Literate Shamanism: The Priests Called Then among the Tày in Guangxi and Northern Vietnam

1. A Personal Odyssey

The immediate context of my interest in the practitioners called Then was a project to document the vernacular character scripts of the Tai-speaking peoples of northern Vietnam.1 My interest in the languages and folk religions of the Tày and Nùng in northern Vietnam began much earlier, in the early 1990s. At that time, I had already begun to collect relevant materials, including reports on the Then in Vietnamese. Then, in 2004, my long-time research associate Ling Shudong 凌樹東 visited a number of different localities in Cao Bằng province, across the border from his native county Jingxi 靖西, and collected texts from mogong 麽公in a number of different districts.2 These were later included in the Mapping the Old Zhuang Script survey volume. Ling’s explorations re-confirmed to me that there was a cultural and linguistic continuum on both sides of the China–Vietnam border, and that an understanding of the cultural geography of northern Vietnam could be of vital importance for improving our understanding of Zhuang society and culture.

With shamanism, my research interest goes back further. I conducted ten years of fieldwork on processional dances and theatre in North Shaanxi, beginning in 1981, and came across healing rituals performed by the shenhan 神漢, one of which I recorded in the Chenghuangmiao 城隍廟in Yulin 榆林.3 Early in the 1990s, I collaborated with Gu Jianguo 顧建國, Director of the Guangxi Arts Research Institute 廣西藝術研究所, to conduct a survey of ritual theatre among the Zhuang and other ethnic groups in Guangxi, covering Laibin 來賓, Guigang 貴港, Hengxian 橫縣, Shanglin 上林, Binyang 賓陽, Qinzhou 欽州, Hechi 河池, and Jingxi 靖西.4 It was a deliberate decision on my part to include the mehmoed and other ritual specialists of Jingxi in this project, partly to address the gender imbalance (all of the shigong 師公 and daogong 道公 were male). Everywhere, though, we discovered and documented the key role played by female spirit mediums or shamans (called mehgimq and yahgimq in most places) in organising rituals for the local community. As part of this project, I conducted fieldwork in Jingxi in 1997, and, together with Ling Shudong, recorded a mehmoed initiation ritual (Lau³ t^hei¹ lang⁴, Ceremony of Tying the Sash).

In northern Vietnam, I conducted fieldwork in three provinces on a number of different religious practitioners. The list includes mogong and daogong, as well as shamanic practitioners like P t and Then. In early 2015, I conducted fieldwork in Cao Bằng, and spent several days attending rituals conducted by a female P t who lived in the outskirts of Cao Bằng city. I attended a Chúc Thọ 祝壽 (Celebration of Longevity) ritual in a village just to the west of Cao Bằng, a house blessing ritual in Hà Quảng 河廣 district, and also a clinic 門診 held by the P t for neighbourhood people with sick family members in suburban Cao Bằng. In late summer of the same year, I returned and conducted follow-up research with the same P t and her family; I also did fieldwork on daogong in Chợ Đôn district in the far southwest of Bắc Kạn province. My direct encounter with the Then began in early February 2016, when I conducted fieldwork in four separate villages in the Tày-speaking area of Jinlongdong 金龍洞 in Longzhou 龍州 county, Guangxi. I recorded five or six ritual performances, including two separate large-scale annual communal rituals Cầu Mùa 求務 (Praying for a (Good) Harvest) and a small household ritual conducted during a baby’s first three months to remove astral obstructions (Giải Hạn 解限). Finally, in August 2017, I conducted fieldwork on Then and other ritual specialists in four separate districts in Lạng Sơn province.5

For fieldwork evidence, the present paper relies primarily on fieldwork in Cao Bằng and Jinlongdong, and secondarily on fieldwork in Bắc Kạn and Lạng Sơn.6

2. The Tày

The ethnic group I discuss here is the Tày. In Vietnam, the Tày are the minority nationality with the largest population, numbering some 1,626,392 people in 2009.7 The Tày are now concentrated mainly in the northern provinces of Cao Bằng, Lạng Sơn, Bắc Kạn, Thái Nguyên, Quảng Ninh, Bắc Ninh, and Bắc Giang, where they have lived for several millenia.8 In Vietnam, the Tày are classified separately from the Nùng, who are also Tai-speaking, but are descendants of relatively recent migrants from China. In Guangxi, however, the Nùng (bouxnungz in Zhuang) are lumped together with the Zhuang, while the Tày population refer to themselves as Bu-Dai 布傣, a transliteration of the native term. If one confines ones reading to Chinese-language accounts, though, this designation means the Tày are too easily confused with the Dai 傣 in Yunnan.9 This is unfortunate: the four major concentrations of Dai people in Yunnan—in Sipsong Panna, in Dehong, in Jinping, and along the Yuanjiang—are all speakers of Southwestern Tai languages, while the Tày are speakers of a Central Tai language.10 In other words, linguistically, Tày is reasonably close to Nùng and to what in Guangxi is called Southern Zhuang.

However, a closer look also reveals some significant differences between the Tày and the Nùng or Southern Zhuang.

3. Tày and Zhuang

These differences become particularly salient in Jinlong in the northern part of Longzhou county, in the far southwest of Guangxi province bordering Vietnam. Tày in Guangxi are heavily concentrated in this area, where they form the majority of the population. Tày are also found in isolated pockets of surrounding counties such as Daxin, Pingxiang, and Fangcheng.

Jinlong is an area of karst mountains. I visited the area in February 2016, at a time which happened to coincide with the annual large-scale communal ritual to pray for good weather and good harvests in the coming year (cầu mùa). I was struck immediately by the difference between Jinlong and the villages further south near the county town.

