The Goddess Vajrayoginī and the kingdom of Sankhu (Nepal)
La déesse Vajrayogini et le royaume de Sankhu (Népal)
p. 125-166
Résumés
The creation of the Newari kingdom of Sankhu and its first king is mythically attributed to the goddess Vajrayoginī. The temple of this goddess is outside the central realm of the kingdom, on a hill nearby. Once a year, during her festival, she descends to the town in the centre of the kingdom. The aim of the article is to relate the inner structure of the kingdom to the shrine of the goddess outside. The first part is concerned with the mythical connection between goddess and kingdom, the second part with the inner structure of the kingdom and the final part with the ritual connection between goddess and kingdom, with reference to the myth. It is suggested in the conclusion that the world-renouncing character of creation-goddesses is a significant structural element contained in the conception of a Nepalese kingdom. But the significance of Vajrayoginī in “Nepal” is not restricted to the (re) creation of Sankhu. The character of the goddess and the practice of her worship depend on the degree of her conceived universality.
Cet article se propose d’établir les liens qui unissent l’organisation interne du royaume néwar de Sankhu au sanctuaire de la déesse Vajrayoginī, localisé à ses marches. Après s’être intéressé aux relations mythiques qui associent la déesse et le royaume, et avoir analysé les valeurs qui président à l’identité du royaume, le développement se propose de décrire l’assemblage rituel qui les réunit. La conclusion suggérera que le caractère ascétique de la déesse est un élément central à la conception du royaume népalais. Si l’importance de Vajrayoginī ne se limite pas au seul royaume de Sankhu, l’extension de son culte montrera différents ordres de références à l’universalité.
Texte intégral


Sankhu is a Newar town situated in the extreme northeast corner of the Nepal (Kathmandu) Valley. It owes its notoriety to the Tibet trade and to its association with the well-known shrine of the goddess Vajrayoginī on a hill nearby. The town’s origin may be dated somewhere in the 7th century A. D. after a trade-route was opened from Kathmandu through Sankhu to Kuti, Sigatse and Lhasa in Tibet. To this trade the town owes its founding and historical prosperity. But Sankhu owes its survival to agriculture which saved it from total economic coilapse after two other trade-routes to Tibet were opened, the first being an India-Lhasa connection via Kalimpong around 1850, the second a vehicular road from Kathmandu through Banepa to Kodari in 1967. Many traders shifted their business from Sankhu to Kathmandu, which strongly reduced the number of actuel residents in Sankhu (only 5 000 in 1978, estimated by the author). The shrine on the forested hill north of Sankhu, where the goddess Vajrayoginī resides, is much older than the town itself. Before it became the temple of Vajrayoginī it was a monastery (vihāra, bahāl), called Gun Vihāra (today still Gun Bahāl), one of the most famous monasteries from the time of the Licchavī dynasty (3rd-12th century A.D.), as an inscription found in Shanku from the time of king Anśuvarman (early 7th century) shows (Lévi 1905, III: 110) . Via this monastery, along with trade, Buddhism was spread from Nepal to Tibet. The word gun is a Kīrantī word meaning ‘ forested hill ’ and is used with that meaning in Newari till today. The Nepalese historian Dhanabajra Bajracharya considers Gun Vihāra as the main Hinayān monastery in Nepal. The ‘ Sankhu-inscription ’ mentions that a group of Mahāsāngik monks were living there. This was a dissident Buddhist group (expelled from the orthodox order) who rebelled against matters of discipline and exclusiveness preached by the orthodox. Such groups gave rise to a new Vehicle in Buddhism, Vajrayāna, in the 8th century A. D. The goddess Vajrayoginī became one of the principal deities in Vajrayāna and in Nepal her cult was centered on the hill above Sankhu. This hill, the Manigiri, is further associated with the Bodhisattva-king Maṇicūḍa who plays a part in the creation myths, as will be shown.1
1This study concentrates on the relationship between the goddess Vajrayoginī in her hill-shrine outside the realm of society and the value of kingship, the kingdom and the town in the ‘ valley of Sankhu ’. The goddess’s significance can be determined by studying the creation-myths and the structure of the kingdom, and is finally confirmed by the annual festival during which Vajrayoginī descends to the town.
1. The creation of the Sankhu valley
2The stories about the creation of the Sankhu valley are in certain respects a variant of the account of the creation of the greater valley of Nepal (Kathmandu valley), as contained in the celebrated Svayambhūpurāna. In one version of this purāṇa it is significantly predicted that the area of Sankhu ’ Through the power of the Ugra Tara [...] will be like a second Nepal This intriguing announcement provides the study of Sankhu with a broad cultural perspective as well as with a comparative element. What should be conceived by the universes of ‘ Nepal ’ and Sankhu? And in which way does Sankhu’s universe resemble that of ‘ Nepal’?
3The Svayambhūpurāṇa contains the description of the different phases of the grandiose creation of Nepal out of a mountainlake, in which the Adi Buddha as Svayambhū in his purest form (a flame) had manifested himself. The first phase is concerned with the draining of the lake through the cleaving of a mountain by Mañjuśrī so that the monsters disappear or can be controlled. The second phase deals with the ’ awakening ’ out of a source of the Adi Buddha’s female counterpart (the goddess Guhyeśvarī) through which the cosmic union of god and goddess is effected. With this act the emergence of a river is connected. After these two phases the dry land invites for settlement. The appointment of a king to rule over the new country concludes the creation. It is interesting to note that creation myths of other Himalayan mountain-kingdoms seem to have the same structure. Bosch (1968) has compared two of these (Nepal and Kaśmīr), an analysis to which will be returned later (see note 6).
4Oral tradition in the Sankhu valley has it that this valley too was originally a lake. Not Mañjuśrī but the goddess Vajrayoginī herself is said to have drained the lake by cleaving the mountain with her sword. The replacement of Mañjuśrī by the goddess appears to be more than an arbitrary idea of the villagers of Sankhu, since there are several similarities between Mañjuśrī and Vajrayoginī which are more than superficial. The attributes with which Vajrayoginī is associated are exactly identical with those with which Mañjuśrī had surrounded himself during the critical act Lévi 1905: 163). These are:
The sword, the fixed attribute of both Mañjuśrī and the Sankhu Vajrayoginī. The two statues of the goddess on the temple-complex are both equipped with a sword. The goddess is also called Khaḍgayoginī.
The two deities on both sides of the gorge (to prevent the hills from ‘ flying back ’). On the Chobar gorge (where Mañjuśrī cleaved the mountain) these are Varda and Mokṣadā. In the case of the Sankhu valley these are Kageśvarī on the Northern hill of the same name, and Kaleśvar on the Cāṇgu hill.
The five Dhyani Buddhas in state.
The lioness and tigress (goddesses). These served as ’ footstool ’ for Mañjuśrī. For Vajrayoginī they are the ever attending deities Siṃhinī and Vyaghrinī.
Prajñāpāramitā. Mañjuśrī carries the Book of Wisdom in his left hand. Vajrayoginī does not carry this Book, but in the Nepalese tradition she is Prajñāpāramitā herself, as the mother of all tathāgatas, ‘ the source of all knowledge and of all mortal Buddhas ’ (Dhooswan Sayami 1972: 40)1.
5By her association with Prajñāpāramitā, the connection between Vajrayoginī and Mañjuśrī becomes clearer, because as such the goddess Vajrayoginī is identified with Sarasvatī. Many Nepalese do not make a distinction between Sarasvatī, the ’ goddess of Learning ’ and Mañjuśrī, the ’ god of Knowledge This relates especially to the important Buddhist shrine behind Svayambhūnāth-hill, the idol of which can be either god. Both are believed to have come from across the Himalayas (Anderson 1971: 234). The iconography of Vajrayoginī does not help to clear this confusion. All the Sādhanamālā (meditation texts) describe Vajrayoginī (and Ugra Tara) as a terrible, awe-inspiring goddess (with a fierce face, protruding tongue, bare fangs, etc., cf. B. Bhattacharya 1958; Wenz 1958). In the beginning of this century, an incident noted down by Lévi (1905) proves that the villagers at that time had the same notion of Vajrayoginī as a terrifying deity since in a somewhat fierce Visnu-image, Lévi says, ‘ The Nepalese of today believe that they recognize in it Vajrayoginī ’ (transl. by Riccardi Jr. 1975 : 41). But the idols in the Sankhu temple do not represent an awe-inspiring goddess. Instead, the goddess is portrayed with an affable Bodhisattva smile, in a ‘ pleasant mood ’ and not naked (like a Tara but not Ugra Tara). In an attempt to combine the role in Creation played by the two deities, one source remarks reconcilingly: ‘It was (Vajrayoginī) who inspired Mañjuśrī to drain the waters of the Valley [...] ’ (Prusha 1975 : 46). Whatever is correct, it will be shown that creative qualities are essentially and unequivocally part of the character of Vajrayoginī.
6In the second phase of the creation of the Sankhu valley the Bodhisattva Maṇicūḍa seems to play an important part. While the original version of the Svayambhūpurāna has only a passing reference to the legend of Maṇicūḍa, a later editor has fastened upon this short reference to treat with Maṇicūḍa in more detail. Thereby he has obviously based himself in one of the different versions of the Manicūdavadāna, and has inserted in turn several new elements which lack in the classical avadāna 2. For this study an interest is taken in one of the additions, namely the one concerning the emission by Maṇicūḍa of the ‘ nine makāras ’, the objects whose name begin with the syllable ‘ ma ’.
7Apart from the makāras-addition the story of Maṇicūḍa is a legend very similar to other such stories which have the benefactions of the principal character as leading theme. Examples are the story of Vessantara and of the Tibetan Dri-med-kun-dan. There is also a Ceylonese variant (Handurukande 1967: xxxi-xxxii). The contents of the story of Maṇicūḍa are as follows. Prince Maṇicūḍa, born with a crest-jewel (maṇi) on his forehead which can effect miracles, is inclined to practice austerities in the forests of the Himalayas. He has to be tricked into accepting marriage and becoming king (of Sāketu). Many appeal to his feelings of generosity. Without fail he gives what people request, including his royal regalia (horse and elephant) and even his own wife and son. He prepares a great ‘ nirargada ’3 in which he sacrifices his own flesh and blood, but is restored to life A king Duḥprasāha (of Ayodhya) requests the crestjewel. Manicūda allows it to be taken out by force4. A river of blood springs from the wound, but again the gods take pity and restore his health. Retreat to the forest after the kingdom has been handed over to his son. In a future life Maṇicūḍa will be the mahābrahmā named Śvataketu, the king of the Dharmamedhā.
