34 8/2015 – Pp 34–50 Journal of Comparative Cultural Studies In

JCCS-a Journal of Comparative Cultural Studies in Architecture 8/2015 – pp 34–50

Frederique DARRAGON

On the ancient cross-shaped towers of Nyangpo and Kongpo in eastern Central Tibet.

Die frühen kreuzförmigen Türme von Nyangpo und Kongpo im östlichen Zentraltibet

ABSTRACT Since time immemorial the Sino-Tibetan Marches in phonetic transcription. Their transliteration have been an invasion, migration and trading corri- is given in brackets when first noted. Chinese dor where aboriginal and displaced tribes, invaders, words are written in pinyin. bandits and traders, all fought and intermingled. In some pockets of this impossible terrain of high mountains and raging rivers are found hundreds of little known towering constructions of unique shapes that I started studying 17 years ago. At the time these amazing towers were routinely destroyed to build new houses, and it was urgent to locate them, map them, document them, date them and glamorize them so as to protect them.

Although I discovered that they could be, based on their architectural characteristics, classified in four groups, there are in fact two types, the star- shaped ones, all located in today’s western Sichuan − a large part of which is traditional Kham − and the cross-shaped ones located in two small regions of traditional eastern Central Tibet (south-east of present-day Tibet Autonomous Region). All are made of vertical interlocking stone masonry pillars with embedded horizontal unpegged wooden beams, an architectural earthquake resistant technology unique to this region and sill used in traditional contemporary houses.

This paper will first list the general results of my study, including a conceptualized comparative architecture survey of local variations and a cross- cultural analysis of ancient towers world-wide. It will then focus more specifically on the cross- shaped towers of Nyangpo and Kongpo which are the oldest group of towers still standing, one tower having been dated from the 4th century and most dating from the 10th to the 13th century. These two well-watered small neighboring regions of rich agricultural lands were originally independent kingdoms reportedly inhabited by wild and ferocious populations, a fact that is difficult to conciliate with the existence of ancient sophisticated towers and with the exceptional role Kongpo plays in the Tibet- an folklore as well as in both the Buddhist and the Bon traditions. I will also glance at local customs and beliefs systems and argue firstly for the need of further comparative ethnology between the different peoples living at the foot of the towers, and secondly for the urgent necessity to protect the intangible local cultures, which are fast disappearing.

A summary of the protection status of the towers and a short mention of the restoration works, by the Chinese government and by my foundation, will complete this paper and a call for further research by Tibetan experts will conclude it.

KEYWORDS Sino-Tibetan Marches, Kham, star-shaped towers, Kongpo, Nyangpo, Tibet, Himalayas, Qiang.

Conventions used in this essay Essential Tibetan terms have been transliter- ated according to the system of Turrel W. Wylie (1959). Common Tibetan terms are presented 34 JCCS-a 8/2015 Frederique Die frühen kreuzförmigen Türme von Nyangpo und Kongpo im östlichen Zentraltibet DARRAGON

KURZFASSUNG Seit jeher waren die sino-tibetischen Grenzregionen SCHLAGWORTE ein Invasions-, Migrations- und Handelskorridor, wo sino-tibetischen Grenzregionen , Kham, kreuzför- einheimische und verdrängte Stämme, Invasoren mige Türme, Kongpo, Nyangpo, Tibet, Himalayas, und Händler gegeneinander kämpften und sich ver- Qiang. mischten. In einigen Gegenden, welche durch hohe Berge und reißende Flüsse gekennzeichnet sind wurden hunderte von wenig bekannten, turmar- tig aufragenden Konstruktionen mit einzigartigen Formen gefunden, welche ich vor 17 Jahren zu stu- dieren begann. Zu dieser Zeit waren diese Türme üblicherweise zerstört, um neuen Gebäuden Platz zu machen. Dies machte es dringend notwendig, sie zu lokalisieren, zu kartieren, sie zu dokumen- tieren, zu datieren, sie bekannt zu machen und in weiterer Folge zu schützen.

Obwohl ich entdeckte, dass die Türme auf Grund ihrer architektonischen Charakteristika in vier Grup- pen eingeteilt werden konnten, gibt es tatsächlich zwei Typen: Der sternförmige Typus befindet sich heute im westlichen Sichuan, das Gebiet, von wel- chem sich heute ein großer Teil im traditionell als Kham bezeichneten Gebiet befindet. Der kreuzför- mige Typus befindet sich in zwei kleinen Regionen des traditionell als östliches Zentraltibet bezeichneten Gebiets (süd-östlich der Autonomen Region Tibet). Alle Türme sind aus vertikal ineinandergrei- fenden Steinpfeilern mit horizontal eingelegten, nicht gedübelten Holzbalken errichtet. Diese Tech- nik gilt als erdbebensichere Bauweise, welche für diese Region einzigartig ist und noch immer in zeit- genössischen Gebäuden Anwendung findet.

In diesem Beitrag werden zuerst generelle Er- gebnisse meiner Studie angeführt, welche eine konzeptionell vergleichende architektonische Un- tersuchung der lokalen Variationen beinhaltet – in Kontext zu anderen frühen Türmen im Himalaya wie auch in einem weltweiten Bezug. In Folge werde ich mehr auf die kreuzförmigen Türme in Nyangpo und Kongpo eingehen, die älteste Gruppe von noch ste- henden Türmen. Einer dieser Türme wurde in das 4te Jahrhundert datiert. Die meisten stammen aus der Zeit zwischen dem 10ten und 13ten Jahrhundert; diese beiden gut bewässerten, an Landwirtschaft reichen, kleinen Nachbarregionen waren ursprüng- lich unabhängige Königreiche, welche angeblich von wilden und grausamen Bevölkerungsgruppen bewohnt wurden – ein Umstand, welcher mit der Existenz dieser alten hochentwickelten Türme und mit der außergewöhnlichen Rolle, welche Kongpo in der tibetischen Volkskunde und in der Tradition des Buddhismus und Bön spielt, schwer in Einklang gebracht werden kann.

Darüber hinaus werde ich einen Blick auf lokale Gebräuche und Glaubenssysteme machen und zuerst für die Notwendigkeit von weiteren verglei- chenden ethnologischen Untersuchungen zwischen unterschiedlichen Völkern, welche am Fuße dieser Türme leben argumentieren und in Folge für die dringende Notwendigkeit dafür, diese immateriellen lokalen, und rasch verschwindenden Kulturen zu schützen. Eine Zusammenfassung des Schutzstatus der Türme und eine kurze Erwähnung der durch die chinesische Regierung und durch meine Stiftung durchgeführten Restaurierungsarbeiten werden diesen Beitrag abschließen und zu weiterer For- schung von tibetischen Experten aufrufen.

JCCS-a 8/2015 35 Frederique DARRAGON On the ancient cross-shaped towers of Nyangpo and Kongpo in eastern Central Tibet

1 Answering, in 2004, INTRODUCTION masonry style, I accordingly concluded that, when a question by Science writer Being an economist by training and a globe-trotter, in the same area, both kinds of towers had been Richard Stone. sailor, polo player, jokey of English thoroughbred built by the same people. as well as an oil painter, a photographer, filmmaker and philanthropist, I suppose that my 17 Furthermore, it appears that today’s inhabitants years long interest in the ancient star-shaped probably are, in a large part, the descendants of the towers of the Sino-Tibetan Marches requires a towers’ builders because this very same masonry few explanations. Starting in the early 1990s, en- style is still used in contemporary traditional local dowed with especially good lungs, wearing local houses (Figs. 3 and 4). garb and armed with basic knowledge of Tibetan and Mandarin I had started crisscrossing Central I therefore circumcised the topic of my research and Western Tibet on my own (some of it on a to the star-shaped towers and the square towers bicycle). A few years later, in 1996 and 1997, standing in the same regions. This paper, after a while traveling in the southeastern parts of Central general review of my findings, will focus on the tow- Tibet and in Kham searching for snow leopards, ers found in two small regions of eastern Central I had stumbled on different groups of remark- Tibet: Kongpo and Nyangpo. able star-shaped towers scattered thorough the Sino-Tibetan Marches. These free-standing tow- Dating the towers ers appeared extraordinary and I soon found out In the late 1990s some of these towers were obvi- that no one knew by whom, or when, or for what ously known to exist but nobody had specifically reason they had been built. In 1998 I decided to studied them. Some Sichuan professors as well as try unraveling this mystery. Gyalrong language expert Marielle Prins1 thought they were the leftovers of the towers built in the late 18th century by the local people to defend DEFINING MY RESEARCH AND ITS GENERAL themselves against the Manchu Chinese Empire RESULTS during the seven years long ‘Gyalrong wars’. I I knew that before use of canons, towers had been knew it was very unlikely because many of the very common all over the world even if only a few towers I had seen while traveling in Draksum are left today. I had already visited more than 100 Tso (Brag gsum mtsho) and Minyag areas were countries and lived in quite a few of them but in hundreds of kilometers away from the region of spite of my extensive travels I had never seen any these two successive wars. The first wood sam- construction similar to these star-shaped towers ples I collected, one from Rimede in Minyag and and I thus immediately realized their uniqueness one from Kala near Draksum Lake turned out to be (Figs. 1 and 2). Originally I was mainly interested respectively from the 12th and from the 14th cen- in the star-shaped towers and not in the tall square tury. These results spurred me to continue. Local finger-towers that stood near the star-shaped ones, people supported my research wholeheartedly as but after collecting wood samples from both kinds, they were touched to have someone from France and getting them carbon dated, I found out that being so concerned about their old towers; the fact they all had been built during the same time period. that − via my foundation − we also repaired a few As square and star-shaped towers exhibit the same schools, roads, water systems and bridges added

Fig. 1 Fig. 2

Fig. 1: Cross-shaped towers of Shug- ba, along the Nyang River. In- itially there were six towers, but only five are still standing, all of them in great need of restoration.

Fig. 2: Three towers, two square and one star-shaped with eight corners, in Rongbrag, Gyalrong (Rgyal rong).

