, but in this group of Plains songs two Sioux melodies (32, 34) have a light drum accompaniment. In song no. 32 the beat changes from to . early in the piece and coincides with the pulse of the melody occasionally. The other love song (34) has an accompaniment which is a steady single pulse () but which does not coinside with the rhythm of the melody. As is typical of instrumental and vocal love songs, the tempi of these Plains melodies are consistently slow to andante. (M.M. = 68–80).

Recalling the structure of Plains flageolet melodies, it was found that those pieces were most often reverting (five examples) and only occasionally progressive or iterative (two examples each). The same distribution of form types has been found to occur in their vocal counterparts; i.e. five pieces have a reverting structure (F4, 29, 32, 33, C2), three are progressive (30, 31, 34) and two, iterative (35, F5). It is noticable that 90% of the Plains vocal love songs in this sample are clearly demarcated into major sections by long-held notes (29, 30, 31, 32, 33, 34, 35, C2, F4). In all but two pieces (29, 31) this note is the tonic or base tone, creating both a structural and tonal clarity in these melodies. It will be remembered that this consistent use of long-held tonic notes was a striking feature of Western Great Lakes and Plains flageolet melodies, and to find it recurring in Plains vocal love songs strongly suggests a close connection between the instrumental and vocal forms. Because the tonic of these melodies is heard so often in prominent positions such as phrase and section endings, tonality in this music is usually very clear. As in Plains flageolet melodies, fifths and octaves above the tonic are important tones of the scale.

The singing style and vocal quality employed in Plains love songs compare close1y to the style of performance of their instrumental counterparts but, because of this, differ markedly from the preceding group of Chippewa and Menominee love songs. It was noted that Plains flageolet melodies were played in a strong, bold manner. Long-held tones were performed with an intense vibrato and quick, sharp grace notes heard an octave above were common. In Plains vocal love songs this same kind of tension occurs in the voice and is indicated by pulsations on longer tones, grace notes, glissandi, rising attacks,and shaprly rising releases at phrase endings. Songs 31, 32, and 34 are excellent examples of this tense singing style. Pulsation and vocal tension are characteriatics of Plains singing style ([Nettl 1954], page 30) but are noticeably absent in the love songs of the Chippewa and Menominee. Only the Kiowa melody (35) resembles the Western Great Lakes style of singing in that it has a drawling, nasal quality and downward glissandi are common. No grace notes or rising releases; i.e., features indicative of vocal tension, are heard. In these Plains love songs, as in Western Great Lakes melodies, a remnant of the structure of flageolet melodies is seen in the brief sharp calls occurring at the end of several pieces (30, 31, 32, 34).

To summarize briefly, Plains vocal love songs resemble their instrmental counterparts in a majority of their characteristics but consequently differ strongly from the typical vocal love song of the Western Great Lakes.

3.3 — Plateau

In the preceeding chapter on flageolet melodies it was noted that musical examples from the Plateau area were relatively few. While a larger sampling of eleven vocal love songs is available for discussion, here, the source of this music remains restricted as before to only two groups, the Flathead and the Nez Perce Indians.⁹ From the discussion of flageolet melodies, it was found that pieces from the Plateau bore little resemblance to those of the Western Great Lakes or Plains, having features that were unique to themselves. It will be shown that many of these characteristics are repeated in their vocal counterparts but that, at the same time, both the melodic line and the singing style are reminiscent of vocal songs from the Western Great Lakes area.

⁹ Two Flathead vocal love songs (36, 37) have been transcribed from Alan P. Merriam's recording, Songs and Dances of the Flathead Indians; eight additional love songs appear on pages 188–191 in his Ethnomusicology of the Flathead Indians; and C3 is a Nez Perce melody taken from Curtis, The North American Indian, volume 8, pages 184–185. See transcriptions for 36, 37, and C3 on pages 136–140 of the PDF version of this thesis.

The tonal ranges of the Plateau love songs fall into two groupings: those with a narrow range of 7–9 semitones (or approximately one-half octave) and those of 12–17 semitones (an octave or more). The scales of these melodies vary widely from tetratonic to heptatonic; howe3ver, four of the eleven songs are pentatonic and in this respect resemble Flathead flageolet melodies. Three of these scales have a tendency towards chromaticism (e.g., 36) and the consistent use of flissandi between notes of narrow interval enhances this feature.