8The section on the makāras is added in the Svayambhūpurāna just before the episode of the extraction of the jewel from the king’s forehead. It was intentionally added to explain the origin of the nine objects, as the title of the manuscript (Paris, Dev 78) shows: ‘ Fourth chapter of the great Purāṇa of Śrīsvayambhū, on the origin of the pond Maṇicūḍa and the nine other makāras . The relevant passage is as follows:
‘ Maṇicūḍa says to Gautama, who had tried to dissuade him from practicing penance, ‘ Gautama, will not liberality, the creeper of plenty, bear every fruit? Look at its greatness’. Saying so, he struck the mountain. There arose a pond in the shape of a hand, full of water possessing the eight qualities. It was known in the world by his name [...]. Arriving at another mountain plain he created Manidhārā. He placed a doorkeeper who had the form of Mahākāla. Together with his pupil, he established the most noble Maṇicaitya and Maṇipurṇā, a yoginī, who was surrounded by faces of lions and tigers. Creating an enclosed place for his couch he meditated on the issuing forth of water. [...] A large amount of pleasing (water) flowed towards the right. As it pleased the mind, it was known as Manoharā. He placed the guardians of the quarters around. Like a second possessor of (things) charming, is the mountain Manicuda ’. (trans. Handurukande 1967: xxx-xxxiii)5.
Later follow the Maṇiliṅga, the river Maṇirohinī and the mentioning of a Maṇināga and the Maṇitīrtha on the confluence of the Manoharā and Manirohinī, which from this point on will be called Manmatī. This river is a tributary of the most sacred Bagmati.
9Thus it looks as if Maṇicūḍa has been given a creative role of great significance. He has prepared the establishment of the kingdom, which, in the text, he announces thus:
’ In the Kali age, there will be (a goddess) [...]. In this place there will be a great king of tīrthas. Through the power of the Ugratārā (this place will be) like a second Nepal. Sage, I shall live in this very place for nine years [...] ’ (op. cit., xl-xli, see also de La Vallee Poussin 1894: 310).
The significance of the last remark is mysterious, but the announcement links the Bodhisattva to the origin of a kingdom and a king. Without a doubt this kingdom is to be the kingdom of Sankhu and the king Śankara Deva, as will become clear later.
10It is this creative power (to bring forth a goddess, rivers, a fountain, a caitya and the like), which distinguishes the Bodhisattva Maṇicūḍa of the Svayambhūpurāna from the same king of the avadāna. To ascribe a creative role to a Bodhisattva or Guru is not unique, as the example of Mañjuśrī shows. In the case of Mañjuśrī there is a direct explanation since Mañjuśrī can be linked to Prajāpati, the worldcreator. This follows from a Tibetan story which mentions that in the beginning of the creation of this world Mañjuśrī had from his body a huge turtle emanated, which he later killed while it was floating on the primordial waters. The fact that Mañjuśrī killed his alter-ego, the turtle, in the creation process reminds one of the theme of self-immolation, according to Bosch (1968: 244-245)6, which is known from numerous myths and rituals of the Vedic period (i.c. the Purusasūkta and the Brāhmanas), in which the sacrificer is also the substance of sacrifice. Since a creative function is also ascribed to Maṇicūḍa one is very much tempted to relate this quality to Maṇicūḍa’s self-sacrifice during the nirargada, in which, as was mentioned before, he sacrified ‘ his own flesh and blood ’.
11In respect to the second phase of the creation of Sankhu valley, the awakening of the Adi Buddha’s female consort (connected with the emergence of a river), a closer look to the created makāras by Maṇicūḍa is necessary. In the quotation from the Svayambhūpurāna given before, it is told how after a pond, a tap and a ‘ doorkeeper ’ ‘ the most noble Maṇicaitya and Maṇipūrṇā, a yoginī who was surrounded by faces of lions and tigers ’, were created. There can be no doubt about the exact identity of the Maṇicaitya and the Maṇiyoginī in view of the accounts in the chronicles (vaṃśāvalī) and the temples to be found today on the Maṇigiri above Sankhu. The qualification ‘ most noble ’, referred to in a purāna dealing with the sanctification of the land by some holy manifestation of Svayambhū already points in the direction of the Ādi Buddha Svayambhū. Indeed, the Manicaitya is an earlier name for the Svayambhūcaitya called Padma Giri Dharma Dhattu Caitya which can be found in the smaller pagoda next to the ’ Red Faced Goddess ’ – temple of the Vajrayoginī temple complex (originally caitya referred to a mound or tree, later to a memorial shrine; in Nepal a small stone Buddhist monument with the shape of the big Nepalese stūpas, here referring to a brass specimen with the shape of the Svayambhū stūpa). A replica of this caitya, carried on a separate khat (palanquin), accompanies the goddess down to the town during her festival. The origin of this caitya is attributed to king Māna deva (464-505 A. D.), which is confirmed by several references in the vaṃśāvalīs, e.g.:
’ Śrī Manadeva undertook meditation at Gun Vihāra. By virtue of his meditation a huge caitya came into being and was thus founded ’ (Gopāla Raja vaṃśāvalī).
Maṇipūrṇā or Maṇiyoginī are earlier names for the goddess Vajrayoginī. This goddess plays an important part in the vaṃśāvalīs in connection with pre-Malla kings (mythical as well as historical). She has become one of the principal deities of Tantric Vajrayāṇa Buddhism and was brought over to Nepal from north east India where Vajrayāṇa developed in the 8th century A. D. In Nepal she attained much popularity and is often mentioned together with Śri Svayambhū. They form a pair, like their temples on the Vajrayoginī complex. In Nepalese Buddhist sources their connection is explained as follows:
‘ Our tradition regards Vajrayoginī as the Sakti (prajñā) of Vajradhara – the Adibuddha who is the originator of five Dhyani Buddhas. He is the highest deity of the Buddhist pantheon of Nepal. ’ (Dhooswan Sayami 1972: 40).
In the local tradition of Sankhu the two are also connected in this way: the goddess is considered the spouse of Buddha/Svayambhū/ cibadiyu (caitya god). In the Hindu version of the story (in the Nepāla māhātmya of the Skandapurāna) it is also the śakti of the highest god who appears, Girijā (‘ mountain-born ’, epithet of Pārvatī), śakti of Śiva:
‘ Als die Girijā darauf mit der Askese des (Janārddhana in the form of) Buddha zufrieden war, offenbarte sie sich unter dem Namen (Form) Vajrayoginī. ’ (Uebach: 59, 1. 56).
Thereby the river Manimatī (Nepālamāhātmya) or Manoharā (Svayambhū purāna) also appeared. And this completes the creation of the valley of Sankhu.
2. The creation of the kingdom of Śankhapur
12Apart from the references to the founding of Sankhu in the chronicles there is a Buddhist avadāna (lift. ‘ glorious acts ’) dealing with this subject. It is named Maniśaila Mahāvadāna. The original text, the date of which being unknown, is probably kept by the Sankhu Vajracāryas. One of them, Barnabajra Bajracharya, has published his own version (Banepa 1962, in the Newari language)7. In the second volume of this work, the 7th and 8th chapter are dedicated to the origin of the town. In these chapters a number of instructions issued by the goddess Vajrayoginī are mentioned to which the kingdom has to conform, like the ‘ good omens ’, the ‘ holy places ’, the shape, the water-systems, gates and roads. Whether these instructions were present in the original text, or whether they were added later (on basis of the actual situation) can not be assessed, but at least they are presented as a totality by a Vajracārya priest. The whole system or collection of systems includes elements which were not considered in other studies concerned with Nepalese townplanning. But on the whole the structure of the Sankhu-’ kingdom ’ does not deviate essentially from the structure of other kingdoms in the Kathmandu Valley.
13In the above mentioned avadāna the first announcement of the future kingdom of Śankhapur takes place near a ghat (cremation ground). It is told how a low-caste girl, Kutilmati (who was freed from her sins committed in a former life by the goddess Vajrayoginī) dies while resting near a place called Tārātīrtha. Since she was married to a high-caste person the problem is raised by Kutilmati’s husband on the one side and her father on the other side on which ghat to cremate the corpse. The gods of heaven interfere and recommend another burning-ghāt nearby, called Papomukti. This place, the text says, is very important, esteemed and auspicious, where Tara comes to play, feast and dance in the company of Siṃhinī and Vyaghrinī, where people will achieve success after invoking the goddess, and where the goddess gives boons to the people, saying: ‘ a more holy ghat is nowhere to be found in Nepal ’ (this place is probably on the location of the present temple complex which is indeed near Tārātīrtha). There and then, in the presence of the mourning crowd, the announcement is made: ‘ In the first part of the Kaliyuga, in the pleasant forest of Maṇiśel, a place inhabited by tathāgatas (‘ those who have attained the highest state of perfection ’), out of the roots of a holy tree king Śankara Deva will be born. The king will be crowned in front of Śrī Ugra Tara Vajrayoginī Devi and on the confluence of the two sacred rivers Manumatī and Maṇirohinī the kingdom of Śankhapur will be established ’ (followed by a prayer).
14In the following chapter the birth of the king and the founding of the kingdom are presented in more detail. The text is self-explanatory and will be quoted in full.
The Buddha said (to his disciple): Listen carefully how Sankhu was created. The goddess Vajrayoginī appeared in the forest. For her worship she created a priest, Baca Siddhi. He served her properly.
In the year 1801 (of the Kaliyuga), on the third day of the waxing moon of Phalgun, the goddess went for recreation to the forest. One of her necklaces, named Nāgāpundala, fell down. Water spouted up. On that place she created a sacred pond which was called Manikundala tīrtha. She went on and planted a seed, rasa utpaña (royalty-creating seed) on the root of a holy tree.
15(A much later Kaliyuga priest, Jog Dev, is then introduced. He is instructed to go and look for the newly-born royal baby) .
Jog Dev found the baby, which was fed by liquid dropping from the branches of the tree into its mouth. He presented the baby to the goddess. Jog Dev found the baby blessed with all the 32 good qualities of the world. The goddess crowned the child in a brief ceremony. The king was taught all sciences by the priest and through the blessings of the goddess he mastered all sciences of the world. He grew up into a handsome man like the moon in its waxing phase.
One day, when Jog Dev was performing the daily ritual the goddess was pleased. She told Jog Dev that to establish the king a kingdom had to be created. Jog Dev had the idea of letting seven villages merge into one township (follow the names of the villages).