Figs. 3 and 4: Masonry style in Gyalrong: top of 14th century star-shaped tower of Dzonggak (Rdzong ‘gag), and two houses in Rongbrag; the dark house is about 200 years old while the white one was built a few years ago. Fig. 3 Fig. 4 36 JCCS-a 8/2015 Frederique Die frühen kreuzförmigen Türme von Nyangpo und Kongpo im östlichen Zentraltibet DARRAGON to the locals’ friendliness. I was surprised and sad- when they were built. The towers located in the 2 I am thankful to Prof. dened by their general lack of cultural self-esteem. Qiang minority villages (Fig. 6) in the Min Moun- Chen Zongxiang for making available to me the broad re- Those who were Buddhist or Bon often considered tains are impossible to date because they stayed search concerning the towers themselves as second class Tibetans; the Qiang, in use until early 20th century due to the constant he had done in ancient Chi- who still have a sort of animist-shamanist religion, inter-villages feuds; as a result they have been con- nese texts, and to some of my students and volunteers were often hoping to move to town and become tinually renovated and beams from many different or employees of my founda- more Chinese. All were completely unaware of the periods are found in each tower. Nevertheless some tion who, over the years, uniqueness of these constructions. are possibly very old as these very towers were have helped me with research in Chinese and translations. already mentioned, in the Eastern Han Dynasty Very tall free-standing stone During the 17 years of my research I have been (25–220 AD) Annals as built by the Ran Mang Tribe towers are mentioned in Chi- on innumerable field trips (I even bought a house in the then Wenshan prefecture3. Sun Hongkai nese ancient books dealing with western areas inhabited in one of the towers’ regions) and spent countless (1989: 95) explains that “Historians believe that by ’barbarian‘ tribes. The first hours in libraries or discussing with historians, an- the present day Qiang people are the descendants mention was in the Treatise thropologists, archeologists, architects and stone of the Ran Mang Tribe”. about the Nam Man Xi Nan Yi of the Eastern Han Annals (Hou builders. I have also sent to carbon dating − to a Han Shu), it was followed by top laboratory in the US − more than 100 sam- Mapping the towers numerous cites in most of the ples of wood that I collected from the now broken I have now mapped practically all the existing history and geography books including those of the Sui and beams still found inside most of the towers. Obvi- towers (more than 1000 if one also counts the rec- Tang Dynasty and until the ously what is dated is the time when that wood ognizable ruins) and have established that based Qing Dynasty. But if these re- sample died, but by selecting small beams tightly on their architectural characteristics they can be marks were numerous, each time they only consisted of embedded in the masonry (Fig. 5), one can be sure classified in four groups. The towers territory rep- a simple description in a few that this specific beam: resent about one third of Texas, it extends from lines. today’s Khrochu4 in the North, to Mili (Rmi li, Mu li) 3 Present day Min River’s • had not been replaced (if it had been replaced, in the South and from Maoxian in the east to an- it consequently would be newer than the build- watershed: the eastern coun- cient Nyangpo and Kongpo in the west. As I started ties of Aba Tibetan and Qiang ing construction date) researching the history of the different regions, I Autonomous Prefecture. • nor had it been stored during many years became to realize that each group of towers was lo- 4 I want to thank Sonam fore the construction of the building (if it had cated in a geographical pocket that corresponded Wonjal, John Bellezza and been stored or used in a previous building it roughly to the ancestral lands of an ancient tribe Guillaume Jacques for having provided me with the Tibetan would consequently be older than the building (Figs. 7 and 8). Three of these regions are in to- versions of local names in one wishes to date) day’s western Sichuan5 and can be described as: Kham; all the eventual mis- takes are mine. • the lands occupied by the contemporary Qiang It is also important to select small beams because, in the Min Mountains (i.e. Lixian, Maoxian and 5 I use the term Sichuan firstly, in well forested areas small beams are un- here because there is no oth- Wenchuan counties); likely to be stored for long, secondly, as during each er way to describe this whole region. It is a contemporary growing season trees add a new growth ring just • the region still called Gyalrong today, that of administrative term which under their bark, the dating of a sample originating the famed 18 kingdoms of Gyalrong of the has no relation whatsoever from the center of a large trunk (i.e. hundreds of past; with territory divisions at the time when the towers where years older than the part below the bark) will yield • the territories extending principally east of the built. The towers’ territory dates preceding the construction time by hundreds middle lower reaches of Nyagrong River and cannot be labeled as Kham or of years. Eastern Tibet given more than some of its tributaries, from Mili in the south 15% of near complete towers to Draggo (Brag ‘go) in the north, and from are located in Qiang Minority’s The 82 towers (and more than 100 samples) that Shongpa in the east to Tongpa (Stong pa) in villages, and that the contem- I dated by carbon dating were built from the 4th to the west. porary Qiang, called ‘Qiang Zu’ th th (see note 11) are neither Bud- the 17 century, most of them being from the 11 dhist nor Tibetan. to 14th century. In spite of my careful selection of These are part of the traditional lands of the cel- samples, a specific tower could eventually be erro- ebrated Minyag (Me nyag) tribe (called Dangxiang neously dated but so many similar results reinforce Qiang in ancient Chinese texts) who established each other’s; the overall time bracket of the carbon many kingdoms (Stein 1951: 225 to 228) during dating results is coherent with the limited written its near 2000 years history (Van Driem 2001: 450) evidence2. including the last one, Chala, whitch endured well into the 20th century. I have included Zhaba (Dra pa) It is not possible to guess the age of a tower by its in this group because recent studies have demon- construction style since within each of the regions strated close links between the Zhaba language and the towers are nearly identical, independently of the two Minyag languages (Sun, 2007 and personal

Fig. 5: A ruined tower in Trendzi (Khran ’dzin) at 3657 m of altitude. As all the outside squar- ish stones have been looted to be used in new houses, this ruin allows us to see the sophisticated arrangement of unpegged wooden beams of the Minyag towers. Towers from other regions have less numerous beams. The roundish river-bed stones used in the rest of the wall have been left on site. As the small round beams date unquestionably from the time of the tower’s construction, they offer an indisputable possibility to date accurately this tower. The carbon dating yielded a calibrated result of AD 1030 to 1250.

Fig. 6: Yingzui fortified Qiang village of the Min Mountains. Fig. 5 Fig. 6 JCCS-a 8/2015 37 Frederique DARRAGON On the ancient cross-shaped towers of Nyangpo and Kongpo in eastern Central Tibet

6 In that same paragraph, Hazod indicates that “a small piece of wood found in one of the towers” gave a radiocarbon of 12th−13th century but he reasonably adds that it is probably not reliable as it is known that these structures have been repeatedly rebuilt after the Langdarma (Glang dar ma) period.

Fig. 7

communications 2009). The Zhaba towers are fairly Since time immemorial the Sino-Tibetan Marches similar to the Minyag towers but show some local were an invasion, migration and trading corridor variations. where aboriginal and displaced tribes, invaders, The fourth region, composed of the ancient king- bandits and traders, all fought and intermingled. doms of Nyangpo and Kongpo is in eastern Central The history of the whole area is still wrapped in Tibet. mystery. As there are no local written languages, information dating from the time of the towers can With the exception of the contemporary Qiang, all only be found in some Tibetan books mainly con- the native people now living near the towers have cerned with religious matters or in ancient Chinese been assimilated to the Tibetans, either by con- texts dealing with frontier regions. Aforesaid, in quest at the time of the Tibetan Empire or because these Chinese books tall towers and/or ‘tower cul- they became Bon or Buddhist later on, or possibly tures’ started being mentioned nearly 2000 years still, because their native animist-shamanist religion ago. Although the construction of such monuments was similar to the ancient unsystematized Bon, or definitively predates the diffusion of Buddhism and to what is called the “Nameless Religion” by Stein of systematized Bon in the concerned lands, there and the “Folk Religion” or mi chö (mi chos) by is not a single mention of them in any book written Tucci. These people exhibit many indigenous local in Tibetan. A Gyalrong lama once explained to me customs but their native tongues, although gener- that he had once read about pre-empire aligned ally belonging to the Qiangic language family (Sun towers standing every 10 li. As he was questioned Hongkai 1989 and 2007; G. Jacques and N. Tour- by R. Stone, writer of a 2004 Smithsonian article nadre, personal communications 2001 to 2009), all about my work, Eric Mortensen expressed a similar include a large Tibetan vocabulary. Kham Tibetan idea. Possibly they were evoking the aligned free- is broadly spoken and slowly displacing the native standing towers that local people consider to have tongues. In the eastern part of the towers regions been erected before the Tibetan Empire as those Mandarin Chinese is dominant as only about 30% of of Yarlung Sokha (Yar lung Sog kha) (Gyalpo et al. Qiang people are able to speak their own language. 20006: 206). The towers of this study are never Other non-Buddhist minority peoples, including aligned but generally in small clusters of three, many speaking languages belonging to Loloish fam- sometimes up to six towers. ily, also inhabit this corridor but in regions where no star-shaped towers are standing. Because of the lack of written sources, and because vernacular architecture is defined by local It cannot be excluded that similar towers existed environmental constrains, social organization and in other regions, more specifically in the now well- needs of the concerned society, it appeared to me developed ones or in areas where clay is inexistent early on that the best way to understand both the and the soil quality does not allow constructions of uniqueness and the possible functions of these stone and mortar to last. ethnic corridor towers was to compare them with other ancient towers from other parts of the world. Understanding the towers Defensive towers were very common during the It is important to clarify that quite a few differ- Middle Ages, but once more efficient war technolo- ent towers exist in other regions of Tibet and/or gies had made them obsolete, most were destroyed southwest China and they are not included in this to reuse their land or their materials; these tow- paper, they are ers were rarely very tall. The first very tall tower (around 120 m) was a beacon, the lighthouse of • the numerous ordinary small square towers of Alexandria, set on Pharos Island and built in the regions where there are no star-shaped towers; 3rd century BC. It was emblematic of Alexandria’s • the earthen towers, as those of Buwa, greatness, as it was represented on the city’s coins Chaktreng, Lhodrak (Lho brag) among others; and was considered one of the Seven Wonders; Fig. 7: • the ‘grooved’ towers found in Chims (Mchims) numerous builders tried to imitate it. Today, the Map of the four tower regions. (see Fig. 14) and Lhodrak. tallest towers are all office buildings wanting to 38 JCCS-a 8/2015 Frederique Die frühen kreuzförmigen Türme von Nyangpo und Kongpo im östlichen Zentraltibet DARRAGON

7 Another technology with similar advantages is the construction by reduced tiers. Some towers combine both technologies as the 12th century Minaret of Jam.

Fig. 8

symbolize the prestige of a company, city or coun- to the local environment and not in any specific try. Most of the ancient flamboyant free-standing compass direction. towers of the past were also some kind of status symbol even if they had been built as beacons, Mutually unintelligible languages are numerous and tombs, victory statements, or family towers. Pope the towers are designed by native words specific (1965: 96) writes: “perhaps the antecedents of all to each local language (Sun H.K. 1989: 94 to 101; these towers were the cains and poles erected by Jacques, personal communication 2015; and per- earlier Central Asians to commemorate their vic- sonal observation). In regions where people speak tories. […] The inscription on the tomb tower of either Lhasa or Khamba language, the towers are Mu’minin Khatun at Nachshirvan emphasises the called khar or dzong khar. time-defying intent of these structures: ’Everything passes, may this remain’ ”. Some towers were also Towers technology built for religious purpose but less than it would Some characteristics that have recently been con- now appear to our Judeo-Christian eyes. In a 2004 sidered to be typical of the Sino-Tibetan Marches personal communication with Prof. Emeritus Grabar, towers are in fact quite common in worldwide ar- he indicated that many towers currently used as chitecture (Figs. 9 and 10): minarets were in fact built as beacons for caravans traveling in the desert. Early temples were often • Slanted outside walls and walls thicker at the very large, as were the ziggurats or the Roman bottom than at the top. This is a technology and Greek temples, but they were rarely in the routinely used in ancient buildings, it adds sta- form of towers. bility by lowering the center of gravity, and in very tall buildings it also prevents the lower ranks of stones/bricks to be crushed by the It has been impossible to find a geomantic prin- 7 ciple justifying the Sino-Tibetan Marches towers’ weight of the rest of the construction . distribution. Some towers are located in strate- • Sacred or decorative white stones or white gic places as entry points to valleys or mountain paint imitating white stones. tops. The others are built in regions of good ar- • Little corner gablets. These are commonly able lands but to avoid using the limited arable used from Yemen to England, both as a deco- acreage, they are often, as the villages them- ration and as a way to ‘pin’ down the corners selves, built right at the foot of a mountain or of flat roof constructions. Fig. 8: on a stony ridge. Doors are located in relation Tower Typology JCCS-a 8/2015 39 Frederique DARRAGON On the ancient cross-shaped towers of Nyangpo and Kongpo in eastern Central Tibet

8 It was used in the 14th−16th centuries Meteora monasteries in Greece and possibly in some south Hima- laya constructions (Kimmet, personal communication 2013) although not in the kath-khuni buildings of Himachal Pradesh; the kath-khuni technology is a wooden architecture using some stones (Handa in 2011: 100), while the Sino-Tibetan Marches towers technology is just the opposite: a stone architecture using some wood.