For Jinlong, we are fortunate in having a report based on the early 1950s investigations into the social history of the area. According to that report, the ancestors of the current Tày population began their migration from Hải Dương in the Red River delta sometime around 600 years ago. The migration of these people took them through Thái Nguyên and eventually to Hạ Lang district in the eastern part of Cao Bằng, whence they moved east into the Jinlong area.11 Once in occupation of Jinlong, they forced the local inhabitants to work for them as indentured labourers and field hands; most of this population were Nùng. There is, thus, a vast social difference between Tày and Nùng in this area, with the Nùng forming part of an underclass and the Tày monopolising ownership of the best land and other forms of wealth. The Tày are literate, while the Nùng have a culture which is orally transmitted.

This divide continues to this day. Recent reports indicate that Tày and Nùng did not inter-marry at all until sometime in the 1950s,12 and the Tày proudly retain traditional costumes, traditional housing, and traditional customs. Traditional customs include teeth-staining (also found among the Thai Lue and other Thai groups much further west) and betel-nut chewing.(13 They also retain unbroken their performance culture (song, instrumental music, and dance), their writing system, and their literary traditions. Further to the south, the Nùng around the county town of Longzhou have a quite different traditional culture. Among the Nùng villages, as in other Zhuang areas, there is the custom of holding seasonal song festivals (“song markets”), but the song repertoire is orally transmitted and written song texts are hard to find.

The Jinlong Tày managed to retain so much of their local power vis-à-vis other groups because a contingent of 500 young Tày men joined the Red Army in early 1930, taking part in the uprising at Longzhou (Longzhou qiyi 龍州起義) and the subsequent establishment of a Right River Soviet Area in western Guangxi under the leadership of Deng Xiaoping.14 The Jinlong Tày in this way came to have excellent connections with the provincial Chinese Communist Party (CCP) leadership, a crucial factor in resolving any number of difficult issues, such as preferential access to middle and higher education. The loyalty of the Jinlong Tày to the CCP and to the Chinese state is, thus, more or less unquestioned, in spite of their location right on the border with Vietnam and their Vietnamese origins.15 Another indication of this favoured status is that the art of playing the two- or three-stringed lute, the primary musical instrument of the Then, was among the first batch of proposals to be granted Intangible Cultural Heritage status by the Guangxi government in 2006.16

On the other side of the border, the Tày and Nùng peoples became active participants in the Indochinese communist movement from an early date. Ho Chi Minh returned to Vietnam in early 1941 after 30 years in exile to the Bắc Bộ base area in the northwestern part of Cao Bằng. The population of Cao Bằng province at that time was 80% Tày and Nùng, and the newly established Viet Minh relied heavily on the support and participation of Tày and Nùng villagers, who had a long history of insurrection. Tày men also were among the earliest leaders in the Indochinese Communist Party beginning in the 1930s, and provided support for political training classes in Longzhou across the border. Beginning in 1941, Cao Bằng was also the site of a new initiative to foster local pride, and consolidated it by initiating a “culture-teaching movement” in the areas under the party’s control.17 A substantial amount of this cultural activity would have been conducted in the local languages, Tày and Nùng. Thus, Tày and Nùng villagers formed a core support group for the Vietnamese communist movement, and earned for themselves a right to a seat at the national table. It is probably no accident that scholarship on the Tày and Nùng got off to an early start in post-1954 Vietnam, and was well supported.

4. A Brief Review of Research on the Then in China and Vietnam

The salience of the Then among the religious practitioners in northern Vietnam attracted the attention of scholars there from an early date. The first major conference on the rituals, music, dances, and social position of the Then was held in 1975,18 just after the end of the war with the United States (US), and this was followed up by studies on various aspects of Then rituals and performance culture. Most of this work was conducted by scholars attached to the Institute of Popular Culture and the Ethnographic Museum of Vietnam. A thorough overview of this research is given by Nguyễn Thị Yên in her monograph on the Then, which was based on her doctoral thesis.19 There are now available many book-length studies of specific Then rituals, and annotated editions of quite a few liturgical texts.

Overall, the quality of this scholarly output is high. Vietnamese scholars learned from French ethnologists, as well as Soviet scholars, and they were aware of some of the international dimensions of their research topic, including Russian scholarship on shamanism.20 A major point of difference with much scholarship from the People’s Republic of China (PRC) is that the Vietnamese are evidently allowed to describe minority cultures in terms of their own internal cultural logic and from the point of view of religious practitioners and local people, rather than filtering everything through the lens of the dominant majority culture.

By comparison, much of the PRC scholarship on the Then is oddly off-centered. The performance culture of the Then in Longzhou attracted the attention of local and provincial scholars early in the Reform Era, and an investigation on the Then was conducted as part of the Cultural Compendium surveys of traditional cultural forms during the 1980s. A report on the dances of the Then was included in the Guangxi provincial volume on folk dance, under the title “Tianqin wu” 天琴舞 (“Dance of the Heavenly Lute”).21 The forms of performance, or rather their constituent elements—song, dance, vocal music, and instrumental music—remain one of the main foci of scholarly attention in China, but the actual content of the performances and ritual structure remain under-described. One recent encyclopaedia-style handbook of regional culture in the southwest of Guangxi devotes a great deal of space to the description of the two-string lute itself, mentioning in passing that it had its origins as a musical instrument used in wu 巫 performances, which are not described. The impression given in such accounts is that “Tianqin” rituals, as opposed to performances in non-religious settings, are either a thing of the past or are insignificant. Trying to make sense of the Then on the basis of these Chinese-language accounts would be next to impossible.

There is some PRC scholarship that is much better than this. One report published in Minzu yanjiu was based on intensive fieldwork in Jinlong, and provides some specific information about ritual names given to ordinands, numbers of Then practitioners, and relationships between the Then and the general populace. It also points out the general preference among ordinary people to engage Then practitioners whose parents and grandparents were also Then, since these Then were seen as having access to some generations of accumulated knowledge and experience. Even in this article, however, there is an emphasis on a supposed secularising and modernising transition, away from traditional rituals to “entertaining people”. The fact that Then rituals entailed journeys up into the sky is mentioned only in a single sentence.22

However, we know also that scholars can produce useful work and provide useful information even if their explanatory frameworks are flawed. It is worth commenting, though, that recent Chinese scholarship treats the “Tianqin” practitioners solely as a China-based phenomenon, and does not look at any of the Vietnamese studies on the Then. The scholarly gaze stops at the border, even though the border is but a few kilometres away, and even though local people travel back and forth all the time. By contrast, at least some Vietnamese scholars incorporate Chinese scholarship within their purview, and at least one Vietnamese scholar conducted fieldwork in Jinlongdong, as well as in the northern provinces of Vietnam.