The goddess said: The kingdom has to be composed of all the good omens. Holy places have to be assembled and be part of the kingdom. Three northwards tentacles (roads?) of the town have to be made. The shape of the town should resemble a conchshell. Around the town there should be gutter-canals. The town should also be encircled by water on three sides. The town should consist of eight wards and roads in all eight directions. There should be eight bahāls (’ monastic ’ complexes) and the town should be raised on the south. There should be eight (aṣṭa) mātṛkā (pīṭhas) outside the town in the eight directions and eight inside the town. The capital should be in the centre of the kingdom [...]. The burning ghāts should be on the north. The king of this kingdom should be Sankar Dev. The town will be called Śankhapur. Then the goddess disappeared.
There follow the names of the different systems introduced by the priest: 8 ṭols, 8 bahāls, 9 ponds, 9 vihāras, 12 tīrthas, 8 mātṛkā pīṭhas outside the town (+ Siṃhinī and Vyaghrinī pīṭhas) and 8 inside the town, 4 gates in the cardinal directions and 4 resthouses (sataa) in the town where the goddess is to be received during her festival. The last part of the chapter concerns the celebration of the festival.
16Historically, the origin of the town Sankhu is indeed attributed to a king Sankara Deva, the Thākūri (there have been two other kings with the same name), son of Bar (Vara) Deva. Sankara Deva reigned over the Valley in the first quarter of the 10th century A. D. The king was named after the Hindu reformer Saṃkarācārya. ‘ Śankara Dēva reigned 12 years. This Raja built a village in the shape of a sankha, or shell, and named it Sānkhu. It was dedicated to Ugra-tārā-devī ’ (Wright’s chronicle: 153). It is not known whether Sankara Deva himself has, during those 12 years, ever lived in Sankhu (his father lived in the palace in Mangalbhat, Patan). In any case there used to be a palace in the western central part of Sankhu, where at least royal representatives from the courts in Kathmandu and Bhaktapur (alternately) have put up residence. All indications lead one to assume that Sankhu had a royal status. One Malla king is mentioned to have lived in Sankhu for some time, according to the chronicle (Wright: 206): ‘ This Raja (Sūrya Malla) took Chāngu Nārāyana and Sankhapur from the Bhātgāon (Bhaktapur) Raja. He went to live at Sankhapur, and in order to please the goddess Bajra Joginī, he instituted her rath jātrā. He lived there six years, after which he returned to Kāntipur (Kathmandu) and died. To this Malla king the temple of Taleju in the south of Sankhu (now totally ruined) must be attributed. Sankhu has never been a fully sovereign kingdom. But in the Nepal Valley there used to be the institutions of dvairājya and ardharājya, territorial structures in which the kingdom was divided in two. The two kings were either two brothers, or father and son, or maternal uncle and uterine nephew (Toffin 1979a : 77, n. 32). By such arrangements many settlements obtained royal status.
3. The structure of the kingdom
3.1. The valley
17The map (Sankhu Valley Nepal) shows the valley of Sankhu as part of the greater Kathmandu Valley with indicated in this valley (what I call) the outer ‘ circle ’ of mātṛkā pīṭhas. The pīṭhas mark the border between the sphere of human society (habitation and fields) and nature (forests). The first sphere is associated with ordered or sanctified space, the second with chaos, danger and wild animals. The Vajrayoginī temple-complex is not within this circle. Thus it maintains its status as a shrine outside the realm of society, which is in conformity with the yoginī character of the goddess. The main significance of the Vajrayoginī festival is the descent of the goddess from the hill into human society, crossing the mentioned border. This border is ritually stressed on two different places:
On the location of Kholaghal (‘ down, deep place ’) • This place is situated north of the town, where the pilgrim’s way from the town to the temple is about to cross a stream. On the other side of this stream starts the steep ascent, the beginning of the actual hill. While on the first night of the Vajrayoginī festival the statues of the four deities are being carried down by the Newar population of the town, it is the Tamang population from the hills which on the final day of the festival takes them up again as from this place Kholaghal. Before the statues on their respective palanquins are given over to the Tamangs they are tossed and dragged forwards and backwards before they finally pass a gutter. Then the Putuvār blow their long horns, a sign that there is no way back (for Putuvār horn-blowing, see 3-2.g). Informants consider Vajrayoginī as the daughter of the town and they compare the spectacle on this place with the moment when a daughter is given in marriage and carried by her family up to a certain point where she is taken over by her inlaws, this being a dramatic moment during the marriage ceremonies.
On the location of Mahākāl. This place is just below the final steps leading to the gate of the hill-temple complex.

Map. i.
There are a resthouse, a statue of Ganeś, a watertap (dhārā), the Mahālakṣmī pīṭh and a huge triangular stone, pointed upwards which represents the Bhairava Mahākāla, the ‘ doorkeeper ’, said to be created by Manicūda. Bloodsacrifices by most of the jātis of the town and those by the Tamangs from a village nearby (Gumārcok) are made here, for Mahākāl. In the chronicles there is a story of historical significance connected with this place. Śamkarācārya is said to have come to Nepal to destroy Buddhism. According to the story, he killed many bauddhamārgis, destroyed their books and then went to the Manichūra mountain to destroy the Buddhists there.
Six times the goddess Mani Jõginī raised storms, and prevented his ascending the mountain, but the seventh time he succeeded. He then decided that Mahākāla, who was a Buddha and abhorred hinsā, should have animals sacrificed to him. Mani Jõginī or Ugra Tārinī was named by him Bajra Jõginī. Having thus overcome the Buddhists, he introduced the Saiva religion in the place of that of Buddha. This ends Shankarāchārya’s triumph over the Bauddhamārgīs of Nepāl (Wright: 119-120).
18There is a local version of this story which accounts for the importance of the place Mahākāl as a borderline. In this version the encounter of the Śaivite reformer and the Buddhist goddess is also mentioned. They are said to have engaged in religious debates on the moutain (Vajrayoginī being here identified as the goddess of wisdom), with ’ books piled up between them ’. The outcome was inconclusive after which it was decided to let the border of Hinduism and Buddhism henceforward be established there. North of that border (i.e. Mahākāl) Buddhism was to prevail, south of that border Hinduism. And indeed, the goddess remained under the care of the Vajracārya priests. Nowadays, while bloodsacrifices take place for Mahākāl, these are prohibited on the temple-square where there is even a leather-shoe-prohibition (although not all the offerings presented to the deities within the confines of the temple-complex are purely vegetarian).
19We shall not at present identify the lines connecting the eight mātrkā pīṭhas with the borders of the kingdom. The mātṛkā define an area ritually but not necessarily politically. They may be guardians of the directions (like Dikpāla or Lokapāla) who protect the inner area from all calamities from outside (Toffin 1981: 61) but the spiritual aspect of these ‘ outer ’-mātṛkās seems to be more important than the military aspect (in contrast to the ’ inner ’ asta mātrkās who are invoked by Karmācāryas during the kṣatriya festival of Dasaī-Durgā Pūjā). The Vajrayāna concepts about mātṛkā goddesses differ from the Hindu tradition (e.g.they are certainly not śaktis of gods). The significance of Varajayāṇa mātṛkās must be considered on different levels. On the individual psychic level each mātṛkā represents a human vice and by venerating the mātṛkās one can control those vices in oneself. For the initiated priest the mātṛkā stand for the elements (tattvas) of the world corresponding with those of the human body, and by venerating three circles of aṣṭa mātṛkā pīṭhas (24) around the country, the devotee can experience how the world he inhabits is mirrored in his body (according to Gutschow et al.). This makes the pīṭha pūjā pilgrimage also a way into the pilgrims own self (Gutschov and Bajracharya 1977: 9). In a maṇḍala the border with the eight cremation grounds symbolise the eight aspects of individual consciousnous, vijana, which tie the individual to the mundane world of the cycle of rebirths, saṃsāra. Each of the cremation grounds (śmaśāna) is conceived of as consisting of a mātṛkā, a bhairava, a nāga, a tree and a kind of flower. In connection with the maṇḍala of the kingdom (see 3-2.h.) it is relevant to note that physically, at least some of the mātṛkā pīṭhas around Sankhu correspond with real cremation grounds (Mahālaksmī) mentioned in the myth as Papomukti, ‘ the most holy ghat of Nepal ’; Brahmāyanī, the former ghat of Sankhu; Rudrāyanī, the present maśān ghāt; Vaisnavī, an untouchables’s ghāṭ; the others are also suitably located near a river or stream). Finally, the mātṛkās are mythically connected with the struggle between gods and demons for the Tree of Wisdom (or Riches). To win the tree the gods take the form of mātṛkās, in which the powers of Kālī and Bhairava are thought to be combined. The mātṛkās are then conceived of as being partly male (the upper hands) and partly female (the lower hands). Through this transformation the demons can be defeated or their bad qualities appeased (personal information from Mr. Badri Ratna Bajracharya, Kathmandu 1979) .
20Considering these arguments about the mātṛkā pīṭhas in connection with the border of the kingdom it must be concluded that these do not coincide. The relation between the area of the shrine in the forest (outside the sphere of society) and the inner area bordered by the mātṛkās is as the classical Indian relation within the kingdom between the forest (āraṇya) where the king should not go, and the village (grāma) where the king is to reside, not unlike the Roman concepts of imperium militiae and imperium domi (after Versnel, cf. J. C. Heesterman: Two Types of Spatial Boundaries, forthcoming). The outer border of the kingdom is not territorially defined. The borders of the settled, civilized area of Sankhu used to be crossed on the trade route par excellence. The sovereignty of the king was defined by the degree and the distance of his control of this route. In principle this reached as far north as where Tibet’s power was met. There used to be no other ‘ kingdom ’ between Sankhu and Tibet (a folk-etymology of the name Sankhu is Sako : ‘ land/town below Tibet ’).

Map. 2.