Fig. 9 Fig. 10

However, it is perfectly justified to study the had obtained. Many kinds of local stones, including Sino-Tibetan Marches towers separately from oth- slate and granite, are used; but sandstone, when er towers in China or Tibet. Indeed, they are truly available, is the preferred material, more specifi- distinctive because of a set of characteristics unique cally for the part of the wall facing outside of the to them (and to the local houses – Figs. 11 and tower. All the stones are uncut but for the corner 12), which also enable them to resist earthquakes: stones that are occasionally roughly dressed. Gen- erally easy to collect round river-bed stones are • They are made of interlocking vertical tall used for the part of the wall facing the interior pillars, woven together, leaning into and but- of the tower (see Fig. 5). Some Gyalrong masons tressing each other (Fig. 13). A totally different stress that a relative hardness of the stones needs architectural concept from that of the ancient to be taken in consideration. The mortar is mud Muslim towers of Persia, India or Afghanistan and clay. Surprisingly, there are no round towers (Fig. 14), which in fact are round towers with in the regions where star-shaped or cross-shaped ridges (Darragon 2009b: 27). towers are found. • Unpegged horizontal wooden beams are embedded in the masonry to add flexibility (see Fig. 5). Although common in adobe construc- A CASE APART: CROSS-SHAPED TOWERS OF tions, this is very rare in stone masonry.8 NYANGPO AND KONGPO Round or square short towers are a familiar sight in The species of timber used vary depending of the Central Tibet, as they are often attached either to regions, more common are spruce, fir and larch, monasteries or to castles. But, apart from the ones but in Nyangpo all the samples collected are of already mentioned in Yarlung Sokha, a good num- larch. This information was given to me by Prof. ber of free-standing towers are found in Nangdzong A. Bräuning whom I had invited in 2004 to a six (area more or less similar to ancient Chims) and week tour of the towers in the hope of creating a in Lhodrak (also home to the very famous square dendrochronology data-base. He was not able to Milarepa tower); quite a few of these towers are get the funding for the AMS wiggle matching carbon of a ‘grooved’ shape (Fig.15) apparently unique to dating until late 2009, but the results, once out, these regions (Darragon 2009a: 4). were in the same time-brackets than the ones I Fig. 9: 18th century Riverton Pike in England.

Fig. 10: This Yemeni contemporary village is a good example of slanted wall architecture, also with little gablets on the corners of the rooftops and white stones for good luck.

Fig. 11: Contemporary houses in Minyag and Kongpo/Nyangpo as well as Gyalrong rely mostly on embedded beams to resist earthquakes; here the inside of a house in Kongpo.

Fig. 12: Fig. 11 A Gyalrong house which wall has a vertical ridge to reinforce it. This technology is used frequently in traditional Qiang houses and rarely in Gyalrong ones.

Fig. 13: Twin star-shaped towers of Rimede, Minyag, built in the 12th century and made of enor- mous interlocking pillars. The tops are not the original ones.

Fig. 14: The two 11th century victory towers of Ghazni, Afghanistan. Although their lower tiers appear star-shaped, we can see on this old print that the towers are in fact round. Fig. 12 Fig. 13 Fig. 14 40 JCCS-a 8/2015 Frederique Die frühen kreuzförmigen Türme von Nyangpo und Kongpo im östlichen Zentraltibet DARRAGON

9 R.A. Stein does mention these towers many times. As a matter of fact, it was Michel Peissel, a friend, admirer and student of Stein, who told me about the Minyag towers. At the time I had only seen the Rongbrag and Shugpa ones. Sadly Michel left us in 2011, I am very grateful to him for having taught me how to film and helped me produce my tower documentary.

Fig. 15 Fig. 16 Fig. 17 Fig. 18

Remarkably no star-shaped towers can be found was also the first person to notice that although outside of western Sichuan, which would seem to the cross-shaped towers had always been consid- confirm that this technology developed locally in ered (including by the local government) to have 12 the ethnic corridor, and − as proven by mentions corners, a few of them only had eight corners (Dar- in Chinese ancient texts − that happened before ragon 2005: 118); contrarily to the star-shaped the Tibetan empire and the ‘tibetanization’ of many towers, which always have the same number of indigenous populations. But then how could we ex- inward-pointing corners and outward-pointing plain the existence of cross-shaped towers? Not ones, the cross-shaped towers with 12 outward- only these cross-shaped towers are both similar pointing corners have eight inward-pointing ones, and different from the Sichuan ones but they only and the towers with eight outward-pointing corners exist in two small regions of eastern Central Tibet, only have four inward-pointing ones. Consequently, each with a rather extraordinary, although little whereas the star-shaped eight-corner towers are known history: the ancient kingdoms of Nyangpo the most stable design, the cross-shaped eight- and Kongpo. corner towers are rather unbalanced and all of them are in an advanced stage of ruin; one needs an ex- How do the Nyangpo and Kongpo towers dif- pert eye to notice that they only have eight corners. fer from the ones in Sichuan? From far away the cross-shaped towers are quite Notwithstanding these essential structural differ- similar to the star-shaped ones and I was the ences, the towers of eastern Central Tibet are very first person to notice their differences (my 2003 similar to the ones standing in Sichuan as all of them documentary and my 2005 book). The few per- were constructed following the same architectural sons who mentioned both kinds had probably only technique, unique to the Sino-Tibetan Marches, of seen one and heard of the other, as Stein9 (1951, interlocking stone masonry pillars and imbedded 1958, 1961, 1972, 1987), Kingdon-Ward (in Cox horizontal wooden rods. There is no doubt that the ed. 2001: 179 and 180) and, more recently, Hazod inhabitants from the different regions were, if not who suggested in 2006 that people from Eastern belonging to the same ethnic group, in rather close Tibet must have settled in Nyangpo. contact. This is also proven by a number of simi- larities between the different types of towers: the It is very unlikely that the same people would have cross-shaped towers have foundations and have built the cross-shaped towers and the star-shaped the door at floor level, thus on that aspect they ones because both were built during the same time resemble the Qiang and Zhaba towers but differ period but using a somewhat different structural from the Gyalrong and Minyag ones (which have model (Darragon 2003: documentary film; 2005: little or no foundation but a three to five meter 35 and 118, see typology in this paper). Both kinds solid bottom part). Like the Qiang and the Gyalrong are made of interlocking pillars but the pillars them- towers, the ones in Kongpo and Nyangpo have few selves are of different shapes (Figs. 16, 17, 18 and openings, these are usually only small vertical slits see Fig. 8) resulting in two types of structurally located in the middle of the broader sides. As the different towers: the cross-shaped towers have Qiang towers, the southeast Tibet ones also have only 90° corners while the star-shaped towers’ rather little embedded wood. Yet, aforesaid, some Fig 15: Front of a Nangdzong tower outward-pointing corners are acute angles and the embedded wood is still a characteristic of today’s displaying a vertical groove inward-pointing ones are obtuse angles. Minyag, Gyalrong, Kongpo and Nyangpo traditional above its door. The interiors of the towers are also different; those construction style (Figs. 20 and 21 and see Fig. 11). Figs. 16, 17 and 18: of the Tibet towers reflect the mandala shape of Comparing an 8 corner tower their exteriors (Fig. 19) while the interiors of the Although some Qiang, Zhaba, Minyag and Gyalrong from Gyalrong (Fig. 16) and Sichuan towers are roundish (Figs. 22 and 23). I towers still display all or part of their original tops, two twelve corner towers from Nyangpo - Shugba tower #2 (Fig. 17) and Kala tower (Fig. 18).

Fig. 19: Interior of a cross-shaped tower. The top of this tower is deformed, as most of them are as the roofs have collapsed long ago.

Fig. 20: New Minyag house being built in Tagong.

Fig. 21: New wall in construction in Rongbrag, Gyalrong, in both the embedded beams and planks can clearly be seen. Fig. 19 Fig. 20 Fig. 21 JCCS-a 8/2015 41 Frederique DARRAGON On the ancient cross-shaped towers of Nyangpo and Kongpo in eastern Central Tibet