In this paper, we follow this example and look at the ritual practices of the Then on both sides of the border.

5. The Then: Performance Style and Repertoire

Among the Tày in Vietnam, there are many different kinds of religious practitioners. The kinds that are transmitted by masters who received an ordination certificate are the P t, Then, Mo, and Tạo. Of these, the Mo correspond to the mogong of Guangxi and Guizhou, while the Tạo are quite similar to the daogong. The Tạo, and in some places the Mo as well, are the ones who conduct funerals, and are responsible for conducting the soul of the deceased to the other world. As in other areas to the north of the border, the Tạo make use of texts primarily in Chinese, reciting these texts in a “southern pronunciation”, while the Mo use texts in demotic script. The P t, Then, and the Mo in most areas concentrate on rituals for the benefit of the living, and recite texts in the Tày language along with ritual songs of offering which they learn by heart.23 Both the P t and the Then perform rituals which involve shamanic journeys into the sky.

It is significant that the names P t and Then are both usually written with the same character in their texts, 仸, which is a vernacular variant for 佛 fo “buddha”. The Late Han and Middle Chinese pronunciation of 佛 fo was /but/. With the designation “Then”, the character 仸 was re-interpreted, so that 天 tian “heaven” on the right-hand side is understood as indicating the pronunciation. However, it is more complicated than this, for “Then” is also the general designation in the Tai-Thai languages in Southeast Asia for sky gods. Moreover, sky gods, in turn, represent a particularly well-developed and powerful sector of the Tai pantheon, being closely connected not only with harvests and general prosperity, but also with political power and chieftaincy in the Tai domains.

On the differences in performance style between P t and Then, both shamanic practitioners of Tày ethnicity, Bế Viết Dằng has the following pertinent observations:24

As far as the melodies [of ritual music] are concerned, the Then have different songs in different places, but, in each place, the melody is uniform from beginning to end, with very few variations in the melody, while, by contrast, the songs of the P t are much more rich and varied. Songs such as “Crossing the Sea”, “The Lady in the Moon”, and “Hand-to-hand Combat with Lady Da Dun” are truly beautiful pieces of music.

As for musical instruments among the Then, one uses a lute tính with one, two, or three strings to accompany and direct the singing, and a bunch of hand-bells may be carried on the foot and shaken slowly. Among the P t, the tính is not used, but rather only a bunch of iron or copper chains which are held in the hand and shaken to set the rhythm.

The Then include many passages accompanied by dances, such as the dance “ch u slay” (“audience with the master teacher”) which is quite beautiful, with supple movements, which the professionals perform during festive gatherings. The Then are represented not only in ritual but also in festive gatherings. The P t songs are above all in the Tày language, but the Then, alongside Tày, include words in Vietnamese and Sino-Vietnamese, sometimes in complete phrases. For example,

• In my soul I want to fish for a fish…
• I love to see the moon surrounded by clouds
• I can neither forget nor think about my beloved.
• The rear guard is lined up in rows of three
• The forward guard arranges the horses in front of the officers
• The rear guard with Ngủu Lang drives away the demons.

In the original, these lines are entirely in Vietnamese. The overall impression is that the Then lyrics may be partly based on older P t material, but the Then represent a form of performance which is more artistically developed and less closely tied to specifically ritual functions. There is, however, quite a lot of overlap, and the Then tend to be referred to in quite a number of localities as “P t”.25

Both Then and P t have both male and female practitioners, and they share much of the same repertoire. This is significant in itself, and quite unlike the situation in most of Guangxi, where male and female specialists may collaborate but have quite different ritual traditions, often even in different languages. However, there is still a difference in the performance styles of male and female P t and Then. Male performance tends to be text-based, and male Then and P t recite or chant directly from the relevant ritual manuscripts. Female Then and P t do not do this, but recite their “texts” from memory. Existing scholarship on the Then tends to corroborate this picture. Personal observation in Jinlong, however, and critically important information on the transmission of ritual knowledge suggest a more complicated and more interesting situation. I return to this question in a later section.

6. Are the Then Really Shamans?

Some readers may find it difficult to associate Then ritual practices with shamanism, so a word of explanation is necessary. There is of course a vast scholarly literature on shamanism, and the term is often used without any attempt to provide even a working definition. Scholars who are more meticulous make a categorical distinction between spirit mediumship, which involves spirits speaking through the medium of the religious practitioner, and shamanism, which involves a relationship to the spirit world in which the shaman is more powerful.26 Even so, there are disagreements about how to describe the essential features of shamanic ritual practice.27 Mircea Eliade in his classic work identified trance states as the core feature, but other scholars have expressed discomfort with this idea.28

Under these circumstances, it is helpful to refer back to the region where shamanism was first discovered and documented. In its archetypal form, shamanism is found in Central and North Asia, and in China among the Tungusic, Mongolic and Turkic-speaking peoples. There, a shamanic ritual is structured as a spirit journey, in which the shaman journeys up into the sky, either by climbing the World Mountain or by climbing the World Tree.29 In other contexts, such as funerals or healing rituals, it involves spirit journeys down into the underworld, or out into the uninhabited forest or wilderness, in search of the lost souls of departed family members or the wandering souls of the sick.

For us here, the distinguishing feature of Then ritual is that it is structured as a spirit journey up into the sky and up to the World Mountain, or down to the underworld, or horizontally out into the demon-infested wilderness, and then back again.