All of the vocal love songs in this group have generally undulating melodic lines. Seven of these incorporate a downward trend (these also tend to be songs with ranges of an octave or more) while the four remaining pieces begin and end at about the same level. It will be recalled that the melodic movement of flageolet melodies from the Plateau and of the vocal love songs of the Chippewa and Menominee is very similar. No examples of the more pronounced descending pattern typical of Plains instrumental and vocal love songs occur. The intervals use in these songs are generally small; fourths are the largets intervals employed, (e.g., 36). Again, the vocal love songs and their instrumental counterparts are alike.

Like the majority of love songs both instrumental and vocal, the rhythm of these Flathead melodies is very free. The distribution of durational values is wide and, in the absence of a regular drum-beat, in part creates the unmetered rhythms. Song 36, for example, repeatedly employs thirty-seconds and whole tones in the same short phrase which is freely sung as a melisma on the vocable ‘he.’

Figure 3. Flathead vocal love song (26)

A slow tempo is typical of all Flathead love songs and is characteristic of instrumental and vocal melodies from all the areas examined.

In terms of formal structure, six pieces are progressive and five are reverting (36, 37). More than one half of the Flathead melodies show variants of incomplete repetition. The Nez Perce love song (C3) is an example, unusual in love songs, of a very long piece which is essentially through-composed. There are three sections within the song that can be identified as variations of each other but, unlike most other pieces, no clearly defined sections are observed.

Flathead vocal love songs, like their instrumental versions, do not have long-held tones that provide distinct introductions, mid-section, or endings. In this respect, they are similar to vocal love songs of the Chippewa but differ from both the instrumental and vocal melodies of the Plains.

Tonality is fairly clear in this group fo Flathead love songs even in those that are heavily chromatic; e.g., 36. In song 37, d' is a rather weak tonic since it is not the most prominent tone and occurs only infrequently at phrase endings.

As pointed out previously, it is the Flathead singing style that connects it to vocal love songs of the Western Great Lakes. The drawling vocal quality, glissandi, and falling releases so distinctive of those songs are heard again in the Plateau area.

3.4 — Southwest

Vocal love songs from the Apache, like their instrumental versions, are rare in comparison to musical examples from other areas. Whether this is a modern development or was always the case is unknown. The one song available is in the typical vocal style of the area ([Nettl 1954], pages 22–23) but bears little resemblance to the Apache flageolet melody discussed earlier. Many features, however, are similar to vocal melodies of the Plains.

The range of the Apache love song is one and a half octaves and the scale is tritonic. Melodic movement is in the shape of wide, undulating arcs which cover the full range of the scng. The initial leaps up to the highest point of the melody as well as the consistent use of larger intervals of forths, fifths, and octaves are reminiscent of both the instrumental and vocal love song of the Plains. Like Plains vocal melodies, this Apache song has a fairly regular rhythm, created by the repetition of one rhythmic figure, : however, a rhythmically complex sound results from the alteration of accent on the two beats of this figure. The tempo of this piece is relatively fast in comparison to love songs from other areas (M.M. = 140 instead of an average = 72).

Structurally this melody consists of one phrase repeated several times: AA' AA' / AAA' AA'; the same formal structure as seen in the Apache flageolet melody. A short phrase on the vocables “nya ya ya ηa,” or its variants, functions as an introduction and marks the mid-section and end of the piece. Tonality is very clear since the piece is based on the triad, g-b-d.

A fair amount of vocal tension, seen earlier in Plains love songs, is used in performance of this song. Pulsations occur consistently on the highest tone of the piece, accents are strong and sharp, and glissandi are heard between notes of large interval. The degree of vocal tension used is reflected in the tendency towards the sharpening of the high tones in the latter part of the song. Although it is difficult to conclude whether the vocal tension causes this sharpening, it has been observed that they often occur together.

3.5 — Summary of Charactgeristics of Vocal Love Songs

Drawing together the many characteristics of vocal love songs just discussed creates a picture of considerable complexity. Unlike flageolet melodies whose features combine to form a distinctive instrumental style and, therefore, present a fairly cohesive pattern, vocal love songs, precisely because they are sung, are influenced not only by their instrumental counterparts but also by the typical vocal style of their musical areas. The summary which follows will describe the dominant features of vocal love songs while attempting to correlate them with possible sources of influence. This summary is also provided in chart form on pp. 99-103.

