Centres of Polyphony

East Asia

Japan: Ainus
Arguably the most isolated tradition of vocal polyphony in the world is in eastern Asia, among the Ainus from the northern Japan and Kuril islands. Ainu ethnic, linguistic and even racial origins are much debated. The first Europeans who met the Ainus wrote that they looked like Finns. Later scholarly research confirmed the unique features of Ainu language and race. As arguably the first dwellers of this region (Sakhalin, Kuril Islands and particularly Hokkaido in Japan), the Ainus constitute the substrata of the Japanese.
About 20.000 Ainu live in northern Japan (mostly on Hokkaido) today. Only about one hundred of them can actually speak the Ainu language. After a five-year research program, conducted by the Japan Broadcasting Association in the beginning of the 1960s, about two thousand Ainu traditional songs were recorded. A 600-page volume “Music of the Ainu” containing about 500 transcribed pieces of music was published in Japanese with a small English summary (Kazuyuki, 1965).
The traditional polyphony of Ainu is based on canonic imitation of relatively short musical phrases. According to Kazuyuki (1965, 1975), out of thirteen different genres two are the most important: upopo (round, sung in a canon by elders sitting in a circle), and rimse (round dance, only very rarely sung as a canon). There is an interesting and complex interaction between these two genres: “The way in which these songs are handed down is in disorder today, and many songs are sung equally as upopo and rimse. Even in such instances, however, rimse is rarely sung in canon. Imitation is a significant criterion for judging rimse and upopo. The latter is sung imitatively by several people seated round the lid of a chest, tapping the rhythm on the lid. Usually the lead is taken by the oldest member of the group. The leader turns his face towards a man sitting at his right, and on a cue this man starts singing a beat behind. This continues until the last man sings while the rest go on singing. Thus the whole sound becomes chaotic and the sounds comply with the etymology of the word ‘upopo’, noisy singing like birds twittering. The whole process is repeated several times, and then the leader starts another song in like manner. The falsetto (pon) singing at the beginning of the song is very distinctive, and its effect is like lighting candles one after another in the dark. Upopo and rimse were originally the same as their etymology suggests; they only gradually achieved separate identities. The rimse became a dancing song, and the upopo became an imitation sitting song” (Kazuyuki, 1975:63).  Canonic singing can vary from two parts up to six parts. Here is an example of a six-part canon:
Precise canonic singing is the best known and the most characteristic of Ainu polyphony. As a polyphonic culture, mostly based on canonic singing, the Ainu tradition is unique among the polyphonic cultures. The only other polyphonic tradition that uses canonic singing so widely is the Lithuanian sutartines from the Baltic region. Unlike Lithuanian sutartines, where you hear two-part singing, in Ainu upopo you can hear five- and six-part canonic singing.
Some songs have elements of drone polyphony, as in the following example, which starts as a loose canon and grows into drone two-part singing with dissonant secondal intervals:
Secondal dissonances are even more obvious in the following example, where two women sing in two parts (this is my transcription from the accompanying cassette, from Kazuyuki, 1965):
Numerous influences between Ainu and neighbouring musical cultures had been noted (including Japanese, eastern Siberian peoples, peoples of Kuril Islands, Kamchatka and Sakhalin. Kazuyuki, 1975:66), although the tradition of vocal polyphony seems to be a uniquely Ainu phenomenon in this vast region.

Not much is known about the other (southern) end of Japan – the Ryukyu Islands. According to the available information, group antiphonal singing (uta kake) and call-and-response forms are quite usual (Atumi, 2002:791). Particularly refined and rich are the antiphonal singing traditions on the Amami Islands (between the Japanese mainland and Okinawa). Atumi notes the closeness of uta kake and the antiphonal singing of southern Chinese minorities (the region known for its vocal polyphonic traditions – see below).