By the same token, it is clear that the rituals of the mehmoed in Jingxi county are properly speaking shamanic, as are the rituals of the practitioners called P t or Then among the Tày people in the northern provinces of Vietnam and in Longzhou 龍州 in southwestern Guangxi. Careful comparison indicates that the ritual practices of the mehmoed are similar to those of the Then and P t. Both employ strings of metal bells on chains, held in the hand (or tied to the foot) and shaken to resemble the sound of a horse’s bridle, as a means of propelling the retinue of the ritual practitioners skywards.30 All three kinds of practitioner share a designation in common: in the case of the mehmoed, this is pat³¹, a direct equivalent of P t.31 And the Then in many areas, including eastern Cao Bằng, call themselves P t rather than Then.32 Finally, transmission of ritual knowledge to apprentice mehmoed and also to Then and P t requires both a male and a female teacher, a ‘father teacher’ and a ‘mother teacher’.33

Comparison with the ritual mechanism of Taoist priests reveals a stark contrast. Shigong and daogong do not travel up to the celestial realms themselves, but send messenger gods such as the Meritorious Officers of the Four Divisions of Time Sizhi gongcao 四值功曹 as intermediaries. By contrast, the Then, P t, and moed all travel up to the celestial realms themselves, along with their retinue and cavalcade of soldiers and horses (bingma 兵馬).

Lest it be supposed that the Then undertake their celestial journeys with no recourse to trance, I mention here briefly that both male and female Then employ trance states on their spirit journeys. It is my impression that female Then spend a greater part of the ritual journey in trance than the men do. Nguyễn Thị Yên documented the use of trance states among Then in the ordination ritual in the eastern part of Hạ Lang district, just to the west of Longzhou. The ordination ritual is called Lẩu Then (“Wine of the Then”) in Tày and Lễ cấp sắc (“Ritual for Conferring the Imperial License”) in Vietnamese, and takes place over three days and nights. The presiding Then go into trance early on the morning of the third day, after spending the preceding afternoon and night conducting a series of formal rituals bearing offerings up to all the gates to the palaces in the Heavenly Domain. This is a collective trance (nhập đồng tập thể), the purpose of which is to reward the heavenly guests with food and wine. This collective trance is conducted under the guidance of the presiding Then (ông quan lang), who ensures that all the guest officials from all the palaces in highest Heaven obtain their fair portion of the offerings and enjoy the banquet.34 In this case, clearly, the trance episode takes place separately from the recitation of scriptures.

7. The Journey to the Sky

The feature of Then rituals that is most characteristically shamanic is the mode of communication with heavenly deities. Unlike Taoist priests, who dispatch messenger gods up to the sky, accompanied by written documents (invitations, petitions, and the like), the Then journey up to the sky themselves, and meet the heavenly deities face to face. The rituals of the Then are structured as a journey up into the sky, followed by the conduct of the business of the ritual—delivery of gifts and requests to the highest heavenly deities—and afterwards the return journey to earth. The journey up into the sky is conveyed in the form of narrative verse. The journey starts at the location where the ritual is being held, and progresses upwards, with descriptions of the scenery, dangers, and supernatural beings encountered at each stage on the upward journey. In the Jinlong area, where Nguyễn Thị Yên conducted fieldwork, she found that a complete itinerary includes the following stages:35

• Home 家
• Gate of the Village 村関口
• Mouth of the Waters 水口
• Rice Fields 田地
• Temple of the Earth God 土地
• Turtle Mountain 鼇崖
• Mountain of the Gods 神山
• Graves of the Ancestors 墳墓祖先
• Prefectural Compound 府
• The Tiger General 虎將
• Emperor of Middle Emptiness 中虚皇
• Market of Errands 遣行
• Customs House of Valuables (storehouse) 貴物関 (庫)
• Bridge of Fateful Numbers 數橋
• The Southern Officer and Northern Dipper 南曹北斗
• Babbling Brook (small river) 喜水 (小溪)
• Street of the Hundred Insects 百蟲道
• Cave of the Central Region 中域洞
• Heavenly Thunder 天雷
• Great Ultimate One 太乙
• Waters of the Sea (a calm river) 海水 (漫河)
• Market of the Three Brightnesses 三光墟
• Street of the King of the Country 國王道
• Mountain Valley of Great Flourishing 太華峒
• Valley of Distant Fragrances 遠芳谷
• Buddha who Eats Buffalo (a general) 吃水牛佛 (將)
• Mother who Collects Songs, Mother of Flowers 積歌母、花母
• Waters of Revulsion (big river) 厭水 (大江)
• Prefectural Compound of Bells and Seals 鐘印府
• Prefectural Compound of the Generals 將府
• Street for Reading the Commission (residence of the Immortal Grandmother) 读敕道 (祖母住宅)
• Palace of the Jade Emperor 玉皇宫

These are at least the stations on the upward journey. Not all Then rituals involve the same route up to the sky or the same way-stations. For that matter, for each ritual type, details of the itinerary may vary from one place to another. In the ritual conducted to remove astral obstructions, as conducted by the Then in Na Rí district in the eastern part of Bắc Kạn province, the stages are as follows:36

1

Path to the One Hundred Birds

2

Going down the Path to the Dragon King and the Watery Mansion

3

The Tribute-Bearing Child Emissary from the Northern Kingdom

4

Arriving at the Palace and Inviting the Mother Emissary

5

The Tribute-Bearing Child Emissary Again, an Expeditious Road for Hasty Missions

6

Conveying Orders to the Rowers of the Boats on the Golden Waters

7

Entering the Door of the House of the Worthy

8

Conveying the List of Incense, Flower, Lamp, Tea, Fruit, and Comestibles to Be Offered Up

9

Issuing the Mats Again

10

Setting up the Station for the Crown Prince

11

Welcoming the Generals

12

The Path up the Peak of Su Mi

13

Setting up the Station in the South of the City Wall

14

Setting up the Station at the Gate

15

Setting up the Station at the Official Residence

16

Putting on the Vestments

17

Welcoming Su Mi in Front of the King

18

Breaking Open the Wine

19

Entering the Gate of the Jade Emperor

20

Entering the Gate of the Generals and Presenting Taxes

21

Eating a Repast at Khau Khác and Hau Cài

Admittedly many of these names are opaque, and require explanation. To do this properly even for a single Then ritual is well beyond the scope of this paper. What I can do here is single some of the above way-station names out for comment.