21To conclude this section it must be mentioned that the ’ outer ’ mātṛkā pīṭhas and the Vajrayoginī temple are not totally unrelated. It is true that during the pīṭha pūjā (see 3-2.i) the goddess Vajrayoginī is not included, but during the ceremony of reconsecrating the Vajrayoginī temple images (once every 12 years, for a description of such a ceremony, see Locke: 208-221) the eight mātṛkās and Siṃhinī and Vyaghrinī are ’ informed ’ or ‘ invited ’. A Vajracārya priest makes the pīṭha pradakṣiṇā and offers nuts (gwaye) and a coin (dā) to the mātṛkās in a small vessel (kislee). Sum on Kamal Tuladhar (1979-80: 51) holds that such a gwaye da tayegu ritual is meant to announce a formal ceremony and invite persons or deities to be present. The only pūjā in which ten fierce deities (of the ten directions) are invoked in the reconsecration ceremonies is the pūjā of the dasa krcdhas. They are attracted and then bound with ritual stakes (kilas) to prevent their interfering with the rites which follow (Locke 1980 : 209). As such the mātṛkās have a rather negative relation with Vajrayoginī which has no spatial significance. The lack of a spatial connection could be expected: the Vajrayoginī temple is the focus in relation to which the town is defined, but the mātṛkā pīṭhas are defined with the town (the centre) as focus and not the Vajrayoginī temple.
3.2. The town
22Map 2 (Sankhu) shows the spatial distribution of buildings, shrines, roads and other open space in present-day Sankhu. Although seven villages are said to have merged into one township, a conscious townplanning was applied, as in all other Newar towns. The town was built on an elevated plateau (like Thimi and Kirtipur), not directly on the banks of a river. Its shape is more or less rectangular and has not been influenced by the alignment of a river (as it has been in Bhaktapur). The main street-plan has probably not been altered very much since the founding of the town, neither have the locations of the squares, being the centres of wards (Nepali: ṭol; Newari tvaḥ). Most of the buildings, however, are of relatively recent date, built on old foundations. In 1978 just over 600 houses were counted in the town. The buildings of earliest origin (the bahāls, the palace and several shrines) are almost all reduced to ruins, due to earthquakes and neglect. There were major earthquakes in 1833 and 1934. Wright’s chronicle, in one of the last chapters (p. 269), gives some figures of the damage brought about by the earthquake of 1833: ’ In Sānkhu 236 houses were destroyed, and 21 temples and pātīs. 18 people were killed and 10 injured ’. Such figures are not available for the 1934 earthquake, but local informants estimate the number of casualties and damage as much higher than the former figures. Of the important temples in the town those of Nārāyana, Ganeś-Kaumārī, Bhīmasena and Jyotirlingeśvar were destroyed (the last being the only one which was properly rebuilt). The Vajrayoginī temple was not damaged. The palace and the Taleju mandir were already ruined before the 1934 earthquake. A short description of the town will be given below.
3.2.a. The roads-plan
23The skeleton of the town’s lay-out is formed by two north-south roads, connected at right angles by a number of east-west roads. Along the eastern north-south road most of the shops are situated (the bazār) and this road continues to the north through a gate as traderoute to Helambu and further in the direction of Tibet. The main entrance to the town from the direction of Kathmandu is a gate on the southern side of the western north-south road. In the north, this road continues as pilgrim’s route to the temple of Vajrayoginī. The east-west streets connect, in principle, the squares of the eastern and western wards so that each square is at least a crossroads of roads going into three directions.
3.2.b. The wards (ṭol)
24Territorially, the town is divided into eight wards; four in the west and four in the east. Their names are:

Thus, the town is bisected by ward-boundaries from north to south and divided into an eastern and a western part. The ward-boundaries can not easily be found. They are not indicated by roads or special marks on roads. But every ward has its own Ganeś shrine and the exact ward-boundaries can be established by finding out which Ganeś is worshipped by the inhabitants of every single house. The division in eight of the town is something of an ideal division, connecting a territory or ‘ ritual quarter ’ with a mātṛkā goddess. This is the case in Patan and in certain sections of Kathmandu (Toffin 1979a: 71). In Sankhu not all the locations and names of the aṣṭa mātṛkā pīṭhas inside the town can with certainty be identified. The Vajracārya as well as the Karmācārya priests are not certain about all the names and locations and some have local names only (like inma deo and gola deo). The most important ones are Brahmāyanī (Sālkhā square), Cāmundā (Mahādev Dhokā), Rudrāyanī (outside Sangāhā Dhokā), Vārāhī (Bhau Dhokā), Kaumārī (Īnlā square) and Mahālaksmī-Māheśvarl (aji deo, Cālākhu square). There is not necessarily one mātṛkā pīṭha in one ṭol, but it can be said that there are mātṛkā pīṭhas located in each quarter of the town; three of which are located near the principal gates of the town. The direction of each mātṛkā can also not be established, but all written sources differ in this respect anyhow.

Figure 1 Dual division of Sankhu with pradakṣiṇāpatha passages

Figure 2 Quarterly division with gates (political system)

Figure 3 Quarterly division: two ṭol share one sataa
The identity of the ṭol as a socio-territorial entity is manifested in:
The ‘ way of the dead The route of bringing a corpse to the ghats is a fixed one for every house. The corpse is carried via small lanes through the ṭol to which the deceased belonged to the Eastern North-South road which is followed to Mahādev Dhokā, from where the road continues to the ghats.
Ṭol-festivals. Some festivals are celebrated within the ṭol only, such as the day observed as warding off the notorious demon Ghantākarna (preceding Gai jātrā). In every ṭol two dummies of straw, pots and fruits are erected (one male, the other female) nearby a fire in which a representative from every house burns a sweeping-brush, symbolising all evil from the past year. With great force the male dummy is later pulled over the female figure and further dragged out of the ṭol and the town. In all directions and with great centrifugal force all evil is expelled from the town on this day.
Participation in jātrās. Several town-processions consist of representatives from every ṭol. This is most clearly shown during the week-long Gai jātrā processions in autumn, starting with a procession of cow-masked people (or people carrying a tall structure of bamboo and cloth crowned by an umbrella), each representing a person who died in the town during the past year. Ṭol-groups then move as separate units through the town. The other processions, in which considerable ṭol-pride is involved include the ‘ stick-hitting-dance ’ commemorating victory over Tibet, called ‘ Ghintang-Ghising-Twanga ’, and the other satirical ‘ paddy-transplanting-completion-dance ’. During other times of the year the Nava Durgas and the Devīs also dance in every ṭol. Musicians of the Putuvār caste blow their long horns (Newari : ka) when physical borders such as ṭol borders are crossed. All ṭol ceremonies start with the invocation of the ṭol-ganeś.
3.2.c. Topography and shape of the town
25In other towns of Kathmandu Valley (Kathmandu, Bhaktapur, Panauti, etc.) the major territorial dualism of those towns is expressed in a high-low or uppertown-lowertown division (Gutschov and Kölver 1975: 48; Toffin 1979a: 69, 70, 1981: 60). While there is a very clear dualism in Sankhu (see Figure 1, and the ṭol-division in 3-2.b) the author did not hear of such a classification or terminology in Sankhu. In any case, the West is the higher and the East (closest to the river) the lower part of the town. Topographically, the Northern and Western parts of the town are higher than the Eastern and Southern parts. Although no exact measurements have been done, this can be concluded from the alignment of small water-gutter canals. These bring water into the town from the North (pilgrim’s road), then branch on the first main square (Dhõla) where half the amount of water is carried further to the south and the other half towards the east. It is interesting to note that informants connected the name Sankhu (from śaṇkhā: conchshell), not with the visual shape of the town, but with the fact that all water which enters the valley and the town from the North is ‘ poured out on the South, like in rituals priests pour water out of a conchschell ’. This explanation is typical for Newar agriculturalists who are daily concerned with rituals and with water for irrigation purposes (especially for the winter crop wheat and potatoes).
3.2.d. The palace (layku)
26The palace used to be situated on a fairly high location in the central western part of the town near the Kaumārī pīṭha. Presently on this place there is a Panchayat office, a milk-collecting centre, a ’ park ’ with a platform (which people consider the ‘ coronation platform ’) and a small Visnu-shrine in a hollow place on the base and between the roots of a huge (very old) Ficus tree (ficus religiosa).
27The palace is conceived of as the centre, not only of the town but also of the universe, and ideally identified with Mount Meru, encircled by seven oceans (courtyards) and seven concentric circles of sacred mountains (towers and temples) (Toffin 1979a: 68). The ancient palace of king Amśuvarman (7th century A. D.) was likewise identified with the mountain Kailāsa as the name Kailāsakuta indicates. Except for the mentioned Kaumārī shrine there are at present no temples attached to the area of the palace. It may be significant that the temple of the royal (Malla) goddess Taleju is located a big distance away from the palace (unlike the situation in Kathmandu and Patan). This could indicate that another deity (or Kaumārī) formally belonged to the palace grounds.
3.2.e. The bahāls and the Vajracāryas
28In the town all but two bahāls have vanished (a bahāl, in Newar Buddhism is a ’ monastic ’ complex housing the families and shrines attached to the ‘ monastery ’). However, during the ‘ bahāl pūjā ’ the Vajracārya priest still traverses the town and performs a small ritual near the old sites, now overgrown with weeds or turned into garbage heaps One bahāl was extended time and again and has several courtyards. The names are: i. Gun (Vajrayoginī temple complex), 2. Duchhe, 3. Thaathu, 4. Om, 5. Opi, 6. Sui, 7. Yā, 8. Mansu, 9. Ko.
29Those Vajracāryas who do not live in one of the two remaining bahāls have houses which are scattered over the town. There are five (some claim seven) lineages of Vajracāryas in Sankhu, all of them being branches of one clan with ancestor Bācā Siddhi, believed to be created by Vajrayoginī. The names of the lineages were given as: I. Jitauri Deva, 2. Amṛta Prabhaa, 3. Siddhi Bara, 4. Caitya Muni, 5. Kara Bīra.
3.2.f. The gates and walls
30The gates lie on the extreme sides of the North-South roads. All of them have been given a special designation or function and a classificatory direction. Dhõla or Vajrayoginī Dhōkā, the North gate, is for ‘ bringing in and taking out the goddess during her festival ’ (Bajracharya 1962: ch. 8). Sālkhā or Mahādev Dhokā, the East gate, is for ‘ taking out the corpses to the ghats ’. Just outside this gate there are many small, open votive shrines, a pond and the graceful Jyotirlingeśvar temple. The South gate, Sangāhā Dhokā, is for ‘ sending away the brides ’. Bhau Dhokā, the West gate, is for ‘ bringing in the brides ’ (these customs are still followed). Near the last gate, just inside the town, there is a Bhagavatī shrine.
31By their association with the four cardinal directions the four gates bring about a structural division of the town into four quarters (North, East, South, West, see Figure 2), whereby two wards are grouped together into one quarter. The political division of the town into three Village Panchayats follows this division. The East quarter (Sālkhā and Cālākhu) is part of Bajrayoginī Village Panchayat, the South quarter (Dugāhiti and Sunṭol) is part of Sunṭol Village Panchayat. The whole ’ upper town ’ (West and North quarters) is the Pukhulachi Village Panchayat, but a further political division would certainly divide these two quarters.