The average range for vocal love songs is 13 to 15 semitones, or about one octave. This is comparable to the ranges found in flageolet melodies of the Western Great Lakes, Plains, and Plateau areas and, therefore, strongly suggests the influence of the instrumental song on the vocal. In addition, however, there are a number of vocal love songs from the Western Great Lakes and Plains areas which have unusually large ranges of 23 to 27 semitones (approximately two octaves) and in this they are similar to vocal songs of the Western Great Lakes region which typically have very large ranges ([Nettl 1954], page 25). Some vocal love songs of the Plateau, which do not fit either of these patterns, have a consistently smaller range of seven to nine semitones; while the Apache love song, sole representative of the Southwest, has a range in keeping with the typical vocal song style of its area ([Nettl 1954], page 22).

In terms of scale, vocal love songs from each of the four musical areas correspond closely to the most commonly used scale of flageolet melodies from their area. That this scale is also the most frequently used in typical vocal music of each area makes it impossible to indicate the source of influence. Thus the pentatonic scale is the most important in vocal love songs from the Western Great Lakes and Plateau, and the tetratonic for sonss from the Plains. The tritonic scale, occurring in the Apache love song, is typical of vocal music from the Southwest ([Nettl 1954], page 22).

An undulating melodic line with gradual descent is common to most of the vocal love songs from the Western Great Lakes, Plains, and Plateau areas and can be considered a distinctive characteristic of this type of song. The initial octave leap and straight descent typical of Western Great Lakes and Plains flageolet melodies does not occur in their vocal counterparts, although some large intervallic beginnings are seen in Plains vocal love songs. Terracing, which was notably absent in flageolet melodies, occurs in a few Plains vocal love songs. This indicates an influence from the typical vocal music of the Western Great Lakes and Plains where melodic movement is almost always of the terrace-type ([Nettl 1954], pages 25 & 27).

The intervals used in vocal love songs are smaller than those common to flageolet melodies. In the latter, wide intervals of a fourth, fifth, and octave are prominent and reflected ian instrumental style, whereas in vocal love songs step-wise movement predominates and melodic lines are generally smoother. In this sense, vocal love songs show a closer relationship to the typical vocal music of their respective areas. However, two exceptions to the use of small intervals in vocal love songs occur: the first in songs fromt he Plains, where some melodies using wide intervals are similar to their instrumental counterparts, and the other in the Apache melody which, because of its triadic scale, consistently makes use of wide intervals.

Unlike flageolet melodies which very frequently have a free rhythm, vocal love songs display more metrical regularity. An underlying duple meter is common in vocal love songs from the Western Great Lakes and Plains areas. The use of isorhythmic material, seen often in the typical vocal music of the Western Great Lakes but not in the Plains area ([Nettl 1954], page 26), also occurs in some vocal love songs from these two areas. As exceptions to the general rule that flageolet melodies are rhythmically free and vocal love songs more regular, it should be noted that Plains flageolet melodies tend toward rhythmic regularity and the vocal love songs of the Plateau are free.

Related to rhythm is the distribution of durational values. In general those vocal love songs that are restricted to a narrow range of note values (i.e., from the Western Great Lakes, Plains, and Southwest) have a more regular rhythm and those with a wide range are more likely to be rhythmically free (Plateau). Dotted rhythrms are seen in both the vocal love songs and typical vocal music of the Western Great Lakes area. With regard to the distribution of durational values, vocal love songs appear to correspond more closely to the typical vocal music of their respective areas than to flageolet melodies which, it will be recalled, have a predominance of long-held tones that function as introductions, section dividers, and endings. A slow tempo is one characteristlc which is uniform for both instrumental and vocal love songs and this feature sets the genre apart from most other Induan music.

In terms of formal structure, vocal love songs show similarity to their respective instrumental counterparts rather than to the typical vocal music of each area. Thus, instrumental and vocal love songs from the Western Great Lakes are most often iterative, while typical vocal music of the area is progressive; Plains instrumental and vocal love songs are reverting and, in this case, the typical vocal music is as well; and finally, both the Apache flageolet melody and the vocal love song are iterative. The only exception occurs in the Plateau area where the majority of vocal love songs are progressive and the flageolet melodies, iterative. Stengthening the formal similarity between instrumental and vocal love songs, some vocal love songs; i.e., from the Plains area and Southwest, regain, but to a lesser extend, the long-held tones typical of flageolet melodies that function as introductions, structural dividers, and endings.