8. Indigenous and Imported Religious Elements

At the highest point in the heavens is the palace of the Jade Emperor (stage 32, 19). This deity, or at least the name, is clearly of Chinese origin. The Jade Emperor may originally be a deity promoted by the Song emperors, incorporated into Taoism, but the cult became so widespread as to form a standard fixture in Chinese popular religion.37 The same is true in Vietnam. Also of Chinese origin, but more identifiably Taoist, are the Three Brightnesses (Tam Quang 三光)—the sun, moon, and stars. Special reverence to the Three Brightnesses is a particular feature of the Tianxin zhengfa 天心正法 tradition, starting in the Song dynasty, and the San Guang continue to feature prominently in the southern Taoist schools of later centuries.38 In the Then journeys to the sky, the Tam Quang themselves do not put in an appearance; rather, the Tam Quang Market is the place where the Then and their entourage purchase the pigs to be offered at the court of the Jade Emperor. Thus, the extent to which Chinese-derived deities actually play an active role in the ascent of the Then—as opposed to having their names appended to particular stages in the journey—is a moot point, and one that needs to be investigated further.

It is also noteworthy that, in common with southern forms of Taoism, the Then have a militarised religion, and make the journey up into the sky with a retinue that includes generals (tướng 将) and spirit troops (binh mạ 兵馬). In the Jinlong area, the domestic altars of high-ranking Then practitioners have flags of command hanging up in rows behind them, indicating the number of troops and horses the Then has at his or her command.

Other figures in the sky realm are Buddhist in origin, such as the Old Buddha, the Old Buddha’s Mother, Avalokitešvara, and quite possibly a number of the yakšas. Many elements in Then ritual are based on Buddhist templates, such as the offerings of flowers and lamps, and the general layout of the offerings.

Some place-names can also be shown to derive from Chinese sources. Turtle Mountain (Aoyai 鼇崖, stage 6) is probably a reference to Aoshan 鼇山. Aoshan also appears in the ritual texts of the Zhuang mogong, and in Zhuang mythology was a mountain so high that it was the last stop before one reached the realm of the immortals. The name comes from classical Chinese literature. The ao 鼇, an enormous turtle that supported a mountain on its back, appears in the Liexian zhuan 列仙傳, where it associated with the mountain of Penglai 蓬莱. This and references in other early Chinese writings meant that the ao came to be associated with the ascent to heaven, success in imperial examinations, and elevation to high office.39

Other place-names, however, indicate the presence of what appears to be a pre-sinitic and pre-Vietic geography. The Peak of Su Mi 首眉, found at stage 12 in the Giải Hạn ascent to heaven, is found throughout the Tai-speaking areas of Guangxi and Guizhou, over a very wide area. The name takes on various forms according to local dialect and local script traditions. In central-western Guangxi, the name Couhmiz 州眉 is mentioned in connection with a narrative about the primordial flood, and is understood to be the name of a high mountain. In a Bouyei version of the flood myth from southern Guizhou, the name Cojmiz (written 索密) is explained as the name of an old man who lives on the top of Bolangshan 播朗山, the only peak high enough to escape inundation during the great flood. Even further north, in west-central Guizhou, a suspiciously similar name (站走煤, tɕiaŋ²⁴ ɬeu⁵³ mei¹¹) is found in funeral texts, and is understood as referring to the primordial homeland of the Bouyei.40 All of these variations point to a place or a person associated with Tai flood myths and, hence, a place identified as the location from which the earth was re-populated after the great flood—hence, also the equation with primordial homelands.41

Another name which connects the Then topography with an indigenous Tai substratum is Hanvueng. The palace of Hanvueng is mentioned in Then texts as one of the stages in the celestial journey.42 Hanvueng—the Goose King—is a sky god. In Guangxi and Guizhou, at least, he was originally the son of a local lord who was murdered by his step-brother. He then flew up into the sky and became a powerful sky god, exacting rent and yearly sacrifices from the domains below on earth. Hanvueng legends are found throughout western Guangxi, Guizhou, and eastern Yunnan.43

My previous research demonstrated the existence of a Tai religious geography and a common poetic rootstock extending across vast distances in Guangxi, Guizhou, and Yunnan, in spite of dialect and local cultural differences.44 Here, we can begin to see that this indigenous Tai substratum extends down into the Tày communities in northern Vietnam, including areas well to the south of the Vietnam–China border. Among the Then, at any rate, this Tai religious geography was re-structured as the landscape of a shamanic journey up into the sky.

9. Comparisons with Other Southeast Asian Traditions

The stages in the celestial journey in the Then tradition bear comparison with other shamanic traditions in Southeast Asia and China. For a start, it is clear that the Then journey resembles that of the mehmoed in Jingxi in southwestern Guangxi. Both start the journey from the domestic space, go out through the fields, visit the temple of the Tudigong 土地公, the god of the locality, go past the outer fields and lower hills, and then progress onwards up into the sky.45 Both traditions include Crossing the Seas as a stage in the journey.46 Detailed comparison will have to await a further study. Comparison with other Tai shamanic traditions such as that of the Lao-Tai of northeastern Laos would also be instructive. The practitioners there are also called moed (mae mot). Findly recently published a detailed study of their practices, and discusses the routes of their shamanic journeys in some detail. The particular routes she documented, however, come from funeral rituals, and the purpose of the journey is to conduct the souls of the deceased to the place in the land of the dead which is appropriate for their individual circumstances. Both the journey and its destination are different from the celestial journeys I discuss in this article, although the Then also conduct similar rituals for the dead.47 Shamanic rituals are also widely found among other ethnic groups in Vietnam.48 Similar practices are found among the Thai groups in northwestern Vietnam and elsewhere. Some details of the ritual procedures of the White Thai Mé Môt and a range of other religious specialists is given in Robert’s ethnography.49 Such practitioners are called either bà Then or Mo Một in the northwest, and they play a principal role in the core ritual business of large-scale communal festivals among the Thai.50