10 I surveyed 44 towers (or recognizable ruins) in Nyangpo, collected 30 samples from 1998 to 2009, and as some towers were dated more than once, this resulted in 23 towers being dated; only three of which could have been built as late as the 14th century and one as late as the early 15th century. In Kongpo I saw one tower and eight barely recognizable ruins, I collected two samples, both in 2008 and this resulted in one tower and one ruin dated. All the radiocarbon dating analysis were carried out by Beta Analytic, Florida, USA, world largest radiocarbon dating laboratory, accredited Fig. 22 Fig. 23 Fig. 24 with the highest level of recog- nized quality attainable (PJLA the roofs of the Nyangpo and Kongpo towers have however, all the accessible beams had already Testing Accreditation #59423). All the detailed results, with all collapsed. This earlier collapse of the rooftops is been sawn and collected for fire-wood. But in photos and GPS locations, in line with the findings that tower use, and tower the last decade the ‘New Village Policy’ combined will soon be uploaded on construction, stopped earlier in Kongpo/Nyangpo with local’s accrued wealth generated by Yartsa Academia.edu. than in other regions. These two regions of eastern gunbu (Dbyar rtswa dgun ‘bu) collection (personal Central Tibet being well watered, local houses have observation from 2004 to 2011) resulted in the pitched roofs instead of flat ones. If the towers’ construction of new permanent settlements and the roofs had been of the pitched ‘saddleback’ kind, two large towers (one possibly dating from the 7th it would have made the towers unpractical to use and the other from the 10th century from samples as beacons. we collected in the ruins) were dynamited in 2007 (Fig. 28). The government is inefficient at protect- The size of the towers varies broadly within each ing towers in very remote areas. region. The outside diameter of the bottom part Fig. 22: ranges from six meters to more than eleven, the Of mandalas, mu ropes, garudas, demons and Round interior of the restored Rimede west twin star-shaped original height is difficult to appreciate given that towers tower. The ladder, made of a very few towers have their original top part. The One cannot fail to notice that the cross-shaped tow- carved trunk, is of the tradi- tallest ones are said to have reached seventy ers’ horizontal plan has the shape of a mandala tional Himalayan type still in use today. meters, but today only one tower, standing in Gyal- similar to that of many pillars in monasteries and rong, nears fifty meters of height. The narrower of the bottom of many chortens including the re- Fig. 23: part of the wall at the bottom of the largest towers nowned Gyantse Kumbum (Rgyal rtse Sku ‘bum). Floors are made of supporting large wooden beams covered can be close to two meters thick. Some towers still The towers’ ‘mandala’ shape can also be compared first with planks, then with have parts of their original wooden floors, these are to that of some temples in Himachal Pradesh. But twigs then with earth, as it can similar to those of the traditional houses, and were there is a major difference between the cross- be seen during the restoration of the Xizongka 15th century linked by ladders (Figs. 22 and 23). shaped towers and these religious constructions: star-shaped tower. Nearly half the lack of “swelling dome-shape of the anda or of the tower had collapsed. I have counted 53 towers in eastern Central Ti- ‘egg’ also called garbha or ‘womb’ that is the swell- Fig. 24: bet, 25 of which were dated; they are the oldest ing form of any stupa, anywhere […] even more In 2006, we collected a sample group as most of them have been built between universal for stupa construction than the mandala from a beam still tightly em- the 10th and 13th century10. The most ancient shape” (Bernier 1997: 42). Although we do not bedded in the masonry. It is a small round beam and one can tower (Figs. 24, 25, 26 and 27) still standing was know what kind of top the Kongpo/Nyangpo towers see that very few outside rings erected in the 4th century in the higher reaches of had and although their outside walls are slanted, are missing. the Nyangchu River, at 30°15.226m N/ 093°05.119m these walls are perfectly straight, and none of the This sample was first dated by beta-analytic in the USA as E, near the largely new village of Tongo. towers presents even the beginning of the rounded sample #222957 then again shape of the anda. in 2008 as #239600 yielding This tower is one of the smallest. Contrarily to all similar results of Cal AD 220 to 400. the other cross-shaped towers whose loopholes are As Ramble explains in 2007, imposing or superpos- Later, a more precise AMS wig- all centered in the middle of the pillars, this tower ing a mandala shape on a natural site is a way of gle matching (done by Prof. A. also has small slits positioned in its inward-pointing taming it; building mandala-shaped constructions Bräuning in Germany) yielded a result, for the most outside corners (Fig. 26). certainly exemplifies the same principle. But the rings of AD 318 +/- 40 years. mandala shape is also a representation of the four It is the tower located at the highest altitude compass directions, embodying cosmic harmony. Fig. 25: Local people have started loot- (4000 meters above sea level) and consequently It is not necessarily a religious concept as it also ing the corner stones from the the less populated area. Until recently that tower exists, not only in Indian cosmology and ideals, but bottom of one corner of the and two other ones nearby had been safe because in Chinese philosophy as well as in other civiliza- back of the tower. very few sedentary people lived around them; tions then often labelled ‘Andean cross’. Because Fig. 26: Prof. Li Chunxia and the author in front of Nyangpo oldest tower. In the green circles, one can see the loopholes, some in their usual position in the middle of the broader sides and some, unusually, in the inward pointing-angles.

Fig. 27: Ruin of one of the two large towers that were standing north of the oldest tower.

Fig. 28: We found dynamite wraps in the ruins of the two large towers nearby which were destroyed in 2007. Fig. 25 Fig. 26 Fig. 27 Fig. 28 42 JCCS-a 8/2015 Frederique Die frühen kreuzförmigen Türme von Nyangpo und Kongpo im östlichen Zentraltibet DARRAGON any human endeavor has a spiritual dimension and Chinese scholars have studied in detail the Qiang 11 Regarding the pos- because the towers (both star-shaped and cross- towers and superficially the Gyalrong towers, and sible origin of the mu rope, I must differ from Aris (1980: th shaped) resisted earthquakes, they were also, in they started to collect legends in the early 20 cen- 125) who, when extrapolating a way, a lay and vernacular symbol of cosmic har- tury. Aforesaid, the first towers to be recorded by Stein’s few words in Tibetan mony. Chinese scribes were those of 2nd or 3rd century Civilization, attributes it to ‘Ch’iang mythology’ (Qiang, Ran Mang tribe of the Min Mountains; this tribe’s as it is now written in pinyin As a suggestion for the 2004 Smithsonian article descendants, today’s Qiang minority people still live was spelled Ch’iang or K’iang already mentioned above Horleman (who had seen in the Min Mountains, still have towers and still in the past). In “Les K’iang des Marches Sino-Tibetaines”, the towers in my documentary at the Discovery have an ancient animist/shamanist religion, pos- Stein clearly explains that the Channel premiere in 2003) hinted that the tow- sibly similar to ancient Bon12 but unconnected with Tibetan words mu and phya ers could have been a substitute for the mythical the towers. The section about Wenshan Prefecture are Qiang words designing the sky (mu pya); but there 11 mu ropes (dmu thag) to get back to the sky, the (in the Book of Sui) records that “the wife of the is no tradition of dmu rope in same idea was also expressed by one Gyalrong last king of the Chen Dynasty (538−598 AD) gave the Qiang mythology, as the lama talking to Bellezza who saw the towers for the birth in the Su village […]; the inhabitants built Qiang are known to burn their dead (Graham 1958: 41 and first time in 2009. There are more than 1000 tow- a seven story blockhouse for the event” already 42; Stein 1958: 4, based on ers still standing, in diverse stages of ruin; in their expressing the status symbol character of some very numerous such explana- heydays they must have been counted by the tens Qiang towers in the 6th century. tions in Chinese Annals; as a matter of fact the burning of of thousands; in 1924, Kingdon-Ward counted 200 their dead is also specified, towers in Kongpo alone. In ancient texts the mu Tower construction appears to have stopped around in the Eastern Han Annals, ropes were used by the dead kings prior to Drigum the 15th century in eastern Central Tibet and a bit as a characteristic of the Ran Mang tribe). The term ‘Qiang’ Tsenpo’s (Gri gum btsan po) death, it is hard to later in Minyag, Zhaba and Gyalrong. If their towers invites some clarification. In explain why towers, as mu ropes substitutes, would had been sacred, the inhabitants of these regions, the article mentioned above, have been built in such numbers, at such expense all of whom are deeply religious, would certainly Stein aims at elucidating and proving the connection be- and so late in history. Macdonald (1971: 294 and not have forgotten these monuments’ sacredness tween today’s Qiang minority 295 note 385) mentioning Stein, explains that the from the 16th century to today. Western travellers (‘Qiang Zu’), ancient Qiang mu ropes are also a way of communication with the and ethnologists of the two past centuries did no- tribes (also called Di-Qiang tribes), and Tibetans. This sky, used by later Bonpo priests, but she does not tice some of these towers; they also mentioned is a case that has long been seem to imply that they exist in any tangible form. that the local inhabitants knew nothing of them. made by Chinese scholars, al- Admittedly the mu rope or ladder is a common though some did disagree, as Wang Mingke who argued, in myth in Himalayan populations as explained by Aris All the above, added to the fact that the towers are the late 1990s, that the Qiang (1980: 126) describing the Aka origin myth. Ram- never mentioned in any book written in Tibetan Minority had been invented by ble (1997: 91) also evokes ‘spirit ladders’ made of convinced me that the towers topic of my research the Chinese government in the 1950s. Wang Minke’s claim has sticks for spirits and/or dead people. But if towers have no connection to religion. That came as a been contradicted in the last were actually built as sky ladders, they certainly surprise to me because at the beginning of my re- decades by the identification did not need to be so massive, have an exploitable search, based on the sacredness of the number of the Qiangic language family and by DNA testing supporting interior and often feature large rounded loopholes, eight for both Chinese and Tibetan Buddhists, I the view that the Qiang Minor- perfect for looking outside or firing arrows (Fig. 29). was under the impression that the towers were ity is the oldest population in religious edifices. Quite surprising too is the fact the family as Qiang have the greatest genetic diversity of that towers are often found in Bon strongholds Sino-Tibetan Himalayan popu- even if there are also other Bon strongholds with- lations (Kang L.L. et al. 2011: out towers. Some square three to four story tall 92 and 97; Deng et al. 2004: 346). Concerning the connec- temples exhibiting ancient painted murals (dating tion between today’s Sichuan from the 13th or 14th century for the earliest one: Qiang and the Tibetans of that of Shalake in Rongbrag) do exist in Gyalrong, Central and Western Tibet, they do share the same, fairly Minyag and Zhaba. In Chinese these constructions rare, Y Chromosome Haplo- are sometimes also called ‘diao’ as the towers; still type D-M174 (YAP lineage), they are never designated by native names but but they must have separated about 6000 years ago (Deng et always called lhakhang, indicating their religious al. 2004: 346) because each use. Their walls are usually thinner than those of belongs to different subclades the towers. (Shi et al. 2008: 5, Table 2). Studies of other haplotypes demonstrate a more recent In general, and more specifically in Sichuan, small gene flow between Ti- tales attached to the towers are few because betans and populations of the Fig. 29 Sino-Tibetan Marches (Shi et patriarchal Lamaism has been slowly erasing the al. 2008: 6). 6000 years ago ancient often matricentric indigenous myths13. was prior to any mention of As in Gyalrong where the prayers addressed to Qiang tribes in Chinese writ- The Sichuan Han scholar of Tibetan studies Prof. ings, as these started about Shi Shuo, who became interested in the towers Mudo (Dmu rdo) sacred Mountain still call it by the 3500 years ago. At that time fairly recently, has also been routing for a religious name of the native goddess Sar Rgyal mo rong the ancient Qiang were pos- Dmu rdo (Van Driem 2001, and personal obser- sibly not builders but herders origins of the Han dynasty period towers based as the Chinese character for on rather flimsy justifications. I answered his two vation) while today’s lore often considers Mudo their name suggests (Stein presentations regarding this topic in a 2011(a) pa- as a male mountain, the ‘devil’s stone’ (G.yung and most Chinese scholars), and Ren 2011, and personal observation). I started we can only speculate where per “Are Towers Needed to Reach God in a World and when and eventually from of High Mountains and Omnipresent Spirits”. Some collecting legends in the late 1990s; I will only re- whom they acquired their re- of my 2011a papers’ arguments are included in cord here the ones relative to the Kongpo/Nyangpo nown building skills. These towers, and as I explain in my 2005 book, some of skills are not limited to stone this paragraph. buildings as the Buwa earthen these legends are also mentioned by Kingdon-Ward towers have demonstrated: in Even if the towers had never been studied, dated, and Chan. Towers are said to have been: spite of being located barely or even mapped as a group, as explained earlier, 50 km from the epicenter of • built by the indigenous Kongpo people to de- the devastating 2008 earth- they had been briefly mentioned for nearly 2000 fend themselves against the Tibetans; quake, they hardly suffered years in Chinese books. All the Chinese texts com- from it. • built by ‘demons’ headed by Dü Achum Gyelpo, menting about the towers’ uses had listed them as 12 I am not knowledgeable defense, storage, beacon and status symbol, but Ling Gesar’s enemy, this is why they are called enough to draw in depth com- religion was never cited, although it was some- dükhang (bdud khang) (demons’ houses); parisons but Huber … (continue following page) times stated that these towers were similar to • built by the Tibetans, 200 years ago, as a de- “Buddhist pagoda’s because one entered them by fense against the wild ‘Poba’ (the people from Fig. 29: their base”. More recently quite a few excellent Pobo, a region east of Kongpo); Loophole of a Zhaba tower. JCCS-a 8/2015 43 Frederique DARRAGON On the ancient cross-shaped towers of Nyangpo and Kongpo in eastern Central Tibet