As in flageolet melodies, the tonaity of vocal love songs is generally strong and unambiguous. Fourths, fifths, and octaves are still important tones of the scale but are not as prominent as in flageolet melodies.

Finally, a discussion of the vocal quality with which love songs are rendered shows two clearly distinct types. The first is a nasal, drawling vocal technique used by the Indians of the Western Great Lakes area to imitate the sound quality of the flute. Rising glissando attacks, glissandi between wider intervals, and rising and falling releases are common to both instrumental and vocal love songs in this area. This style of singing is also heard in the performance of love songs in the Plateau area, although it is not known whether this is conciously in imitation of the flageolet.

The second vocal style used in singing love songs is found on the Plains and in the Southwest. In most ways it compares to the typical singing style of these areas with its characteristic use of pulsation and vocal tension. This vocal tension has, as an instrumental counterpart, the intense vibrato tone with which the long-held tonic notes of flageolet melodies are played. In both instrumental and vocal performances, this tension creates grace notes, sharp accented tones, glissandi between notes, and the sharp rising release of tones. While the first singing style described is said to be in imitation of the flageolet, it is very doubtful that the performance style of Plains flageolet melodies has influenced vocal technique since the vocal style of this area is exremely pervasive and is heard in the performance of almost all songs.

Editor's Note: The following table depicts the information in the tables on pages 99–103 of the original thesis. However, some of the relationships between elements in neighboring columns (represented by double-headed arrows) could not be represented on this web page. Please see pages 99–103 of the PDF version of this thesis.

Transcriptions

Chapter 4 — Love Songs with Instrumental and Vocal Versions

In the precceding two chapters, reference was occasionally made to melodies which had both instrumental and vocal versions. Although the close relationship between flageolet melodies and vocal love songs has already been established and discussed in some detail, a study of these ‘pairs’ of melodies would not only further reinforce this connection but wuld also be informative in determining what featurss combine to create the ‘vocal’ or the ‘instrumental’ style of the same piece.

Four pairs of songs are available for this type of comparison:

• Chippewa flageolet melody (1) and Chippewa love song (27)
• Sioux flageolet melody (8) and Sioux love song (29)
• Sioux flageolet melody (9) and Sioux love song (30)
• Kiowa flageolet melociy (15) and Kiowa love song (35)

Other writers ([Kurath 1956]; [Densmore 1932a]) have mentioned that, after playing a flute melody, the performer would also give a vocal rendition of the same piece, but in neither case has the music been published. All four sets of songs are performed by men since, it will be recalled, only men played the flageolet even through it was permissable for both men and women to sing love songs.

When transferred form instrument to voice, the tonal material of all four pairs of melodies remains essentially unchanged. Thus, the scales of the Kiowa and Sioux melodies are unaltered while, in the case of the Chippewa melody, the vocal version is slightly simplified by a reduction of its scale from heptatonic to pentatonic. The Kiowa and both Sioux melodies retain their ranges of one octave but the high coda endings of the instrumental meodies (8 & 9) are omitted in their vocal versions (29 & 30). The range of the Chippewa vocal melody (27) is slightly expanded from a major seventh of the instrumental version to a major ninth. This increase might possibly be due to the general influence of vocal music from the Western Great Lakes area where songs tend to have larger ranges than average ([Nettl 1954], page 25).

The change from instrumental to vocal idiom is most clearly seen in terms of altered melodic lines. In all four instances the melody has been ‘smoothed out’ and made less elaborate in the vocal versions. Melodically non-essential leaps, turns, grace notes, and other ornamental devices which are typical of flageolet melodies are lacking in their vocal versions. The texts of all four vocal songs are related to their melodic lines in simple syllabic fashion. The following examples illustrate the degree of simplification that occurs.

Figure 1. Opening phrase of Kiowa instrumental melody (15) -top- and Kiowa vocal melody (35) -bottom

Figure 2. Kiowa instrumental melody (15) -top- and Kiowa vocal melody (35) -bottom

Closely related to the smoother melodic line of the vocal songs is a decrease in the use of wide intervals. Although the leaps of a fifth (Chippewa 1, 27), sixth (Sioux 8, 29) and octave (Sioux 9, 30) which occur at the beginning of the pieces are retained, internally the larger intervals have been deleted. The following example shows how the opening phrases of the instrumental version of Sioux melody (9) have been altered in the vocal song (30).

