The musical traditions of Native Americans are stereotypically regarded as monophonic. Despite the presence of strong group-singing traditions in most American Indian traditional music, monophonic (solo, unison and often loosely heterophonic) singing predominates.
At the same time, unlike many of the other “monophonic” regions of the world, where the study of dominant monodic music totally pushed away any study of the elements of polyphony, quite a few American scholars (mostly musicologists and ethnomusicologists) contributed to the study of elements of part-singing among North American Indians (see the survey in Nettl, 1961).
According to the available sources, mostly summarized in a concise and very informative article by Bruno Nettl (1961), information about the polyphonic singing styles of different Indian peoples is quite abundant. Although some regions and peoples lack any references to polyphony, solid information is available on other regions and peoples. Most importantly, this information questions the existing general stereotype about the general monophonic nature of American Indian vocal singing traditions.
Among different regions of North America two regions are particularly rich in information about vocal polyphony: (1) the Northwest Coast, and (2) the East Coast.
Plenty of information indicates that the Northwest Coast Indians (particularly the Nootka and Salish) were familiar with a part-singing tradition, and particularly often used drone polyphony. Drone could be the highest part (as among the Makah) as well as the lowest part (Salish Indians). Makah used a so-called “metal pitch”, a drone that sounded on top of the melody. According to Densmore’s informants, the “metal pitch” [high pitch drone] was sung by individuals, mainly women, who “either did not know the song or wished to improve its quality” (Densmore, 1939:130). It is very interesting that the Makah would use a “metal pitch” drone to accompany a stranger when he would sing his own song for them. Densmore provides some very interesting information about the additional harmonic tones that Salish singers used to place during the long notes in the melody (Densmore, 1943:31). Abraham and Hornbostel transcribed this example of drone polyphony from Thomson River Indians:
Singers were divided into two groups (so both parts were sung in unisons). Some of them sang a melody, and others occasionally repeated the same note, a third lower than the main tone of the melody (Abraham & Hornbostel, 1922:32). 
According to Roberts, Nootka had parallel three-part singing traditions, with men singing a melody and women singing together with them, some singing a fourth, and some an octave higher. They had special terms indicating the relationship of the parts. She indicates that Nootka have known polyphony for a long time. Polyphonic elements in the singing of the Nootka and Kwakiutl were also discussed by Ida Halpern at the “Centennial Workshop on Ethnomusicology” in 1967: “There are slight polyphonic tendencies noticeable. However, to speak of polyphony when there is a slight momentary discrepancy of pitch should not be considered true polyphony, but unintentional polyphony. Some of the Nootka songs, however, have a more truly polyphonic feeling than those of the Kwakiutl” (Halpern, 1975:25). Later she notes that polyphonic moments “always occurred the same way in the same song” (pg. 42).
According to Keeling vocal counterpoint is particularly characteristic for the Californian Indians: “Throughout this area the various tribes also perform secular dances such as Shakehead Dance or the Pomo Ball Dance, and these illustrate the basic style of group singing in the region. This music resembles the Northwestern style in that it is contrapuntal , but the nature of the counterpoint is very different. In north-central California, the solo part is augmented by another basically rhythmic part known as “the rock.” This is an accompaniment, sung by one person, and the singing itself is called “holding rock”. In addition to the soloist and the “rock” there is rhythmic beating of clapsticks (played by the singer) and often the sound of whistles blown by dancers, so that the whole musical texture is fairly complex” (Keeling 2001:415-416). “…The presence of vocal counterpoint as a basic element in public singing is one of the most distinctive characteristics. While certain types of polyphony have been noted among tribes of the Northwest Coast, true rhythmic counterpoint is much more prevalent in California than in other regions of North America. Contrapuntal styles have been described here among the Northwestern and north-central California nations, but other types of counterpoint are also evident in historical sources and archival recordings of music pertaining to certain earlier rituals of Southern California groups” (ibid :418).
Another important region for indigenous polyphonic traditions was the Eastern Coast. In the Northeast, drone was mostly used. Roberts mentions polyphonic songs with several singers singing a main melody and others carrying a drone (Roberts, 1936:7). Menomini also used a drone “to help the singers” (Densmore, 1939:26). In another interesting case, “ Female informant, wife of a medicine man, helped her husband to sing by standing beside him and singing a drone above his melody” (Nettl, 1961:359). Therefore, drone could be a top part as well. The Delaware and Fox Indians also used drone in their traditional songs (Baker, 1882:14).
Information about polyphony from Southwestern USA is more fragmentary and only mentions a couple of tribes from remote regions: the Yaqui women sang a very high drone (octave and fifths above the tonic) (Fillmore, 1889:306). The Papago women also used a drone above the melody “for the space of three or four measures” (Densmore, 1929:14). Call-and-response with the soloist and chorus “stepping” on each other (or overlapping) in a dance song of the Creek Indians has been eloquently described: “It is … difficult to divide between where the leader stops and the chorus comes in . . . The more animated the dance becomes, the more merged and rapid are the parts. The effect of this is . . . bordering almost on harmony” (Speck, 1911:162. Cited from Nettl, 1961:360). The Cherokee and Shawnee are also documented to have the same kind of responsorial singing with parts “stepping” on each other (Nettl, 1961:360).
Information on Plain Indian polyphony is also available. And again, we see the use of drone, sung by women of the Pawnee Indians from the northwest and south west of the region (Densmore, 1930:659). According to Helen Roberts, the Oglala Sioux sang in parallel thirds and simple imitation-like call-and-response where the chorus starts while the soloist is still singing. Her general remark is “choral singing was common among the Oglala Sioux” (Roberts, 1936:7). The Arapaho are also noted for such cadencial two-part singing.
Nettl also discusses the very interesting tradition of the use of the instrumental drone in this region. Drumming on the kettledrum, partly filled with water (and accordingly specially tuned) accompanies singing with a steady instrumental “rhythmic drone”. It is very interesting that according to Nettl (who transcribed such songs from the Arapaho Peyote), the kettledrum is tuned to either the tonic of the melody, or a fifth below the tonic (Nettl, 1961:358). Polyphonic elements in Peyote songs are also mentioned by Gooding (2001:445).
Summarizing the information on polyphony among North American Indians, Nettl writes that out of six musical areas, identified in his seminal work on Indian music (Nettl, 1954), “only two, the Great Basin and the Athabascan, one an exceedingly simple style and the other represented only by the Navaho and Apache, lack references to Polyphony” (Nettl, 1961:360). In the concluding part of the article Nettl discusses different possibilities of the meager distribution of vocal polyphonic traditions among North American Indians, and suggests two different models: (1) vocal polyphonic traditions among American Indians existed earlier much more widely, and the isolated regions with elements of part singing are only the remnants of this old tradition, or (2) by the time of the first contact with Europeans American Indians were “on the threshold of developing an elaborate polyphonic style” (Nettl, 1961:362). I do not want to discuss these very interesting historical issues in this part of the book, dedicated to a review of the information available from various sources. We will discuss this topic in the second part of the book, in a special section (or “case”), dedicated to the vocal polyphony of American Indians.