… but Huber (2013, I am grate- • the ‘camping grounds’ of Songtsen Gampo’s Nyang River, ancient Nyangpo and Nyangpo ful to Tim Bold for sending me (Srong btsan sgam po) army; kingdoms this paper) does list similari- ties between the Qiang shibi • used for trapping large birds including garu- As one can see on the map, towers are found in priest-shamans’ symbolic para- das; great quantities in the Nyang River watershed. phernalia and that of the shen Mountains being natural frontiers ancient kingdoms (gshen) priests; Opptiz (2007: • used to channel smoke intended to call the in mountainous regions often corresponded to a 171) has found that the Qiang rain in times of drought. shamanic drums closely re- river’s watershed, it would consequently appear semble those of North Mongolia logical to broadly define the territory of ancient and of South Nepal. Mathieu 15 (2003: 97 to 142) argues Most of these shed little light on the towers’ pos- Nyang po (also called Myang po and Nyang yul, convincingly that the Dongba sible ‘raison d’être’. Three of these folktales are Hazod in Dotson 2009) as the Nyangchu’s and mid- religion of the Naxi, also be- incompatible with the actual age of the towers dle Nyang River’s watersheds16. Very little has been lieved to be of Qiang origins, is related to ancient Bon. In my and the two last ones are clearly ‘imported’ leg- written about the Nyangpo kingdom and Karmay, 2011b presentation I describe ends, the one about the garudas could have been as I believe, was a pioneer while attempting to map a 2000 year sacred three- imported from Yarlung Sokha or elsewhere in it in 1987 on his way to Kongpo. In a 2006 IATS legged guozhuang unearthed in the Min Mountains, which Central Tibet where the same fable exists (Gyalbo presentation following his first trip in the region, is engraved with drawings of et al. 2000, and Hazod 2014, comments to this Hazod draws a map of a fairly small Nyang po cen- three human (spirit-medium paper) and the one about ‘calling the rain’ prob- tered on Draksum Tso; but, based on ancient texts like) faces with feathers on their head (Zheng 1946: Plates ably comes from a drier region because Kongpo and locals’ interviews, he deals at length with the I and XIII); this could possibly and Nyangpo are rather rainy areas, lichen even connection between this region and the Mon. A con- relate to the feathers − ‘bird grows on trees. However, they do indicate that the nection that I had hinted about in my 2005 book, horns’, or grass, ‘planted’ on the crown of the head in Bon towers are not Buddhist monuments, as ‘demon’ we will come back later to that topic. and Tibetan rituals, more spe- is the usual epithet given to non-Buddhist popula- cifically those of the Drigungpa tions14. The two first legends suggest local people’s By all means understanding Nyang yul/Nyangpo (‘Bri gung pa) (Huber 2013: 279, note 52) located in Meldro resistance at being run by Tibet and/or becoming is not easy. To complicate matters, the river that Gungkar, just west of Nyangpo. Buddhist. A few Draksum Tso people do believe that comes from the north and pass along Nyangpo “the towers were built by their ancestors”, without village (and near a mountain home to the local 13 In relation to the Ran Mang, the Eastern Han Book volunteering any specific use. deity Nyang Lha as located by Hazod) is called also mentions that power is Nyangchu by the locals but is not really the source transmitted through the fe- Today many of the people touring Tibetan cul- of the Nyang River (see note 16); in fact the Nyang male line and women are held in high esteem. tural areas, be them Chinese or Westerners, are River (i.e. Nyang chu but also called Gyamda chu) Buddhist, and folktales are being reformulated to originates in the Mila (also Mira) Mountains near 14 There is also a “demon please the religious tourists. I recently heard a new the Kongpo Pala pass and flows west (Karmay and mountain” west of the confluence of the Tsangpo and the version of the towers’ usage: a nesting place for personal observation) to join the Nyangchu River Nyang River. During my 2008 garudas. in a tiny outpost called Gyamda about 20 km up- visit to the Chims tombs of stream of the Chinese township of Gongpu Jiangda Leb (Sleb ri), locals informed me that it is only recently (Tib. Kongpo Gyamda). The north side of the mid- that people started living in CAN THE TOWERS HELP US DRAW dle Nyang River’s watershed is composed of three that valley, in the past these NYANGPO’S AND KONGPO’S BORDERS AND very different valleys, which are, from east to west: impressive constructions were called “demon tombs”, well in ILLUMINATE THESE KINGDOMS’ HISTORY? Nyangchu, which seems to have been deforested line with the fact that early Eastern Central Tibet towers were built during at long ago17, Drugla where some forest coverage still Chims people were certainly least an 1100 year span in the course of which exist and Draksum, home to both a rich soil and not Buddhist as they were said to be “frog and fish eaters”. borders certainly experienced wide variations. How- superb forests. The Draksum Valley is graced by ever, the fact that the towers cover such a limited the famed and stunning turquoise colored Draksum 15 For the sake of clarity I territory must be meaningful, and their existence Lake. In his 2006 presentation Hazod evokes the will use Nyang po for the kingdom during the Tibetan Empire could provide new clues to be added to myth collec- three ‘trade doors’ of Draksum Tso, but Nyangpo’s and Nyangpo the rest of the tion and textual information to possibly shed light two important trade routes from Chamdo to time. on the local history. Lhasa, pass either through Drugla or Nyangchu 16 As we will see later, the part called Nyangchu by the local people is not the upper Nyang River.

17 Wangdu (2006) recalls texts dating from the early Qing dynasty indicating that fire-wood was already scarce at the time.

Fig. 30: Map of towers in Nyangpo and Kongpo. Fig. 30 44 JCCS-a 8/2015 Frederique Die frühen kreuzförmigen Türme von Nyangpo und Kongpo im östlichen Zentraltibet DARRAGON

Valley. However, as numerous impressive tow- of the Tibetan Empire and Kongpo the rest of the 18 Because in this paper I give very many times well ers stand in the Draksum Valley, their existence time) are even more elusive than that of ancient deserved credit to Hazod, I is possibly best explained as local trading posts. Nyang po. Today quite a bit of confusion springs cannot refrain from also adding From Nyangchu Valley there is also a well-used from the name of the Chinese administrative county that he occasionally used some of my towers data without giv- trail to Meldro Gungkar (Mal gro Gung dkar). As called ‘Gongpu Jiangda’ (Kongpo Gyamda) which ing me credit. large batches of fairly rich agricultural land and a in fact only encompasses the watershed of the good number of towers are also found in the south- high and middle reaches of Nyang River. The lower ern part of the middle Nyang River watershed, one reaches of the Nyang River and the part of the would suppose ancient Nyang po and later Nyangpo Tsangpo watershed near its confluence with the to both have encompassed that area. From at least Nyang River are now Milin and Linzhe counties. three of these valleys fairly accessible southern But Kongpo Gyamda was in fact created long ago, passes lead to the northern bank of the Yarlung when, as explained by Wangdu, the administration Tsangpo, which is usually considered as being part of the 5th Dalai Lama included Nyangpo into Kongpo of Kongpo (Fig. 30). as one of its four dzong. This is why Kongpo is the location of the part of the Gesar Epic which takes In his 2009 Tibetan Empire map, Hazod includes place near Draksum Tso and why Kingdon-Ward, the three main valleys of the northern part of the who visited the lake and the towers area in 1924, middle Nyang River’s watershed in Nyang po but also calls the whole region Kongpo. Hazod was not the ones of the southern part. However, as mistaken when he stated that the “Kongpo towers practically all the towers still standing that I have should be called Nyangpo towers”, not only because dated (to the exception of one or possibly two lo- as we will see, some towers are definitively outside cated north of Nyang po village) were built after the of Nyangpo, but also because, as explained above, Tibetan empire, we cannot possibly try to delimit Nyangpo was actually part of Kongpo since the time ancient Nyang-po based on the still existing towers. of the 5th Dalai Lama.

In years following his 2006 presentation, Hazod, In 2008 I was astonished to see, from a large bridge with whom I sometimes had a conversation, has crossing the Yarlung Tsangpo River, a cross-shaped written at length about the Nyangpo-Mon connec- tower standing in a small village and in full sight on tion. Writings on which I have relied broadly18 as I the north bank of the river. I had already been on can only speak basic Tibetan and cannot read it (but that bridge, which crosses the mighty Tsangpo in I usually travel with an old friend of mine, Penba front of the Milin airport, a few times; but the tower Tashi who is also fluent in Chinese and English). had previously been hidden from view by another I need to explain here the reasons for the publica- construction that, apparently, had just gotten torn tion of my 2005 book as bilingual Chinese-English down (Fig. 31). Right after that discovery I went book. This was not intended as an academic book to search for towers in the other villages along the although it did represent seven years of research. rest of the north bank. Having heard that it was I was mainly aiming at protecting the towers and forbidden to destroy them, people were originally that could only be achieved by a close cooperation reluctant to speak but they finally open up and I with the Chinese government and the involvement discovered close to a dozen of ruins. I was told of Chinese scholars. that all of them had been destroyed “decades ago”. As the whole text had to be authorized by the Chi- nese censors, I had to use, in the English part, Some local officials explained that many towers the pinyin version for all the words. As most of my could have collapsed in 1954, as the region was publications were primarily meant for a Chinese rattled by a very powerful earthquake. But such audience I generally continued to use pinyin in their towers have well proven their resilience to earth- English versions. It was important for my book quakes and I believe that ignorance, neglect and to be sold in ordinary Chinese book shops, as mainly development were to blame. books mentioning Tibet and written by westerners are usually only sold in special international The tower still standing in Seme had been saved book shops. My book and the rest of my promotion from destruction by a small shrine erected next to campaign did achieve their goals and did get these it. Folks had refrained from blowing up the tower towers known and protected. because they feared damaging the shrine. So they were letting time, wind and rain, slowly erode the Kongpo, Rkong po, and the discovery of tow- tower until it would become feasible to dismantle ers along the Tsangpo River it by hand. When I first saw the tower it was still The borders of ancient Rkong po (also called Rkong fairly complete even if some of its corner stones yul, Hazod 2009, and, for the sake of clarity I will had already been looted. use Rkong po to designate the kingdom of the time

Fig. 31: Seme tower in Kongpo as seen from the Milin Bridge crossing the Yarlung Tsangpo.