32Apart from the main gates some old informants have mentioned four other ’ subsidiary ’ gates, all located in the East and West sides of the town (sometimes two in one ṭol) . It is likely that ideal planning and practical ordering have been in conflict here. On the Eastern and Western sides of the town lie an important number of fields and it stands to reason that there were in all times, like today, short access-lanes to these fields. Nowadays the untouchable’s (Kusle, Poḍe) quarters in the East and the West are also connected by lanes with the bigger roads and so they have de facto become part of the town (termed as Kusle ṭol, Poḍe ṭol), at least spatially. Socially they are literally outcastes, as is most clearly shown during Gai jātrā: the dead persons of the untouchable quarters are not represented in the jātrā.
33The town used to be fortified but there is no evidence that it was completely surrounded by wall. Presently there are no walls to be found. The only site where a wall is claimed to have existed (by village informants) is on the Eastern side, between the Eastern wards and the untouchable’s quarters (which leaves the latter ‘ impure ’ category indeed outside the ‘ pure ’ town).
3.2.g. Social composition of the town and brief socio-economic outline
34Table I shows the social composition of the town, consisting of 16 endogamous Newar castes (jātis) and a few families of Indo-Nepalese origin. The names are arranged in hierarchical order. Between the castes of which the numbers are connected by a line no clear hierarchy can be established. The Brahmans (Dya) come from Patan (Tadkan chowk). Families of two Brahman lineages live in Sankhu in turns, one year each time (4-year cyclus) to render festival services and act as family priests. They live in Balampu chowk in the Northeastern part of the town. The festival of Mādhava Nārāyana (in the month of māgha) provides them with food for three months a year. They have also land near Sankhu. The majority of the population, the Śrestha, live all over the town, except in the lowest quarters. They share the whole centre with a few Vajracārya and Pum families. Three compact Jyāpu quarters lie in the Eastern and Southern part of the town. Two Sāymi quarters in the South and the West. The untouchable castes live outside the ‘ walls ’ of the town, as was mentioned, except several Jugi families. Further on, near the Salinadī river live a group of Vaisnava samyāsin and one Kānphaṭa yogi. In the village Southwest of Sankhu, Salamgutar, live Chettris originating from a locality south of Lalitpur. Lapsifedi, Northeast of Sankhu, has a mixed population of Tamang, Brahman, Chettri Thākūri and Newar. On the Maṇigiri Ghumārcok is Tamang and Bhulbu is Chettri (Karki clan).
Table 1

Sankhu castes (jati)
Indo-Nepalese castes: Kāmi, (i), Śārki (I), Gāine (I), Dāmāi (I)
Figures connected by lines: between these castes no hierarchy can be established.
35Śrestha and Jyāpu are the biggest landowners. The lower castes own little or no land. Some of them are tenants, others work as paid agricultural labourers. Their ritual occupation results in no more than a symbolical extra income, mostly in food. The holdings of those who own land are usually small. About 50 % of the households is self-sufficient in food (especially rice), 30 % has rice for about eight months per year and 20 % not more than four months per year. The people with capital tend to be money-lenders and thus bind the poor to them. The rich people have invested in rice mills; handlooms give others an additional income. Except for a little trade Sankhu is predominantly an agricultural community nowadays. Many families depend on remittances from family members who have gone to live in Kathmandu. Economically as well as politically Śhrestha dominate the town. As was mentioned, the political authority in Sankhu is now shared by three heads of the three Panchayats with their councils. In the past this was not so. During the 19th and early 20th century there was one community chief (dvāre). In this person both political as well as ritual authority was vested. He was, so to say, the local ‘ king ’.
3.2.h. Compilation of data; the town as a maṇḍala
36In many ways the lay-out of Sankhu conforms to ideas on Indo-Aryan town-planning as contained in the Śilpa Śāstras (with its subdivision Vāstu Vidyā, 5th century A. D.) which in turn go back on Kautilya’s Arthaśāstra (4th century B. C.; cf. Dutt 1925). This applies especially to the idea of a walled fortification with four gates in the cardinal directions, a geometrical street-plan of North-South and East-West roads at right angles, and separate quarters of the town being reserved for particular jātis (professions). As underlying structure, most Newar towns, like Sankhu, have adopted a number of these principles, diverging from the ideal groundplan here and there depending on the physical environment. This ideal groundplan is ’ identical with geometrical figures drawn on sacrificial altars ’ (Dutt 1925: 7). Such a conception comes close to the idea of the townplan as a cosmic map like the man data or yantra (which are Vedic in origin). The maṇḍala (in casu the square vāstupuruṣa maṇḍala), like the town, has also a structure consisting of a central ‘ palace four ‘ gates ’ in the cardinal directions, four ‘ quarters ’, with gods in all eight directions and encircled by cremation grounds and holy mountains. Bernier (1979: ch. 6) has described this maṇḍala as the structure underlying the Nepalese pagoda temple-plan. A crude town-maṇḍala is depicted in Auer and Gutschov (1974) for Bhaktapur (cf., also: Toffin 1979ā: 73; MacFadyen, Vogt 1977).

Map. 3.
37To compile all information about Newar townplanning presented by various scholars, it is possible to design a hypothetical ideal maṇḍala for ’ the Newar town ’ in general. For such a design the following elements from the description of the town of Sankhu are relevant:
the palace in the centre of town (see: 3-2.d),
the town’s dualism (see: 3-2.b and Figure 1),
the eightfold territorial division in ṭol (see: 3-2.b) and the ‘ inner circle ’ of aṣṭa mātṛkā (see: 3-2.b),
the four gates and the corresponding division of the town into four quarters (see: 3.2.f and Figure 2),
the ’ outer ’ circle of eight mātṛkā pīṭhas, corresponding with eight cremation grounds, and
the pradakṣiṇā patha, to which attention will be drawn in 3.2.i (see Map 4).
38For the resulting maṇḍala diagram, see Map 3 of which the corresponding text is self-explanatory in view of the description given above. The sanctuary of Vajrayoginī is not contained in this cosmological model. It is simply not defined in relation to the centre of the kingdom or the town. It transcends the whole idea of kingdom or territory.
3.2.i. Pradakṣiṇā and pradakṣiṇāpatha
39Tantric priests, during their meditation, can invite deities to take their places in the maṇḍala. The maṇḍala of the town comes to life during the invocation of the deities by worshippers and priests. Very seldom a single deity is worshipped exclusively. Usually the devotees make a pradakṣiṇā through the locality, through a certain part of the town or through the whole town, depending on the occasion, but each time circumambulating the centre. The most ordinary and restricted pradakṣiṇā is performed daily, in the early morning, by the women of every house, to worship the ṭol-deities (a gaṇeś, a bhairava, a nārāyaṇa, and often including a mātṛkā, a nāga, a caitya, a linga, etc.). During more specific occasions large groups of devotees, accompanied by a priest, circumambulate the centre of the whole town to invoke all, or a great number of town-deities. An example are the daily pradakṣiṇās during the month gŭla (Nepal Sambat) by members of a Sāymi guṭhi (called dafa), who visit the shrine of Vajrayonigī and on the way back traverse the town, worshipping at every single shrine of the town (about 60 in number), accompanied by a Vajracārya priest. A similar pradakṣiṇā is made during Krsna jātrā by small girls, led by the Brahman and a Karmācārya. And finally, there is the ’ outer circle ’ asta mātrkā pītha pūjā in autumn, accompanied by a Vajracārya priest, a pradakṣiṇā lasting approximately nine hours. Furthermore, there are the festivals and the jātrās during which the deities themselves make a pradakṣiṇā through the town. While the former mentioned groups of devotees may follow different routes through the town, the deities always follow the pradakṣiṇā patha, the auspicious way through the universe of the town (see Figure 5), in a clockwise way. To the author’s opinion the pradakṣiṇā patha does not separate a ’ right ’, ‘ pure ’ inner space (the centre of town) and a ‘ left ’, ‘ impure ’ periphery (as was argued by Gutschov and Kölver 1975: 21, 42, and Toffin 1979a: 71). This is not so because the pradakṣiṇā patha does not go around, but goes through the settlement, around the centre. It is the ‘ walls ’ which separate ‘ pure ’ from ‘ impure ’ by excluding the untouchable’s quarters from the town. Later, Gutschov (1980: 62) seems to correct his earlier view in putting it that the pradakṣiṇā should be understood as an integrating force: ‘ Linking the characteristic squares, which represent historical, social and spatial heterogeneity, the pradakṣiṇā could [...] be understood as a means of overcoming differences. Unity is a result achieved through ritual ’. The sense of unity of the town and the sense of individual citizens being a part of this unity is also expressed during the life-cycle rites. On different occasions individual citizens make the whole pradakṣiṇā through the town, following the pradakṣiṇā, patha clockwise (a baby on the occasion of its first feeding with rice, carried by its maternal uncle; at marriage by the ‘ incoming ’ and ’ outgoing ’ brides; at the auspicious age of 77, in a rath ; and finally, after death, when the deceased is carried via this road for the last time, now counter-clockwise, on the way to the ghāṭs).

Map. 4.
40Most of the pūjās and jātrās have a significance for the town as a whole. They can be religiously classified as more specifically Hindu, or more Buddhist, more Śaiva or more Vaiṣṇava depending on the officiating ritual functionaries and the nature of the sacrifices. A broad division in the religious life of this Newar community can best be made according to the category of priests. This is as follows:
The Buddhist Vajracārya priests (gubhāju), the keepers of the shrines of Vajrayoginī and Basundharā Devi. They are officiating at the jātrās of these goddesses. They provide a yajamāni services for the jātis Jyāpu (farmers), Sāymi (oilpressers) and Nau (barbers).
The Hindu Brahmans, specialists of the Vedas. They are officiating during the festivals of Mādhava Nārāyaṇa and Kṛṣṇa (Deo Brahman), Jyotirliṅgeśvar and Narasiṃha Deva (Bhatta Brahman) . They are family priests for all other pure jātis.