10. Then Texts and Literacy

We should not be too surprised about the phenomenon of literate shamanism among the practitioners called Then. As is well known, shamans among the Manchu clans in Jilin and the Manchu-speaking Sibe of Xinjiang kept “for centuries” in their possession “handwritten ceremonial books”.51 These were called shen benzi 神本子 in Chinese.52 Mihály Hoppál’s judgment was that these written texts “almost certainly contributed to the continuous preservation of the Manchu shaman traditions”.53

Texts possessed by Then practitioners include both liturgical texts for recitation during rituals and other narrative texts and song texts performed in non-religious (or non-ritual) contexts. There is no sharp division between these two general categories. Particular segments taken from shamanic journey texts, like the visit to the kingdom of One Hundred Birds, can also be the basis for a longer, more elaborate, and more artistic form that can be sung independently as a performance item in art song. The Then in Jinlong also have a third category of texts, dictionaries of the Tày vernacular script (chữ nôm Tày).

The Then conduct a wide range of rituals, including both communal rites and rituals for individuals and families, encompassing regular rituals like life-cycle rites and seasonal rites, and rituals to ward off affliction and disease. In the Jinlong area, the Taoist priests are left in charge of funerals, leaving the Then to cover a wide range of rituals for the living. Small-scale Then rituals may be performed by individual priests or by a small number of priests, while major communal sacrifices like the Cầu Mùa are performed by as many as 8–10 priests. Male Then, at least, have ritual texts written in the Tày vernacular script for these rituals.

Many of these texts are of very considerable length, up to two or three thousand lines. The Cầu Mùa (“Praying for the Harvest”) text from Jinlong has some 900 lines, and takes about 3 hours to recite. Recitation of the longer texts takes place over a period of around 24 hours, typically starting in the afternoon of the first day and then continuing throughout the night to either daybreak or mid-morning of the following day.

The script (chữ nôm Tày) is a character script based on Chinese, similar to the Old Zhuang script. Like the Zhuang script, the Tày script relies on borrowing either the sound or the meaning of the original character, or, in the case of Han loan words, borrowing both the sound and the meaning, suitably modified to a local form of pronunciation. In the case of Tày, the modified form of pronunciation is often similar to Hán-Việt, the Chinese pronunciation standard current in Vietnam.

Then texts themselves are mostly written in a form of seven-syllable verse. with patterns of waist-rhyming very similar to the ritual texts of the Zhuang, Bouyei, and other Tai-speaking peoples. One can tell immediately that the verse forms are not Han Chinese, nor for that matter are they like Vietnamese (Kinh) verse, where alternating lines of six and eight syllables are the most common pattern. The texts also evince what I call pervasive parallelism, wherein the ideas and grammatical word classes of one line are echoed in a second line; this phenomenon is also found in Zhuang ritual verse.54

There are some significant differences between the recitations of the Then and the general recitation practices of the Zhuang mogong in western Guangxi. One of the most important of these is that the mogong recitation is based on “performative literacy”, and the mogong may not necessarily be able to “read” their texts in the conventional sense.55 That is, throughout much of the Zhuang-speaking highlands in Guangxi, the act of reading from a ritual text is most often an oral performance, and the specific form and content of the recitation itself is orally transmitted. Vernacular priests typically come from families with priestly traditions, or from villages where there are vernacular priests. Typically, as small children, they accompany older relatives to rituals, and gradually learn to recite a good part of the repertoire from memory. Later, when they apprentice themselves to a master priest, they are given more systematic instruction, but with the form of recitation always based on the performed oral version. Later, at the point of their ordination or just before, they are given ritual texts belonging to the master priest and told to make copies for their own use. They, at this point, went to a village school or had family-based instruction in Chinese, and they copy the texts faithfully, character by character.

What this means is that the textual transmission and the transmission of the recitation take place separately, at different stages in a priest’s life, and through different modes of instruction. In performance, priests learn to turn the pages at more or less the right time; however, otherwise the recitation of the text during a ritual takes place without any act of actual reading. Over the generations, various discrepancies arose between the text as recited and the text as written, but these discrepancies are not usually noticed by the priests themselves, unless they happen to have an unusually high level of literacy in Chinese. It frequently happens in our interviews with Zhuang vernacular priests that the priest is unable to comment on the meaning of verses taken from the middle of a passage, or unable to give any more than a general meaning to a line of verse. This is not a sign of backwardness or lack of sophistication, but rather perfectly normal in village society in this part of southern China.56

Among the Then, the recitation is based directly on the act of reading the text. That is to say, the Then reads each line as he recites it. In talking with Then from Jinlong, I discovered that they were actually able to recite lines and pronounce individual characters out of context. Mogong usually cannot do this, and have to start over at the beginning of the text or the beginning of the section.

There are two factors that account for this. The first is that the Tày vernacular script is a much more standardised script, with one written form usually corresponding with one spoken word, and vice versa, while, in mogong texts, there may be as many as five or six different ways of writing the same word. The second factor is connected with the first; the Then possess dictionaries of the vernacular Tày script. These dictionaries are pre-modern dictionaries, patterned after the Er Ya 爾雅, and come in two different versions: Tày-Việt-Chinese and Việt-Tày-Chinese.57

How is it that the Jinlong Tày come to have these dictionaries, and how old are they? Do the Tày and the Then in other areas also have them? These are currently still questions to be answered by further investigation.58

One historical circumstance, however, is definitely relevant. That is, until 1927, the local schools in the Jinlong area taught students not in the medium of Chinese, but in Tày, using the local Tày vernacular script. The entire pre-modern primary school curriculum, including the Sanzijing 三字經 (Three Character Classic) and so on, was taught in Tày. In 1927, local officials came from the newly established county administration and forced local people to abandon Tày and adopt Chinese as the medium of instruction.59 Was this government action one of the factors that led local people to throw in their lot with the Communist Party?