Alan Lomax united China with most of the countries from the Middle East to East Asia under the collective name “Old High Cultures” (Lomax, 1968). These “Middle East – East Asian Monophonic Belt” cultures are based on sophisticated vocal monophony, deep traditions of professionalism, highly developed musical instruments, instrumental ensembles of ancient origins, and sophisticated theoretical works about music. Despite this closeness, there are a few extremely important differences within this group. Most importantly, the scale systems (as I mentioned earlier, the “DNA of musical cultures”) are totally different. East Asian musical cultures are mostly based on pentatonic anhemitonic scales (scales without the semitone, the smallest interval for tempered music, although pentatonic with semitones are also present, most notably in Japanese music), while Middle Eastern music is known for the extensive use of semitones and even smaller intervals. The use of specific melismatic embellishments in Middle Eastern music is another important element that is absent in East Asian music. Even without mentioning other differences between Middle Eastern and East Asian musical traditions, it is clear that the group of “Old High Cultures” consist of at least two principally different groups of cultures – (1) Middle Eastern group of cultures (from North Africa to North India and Tajikistan) and (2) East Asian groups of cultures (China, Japan, Korea and part of the Southeast Asian cultures). And still, with regard to dividing the world’s musical cultures into monophonic and polyphonic groups, Lomax’s suggestion of uniting these cultures in a group “Old High Cultures” is justified.
China, geographically one of the biggest and the most populous country of the world, is home to a large number of ethnic minorities. The majority of Chinese are Han. According to the available information, although mainstream Chinese (Han) folk music is generally monophonic, interesting forms of group singing in call-and-response forms are an integral part of their traditional singing. For example, work songs – haozi, a loud outdoor genre accompanying different types of work – are performed by a leader’s call and group response. Another group song is tian’ge – a responsorial harvest song from southern China, connected to the different processes of working on rice fields. Tian’ge helped to maintain the working rhythm and kept the workers’ spirits up. Sometimes workers would just listen to the specially invited professional singer while working, or the “landowner would arrange for two competing groups of singers, who would take turns, trying to cap each other” (Jianzhong, 2002:153). According to available information, group singing among the Han is always monophonic (unison, or social polyphony only. For the elements of heteropony in Han classical and folk music see Mok, 1966). Therefore, China is generally known among ethnomusicologists as a classical country of vocal monophony.