Fig. 32: Unusual lintel of Seme tower. Fig. 31 Fig. 32 JCCS-a 8/2015 45 Frederique DARRAGON On the ancient cross-shaped towers of Nyangpo and Kongpo in eastern Central Tibet

19 This fable is clearly an- It was identical to the Nyangpo towers but for one Epic the Kongpo’s demon is Dü Achun Gyelpo (Dü drocentric, misogynistic (Gyalbo et al. 2000: 56), and a promo- difference: the lintel above its door (Fig. 32) whose A khyung Gyel po) and Ling Gesar is a Buddhist tion of patriarchal values and beams had a somewhat stair-like configuration in- champion who fights him. As Ramble writes (in political organization, it recalls stead of being set on one single level as were the 1997, note 98, page 198 citing Stein) “one version both the Eve and the Pandora myths in which the overly cu- lintels of all the other towers. Possibly that would of the Tibetan account of the appearance of man rious and sexually enticing make the door harder to ram. It would seem that has it that the first parents − the monkey and the females need to be controlled this kind of lintel was a local Kongpo variation as demoness19 − had their critical meeting at Bresna by men and religion; a similar narrative is that of the original another Kongpo ruin appeared to have had such in Kongpo”, even if the current tradition sets this demoness of Tibet whose sin- a feature; yet, all the ruined Kongpo towers were meeting in Yarlung. But more importantly Kongpo is fulness, as J. Gyaltso in 1987 just barely recognizable and, with the one excep- the site where Nyatri (Gnya’ Khri) had, in Karmay’s demonstrates masterly, must be contained before the coun- tion mentioned above, the lintels had completely words (1998: 222) “to abandon his culture of the try can be civilized. Based on disappeared. I was able to collect wood in the Seme gods and adopted that of a man to become the my personal research I have tower − sample 254459 which yield a calibrated first king of the black-headed people”. It is also the established that the monkey 20 ancestor is an extremely rare result of AD 1320 to 1350 and 1390 to 1440 − as location where new funeral rituals were invented , myth world-wide but today is well as in one of the ruins, sample 260786 which leading indirectly, so it is said, to the re-establish- common, albeit with very dif- yield a calibrated result of AD 1240 to 1300 and ment of the Yarlung dynasty21. Not only had the ferent scenarios, to 17 of the 56 Chinese official minorities. 1370 to 1380. Both are two of the most recently rituals to be reformulated (possibly because of a Writings by Stein (1958: 5) built cross-shaped towers. change of ethnicity/beliefs of the ruler?)22 but this and by Eberhard (1968: 51 to was following a ’ransom‘, a price to pay to allow 53) confirm the ancientness of these creation myths and in I could not find any tower or any ruin of tower on the Yarlung dynasty to continue. The myth here the “History of the Northern the south bank of the Tsangpo but that region is evokes a total paradigm shift of major importance. Dynasties” of the 7th century much more developed, with many roads and towns, Nevertheless, we know that tombs are found all it is written that tribes de- scending from the San Miao so possibly the towers, if they ever existed, had over Tibet, possibly many of them do predate the — considered today to be an- been destroyed ages ago. Local inhabitants, most Yarlung tombs although most have not yet been cient Qiang — say that they of whom are not native, did not remember any dated reliably (Aldenderfer, 2013: 294 and 313). were “originating from the mi monkeys”. Gyalbo, Sørensen tower but were eager to speak of the “ruins of the and Hazod (2000: 51, note Kongpo king’s castle” (where I could not find any But, surprisingly, this ‘mythically charged’ region 40), state that the monkey wood); it is not located in Bresna, the traditional was at the same time referred as inhabited by “bar- of the rocks tale literary form dates from the 11th century center of Rkong po, but in Rabkha. We were also baric populations who practiced incest and murder” (adding that it had probably told that the “tombs of Kongpo kings were in Rab- (Karmay 1998: 216; Ramble 1997: 144–146), and been formulated in the later kha”, but although we searched the region for a full even were “semi-human cannibal beings” (F.W. phase of the dynastic period) while Vitali (2003: 37, note day we could not discover any tombs. Not that I Thomas 1948: 15). Partick Mansier adds, in his 1) mentions its “overtly Bud- am any expert to recognize collapsed tombs. Some 1983 thesis, that many of the inhabitants of pe- dhist treatment”. Therefore I people said that they had been desecrated and de- ripheral regions reported to still be qualified of la am inclined to believe, as I argued in a 2008 presenta- stroyed long ago, a few also mentioned a “king of lo (kla klo) –barbarians- by the people from Cen- tion in Shanxi, that the widely Kongpo statue” at an uncertain location. tral Tibet. This reputation endured until recently: popular monkey/demoness in 1997, during my first trip to Draksum Tso, I was patriarchal Himalayan Mon- key Creation Myth is in fact a The Kongpo riddle and the connection informed by the Chinese driver of the old jeep I had Buddhist reformulation (“mod- between the existence of towers and the re- rented in Kongpo Gyamda not to accept any drink eled upon passages found in gion’s history from the locals “who have the habit of poisoning the Rama” as explained by Gyalbo, Hazod and Sørensen Books written during the Sui and the Tang Dynas- outsiders”. Apparently quite worried he had insisted (2000: 54, note 50) of a more ties describe numerous towers in the Min, Dadu and that I buy a few bottles of water before setting out. ancient less mysoginistic mon- Yalong Rivers’ watersheds, the same regions where key/princess Qiang myth that was probably the original plot towers are found today, yet, even after having dat- Sørensen and Hazod (2005: 222) argue quite con- for the monkey ancestor re- ed 54 Sichuan towers, I did not find a single one vincingly that the region of eastern Central Tibet lated legends of present-day from these times; the earliest towers still standing was chosen because it represented the most favora- Chinese minorities. in Sichuan date from the 11th century. The older ble natural environment and because the Tibetan 20 If I understand well one towers must have collapsed, or, once in poor condi- historiography identifies the monkey descendants of my communications with tion, they had been destroyed to build new ones. with the Lhopa and Monpa hunter-gatherers; the Samten Karmay, the word ’rope‘ is not mentioned in an- route from Kongpo to Yarlung would then describe cient texts, only a sort of mu We have already established that, as there is only the transition from savagery to civilization. But ’rope concept‘, consequently, one tower dating from the Tibetan empire period, these people were certainly not hunter-gathers could it be possible that earlier kings were in fact cremated these constructions cannot help in defining ancient (probably just ’foreign‘) since they knew how to and the ’rope‘ would have been Nyang po’s frontiers. The case of Rkong po/Kongpo build stunning towers. Also Rkong po was possi- a metaphor for the smoke is more complicated still: firstly the Kongpo towers bly independent as it is not even included in the carrying what was left of the body to the sky? Smoke is said are of the most recent ones and they only exist geographical and military divisions of the country “to escape from the Door to in a small part of Kongpo, (rather similar to the belonging to the Gyo ru division (Karmay 1998: the Sky” of tents and ancient Kong po stod as defined by Hazod in 2006) but 214, referring to Uebach). Karmay attribute this houses, as it often represents the ‘Axis Mundi’ connecting more importantly it seems that no towers ever to a special relationship with the Tsangpo and the the three cosmic zones (as stood in the lands corresponding to ancient Rkong 9th century Kongpo inscription does appear to con- explained by Eliade in 1975) po. Even if there has been quite a good deal of firm close ties between the Rkong po kings and the common to many Asian my- thologies. ’narratives migration‘ in and out of it, there is no Yarlung dynasty. Then, were the ancestors of the doubt that Rkong po/Kongpo has an exceptional Tsangpo the towers’ builders? On the other hand, 21 Or possibly was it that place in the Buddhist and the Bon Traditions as legend has it that Songtsen Gampo was born in the Yarlung Civilization as such in fact only started at that well as in the Tibetan folklore. Then, unless towers previously mentioned Meldro Gungkar, a place time? Cuevas, in 2003, com- were only related to trade and trade routes (and where there are no towers; and Stein (1972: 79) ments “the myth of the Emperor in this case they might not even have been built writes of Kongpo becoming accessible when, in the Drigum’s death points symboli- th cally to this great political and by Kongpo people) it is hard to correlate the re- 15 century, Thangtong Gyalpo (Thang strong rgyal cultural transition, the founding markable prominence of this small kingdom with po) “opened the route to Kongpo” by building a of the Tibetan Empire, which had the lack of towers in what we consider to be its bridge; the exact time when tower building stopped culminated with the introduction of an entirely new ritualized ancient territory. in Nyanpo and Kongpo. perspective on death and the dead”, before quoting Haarh In the Bon Tradition Kongpo is home to the demon As I am not an expert, I will not get in the details “this change of the concept of the mutual relation between Khyabpa Lagring (Khyab pa Lag ring), the most of all the narratives, myths and legends attached to Living and Dead presented ferocious opponent to Shenrab (Gshen rab) the Kongpo and Nyangpo, but I will list a series of co- the spiritual background for said founder of the Bon religion, while in the Gesar incidences and replicate names in different places. the rise of the Dynasty”. 46 JCCS-a 8/2015 Frederique Die frühen kreuzförmigen Türme von Nyangpo und Kongpo im östlichen Zentraltibet DARRAGON