At any rate, it is worth emphasising how unusual this local pre-1927 education in the vernacular language and script was in the Guangxi context. Indeed, I found no trace anywhere else of anything similar; invariably, Chinese was the medium of all pre-modern primary instruction throughout the Zhuang and other minority areas, at least in Guangxi. In all probability, Jinlong was an exception to the general rule because they were migrants from Vietnam. By migrating into the Jinlong area when they did, the Jinlong Tay managed to avoid the consequences of French colonial rule, including the adoption of the romanised Vietnamese script (quốc ngữ, “national language”) in the schools and the abandonment of teaching in the Chinese script. Of course, it is also worth emphasising that Tày could serve as a medium of instruction in the schools because it had a relatively standardised script, and because there were dictionaries.

11. Female Practitioners and Literacy

Were girls, as well as boys, educated in the Tày vernacular script? Perhaps in some localities they were. Judging on the basis of what I saw, it is certainly true that female Then in large-scale public performances do not recite from their texts, whereas male Then certainly do. Male P t and Then also use ritual manuscripts as a basis for their recitations during small-scale domestic rituals as well, at least in the Jinlong area. Documentation on Then from elsewhere, primarily Lạng Sơn province in Vietnam, tends to suggest that female Then conduct rituals and perform a repertoire which is orally transmitted.60 However, at least in Jinlong, one can see piles of ritual manuscripts which female Then place on the mat beside them during the ritual. Evidently, the women possess their own copies of ritual texts; it is just that they do not open them and read from them in public in the way the men do. What I infer from this is that the female Then in the Jinlong area may well be able to read ritual texts, and are not illiterate. Perhaps one could say, turning this around, that their ritual performance is that of an oral culture on the surface level, but that this orality is performative in nature. In other words, while the male vernacular priests such as mogong often evince a kind of performative literacy, acting out the process of reciting from a ritual manuscript, female Then in this locality evince a form of performative orality.

Moreover, for both male and female Then apprentices, the way in which ritual knowledge is transmitted and received is highly reliant on ritual texts in any case. Throughout the northern part of Vietnam and also in Jinlongdong, both male and female apprentice Then are required to have one male and one female master teacher. We know that the mehmoed in Jingxi and other localities in southwestern Guangxi also become the disciples of both an elderly female master and a Taoist priest; there, however, it seems that the master–disciple or teacher–student relationship does not entail the female apprentice mehmoed mastering any part of the Taoist liturgy, recited in Chinese.61 With the Then, the situation is different.

Key information about the actual mode of transmission is provided by the Vietnamese scholar Vi Hồng, who reports on the situation in Hòa An district in the central part of Cao Bằng. Vi Hồng himself came from a family with a long tradition of Then practice over many generations: his paternal grandmother was a high-ranking Then priestess well known throughout the central part of Cao Bằng, and one of his aunts was also an ordained Then practitioner. On the way in which ritual knowledge is transmitted, he reports that each Then has one “mother teacher” (slay mẹ) and one “father teacher” (slay cha). The “mother teacher” is usually a senior practitioner with long years of experience in the art of the Then: a bà or bà Then.62 The “father teacher” is a senior male Taoist priest. The way the “mother teacher” imparted ritual knowledge to her “children” was through oral instruction, since in this area girls traditionally did not go to school and could not read. The “father teacher” taught the “children” the entire set of lyrics and formulae as written in the ritual handbooks. That is to say, the girl apprentice Then received from her male “father teacher” a direct transmission of the contents of all the ritual texts needed for the recitation of texts during Then performances. To put this another way, female Then did not receive their ritual knowledge by means of a canonical oral transmission, by word of mouth from one generation to the next, but received the text-based knowledge necessary for the performance of Then rituals directly from the recitation of their “father teacher”, based on his direct reading from the texts, which they then committed to memory.63 Female Then in this area received the transmission of ritual knowledge orally, but this orality was, as it were, only one generation deep. Just how widespread this particular practice was in the Tày-speaking regions of northern Vietnam is, again, something that is yet to be investigated in detail. We know that in any case there were wide regional disparities in the numbers of male and female Then practitioners; in Hạ Lang district, just across the border from Jinlongdong, the Then tradition was predominantly or exclusively in the hands of males.64

We know in any case that female P t elsewhere in Vietnam are often literate in written Vietnamese and are able to consult almanacs and other written sources. I have seen P t in Cao Bằng conduct day-long consultancy sessions, solving problems for a wide range of clients. The primary reference work was a traditional almanac, or rather a Vietnamese-language version of one. If the problem was a sick family member, the clothes of the sick person were brought to the healing session, and, after discussion of the circumstances, the P t would briefly flick into a trance state, consult the almanac, and then provide a written script for her clients, based on the particular astral impediment that she had discovered. Of course, most of the clients were also women, and they could also read their prescriptions.

12. Literacy North and South of the Border

How is it that there comes to be such a different pattern on both sides of the China–Vietnam border? After all, the languages, peoples, and general social and cultural patterns are either the same or very similar, and there was plenty of cross-border movement in both directions for centuries if not millennia. I will try to answer briefly.