Now if we look at the musical traditions of Chinese minorities, we will find plenty of traditions of vocal polyphony that are mostly unknown to western scholars. Almost half of the Chinese minorities have been documented to have vocal polyphonic traditions. If we take into account that the number of Chinese minorities is over fifty, and that their combined population numbers around 100 million, we may get an approximate idea of how rich (and how unknown to western readers) these polyphonic traditions can be.
Let us listen on this matter to one of China’s leading ethnomusicologists, Mao Jizeng, from the Central Institute for Nationalities from Beijing: “Although some foreign scholars believe that China has only monophonic folk song, in fact more than twenty minority peoples in China have polyphonic songs. [Another Chinese scholar, Qia estimates a more precise number – 25 minorities, See later. J.J.] Some folk singers have a saying that indicates a deep-seated tradition of polyphony: ‘A lamp has two wicks, because if there’s only one the light isn’t bright enough.’ Polyphonic songs can be divided broadly into three groups: (1) part singing, found among the Zhuang, Yao, and Miao; (2) songs in which a continuous bass line with a fixed melody is combined with the main melody, found among the Dong and Mulao; (3) folk songs in canon or round form, such as the songs of the She. Looked at as a whole, the forms of the various musical lines are basically integrated, and their melodic movements, rhythms, supporting notes, and finals are the same … The frequent appearance and prominent use of major seconds and the special progression of major seconds resolving to unisons are important characteristics of some polyphonic folk song” (Jizeng, 2002:449).
A very detailed and informative special section on multi-part music among Chinese minorities has been written by Shen Qia, another leading Chinese ethnomusicologist from the China Conservatory of Music, Beijing, for the Garland Encyclopedia of World Music. Here is most of the text of this very important section for our topic: “As far as we know at present, apart from the Han, twenty-five minorities have multi-part music. These include the Zhuang, Buyi, Dong, Maonan, Mulan, Dai and Wa of the Yue-Pu system; the Miao, Yao, and She of the Miao-Yao system; and the Qiang, Yi, Hani, Lisu, Naxi, Bai, Lahu, Iinuo, Tujia, and Jingpo of the Di-Quang system. Eighty percent of them are distributed within Guangxi, Guizhou, and Yunnan regions; the others are found in the Aba district of Sichuan, in southwest Hubei, western Hunan, and the Ningde region of northeast Fujian. Following are some important forms of multi-part music.
“Galao. The dong form galao ‘ancient and solemn suite’ is sung in a southern Dong dialect and flourishes in the countries of Liping, Congjiang, Rongjiang, and Sanjiang in the southeast Guizhou.
“Wennai. This is a generic name given by the Buyi to their multi-part folk songs. The Buyi wennai are often drinking songs for wedding celebrations. There are also some wennai called “double songs”. Wennai are mainly found in Libo, Sandu, and Dushan in south Guizhou.
“Huan, bi, xi, and liao. These are all names given to shan’ge folk songs in different Zhuang areas. These localized terms cover many different melody types, of which multi-part songs are only one:
“• Huan are found along the You, Hongshui, and Liu rivers in west and central Guangxi. They include many two-voice forms. Gumeihuan, named after Gumeitun, its place of origin, is now current in the countries of Tianyang, Debao, and Bose in western Guangxi. Huanleng is found in Shanglin in central Guangxi. Huanliu is named after the vocables used, and is found in Pingguo Country in west-central Guangxi. Huanya ‘song of the Buyayi’ is so named because of local people, the Buyayi branch of the Zhuang, call themselves the Ya; this last type of song is current in Tian’e, Guangxi. In the three-voice genre sandunhuan, current in Shanglin and Mashan countries in Central Guangxi, the middle voice is the principal one. The four-voice genre huanyue originated as a duo singing style in shamanic rituals; it too may be found in Shanglin and Mashan countries.
“• Bi are found mainly in northern and northwestern Guangxi. They are variable in form, and the melodies differ from place to place. Some forms of bi are multi-part. For example, in Luocheng there are binongnai, a women’s duo song; and binongniang, a men’s duo – both sung antiphonally. This country also has biyewan, sung at night; bijiang, a narrative form; biyou, a fast bi; biyan, a slow bi; bidan, a bi with regulated vocables; and bifuyin, bisangye, and bimaiwei¸ among other forms.
“• Xi are found in Fusui, Daxin, Ningming, Longzhou, Jingxi, Debao, Napo, Chongzuo, and Tiandeng in Southwest Guangxi. Xi include many melody types, in which some are multi-part shan’ge folk songs.
“• Liao are also called huanliao, from huan ‘song’ and liao, a vocable – thus this class of songs is named for its vocables. Huanliao can be divided into three categories: huanlei (long song), huanding (short song), and huanzhong (medium-length songs). They are found mainly in Tiandong and Pingguo in west-central Guangxi.
“These multi-part songs include some characterized by harmony and some characterized by polyphony. Polyphonic songs are the most common and may use heterophony, imitation, or nonimitative polyphony.
“Over the past decade or so, Chinese scholars have collected a wealth of multi-part music and researched it in depth. The accumulated material proves that the multi-part music discussed here is an ancient and indigenous form of folk song, owing nothing to the multi-part music introduced by Western missionaries in the nineteenth century and the early twentieth century” (Qia, 2002:489-490).
To this amazing variety and richness of traditional polyphony, present among Chinese minorities, we could add the relatively recent (1995) fieldwork discovery of a quite sensational polyphonic tradition, made by a Chinese scholar Zhang Xingrong from the Yunnan Art Institute. In the mountain village of Puchun (a village without electricity and roads) of the Hani people, located on the borders of Honghe, Luchum and Yuanyang counties, they recorded five-part a cappella “Bridal Laments” and eight-part accompanied “Rice Transplanting Songs”, where all the parts are performed by individual singers and singing is also accompanied by traditional instruments – an end-blown flute labi and three-stringed plucked lutes lahe.
Unfortunately almost no information is available on traditional polyphonic singing in Tibet, although scholars note the presence of the tradition of table songs performed in big communal groups, and rich traditions of antiphon singing in Tibetan dance-songs, including antiphony between soloist and group, or between the men’s and women’s groups (Jizeng, 2002:477, 483). On the other hand, the famous Tibetan monks’ chant is in fact a group singing (chanting), and should be considered among the different types of group singing, although it is hardly an ordinary type of vocal polyphony.
In his widely known book “Work and the Rhythm” Carl Bucher gives an interesting example of two-part drone polyphonic singing from Tibet. This is tongskad, traditional work song:

Taiwan is another very important region of traditional polyphony, and, as a matter of fact, one of the centres of study of vocal polyphony in Southeast Asia (I am referring to the fact that Taiwan hosted a special conference on traditional polyphonic singing in 2002. For more information about this and other conferences fully dedicated to the research of traditional polyphony see the special Appendix at the very end of this book).
Several indigenous peoples live in Taiwan: the Atayal, Amis, Bunun, Paiwan, Saisat, Tsou, Rukai, Puyuma, Shao. Several of them practice vocal polyphonic singing.
Atayal is one of the biggest aboriginal groups in Taiwan (numbering around 100 000). They mostly sing solo, although during the wedding ceremonies group singing, featuring two-and three-part canonic polyphony occurs.
Ami is the most numerous group in Taiwan (around 150 000 population). Their polyphony is mostly based on two-part singing, with the low part performed in unison (or solo) and the high part singing a wide-range pentatonic scale-based melody. Occasional three- and four-part sections also appear. Millions of western listeners have heard traditional Ami two-part polyphony without realizing they were listening to a Taiwanese traditional song, because when the British rock-band “Enigma” used the recording of Amis’ two-part polyphonic harvest song (recorded in Paris in 1988 from native Taiwanese singers) as the main melody for their 1994 worldwide hit “Return to Innocence”, they did not acknowledge the source of the song, singers, the original recording and even ethnic origin of the song.
Bunun are the most mountainous people on Taiwan, living in the central Taiwan mountain ranges. Their polyphony is the most “harmonic”. Drone has a prominent place in their singing. For example, in a song depicting the growth of the seed from the ground, Bunun singers use a unique gliding-up drone, starting very low and gradually going up, while other parts also go higher modulating, but without gliding. In Bunun songs you hear the sound of full chords most of the time. Drones are usually in the top or the middle of the texture. Harmony can be based around the major triad, or more adventurous sets of harmonies can be used as well, as in the next example (from the cassette, my own transcription), ominously called “Song of the Victory of the Head Hunters”. Here is the first half of the song:
The Paiwan live in the southern part of the island and number around 70 000. In their traditional singing they use two-part drone polyphony, and small-range melodies. Unlike Bunun, who uses drone on top of the multiphonic texture, the Paiwan use drone as the lowest part. Secondal dissonances are also used in Paiwan polyphonic songs.
The Rukai are close ethnically and geographically to Paiwan, and this is clear in their singing style as well, although Paiwan polyphonic singing seems to retain more traditional elements than Rukai polyphonic singing.
The Saisiat are one of the smallest groups numbering fewer than 10 000 people in Northwestern Taiwan. Their traditions are deeply influenced by the neighbouring Atayal, the most numerous and influential Taiwanese mountain people (see above). Regarding polyphonic singing, according to available data, they did have a tradition of vocal polyphony which was lost during the 20th century: “…at one time there was singing in parallel fourths, but no-one sings in this way anymore” (Tsang-houei, 2002:525).
The Tsou people are also very few (under 10 000). They live in the high mountains of the Central Ranges and are neighbors of the Bunun. Like the Bunun, the Tsou are known for their harmonic polyphonic singing, although two-part singing is prevalent here (unlike the three-part singing of the Bunun people). Unlike any other Taiwan peoples, who use only binary metres, the Tsou also use ternary metres (3/4 and 6/8).