The Hindu Karmācāryas (Ksatriya varna), the highest Śreṣthas who are initiated in the Tantras. They are in charge of the public ceremonies during Dasaĭ (Dūrgā Pūjā), especially those connected with Taleju. They also facilitate the annual Nava Durgā dances. During ritual occasions these priests may have complementary functions. The Vajrayoginī temple is the only sanctuary where priests (Vajracāryas) are permanently present. But with the consent of these Buddhist priests the Hindu Karmācāryas may enter the sanctum sanctorum of the Vajrayoginī temple to perform secret Tantric rituals on certain occasions. Vajracārya and Brahman priests may act together during family rituals. Karmācaryas perform special pūjās during the Vajrayoginī festival (see 4.3) and assist Brahmans during the Mādhava Nārāyana festival. Generally speaking, the Karmācāryas are concerned with Tantric ritual complementing Vedic rites and vegetarian Buddhist rites (see also Toffin 1981: section 7).
41The socio-religious spectre of Sankhu is determined by the three major festivals of Vajrayoginī, Mādhava Nārāyaṇa and Dūrgā, at which the mentioned functionaries play a ritual role respectively. In these festivals is contained the conception of renewing or preserving the kingdom, on a social-territorial plane as well as on a cosmological plane. In all three festivals the goddess Vajrayoginī is invoked in one way or the other. The Mādhava Nārāyaṇa, festival which lasts for one month, is the only Sankhu festival which links the shrine in Sankhu with a number of other Viṣṇu shrines located all over Kathmandu Valley by way of trips made to these shrines by Sankhu participants who carry a small Nārāyaṇa statue with them. The significance of these visits and of the choice of these particular shrines remains to be studied. Dasaï (Durgā Pūjā) links the shrines of all the śakti goddesses together (Bhagavatī, Taleju, the ‘ inner ’ aṣṭa mātṛkās and Kumārī) for the protection of the kingdom and the power of the king. The meat of the buffalo sacrifice is hierarchically distributed among the castes, so that the sacrificial structure serves ‘ à modeler l’ensemble de la société ’ (cf. Toffin 1981, who gives a clear description of Dasaï in the small Newar town Panauti). In this study only one festival, the Vajrayoginī jātrā, will be analysed, in an attempt to link the socio-ritual reality to the myths of creation (being presented in the first part of this article).
4. Vajrayoginī festival
42This spring-festival starts on the full moon of caitra. Caitra is a holy month for the Buddhists in Nepal. On other locations in the Valley, Buddhists pay homage to the souls of their ancestors, especially at Jamacho, the peak of Nagarjung mountain. From this point, on the full moon day of caitra, Vipaswi Buddha is believed to have cast the lotus seed into the Nepal mountain lake. The blooming lotus was later to send forth Svayambhū as the five-coloured flame. The sowing of the lotus seed can be considered the very first act of the creation of Nepal.
The festival of Vajrayoginī lasts for eight days, the names of which are:
1. kwã bijyā : | ‘ day of coming down ’ |
2. sunnyakā: | ‘ day of rest ’ |
3. syañko tyāko: | ‘ day of slaughtering ’ ‘ the more you kill, the more you gain ’ |
4. mū jātrā: | ’ main day of going round the town ’ |
5. mū bijyā : | ‘ day for taking the goddess out (of sataa) |
6. dhū bijyā : | ‘ day of completion of worship ’ |
7. bauyā: | ’ rice offering to the ghosts ’ |
8. thã bijyā: | ‘ day of going up ’ |
43The following observations may be relevant for the purpose of this article.
4.1. Deities involved
44As was mentioned before, there are two ‘ forms ’ of the goddess Vajrayoginī which are locally (in Newari) called hyauñ khwā māju (the Red-faced goddess) and mhâsu khwā māju (the Yellow-faced goddess). The former always remains in the temple. The latter is the procession deity, also called eka jaṭī buddhī mātā (‘Ekajatā learned mother ’). She seems a more peaceful form of the goddess. Around her the festival is centered. The statue is of human size consisting of a heavy wooden mould to which are attached a yellow copper mask with three eyes, two arms, a yellow copper plate (skirt), ornaments and reddish garments. To her right hand (originally in a mudrā attitude) a sword is attached, to her left hand (idem) an unconventional kind of flower. While coming down from the hill the Yellow-faced goddess is accompanied by a huge brass replica of the Dharma Dhattu Caitya (Newari: ciba dyo) and by Siṃhinī and Vyaghrinī (who have the faces of wild animals), each on a separate khat. During the rest of the year these statues are kept in the sanctuary of Gun bahāl, a building which is part of the walled Vajrayoginī temple-complex. All statues are replicas of the statues in the main temples, except for the colour of the face of the main goddess. Normally, the devotees who come to the hill-shrine worship the deities in the main temples first, after which they proceed to Gun bahāl and worship the deities there. The palanquins which are all unroofed and unaccompanied by umbrellas (see hereafter, 4.4) are carried down the hill during the moonlit night. A representative of every house should be present with a torch. The heavy khat of the Yellow-faced goddess is carried by 30-50 men, of all jāṭis. In the town, the statues are placed in a resthouse (Newari: sataa). There are four sataa in the town. Two ṭol share one turn as hosts of the goddess each year. The sataa are arranged in hierarchical order: 1. Dhõla ṭol, 2. Cālākhu ṭol, 3. Īnlā ṭol, 4. Sunṭol. The sharing of one saṭaa by two ṭol results in a grouping of ṭol shown in Figure 3. Through this sharing the mentioned division of the town in four quarters is each time overcome which confirms the idea (put forward earlier) of ‘ integration ’ or ‘ unity ’ brought about by a ritual arrangement (see 3-2.i).
45On the fourth day all deities make a pradakṣiṇā around the town (pradakṣiṇā patha). On this occasion they are accompanied by a statue of Basundharā Devi which is brought from a god-house (dyochem) in the town and joins the deities for the rest of the festival. During the first four days the statues are kept inside the sataa, while they are placed on its verandah during the second part of the festival.
4.2. Ritual officiants
46During the whole festival in the town the deities are under the care of the Vajracārya priests (gubhājus), who daily worship them and present abhiṣeka to them. Eight days before the festival the eldest gubhāju performs a special pūjā in the temple to request the goddess to come to the town. The same gubhāju completes the worship on the final (8th) day of the festival.
47All processions are accompanied by music groups of different jātis, pure as well as impure. Putuvār occupy a special position as they blow their horns at all states of transition, such as at the crossing of borders (town, ṭol) and at the moment when the goddess is taken to or taken from her khat. During these last movements all the ornaments of the goddess are taken off, her body and head are covered with a piece of cloth and funeral music is played.
48A local acāju (Karmacārya), who acts as representative of the government is responsible for all the ornaments of the goddess (especially her head-dress, mukuṭ). On the 4th and 5 th day a jui cakr pūjā takes place in the town, performed by Kānphaṭa yogis (see 4.3, pūjās).
4.3. pujas
49Except for the daily standard pūjās by the gubhājus there are several ‘ special ’ pūjās performed. The very first one is a ‘ night pūjā ’ by Nay, as soon as the deities are established in the sataa for the first time. The main bloodsacrifices on the third day (syañko tyāko) are performed outside the town by individual families or guṭhi-wise. All pure jāṭis sacrifice for Mahākāl on the hill below the Vajrayoginī shrine. The untouchables sacrifice in their localities (for ‘ nasa dyo ’ or Ganeś). The Jugi (Kusle) sacrifice for Gorakhnāth whose shrine is near the river, East of the town. The bloodsacrifices (buffalo, goat, duck) on the 4th and 5th day by the Kānphaṭa yogis are the only bloodsacrifices performed inside the town (on the square in Sālkhā ṭol). Such cakra pūjās are usually performed before all the principal jātrās throughout the Valley. The yogis say that the pūjā is for Gorakhnāth or for the 64 dākinīs and bhairavas. The local people ’ say that if they do not have the pūjā performed the yogis will interfere with the festival by casting spells, etc. ’ (Locke 1980: 442, n. 105).
50On the 5th day there is a special pūjā, ‘ in the name of the king . On the same day individuals practice austerities by lying down with nine oil lamps and pieces of cow dung placed on different parts of the body (mhã mata; cf. Toffin 1981: 61, 63). On the 7th day, at night, a Karmacārya guthi makes an offering to the ghosts. For this purpose rice from all localities is collected and prepared into rice-balls, which are cast away during the night along the main roads of the town (nobody is to see this procedure). This rice-throwing (bau holigo) is preceded by the invocation of an important Bhairava of the town (on a place roughly marked in the pavement in Cālākhu ṭol). On the 8th day, finally, the main priest performs the kalaśa and hōma pūjās before the five deities (for a description of these rituals, see Locke 1980: ch. II). On the 5 th and 6th day after the festival there are two additional pūjās on the temple complex (see 4.4) .
4.4. Significant features
51Eight days before the start of the festival a crier goes around the town, announcing that from that day on till the end of the festival, a. no (foreign, leather) shoes should be worn, b. money lenders should not press for their money, c. no umbrella of any kind should be seen, d. horse nor elephant should be ridden, and e. no ḍolī nor any other kind of carrier should be employed (for regulations a. and b. in Patan, see Locke 1980: 312). Two (c. and d.) items of this announcement clearly concern royal regalia. It seems that the announcement must prevent a public appearance by the king or at least a royal display during the festival.
52However, on mu bijyā, the day after the jātrā of the goddess, a ‘ king ’ does make an appearance. On this day a Jyāpu man called ‘ king ’ goes around the town, barefoot and with a leafed bamboo stalk (instead of an umbrella) carried above his head. This pradakṣiṇā is repeated the following day. This time the ‘ king ’ is accompanied by small boys carrying plates with rice caityas in five colours (white, red, green, blue, yellow). These are later offered to the goddess.
53As with many festivals in the Valley, a ’ royal sword ’ from Hanuman Dhokā (the royal palace in Kathmandu) is carried along with the jātrā. This sword is supposed to be carried by the pradhan panchayat (head of the Panchayat), to represent the king. In Sankhu the sword arrives in the town four days before the start of the festival. On kwa bijyã the sword carrier awaits the arrival of the deities from the hill near the gate in Dhŏla. From the moment the deities arrive in the town the sword precedes the procession. This happens again during mu jātrā. The sword is carried in a cloth sack, it is never exposed. Five days and six days after the festival the sword is carried up to the temple on the hill and remains there during a rājako bijyā and a rāniko bijyā pūjā respectively. These pūjās are performed ‘ on behalf of the Panchayat ’ and are said to be obligatory pūjās, financed by the government and organised by the Panchayat. After the last pūjā the sword is brought back to the palace in Kathmandu.