The reason why I include this part in my paper south of Draksum Tso, villagers will show you the 22 Snellgrove (2003) writes essentially dealing with tangible heritage is that I ruins of Mon Shingtri Gyalpo’s castle, other villagers “This particular story may represent some memorable crisis find astonishing that such towers only exist in two will inform you that it was Gesar’s castle given that when the established rulers small remote kingdoms whose history is so closely, the foot print of his horse is not far. Hazod goes were worsted by an invader and in so many different ways, linked with the most on to explain that the “main seat of Mon Shingkhri from outside”. important Bon and Buddhist Tibetan myths. That Gyalpo” is considered to be near the Tsari moun- 23 There once were six of this would just be a coincidence seems a bit too tain, and that “renowned Tibetan clans come from them grouped by three but the farfetched. this large Mon (forested) area”. Vitali (2003: 50) northern ones were destroyed long ago, I have never seen recalls a claim from the Rgya bod yig tsang that them. As stated by Vitali (2003: 42) all the Tibetan an- “Kongpo was of Monpa origins”. cestral tribes came from the north, but in the myth 24 Obviously Chinese char- acters do not always reflect of the appearance of the first king, Nyatri tsenpo The term ’Mon‘ is considered a blanket term, exactly the local pronunciation, (Gnya’ Khri btsan po) came, in fact, from the south mainly because the different Mon speak appar- and also Chinese cartographers since, when he first descended from the sky, he ently unrelated languages. In my 2005 book I would not use the same name for two different rivers. landed in Lhari Gyangdo, convincingly located in had already noted a possible connection between Kongpo by Stein and Karmay although it was often Kongpo/Nyanpo and Mon or Monpa, based on an- assumed that this mountain was in Yarlung of Cen- cient Chinese texts quoted by Chen Zongxiang and tral Tibet. But, as Stein also explains, the name of on my observation of near identical outfits worn this first king is also spelled Nya Khri. A name that by the Draksum Tso people and the Monpa of the reappears many generations later as the name of Monyul corridor (including the peculiar ’drip-hat‘); one of the sons of Drigum − also Grigum, Drigung I also wrote some about it in one of the papers I − who, depending on the texts, had two or some- presented to the 2009 IUAES in Kunming but there times three sons. After the death of their father, is no space here to debate this topic. Still I would deprived of his divine quality and of the mu rope like to mention a quote by Vitali (2003: 42) where to ascend back to the sky, the three sons left for he states that some ancient Tibetan texts report Kongpo, Nyangpo, and Pobo. The tradition is not re- that “there were three Monpa ancestral tribes of ally clear about which son went where. But Wangdu Tibet: the Mon’s own race, the Minyag at the border writes that both Kongpo and Nyangpo kingdoms between China and Tibet and the Mon of Kongpo”. were each ruled by descendants of Drigum’s sons Today some Minyag are found living together with until the 14th century. Karmay demonstrates that Monpa in Tsona (Minyag Chokyi Gyaltsan (Chos Drigum’s tomb is in Miyul gyi thing (skyid thing), kyi rgyal mtshan), personal communications). The on the east bank of the Nyang River, just before it connection between towers, Mon/Monpa, Minyag runs into the Tsangpo. It is also recalled (by Kar- and Kongpo certainly deserve a more in depth in- may) that the first Tibetan king, after confronting vestigation that I am not capable of carrying out. and overcoming the people from Nyangpo and from Kongpo, went to Mi yul gyi Kyi thing; there is an- To add to confusion, it is said in some Chinese other Miyul north of the Nyanpo village in the valley and Tibetan texts that the original inhabitants of where Hazod has located the tomb of the Nyangpo Kongpo and Nyangpo were the Lhopa (but pos- king. That legend is well spread and I had also sibly Chinese writers just copied that from Tibetan heard of it as I was investigating the towers. The texts). Today’s Monpa, Lhopa and Tibetans do towers (which are in fact only two and not four as share a large part of common genetic material drawn by Hazod on his 2009 map of the region)23, (Kang L.L. et al. 2011: 94 and 95; Qi X.B. et al. are located near a village named itself Drigung or 2013: 1773). This possibly elucidate why the Mon Grigung which sounds both as the ancestral king do not figure as one of the four original people and as the renown 12th century Drigung monastery (mi’u rigs), which, as Vitali (2003: 59) explains: (in Meldro Gungkar) easily reached by a western “did not belong ethnically to sPu rGyal Bod”, and pass across the Mila Mountains. As translated by were: Zhang Zhung, Sumpa, Azha and Minyag. my friend Penba Tashi who I generally traveled Vitali adds that they had “antagonized the sPu rGyal with, the locals relate the name of their village to tribes” and consequently “lost their traditions”. This either the king or the monastery depending of the is true for the formers but not for the later; Minyag person asked. Still, the locals do not pretend that people still do exist, speaking their own language the Nyangpo king buried nearby is Drigum as they and living in small pockets spread through Sichuan consider him to be a Mon king to whom I will re- (Darragon, 2003 documentary film; 2005: 99 to turn later. 111, and forthcoming) where they often, but not always, have built towers. The confusion continues between the two myths, as both the first Tibetan king and one of Drigum’s sons Most of the current inhabitants of both Kongpo and were first in Kongpo or Nyangpo (and/or Pobo), Nyango have immigrated there from other regions before moving on to Yarlung in Central Tibet; the of Tibet. It would appear that few people still speak divine legendary king to establish the Yarlung a native tongue; a thousand of them are living in Dynasty while Drigum’s son went to re-claim his three villages (Fig. 33) north of Draksum Lake (per- father’s ’helmet‘ and re-establish the dynasty. But sonal observation). there are also two Yarlung. The well-known one in Central Tibet and one northeast of Kongpo, in Pobo; The ‘Draksum Tso’ language they speak has only I have been in Pobo but not in that valley which is been briefly studied by Nicolas Tournadre years ago mentioned by Hazod (2009: 177) as Yar lung but its and I have been lobbying the Chinese government river is called ‘Ya Long Tsangpo’ on Chinese maps.24 to protect both the tangible and the intangible heritage of this area. Aforesaid, some Draksum people Ancient Pobo (Spu yul), not listed as an ancient do believe that “the towers were built by their an- principality, leads us to another group of people cestors”; unless the towers were only linked to the often connected − as we have seen − with Kongpo trade routes, the towers’ existence and location and Nyangpo: the Mon. According to a local legend can only be understood in the broader ethnologi- collected by Hazod, that of the ‘Forest Throne’, Mon cal context. Shingtri Gyalpo’s (Shing khri rgyal po) is described as both the king of the 18 tribes of Mon and as one of Drigum’s sons. In Lepa (Rjed pa), a small village JCCS-a 8/2015 47 Frederique DARRAGON On the ancient cross-shaped towers of Nyangpo and Kongpo in eastern Central Tibet

Fig. 33 Fig. 34

DIFFERENT USES OF THE TOWERS 950,000 dollars for their restoration, and quite a It is evident that these lofty earthquake resistant few towers have already been restored. I would say constructions are the testimony of a sophisticated that the situation of towers as a tangible heritage society preoccupied with more than survival, a is rather satisfactory. It is not so of the intangible unique culture worthy of research and protection. heritage; the unique culture of the populations liv- However, defining the exact function of each tower ing at the foot of the towers is fast disappearing is impossible and not of such importance, even if due to globalization and tourism, but mainly due this is the most commonly asked question. In their to a lack of non-menial ways to make a living while heydays, many towers existed, and different towers performing and therefore safeguarding their native had different uses, as indicated by their different skills. In cooperation with Planet Finance we had a architectural characteristics and by the differences 2009−2012 ethno-ecotourism project which was to in location within each local environment. be mostly funded by a 600,000 Euro grant from the EU, unfortunately the project had to be cancelled No relics have been found in any of them. With the because of the 2008 earthquake and political tur- exception of the Qiang ones, the towers have not moil in Tibetan areas. been used for hundreds of years, but children have played in them and probably removed any ancient In cooperation with the PHRCA of Rinpoche Minyag artifact. It is possible that soil analysis would reveal Chokyi Gyaltsen, we have already restored four if the towers have been used for storage, and for ancient towers, we also bought or leased three what kind of goods, but I have not been able to ancient houses that after restoration were given organize it. to the villagers to use as community centers. In cooperation with local people we have also restored The precise utilization of the Brochs, built between one ancient house in Danba and will soon start the 200 BC and 100 AD in Scotland (Fig. 34) is also restoration of two Zhaba towers. still being debated by scholars and most doubt that there was ever a single common purpose for their construction; the list of possible uses includes CONCLUSION defense, control of strategic locations, storage, Even if a few of them were ‘in full sight’ the towers beacon and status symbol; these functions were had failed to interest any scholar until I embarked likely to often be combined and are the same as on my quest. They possibly were considered un- the ones listed, by the Chinese scholars, for the important because not mentioned in any Tibetan Sino-Tibetan Marches towers. As I uncovered that book but more probably because their uniqueness the most flamboyant of these towers were situated can only be understood by a conceptualized com- on ancient trade routes I added defensive storage parative architecture survey of local variations and trading post to the list. cross-cultural analysis of ancient towers world- wide. That also meant crisscrossing for months at a time, by then almost non-existing roads, the huge PROTECTION AND CONSERVATION. swatch of impossible terrain of, as Oppitz (2007: To promote knowledge about the towers and in- 171) writes, “the little known areas of the Sino- terest in their protection, I embarked, starting in Tibetan Marches”. 2001, in a world-wide promotion campaign, starting by filming a documentary (broadcast by Discovery In that same 2007 paper, Oppitz advocates com- Channel, profits were deposited on my foundation, parative ethnography because “so we have, after the Unicorn Foundation, then writing a book, giv- 50 years of Himalayan ethnography, many small ing lectures in different countries including China universes […] but there are few glimpses over and holding many photos exhibits including one at neighboring fences” and he continues “the expres- the United Nations in New York in 2004. I distrib- sion ‘local culture’ has not had the same success uted many reports in Chinese to SACH (Chinese in academia as ‘local knowledge’, even though the State Administration of Cultural Heritage), and local later presupposes the existence of the former”. governments (the bi-lingual version of which I will The Sino-Tibetan Marches towers perfectly embody soon upload on academia.edu). I lobbied both the both local knowledge and local culture and they UNESCO World Heritage Center and SACH; it did are the perfect realm for comparative ethnology result in some towers being inscribed on the World with considerable material culture, as their territory Heritage Tentative List in February 2008, then in stretches, as the crow flies, from 100 km northwest Fig. 33: Tsongo village at the northern 2013 the rest of the Sichuan towers were listed. of Chengdu to 200 km east of Lhasa. No civiliza- end of the lake. tion was created in a vacuum, each is the result Most of the towers are now protected, at the provin- of cross-pollination between different cultures and Fig. 34: Dun Telve broch, Glenelg in cial and even the national level for some of them. choices made by each society regarding what to Scotland. The Chinese government has earmarked about include in their mores. And these extraordinary 48 JCCS-a 8/2015 Frederique Die frühen kreuzförmigen Türme von Nyangpo und Kongpo im östlichen Zentraltibet DARRAGON towering structures are not only unique from an architectural standpoint but in another notable way: they are lay and vernacular, contrarily to most flamboyant constructions.

These towers are more than worthy of the world’s interest and, independently of any political agenda, the people who possibly built them merit our atten- tion; the current inhabitants also deserve to know that they are not ordinary peasants but the descendants of builders of exceptional talent. It is our common responsibility to do our best to safeguard not only the tangible but also the intangible culture of these amazing cultural landscapes.

The two regions of Nyangpo and Kongpo, where unique cross-shaped towers are scattered in lands bursting with myths and folktales of essential importance to the Tibetan Civilization, definitively demand a deeper study that I am not capable of carrying out. This paper, within its limited space is a sort of update of my 2005 book and will be followed by a lengthy detailed manuscript that I am currently writing. Now that the towers are mapped, dated, documented and protected, I hope that Tibetan experts will investigate this ’tower- connection‘ which might illuminate other parts of Tibet’s history and culture.

REFERENCES Aldenderfer, M. (2013). “Variation in mortuary practice on the (2009b). “The Civilization of the Star-shaped towers of the early Tibetan plateau and the high Himalayas”. JIABR, Ethnic Corridor of South-west China”. Presented at th Vol. 1, Issue 1, Montreal, The International Association the 16 World Congress of IUAES, Kunming, proceed- for Bon Research, 293−318. ings published in Chinese in Ethnic Group. Settlement. National Architecture, National Architecture Institute Aris, M. (1980). Bhutan: The Early History of a Himalayan Kingdom, of China (ed.), Kunming, Yunnan University Press, Warminster, Aris & Phillips Ltd. 22−37.