Figure 3. Sioux flageolet melody (9) -top- and Sioux vocal melody (38) -bottom

With regard to rhythm and meter, all four pairs show very little change. Only in the Chippewa pieces is there an alteration of rhythmic figure.

Figure 4. Opening phrase of Chippewa flageolet melody (1) -top- and Chippewa vocal melody (27) -bottom

The distribution of durationa1 values becomes somewhat narrower in the transfer from instrumental to vocal idiom due to the elimination of the quicker, more ornamental figures. For example, a figure such as (1) is reduced to (27) and (9) becomes (30). There is no significant alteration in tempo between instrumental and vocal versions. In two cases, Sioux melodies (8 & 29) and the Kiowa pieces (15 & 35), the tempo is unchanged: the vocal version of Chippewa melody (1 & 27) become moderately faster ( = 80 increases to = 116) and the Sioux vocal song (30) somewhat slower ( = 138 down to = 126).

In terms of formal structure, the instrumental melodics undergo a small degree of alteration in their transfer to the voice. For example, the Chippewa instruemtnal piece is based on one section, made up of three distinct phrases, which repeats three times. The vocal version expands this basic section into four phrases, the fourth one being simply a repitition of the third with the final note altered. This enlarged section is then repeated in the same way as its intrumental counterpart. No new material is introduced in the vocal version of the Kiowa pair (15 & 35) but different repetitions and some variation of existing phrases are made. Thus the form of the instrumental melody, AABB', becomes AABA₄A₄. The long-held tonic note of the flageolet melody which functions as introduction, mid-section divider, and ending is retained in the vocal version in the form of repeated notes. In the pair of Sioux melodies (8 & 29) the introduction and coda, which are typical of melodies but not as common in vocal love songs, have been deleted from the vocal version.

To summarize briefly, it is the melodic line which receives the greatest alteration when a piece is transferred from the instrumental to vocal idiom. There is a general trend towards simplification, with a deletion of melodically non-essential leaps and ornamental devices, but even with these changes, the same melodic outline is clearly discernible in both instrumental and vocal versions. Formal structure is changed to a lesser degree, while other features such as scale, range, tempo, and rhythm remain essentially the same.

Conclusion

The initial impetus for this study of instrumental and vocal love songs derived from a number of questions raised by George Herzog in an article, “Special Song Types in North American Indian Music,” written in 1935. In his paper, Herzog pointed out that “it is not uncommcn to find in the possession of a single group a number of styles, represented in different categories of songs; specific styles that do not seem to have any organic reason for co-existing” ([Herzog 1935a], page 24). Hypothesizing that musical features which perhaps originated in one restricted locality gradually spread, through trade, warfare, and social interaction, to other areas, he then distinguished four types of songs, each of which had a special function, that could be considered intrusions into the existent musical repertoire of a given tribe. One of these was the love song.

In this paper it has been shown that the love song derived, in all probability, from flageolet melodies which had long been used for courting. By first tracing the diffusion of the flageolet from its likely origin in Mexico ([Galpin 1903], page 135; [Sachs 1929], page 214; [Roberts-HH 1936], pages 20 & 25) northward through the continental U.S.A. into roughly the Western Great Lakes, Plains, Plateau, and Southwest areas, it was also possible to delimit the distribution of this special type of vocal love song.

Flageolet melodies from all four areas were then analyzed in an attempt to distill their major charactaristics and determine to what extent this music exhibited a homogeneous style regardless of the musical area in which it occurred. Rrom a study of flageolet construction it was found that the instrument was most commonly made of wood, was approximately 20–21 inches long, 1½ inches in diameter, and had six finger-holes. This standard construction, although not rigid in all details, created an instrument which most often produced melodies with pentatonic scales of about one octave's range. No one melodic pattern emerged as distinctive, although the use of many large intervals, especially octaves, fourths and fifths, within a melodic line was considered to be idiomatic of flageolet songs. Flageolet music was also strongly characterized by a very free, unmetered rhythm and a slow tempo which, together, created a spacious and rhapsodic quality common to the majority of instrumental love songs. Formally, the use of long-held tonic notes, played with a characteristic intense vibrato, as introductions, internal dividers, and endings was also seen as distinctive of flageolet melodies. In addition to this vibrato technique, an idiomatic flute style was created through the use of a number of ornamental devices such as overblown grace notes, glissandi, turns, and trills.