4.5. Interpretation
54Compared with other big national festivals, the Sankhu Vajrayoginī jātrā is not easy to analyse. It looks as if the festival represents cosmogonic actions by the goddess and expresses the ambivalent relation between the goddess and sovereignty. The position of the festival ’ king ’ seems to provide a clue. His role could have been clarified if a local story were connected with his presence, but such a specific story is lacking. Neither do informants have an idea about the significance of the different pūjās and some think that the festival has changed during the centuries, veiling its original meaning. For a proper understanding of the festival the simple question should be: why does the goddess come to the town?
55To answer this question the highlights should be separated from the day-to-day sequence of events. The main events during the festival in the town seem to take place between the third day (syañko tyāko – the caste-less and priestless sacrificial day) and the 7th day (the appeasing of spirits unbound to a specific area). The jātrā of the goddess through the town (on the night of the 4th to 5th day) marks the transition of the deity from a secluded, esoteric goddess (inside her sataa) to a visible, public goddess (on the sataa’s verandah). With this transition the public appearance of Basundharā Devi is also connected. On the two following days, the person assigned festival-’ king ’ makes his appearance. When he has made the pradakṣiṇā, people say that the ‘ king’s offering ’ is made and the main worship is then completed. Informants explicitly mention that ‘ the goddess is ready to receive the king on this day ’. In the myth (Maniśaila Mahāvadāna) there is a hint of what this sequence might indicate. The text says: ’ [...] on the date of the king (i.e. Sankara Deva) ascending the throne the jātrā.of Basundharā Devi was celebrated ’, and also ’ [...] Sankara Deva was placed on the throne and taken around the kingdom ’ (chapter 8). One is reminded here of Indra jātrā in Kathmandu, where on the 4th day of the festival the king worships the goddess Kumārī (Taleju) and receives from her the tīkā (considered a royal reconsecration), after which the king makes (or used to make) a pradakṣiṇā through the town during the two following days (considered a recreation of the unity of the town, cf. Toffin 1979: 62).
56If the episodes from the myth and the comparison with Indra jātrā can be accepted as evidence of the consecration of a king in Sankhu, then indeed the Vajrayoginī festival can be seen as the re-enactment of the creation of the town (through the coming down and the jātrā of the goddess) and the institution of the king by the goddess. This is in accordance with the myth in which the goddess appears as creator of the king and initiator of the kingdom. Following the idea of a cosmological theme, several festival details which would otherwise be difficult to interprete, seem to make more sense now. For a further interpretation of the festival some more remarks concerning the character of the goddess are necessary.
57The goddess Vajrayoginī is, in the context of the kingdom, unconnected with a specific territory, a specific group of people or a specific king, as the word yoginī indicates. There is no reciprocity between the goddess and the people (or kingdom or king): before the initiation of the kingdom she already existed and with the disappearance of the kingdom she will remain. She is, and has always been, solitary like a tiger in the forest, and indeed, as a white tiger she used to appear in visions of the devotees. Although she is the creator of a king, mythical stories in the chronicles show the goddess’s ambivalent relationship with sovereignty, especially with ruling monarchs. Her shrine becomes the refuge for retired kings (Wright: 99, 100) and tradition has it that no ruling king is welcome in her vicinity. Even recently, the burning of the Singhar Durbar in Kathmandu (1973) is connected with a visit which the king of the Nepal had paid to Sankhu three months previously. On the night of the calamity several Sankhu priests are said to have seen flames leaping out of the Vajrayoginī temple. This ambivalence has to be taken into account while explaining the festival king’s performance. With the descent of the goddess and her appearance in the realm of the king something of this ambivalence is bound to be expressed.
58By coming down she is gradually being entered into a relationship with the people (the first one’s being untouchables, Nay) . She is also being entered into a relationship with the domesticated animals (in casu paśu, the sacrificial animals), to an important effect. The sacrifice (syañko tyāko) is still a ‘ free-for-all ’ performance, unstructured, without ritual officiants and without a hierarchical (social) distribution of the prasāda. Through this sacrifice however, the goddess is activated. From her seclusion in the sataa she now comes into the open. By her jātrā through the town she maps out and clearly identifies the seat and the centre of what is to become the kingdom. But there is more. There is evidence, however slight, that the sanctification of the territory by the goddess is followed by several acts representing essential phases in the building of a kingdom, namely the clearing of the forest and the initiation of agriculture. This evidence is as follows:
During the festival big logs of wood (tree trunks) are carried to the town and put near the sataa of the goddess. On the 5th day, after the jātrā of the goddess, these logs are put to fire and they burn during the rest of the festival. Why these trees should be burnt in the town, and particularly on this 5th day, can not satisfactorily be explained but in connection with the act of clearing the forest which is a prerequisite for building a kingdom (cf. the account of building a kingdom, in F. Bhattacharya 1981 : 35, 36).
In the Maniśaila Mahāvadāna it is mentioned that originally there were no ricepounders (Newari: kuti) in the town. After a special request to the goddess by the priest, on behalf of king Śankara Deva (in the beginning of the king’s reign), the goddess granted permission to use such pounders. It is mentioned in the same myth that the grateful king then started a ‘ special worship ’ of the goddess. The worship by the ‘ king ’ during the 5th day of the festival has been directly related to the ‘ special worship ’ mentioned in the myth8. This detail in the myth may be connected with the start of agriculture, especially rice-production, in the kingdom.
The appearance of Basundharā Devi during the jātrā (unexplained until now) can also be connected with agriculture. This goddess, popularly called ‘ Mother Earth’ – goddess is often identified with the goddess of wealth, Mahālakṣmī. Both are considered bestowers and protectors of agricultural produce in Nepal (Anderson 1971: 183).
59This leaves us with the position of the festival ‘ king ’. On the 5th day this ‘ king ’ retraces the goddess’s route through the town. The performance of this ’ king ’, however, is an extremely poor show compared with the grandeur and pomp of the jātrā of the goddess on the previous day. This is largely due to the mentioned prohibitions of royal regalia (note that even the royal sword from Kathmandu remains covered). The ‘ king ’ looks like a mock king except that he is not ridiculed. However, a mock king and a king without regalia would fit into a picture of a period of complete non-integration of the order which usually accompanies representations of cosmogonic acts during Hindu-Buddhist festivals. To be a total order, it must also comprise non-integration or chaos. Through the coming of the goddess the normal order is temporarily disturbed (which is also shown by the ‘ free-for-all ’ sacrifice and the first pūjā by untouchables) and is not restored till after she has left. A king as insignificant servant and suppliant of the goddess is also compatible with the idea of ambivalence in the goddess-king relation. This relation can be precised as follows: their only true relationship can be as bhikṣus. The yoginī, it is true, is the creator of the king, but as soon as the king is installed and his kingdom initiated (marking their very uneasy relationship), the yoginī retreats to the forest again to resume her life as a bhikṣu. There, in the forest, she transcends the mundane world of the kingdom again. Afterwards, the king can only meet her being a bhikṣu himself, such as in case of retirement may be possible, but then he has ceased to be a sovereign.
5. Conclusion
60The question raised in section i on how the universe of Sankhu should be conceived is largely answered by the descriptions of the creation and the structure of the kingdom. But the relation between the universe of Sankhu and that of ‘ Nepal ’ still remains.
61This study has arrived at a conclusion concerning the significance of a yoginī goddess in the creation and renewal processes of a kingdom. The fact that Vajrayoginī is a goddess unconnected with a particular territory, residing outside the sphere of civilization, is hereby a crucial element. How is this situation in ‘ Nepal ’? The part played by Vajrayoginī in Sankhu is played by Guhyeśvarī in the Greater Valley. By the Vajrayānists Guhyeśvarī (Nairātma Devi) is also considered as one of the great, if not the greatest, yoginīs of Nepal (together with the yoginīs of Pharping, Bijeswarī, Sankhu and Phulcok). Her temple, on a forested hillock near the Bagmati-river, upstreams from Paśupatināth, is built over a small ’ bottomless ’ hole or pit, For the Buddhists this is the place where in the beginning of creation the lotus carrying Svayambhū as a flame took root. For Hindus it is one of the 52 pīṭhas on places where sections of Sati Devi’s decomposed body fell after the raging Śiva had carried the body across the face of the earth, following Sati Devi’s suicide in the fire. The monuments and wealth donated to the temple by various kings testifies of the importance of Guhyeśvarī for the kingdom and kingship. Her connection with the royal centre (Kathmandu) is less strikingly enacted than in Sankhu, but is nonetheless represented by an annual jātrā. With this jātrā it is not Guhyeśvarī who moves to the royal palace, but, reversely, it is Taleju (the tutelary goddess of the monarchs) who travels from the palace to the shrine of Guhyeśvarī, at night and during new moon in November. Detailed information about this jātrā is lacking (it is very secret), but it seems that the maṇḍala of Taleju is carried in the jātrā and placed over the holy pit in the Guhyeśvarī temple. After elaborate rituals this maṇḍala is carried back to Kathmandu (communication A. W. van den Hoek) . Thus, while the kings in Kathmandu are yearly ’ reconsecrated ’ by Taleju (through Kumārī during Indra jātrā, see 4.5), Taleju herself receives a ‘ consecration ’ by the goddess Guhyeśvarī. The result in both the cases of Sankhu and Kathmandu is, of course, the same: the kingdom is created and kingship is ’ renewed ’ from outside, in casu, by a goddess from outside. A yoginī is such a goddess, qualitate qua. Apart from the inner structure of the kingdom (as represented in the ideal maṇḍala, see 3.2.h, Figure 4) the significance of the ‘ outer goddess ’ is the most important structural element contained in the idea of a Nepalese kingdom (i.e. what is meant by ’ Nepal ’).
62Although the ‘ kingdom of Sankhu ’ was defined as an embodiement of a universal totality, it is not a closed universe ritually. There are ritual relationships with the more encompassing totality ‘ Nepal ’, apparent from the importance for every single Newar of ’ universal ’ shrines (Svayambhūnāth, Paśupatināth, Matsyendranāth) or ’ assemblies ’ of shrines (sets of 4 or 8 like Ganeś, Nārāyaṇa) and also from the distribution, all over the Valley, of clan-shrines where clan-members assemble at least once a year. Since the centralization of political power (1768) the sovereignty of Kathmandu is also ritually recognized in Sankhu through the arrival, during the Vajrayoginī festival, of the ‘ royal sword ’ and through the position of the local acāju (who used to carry a title with this position, jamdar, ‘ non commissioned officer ’) to whom formal permission has to be asked to start the festival and who carries certain other ritual responsibilities (see 3-2.i, 4.2). Once every 12 years the centre (Kathmandu) also sends Devi dancers to perform in Sankhu (the other 11 years the dancers are locally recruited). Here is a set of different dimensions displaying a clear expression of ritual hierarchy.