South-East Asia
According to the available information, the traditional music of the Viets, the main ethnic group of Vietnam, has some interesting forms of group singing (particularly responsorial forms), although the group singing is based on unison (social polyphony). Therefore we can repeat what is generally known about Vietnamese traditional music, that it is based on a tradition of vocal monophony.
On the other hand, traditions of vocal polyphony are found in the singing traditions of several ethnic minorities of certain regions of Vietnam (Nam, 1988:41). More than 50 minorities live in Vietnam. Some of the minorities live in the lowlands, and others live in the uplands (mountain regions). As is often a case, the earliest population of Vietnam is preserved today as minorities in the mountain ranges (mostly in central Vietnam) (Nguyen, 2002:531). Another mountainous region of the country – the northern part – is mostly inhabited by groups that came from southern China between the 1300s and 1800s. The Vietnamese mountain peoples have different (from the Viets) origins, physical features, languages and musical cultures. Vocal forms of polyphony are documented among the central, and particularly the northern mountainous minorities. According to a very few available examples of Vietnam polyphony, two-part singing is prevalent, and heterophony and drone forms are present. Both of these examples are from Kao-bang province of northern Vietnam:
Unfortunately, the Vietnamese mountain minority’s traditions of vocal polyphony are almost unknown in the western ethnomusicology. Garland’s article on Vietnamese minorities does mention their traditions of group singing, but fails to provide information whether the singing is unison or polyphonic.

According to the available information, traditions of vocal polyphony similar to those of the Chinese and Vietnamese minorities are present in Nepal, and Burma (Myanmar) as well. Unfortunately, these polyphonic traditions have not been studied sufficiently, and the only information I have are works about southern Chinese and Taiwanese polyphonic traditions. They mention the typological closeness of Southeast Asian polyphonic traditions (mostly existing in the forest-covered mountain regions of Southeast Asia) (Tsang-houei, 2002:523).

India has one of the world’s biggest diversity of peoples of different physical types, languages, religious confessions and cultures. This applies to the musical traditions as well. Perhaps one of the reasons for the small national importance of folk music is the sheer diversity of the vocal and instrumental traditions of Indian peoples and ethnic groups, making it very hard for a folk tradition of any one region to become the national idiom of Indian music. Instead, contemporary Indian film music (the biggest film industry in the world) is recognized as the unifying national symbol of Indian popular musical culture. Songs from Indian films are actively put on cassettes and sold in their millions throughout the country. They create a background for social events and are learned and sung by villagers and city dwellers alike (Green, 2000:555).
Not all the traditions and musical styles of different regions and peoples of India have been studied sufficiently. The musical landscape of India is dominated by the magnificent North Indian (Hindustani) and South Indian (Karnatak) classical traditions. Although classical traditions are polyphonic, India is generally known as a country with monophonic vocal traditions, because in both North and South India this polyphony is instrumental (or vocal-instrumental). As a matter of fact it is quite amazing how Indian instrumental music is dominated by drone polyphony, the form of polyphony virtually absent in vocal group singing.
Dominated by the classical traditions of raga both in North and South India on one side, and the massive popularity of film music industry on the other, very little attention has been paid to the folk music of the different peoples of India. The accepted stereotype of Indian music as “monophonic culture” also contributed to the very limited attention of scholars to polyphonic traditions of the different Indian peoples.
The work of Kauffmann and Schneider (1960) about the polyphonic folk tradition of the Assamese people is the only major study of traditional vocal polyphony in India that I am aware of.
Polyphonic singing of the mountain-dwelling Assamese has understandable links with the polyphonic singing traditions of neighbouring mountain peoples of Southeastern Asia. We find here several different forms of part-singing, although the parallel movements of fourths, fifths and octaves seems to dominate:

Another region with documented traditions of vocal polyphony is southern India. Although cultures (and musical cultures) of different regions of southern India (Tamil Nadu, Kerala, Karnataka and Kota) are different, they are still united by a common Dravidian ancestry. According to Lomax, the singing traditions of the tribal communities of southern India contain different forms of polyphony (Lomax, 1968). Interestingly, all four articles dedicated to four main southern Indian regional musical traditions in The Garland Encyclopedia of World Music mention their strong traditions of group singing by both men and women (mostly in call-and-response form) and the popularity of circle dances, but they fail to mention whether they are monophonic or polyphonic. This must be the result of general neglect of scholars towards vocal polyphonic traditions in India. As a matter of fact, the word “polyphony” is absent in the impressive index of this 1000+-page volume dedicated to the music of Indian Subcontinent.

Share it on