Firstly, in Guangxi, the small scale of cultural and linguistic variation is partly a function of Chinese state-sponsored operations over many centuries, including military campaigns, as I demonstrated in some detail elsewhere.65 The western part of Guangxi is indeed an archetypal shatter zone.66 The Nong Zhigao rebellion in the mid-11th century was followed by the establishment of a congeries of very small chiefly domains in the southwestern part of the province, all closely tied to the court and too small to pose a security threat. Political fragmentation led to cultural localisation, because most of the population were tied to the land as serfs, and were not normally free to move. Furthermore, consistently, in response to various rebellions, subsequent dynasties employed detachments of native troops from elsewhere in the province to put down rebellions. These garrison communities were often left in place after the threat passed. Natural disasters also led to populations fleeing many areas and re-locating elsewhere, either in Guangxi or Vietnam. Finally, Chinese pre-modern schooling was in Chinese only, and any use of vernacular scripts took place without official approval or acquiescence. Characteristically, also, in many Zhuang areas, access to schooling was confined to “official families” and to boys, which meant that village culture for ordinary people in most places could not be anything other than orally transmitted.67

In Vietnam, there were the same natural geographic conditions, but a political situation that, by accident as it were, created a relatively uniform regional culture. Everything indicates that this took shape after the Mạc dynasty lords were defeated in their civil war with the Lê and Trính. The Mạc had their capital in the northern province of Cao Bằng for some 70 years, from 1601 to 1677, and the final remnants of the Mạc were defeated in the first decades of the Qing dynasty, after an abortive attempt to link up with Wu Sangui and the Three Feudatories.68 The Mạc while in power promoted schools and civil service examinations in the Việt Bắc (Northern Vietnam) area and, conscious of the strategic weakness of their position vis-à-vis their southern enemies, adopted a policy of allying themselves with the local Tày. When the Lê and Trính were finally successful in taking over in the north, they replaced native chieftaincy with rotating officials, and moved a substantial number of ethnic Vietnamese families—officials and generals—into the area, where they gradually became part of the local population—that is, they became Tày. Quite a number of the prominent Tày intellectuals in the Cao Bằng area come from families that moved to the area around that time from much further south. The fact that Then texts include Vietnamese loan-words and sometimes phrases or entire lines in Vietnamese is a testament to the mixed nature of Tày ethnicity in this area.

The vernacular script (chữ nôm Tày) is, therefore, in its present form, not likely to be very old. If in fact it turns out that teaching in pre-modern village schools was in chữ nôm Tày, as well as chữ nôm Việt and Chinese, this would account for the observable pattern on the ground. It would also explain why chữ nôm Tày is, by comparison with the Zhuang vernacular script, a mature script that can be read without much trouble across a wide geographic area, and basically has a more-or-less one-on-one correspondence between word and graph. There was in fact a vibrant traditional literature in the area; Tày scholars in Cao Bằng discovered as many as 85 titles of literary works in circulation in the region.69

Also, in circulation among the Then and P t were liturgical texts in great numbers. There are texts available for all the major shamanic “roads” that the Jingxi mehmoed travel, and even texts for the mid-autumn shamanic journeying to the moon (Cingj nangzhai). In Guangxi, there are no texts for any of these journeys to the sky. Now, admittedly, we have yet to subject these oral mehmoed songs of Jingxi to a detailed comparison with the Then and P t textual versions. However, even so, the question immediately arises: how might they be related? There are a number of possibilities. One possibility is that the oral songs of the mehmoed are the originals, and the Then and P t wrote them down. Or, perhaps, the Then and P t also originally had an orally transmitted repertoire, which they converted into written form. Yet another possibility is that the Then and P t textual versions are the originals, and the mehmoed committed them to memory and then transmitted them orally. Of course, we must remember that the mehmoed among the Zhuang are predominantly female, while Then and P t among the Tày are both male and female. Thus, the pre-conditions necessary for very similar sets of religious practices to be committed to writing were present in northern Vietnam, but absent in Guangxi.

13. Conclusions

All this points to a broad cultural dispensation among the Tày which is radically different from that of the Zhuang. In Vietnam, there is evidently quite a different relationship between Tày traditional culture and metropolitan political power, and this different relationship is one that encompasses, rather than rejects, traditional religious practices, including shamanic and mediumistic practices.

The vernacular textual traditions also evince widely contrasting patterns. North of the border, the scripts of the mogong, song kings, and ritual masters vary considerably from one place to another; and, in many areas, at least in recent times, the wider village society transmitted its traditional cultural knowledge and corpus of songs orally. Women in particular operate exclusively (or almost exclusively) in an oral medium. South of the border, almost all of this varied cultural content (seasonal and courting songs, ceremonial songs, shamanic journeys, art-song, and long narrative songs) is found in written form. Moreover, much of this broad cultural repertoire circulated quite widely, across many of the provinces in which Tày communities were found. Rather than fragmentation, what one finds is broad regional integration.

This fundamental difference is quite startling. One would have expected something different. Among the Tai-speaking peoples, there was backwards and forwards migration between Annam and Guangxi for many centuries; indeed, the northern part of Vietnam used to be part of the Chinese empire until the end of the Tang. One would have expected that the broad pattern of culture evident in the Guangxi area—of small domains ruled over by hereditary chieftains and consequent variation from one place to another—would be found also in the Tày and Nùng communities south of the border. In terms of cultural geography and ethnohistory, we would expect a comparable level of local-level variation in language, scripts, and vernacular religious practice to extend down as far as the hills on the northern flank of the Red River plain. However, this turns out to be not the case.

The implications of this discovery for our understanding of the cultural and religious history of the Zhuang and other Tai-speaking groups in China are quite profound. For if what we have on the southern side of the border is a coherent, highly developed region-wide culture, literate and integrated, rather than just a southern variant of the fragmentation found on the northern side of the border, then at the very least we have to think of the “total constellation of forces” operating in this broader Tai-speaking borderland region in a fundamentally different way. We are used to thinking of the borderland of Guangxi and northern Vietnam as a peripheral area. On both sides of the border, in Vietnam as well as in China, the mountains are high and the emperors are far away. Naturally enough, we would expect these places to have a low level of culture in metropolitan terms. However, to continue with the notions of centre and periphery, it seems as if the culture of the Tày has many characteristics of a culture of the centre: it is highly unified, strong, with a resilient mode of transmission of complex cultural knowledge from one generation to the next, and it has many of the characteristics of a higher civilisation. For the shamanic traditions of this region, it seems as if we should look for their epicentre not within the borders of China, but in northern Vietnam.