63The significance of the goddess Vajrayoginī and her shrine has different (not wholly mutually exclusive) dimensions too, depending on the relevant context or level. On the purely local level of the town of Sankhu, Vajrayoginī is the kul devata (digu diya) of the Sankhu Vajracāryas, and her shrine, Gun Bahāl, is the central bahāl of a 9-group (8 + 1) symbolizing the totality of the sangha (where all major rituals such as initiations take place). Further, Vajrayoginī is involved in the local town ritual representing the mythical battle between Goddess and demons (the yearly Devi dances in every locality): of the three god-dancers (Devī, Candī, Bhairava) Devi wears the ornaments of Vajrayoginī. She is also invoked during Durgā Pūjā (a festival with the same meaning), the local element of which is expressed in the distribution of the sacrificial meat (see 3-2.i). On the level of the kingdom Vajrayoginī has a tantric aspect (as Ugra Tara playing on ghats. the Red-faced goddess) and a more peaceful aspect (Ekajaṭā, the Yellow-faced goddess), the creator of the king and the initiator of the kingdom as is demonstrated during her festival. On the level of ‘ Nepal ’ (Kathmandu Valley) Vajrayoginī, as Maniyoginī, displays a violent, tantric aspect, accepting human sacrifices (Wright: 97), her temple (periferially located) being the refuge for bhikṣus practising tantrics. Perhaps this is the most essential side of her Vajrayāṇa character (all Sādhanamālā texts describe Vajrayoginī, Ugra Tara, Ekajaṭā as violent deities). On a still more universal level Vajrayoginī is seen as the goddess of Wisdom, which is illustrated by her connections with Mañjuśrī (see 1) and her defence of the Buddhist doctrine (see her encounter with Śamkarācārya, 3. I). Her temple, as seat of the universal goddess, attracts pilgrims from all over Nepal as during the ‘ Nepal’-pilgrimage described in the Nepālamāhātmya.
Bibliographie
Des DOI sont automatiquement ajoutés aux références bibliographiques par Bilbo, l’outil d’annotation bibliographique d’OpenEdition. Ces références bibliographiques peuvent être téléchargées dans les formats APA, Chicago et MLA.
Format
- APA
- Chicago
- MLA
References
Anderson, M. (1971), The festivals of Nepal. London, Allen and Unwin.
Auer G. and Gutschov, N. (1974), Bhaktapur, Gestalt, Funktionen und veligiöse Symbolik einer nepalischen Stadt in vorindustriellen Entwicklungsstadium. Darmstadt, Technische Hochchule.
Bajracharya, Barnabajra (1962), Maṇisaila Mahāvadāṇa. Newari, Banepa.
Bernier, Ronald (1979), The Nepalese Pagoda. Origins and Style. New Delhi, S. Chand and Co.
Bhattacharya, Benoytosch (1958), Indian Buddhist Iconography. Calcutta, K. L. Mukhopadhyaya, 2nd ed.
Bhattacharya, France (1981), La déesse et le royaume selon le Kālaketu Upākhyāna du Caṇḍī Mangala in Biardeau 1981.
10.4000/books.editionsehess.26632 :Biardeau, Madeleine (ed.) (1981), Autour de la déesse hindoue, coll. Purusārtha n° 5. Ecole des Hautes Etudes en Sciences Sociales, Paris.
Bosch, F. D. K. (1968), Indra’s Stṛyd met Vṛtra. Bijdragen tot de TLV dl 124-11, pp. 241-263, ’s Gravenhage.
Bhooswan, Sayami (1972), The Lotus and the Flame. An account of Nepalese culture, Department of Information, H. M. G. Kathmandu.
Dutt, B. B. (1925), Town Planning in Ancient India. Thacher, Spink, Calcutta/Simla.
Evans Wenz, W. Y. (1958), Tibetan Yoga and Secret Doctrines, Oxford University Press.
Gopala, Raja Vaṃśavalī, Nepal National Archives, Ms n° 1-1593.
Gutschov, N. and Kölver, B. (1975), Ordered Space Concepts and Functions in a town in Nepal. Nepal Research Centre Publication, Wiesbaden, Franz Steiner Verlag.
Gustchov, N. and Manabajra Bajracharya (1977), Ritual as Mediator in Space in Kathmandu, Journal of the Nepal Research Centre, vol. 1,
Gustchov, N. (1980), Functions of Squares in Bhaktapur, in Jan Pieper (ed.), Ritual Space in India, Aarp, 17 (London), pp. 58-64.
Hahn, Michael (1979), The play Lokānandanātaka by Candragomin, Kailash, VII, n° 1, pp. 51-67.
Handurukande, Ratna (1967), Introduction to Manicūdāvadāna and Lokānanda. Pali Text Society, London, Luzac and Co, pp. i-xlv.
10.4324/9780429044755 :Heesterman, J. C., Two types of Spatial Boundaries, forthcoming in Festschrift S. N. Eisenstadt.
Levi, S. (1905), Le Népal: étude historique d’un royaume hindou, 3. vol. Paris, Annales du musée Guimet.
Locke (S. J.), J. K. (1980), Karunamaya: The cult of Avalokitesvara-Matsyendranath in the valley of Nepal. Research Centre for Nepal and Asian Studies, Tribhuvan University, Sahayogi Prakashan.
Pieper, Jan (1980), Ritual Space in India: Studies in Architectural Anthropology, Art and Archaeology Research Papers n° 17 (London).
Prusha, Carl (ed.) (1975), Kathmandu Valley: The Preservation of Physical Environment and Cultural Heritage, A protective Inventory. Vienna, Anton School and Co.
Riccardi, T., Jr (1975), Le Népal, part one (English transl. of Lévi, The Nepal), Kailash, III, n° 1, pp. 4-60.
Sumon Kamal, Tuladhar (1979-1980), Gwaye Da Tayegu: An Initial Ritual of the Samyaka Guthi, Contributions to Nepalese Studies, CNAS, Tribhuvan University, Kirtipur n° 182, pp. 47-70.
10.1017/CBO9780511558184 :Tambiah, S. J. (1976), World Conqueror and World Renouncer. Cambridge University Press.
Tevere Macfadyen J. and Vogt, J. W. (1977), The city is a mandala: Bhaktapur, Ekistics, vol. 44 (déc. 1977), pp. 307-309.
Toffin, G. (1979), Les aspects religieux de la royauté Néwar au Népal, Archives de Sciences Sociales des Religions, vol. 48, 1, pp. 33-81.
10.4000/books.editionsehess.26632 :Toffin, G. (1981), Culte des déesses et fêtes du Dasaï chez les Néwar (Népal) in Biardeau (ed.) 1981, pp. 55-79.
Uebach, Helga (1970), Das Nepālamāhātmyam des Skandapurānam (Münchener Universitäts Schriften, Reihe des philosophischen Fakultät, München).
La Vallée Poussin, L. de (1894), Maṇicūḍāvadāna, as related in the fourth chapter of the Svayambhūpurāṇa, The Journal of the Royal Asiatic Society, art. XI, London, p. 310.
Wright, Daniel (1877), History of Nepal, Cambridge; reprint: Nepal Antiquate Book Publishers, Kathmandu, 1972.
Notes de bas de page
1 Vajrayoginī (the temple images) carries in her left hand a lotus-bud. In at least one source Mañjuśrī is also mentioned as carrying a lotus in his left hand (nīlotpala; in the Sanskrit Kēlurak inscription from Java, vs. 8, cf. Bosch 1968: 242).
2 Manicūdāvadāna. The original text must have been written in the 5th century A. D. (Hahn 1979). Later versions show remarkable adaptations, such as the one included in the Svayambhūpurāna.
3 [nirargada ’ (Pali: niraggala), unimpeded, used as ep. of sacrifice in Skt [...] here apparently: yajñā nirargada – ‘ unrestrained ’ form of the aśvamedha (Edgerton: vol. II, 299).
4 Ratna Handurukande (1967) remarks that no sculptural representation of a king or sage giving away a crest-jewel has been found (p. xxxiii). However, there is such a sculpture near the Maṇicūr tank nowadays, though it is not of great antiquity.
5 See also a slightly different version, in: S. Levi 1905: vol. III, note sur les deux planches annexées au premier volume, pp. 168-169.
6 Bosch points out the close resemblance in character between Mañjuśrī and Kaśyapa, the latter playing a similar part in the creation of Kaśmīr (text: Nīlamata) as Mañjuśrī did in the creation of Nepal. Bosch then compares the structure of these legends and brings them into connection with the Vedic myth about Indra’s battle with Vṛtra.
7 I have not been able to obtain the original version. The translations in this section are by T. N. Mali, from the 1962 Newari source.
8 Cf. Santabajra Bajracharya: Sankhu Mai Festival, broadcasted by Radio Nepal, 9 Baisakh 2033.
Notes de fin
1 A first short fieldwork period in Sankhu, in 1978, was made possible by a grant from the Netherlands Foundation for the Advancement of Tropical Research. I further wish to thank Mr. T. N. Mali and Father John Locke S. J. for their guidance in Nepal, and Dr. A. W. van den Hoek, Prof. J. C. Heesterman and Dr. H. v. d. Muyzenberg for their comments. All Maps and Figures in this article were designed by the author.
Auteur
Le texte seul est utilisable sous licence Licence OpenEdition Books. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
Circulation et territoire dans le monde indien contemporain
Véronique Dupont et Frédéric Landy (dir.)
2010
Construire les savoirs dans l’action
Apprentissages et enjeux sociaux en Asie du Sud
Marie-Claude Mahias (dir.)
2011
Politique et religions en Asie du Sud
Le sécularisme dans tous ses états ?
Christophe Jaffrelot et Aminah Mohammad-Arif (dir.)
2012
L’Inde des Lumières
Discours, histoire, savoirs (XVIIe-XIXe siècle)
Marie Fourcade et Ines G. Županov (dir.)
2013
Cosmopolitismes en Asie du Sud
Sources, itinéraires, langues (XVIe-XVIIIe siècle)
Corinne Lefèvre, Ines G. Županov et Jorge Flores (dir.)
2015
L’Inde et l’Italie
Rencontres intellectuelles, politiques et artistiques
Tiziana Leucci, Claude Markovits et Marie Fourcade (dir.)
2018