Bellezza, J. (2009). Newsletter Flight of the Khyung [Online]. Avail- (2011a). “Are Towers Needed to Reach God in a World of able: http://www.tibetarchaeology.com/august-2009/ High Mountains and Omnipresent Spirits”. SUUH In- (Accessed: autumn 2009). stitute Newsletter, 1−9. Soon to be available online at Academia.edu. Bernier, R. (1997). Himalayan Architecture, Canada, Associated

University Press. (2011b). “Cultural Markers of the Civilization of the Star- Shaped Towers”. Lecture at Sichuan University, Book of Sui (Sui Shu compiled in the 7th century), Volume 29, Wen- December 2011. Soon to be available online at Aca- shan Prefecture section (In Chinese). demia.edu. Chan, V. (1994). Tibet Handbook, California, Avalon Travel Pub- Deng, et al. (2004). Evolution and migration history of the Chinese lishing. population inferred from Chinese Y-chromosome evi- Chen Zongxiang, (2004). The Study of Blockhouses in China, un- dence [Online]. Available: http://www.genebase.com/ published. learning/article/24 (Accessed Mars 2013).

Cox, K. (ed.) (2001). Frank Kingdon Ward‘s Riddle of the Tsangpo Dotson, B. (2009). The Old Tibetan Annals, Vienna, ÖAW. Gorges, Suffolk, ACC Distribution. Eastern Han Annals (“Hou Han Shu” compiled in the 5th century), Cuevas, B.J. (2003). The Hidden History of the Tibetan book of the volume 86: Treatise about the Nam Man Xi Nan Yi, Ran Dead, Oxford, University Press. Mang Yi section (In Chinese).

Darragon, F. (2003). Secret Towers of the Himalayas [Documen- Eberhard, W. (1968). The Local Cultures of south and east China, tary film], Discovery Channel. Leiden, E. J. Brill.

(2005). Secret Towers of the Himalayas, Shenzhen, Shen- Eliade, M. (1975). Myths, Dreams & Mysteries, Scranton, USA, zhen Media Group Publishing House. Harper Collins.

(2008). “The Civilization of the Star-shaped Towers as a Leg- Graham, D. (1958). The Customs and Religion of the Ch‘iang, acy of the Ancient Qiang Tribes”, presented in Chinese Washington, Smithsonian Miscellaneous Collections, by Sun Ji at the Conference: Protecting the Heritage Vol. 135-1. after the Earthquake, National Architecture Institute Gyalbo, Hazod and Sorensen (2000). Civilization at the Foot of of China and Construction Bureau of Shanxi Province, Mount Sham-po. The Royal House of lHa Bug-pa-can Tai Yuan, Shanxi Province. and the History of g.Ya’-bzang, Vienna, ÖAW.

(2009a). The Lang Xian Thousand Year Old Grooved Tow- Gyaltso, J. (1987). “Down with the Demoness: Reflections on a ers, SUUH Institute 2009 Newsletter, 1−5. Soon to be Feminine Ground in Tibet”. The Tibet Journal XII-4, available online at Academia.edu. Tibetan and Himalayan Library, 34−46. JCCS-a 8/2015 49 Frederique DARRAGON On the ancient cross-shaped towers of Nyangpo and Kongpo in eastern Central Tibet

Handa, O.C. (2011). “Himalayan Domestic Architecture”. History of Snellgrove, D., Richardson, H. (2003). A Cultural History of Tibet, Science and Technology, Chapter 9, Publisher Infinity Bangkok, Orchid Press. Foundation Princeton [Online]. Available: http://www. Sørensen, P.K., Hazod, G. (2005). Thundering Falcon, Austria, indianscience.org/reviews/Hist%20book%20draft%20 ÖAW. 11_5_05.pdf (Accessed: August 2013). Stein, R.A. (1951). “Mi-nag et Si-hia, Geographie Historique et History of the Northern Dynasties (“Bei Shi”, 659), Volume 96, Bi- Legendes Ancestrales”. Bulletin de l’École française ography 84, Section Di, Tuyuhun, Dangchang, Dengzhi, d’Extrême-Orient, XLIV, Paris, 223−265. Bailan, Dangxiang, Fuguo, Jihu (In Chinese). (1958). “Les K’iang des Marches Sino-Tibétaines”. Annuaire G.yung ‘brug, Rinchen rdo rje (2011). “A Powerful Hero and a de l’École Pratique des Hautes Études, Ve section, Paris, Mountain Deity”. Asian Highlands Perspectives, Vol. 10, 3−15. 73−98 [Online]. Available: http://www.plateauculture. org/writing/dmu-rdo-powerful-hero-and-mountain- (1961). “Les Tribus Anciennes des Marches Sino-Tibétaines”. deity (Accessed: May 2013). Bibliothèque de l’Institut des Hautes Études chinoises, Frederique DARRAGON Paris, Vol. XV. Hazod, G. (2006). “The Land of Shing-khri btsan-po – A Survey Frederique Darragon is of Ancient Nyang-po in Eastern Central Tibet”. Paper French but has lived in many (1972). Tibetan Civilization, Stanford, Stanford University countries and considers her- presented at IATS, Bonn. Press. self a Citizen of the Planet. She holds a master in In- (2007). “The Grave on the Cool Plain: On the identification of (1987). Le Monde en Petit, Paris, Flammarion. ternational Economy from ’ Tibet’s first tomb‘ in Nga-ra-thang of ‘Phyong-po”. Pa- Nanterre University, Paris, pers dedicated to Ernst Steinkellner on the occasion of Sun, H.K. (1989). “A Preliminary Investigation into the Relation- and is also a self-taught oil th ship Between Qiong Long and the Languages of the painter, a sailor who crossed his 70 birthday, Part 1,B. Kellner et al. (eds.), Vienna: the Atlantic, a jockey of Eng- Wiener Studien zur Tibetologie und Buddhismuskunde Qiang Branch of Tibeto-Burman”. Linguistics of The lish thoroughbreds, a polo 70(1), 259−283. Tibeto-Burman Area, University of California, Vol.12- player who kept 2 records for 1, 92−109. 15 years, a photographer, a documentary filmmaker and Huber, T. (2013). “The Iconography of gShen Priests in the Ethno- a philanthropist. In 2001 she graphic Context of the Extended Eastern Himalayas”. (2007). Chinese Languages, Beijing, The Commercial Press created the Unicorn Foun- IITBS Nepalica-Tibetica Festgabe for Christoph Cup- (In Chinese). dation, mainly dedicated to pers, Vol. 1, 263−294. promote education, and in Thomas, F.W. (1948). Nam, an Ancient Language of the Sino- 2004 she co-founded the Tibetan Borderland, London, Oxford University Press. Sichuan University Unicorn Kang, L.L et al. (2011). “Y-chromosome O3 Haplogroup Diversity Heritage Institute to help rein Sino-Tibetan Populations Reveals Two Migration search, document and protect Tucci, G. (1988). The Religions of Tibet, Los Angeles, University Routes into the Eastern Himalayas”. Annals of Human minority cultures of south- of California Press. west China. Since the early Genetics, Blackwell Publishing Ltd/University College, 90’s she has spent more than London [Online]. Available: http://onlinelibrary.wiley. Van Driem, (2001). Languages of the Himalayas, Leiden, Boston, 6 full years crisscrossing rural com/doi/10.1111/j.1469-1809.2011.00690.x/abstract Koln, Brill. China and Tibet. In 1998 she started researching, docu- (Access: March 2013). Vitali, R. (2003). “Tribes which Populated the Tibetan Plateau”. menting, mapping and dating some little known ancient Karmay, S. (1972). The Treasury of Good Sayings, Motilal Banar- Cosmogony and the Origins, Lungta 16, Dharamshala, free-standing star-shaped sidass, Delhi. Amnye Machen Institute, 37−63. towers of the ethnic corridor of the Sino-Tibetan Marches. (1998). The Arrow and the Spindle, Kathmandu, Mandala Wangdu (2001). The Lake Basum Tso, Tibet People’s Publishing To promote the protection of these up to 170 feet tall Book Point. House, Lhasa. astonishing earthquake resistant constructions, she filmed Macdonald, A. (1971). “Une Lecture des P. T. 1286, 1287, 1038, Wangdu, B. (2006) “The History of Jiangda”. The History of Linzhi a documentary broadcast 1047 et 1920”. Etudes Tibetaines Dediees a la Mé- (in Chinese), Beijing, Chinese Tibetan Studies Publish- by Discovery Channel, gave moire de Marcelle Lalou, Paris, Adrien Maisonneuve, ing Company. lectures, held many photo exhibits including, in 2004, one 190−391. Wang, M.K. (1999). From the Qiang Barbarians to the Qiang Na- at the U.N. headquarters in New York and worked in close Mansier, P. (1983). “Lexique et Phonologie du Gyarong de Tsenla”. tionality: The Making of a New Chinese Boundary, collaboration with the UNESCO Doctoral thesis, Paris, Études des Hautes Etudes en Taipei, Institute of Ethnology. and the Chinese government. Sciences Sociales. The towers are now protected Zheng, D.K. (Cheng Te-k’un) (1946). „The Slate Tomb Culture of at the national and provincial levels and many are inscribed Mathieu, C. (2003). A History and Anthropological Study of the Li-fan”. Harvard Journal of Asiatic Studies, Vol. 9-2, on the UNESCO World Heritage Ancient Kingdoms of the Sino-Tibetan Borderland- Naxi 63−80. Tentative List. In cooperation and Mosuo, UK, USA and Canada, The Edwin Mellen with a local NGO employing local masons using traditional Press. techniques and materials her foundation restored 4 ancient Oppitz, M. (2007). “Close up and wide-angle. On comparative towers and 4 ancient houses. ethnography in the Himalayas and beyond”. European Following her thorough mul- Bulletin of Himalayan Research, 31, 156−156. COPYRIGHT tidisciplinary study of these near vanished monuments, All photos are by the author with the exception of she received an award by the Pope, A. (1965). Persian Architecture, New York, George Braziller. Fig. 10, courtesy of Maria Marechaux and Fig. 14, National Architecture Institute courtesy of Andrea Bruno. of China, was granted the title Qi, X.B. et al. (2013). “Genetic Evidence of Paleolithic Coloniza- of “Professor Honoris Causa” tion and Neolithic Expansion of Modern Humans on by Sichuan University and is the Tibetan Plateau” [Online]. Available: http://mbe. now a guest professor at Si- chuan University. oxfordjournals.org/content/30/8/1761.full.pdf+html (Accessed: May 2013). She is presently writing her second book about the towers Ramble, C. (1997). „The Creation of the Bon Mountain in Kongpo”. to re-count her more recent Mandala and Landscape (in Emerging Perceptions in findings; that will include her current research on an an- Buddhist Studies, No. 6) New Delhi, D.K. Printworld cient queendom, the “Eastern Ltd. Kingdom of Women” described in the Chinese Annals, and Shi, H. et al. (2008). “Y chromosome evidence of earliest modern whose people all lived in tall towers. human settlement in East Asia and multiple origins of Tibetan and Japanese populations” [Online]. Available: Contact: http://www.biomedcentral.com/1741-7007/6/45 (Ac- [email protected] cessed February 2013). 86-13880572748 50 JCCS-a 8/2015 Journal of Comparative Cultural Studies in Architecture JCCS-a

JCCS-a 8/2015 51