Having estabished the existence of a typical flageolet style, it was then possihle to compare its features to those derived from a similar analysis of vocal love songs. Proceeding from the supposition that love songs were a result of a transfer from flageolet to voice, certain features of the instrumental melodies were expected to recur in the vocal songs. Although a direct correlation of all features could not be shown, possibly because vocal love songs had also been influenced by the typical vocal style of their musical areas, several significant characteristics were found which definitely indicated their close connection.

Because the vocal melodies were not restricted to the standardizing effect of an instrmnent, their ranges and scales showed more variation and in some cases resembled the typical vocal music of their area. Remnants of the melodic line of flageolet pieces with their wide intervals were retained in many vocal songs from the Plains and Southwest. The free, unmetered rhythms and slow tempo which had been prominent in flageolet melodies was fuund to be one characteristic which set the genre of love songs apart from most other Indian music. The distinctive use of long-held tones as introductions and endings in flageolet melodies also recurred in the vocal love songs and short calls heard at the end of many vocal songs were reminiscent of the sharp grace-note endings of flageolet melodies.

While all of these characteristics revealed important similarities between vocal love songs and their instrumental counterparts, it was in two significant features, vocal quality and manner of performance, that their close connection became clearly evident. The most striking feature of vocal technique was the nasal, drawling tone with which love songs from the Western Great Lakes and Plateau areas were rendered. Several sources ([Fletcher 1893], page 11; [Densmore 1932a], page 208; [Herzog 1935a], page 28) maintain that this technique was meant to imitate the sound quality of the flute. To further enhance this imitation Indian singers would sometimes wave their hand slowly in front of their mouth to interrupt the flow of breath and produce soft pulsations of tones. The unique manner of voicing of love songs also demonstrated their similarity to flageolet melodies. The prominence of glissandi at phrase beginnings and endings and between wider intervals as one distinctive feature of flageolet melodies which had been taken over into vocal love songs. On the other hand, vocal love songs from the Plains and Southwest were performed in a manner typical of most of the vocal music of those areas. Thus, vocal tension, pulsating tones, and a large number of grace notes were common in these love songs, quite in contrast to the vocal technique of the Western Great Lakes and Plateau.

Having isolated and analyzed numerous technical features of both instrumental and vocal love songs from several groups that are widely separated both culturally and geographically, do these individual parts synthesize into a unique and separate genre? Can a conclusive statement be made regarding the existence of a musical style which cuts across usual musical and cultural boundaries? With very little qualification it can be said that the features of flageolet melodies combine to form a distinctive instrumental style which is basically the same throughout all the areas discussed. The question of one distinct style for vocal love songs is, however, more complex since the features of these songs, when examined individually, reveal influences from both their instrumental counterparts and from the typical vocal style of their musical areas. This picture is further complicated by the coincidental rise of vocal love songs as a favourite form of expression and the influence of European culture. From this intricate web of influence and cross-relations, this study has attempted to extract enough evidence to further validate Herzog's theory of special song types by showing that vocal love songs are, indeed, linked in numerous and significant ways to their instrumental cuunterparts and that together they form a recognizably separate and unique genre.

List of Songs on Accompanying Tape

Tbe melodies on the accoonpanying tape have been grouped into three categories: 1. flageolet melodies, 2. vocal melodles, and 3. four pairs of melodies which have both instrumental and vocal versions. They are heard in the following sequence:

1. Winnebago flageolet melody (6) — Western Great Lakes
2. Meskwaki flageolet melody (7) — Western Great Lakes
3. Sioux flageolet melody (9) — Plains
4. Kiowa flageolet melody (16) — Plains ([Cozad-E 1964])
5. Flathead flageolet melody (17) — Plateau
6. Apache flageolet melody (19) — Southwest
7. Chippewa vocal love song (20) — Western Great Lakes
8. Chippewa vocal love song ( 22) — Western Great Lakes
9. Chippewa vocal love song (23) — Western Great Lakes
10. Sioux vocal love song (32) — Plains
11. Kiowa vocal love song (35) — Plains
12. Flathead vocal love song (36) — Plateau
13. Apache vocal love song (38) — Southwest
14. Chippewa flageolet melody (1)
15. Chippewa vocal love song (27)
16. Sioux flageolet melody (8)
17. Sioux vocal love song (29)
18. Sioux flageolet melody (9)
19. Sioux vocal love song (30)
20. Kiowa flageolet melody (15)
21. Kiowa vocal love song (35)