1Contrary to general practice, I shall not present here the conclusions of a study, but rather the questions that arise at the outset of an investigation. The investigation here is into the origins and meaning of the mdos rituals. It was shaped by research conducted together with Samten G Karmay on the ritual of the Deer with Large Antlers (Sha ba ru rgyas) within the framework of the Centenary Colloquium of the Fifth Section (Religious Sciences), at the École pratique des Hautes Études in September 1986 (Blondeau and Karmay 1988). The Deer with Large Antlers ritual has been cited by Buddhist sources, since at least the 14th century, as the prototype for the bloody rituals of the ancient Bönpo. It is the performance of this ritual that is said to have convinced King Khri srong lde btsan to ban Bön and expel Bönpo practitioners. In point of fact, the Bönpo ritual that has come down to us is a mdos ritual in which there is no trace of a real sacrifice. If we are to accept the view of the Bön tradition, the text is a gter ma whose ‘discovery’ dates back to the 10th or 11th century. Although we are unable to confirm such antiquity, it is certain that its style and contents are archaic and scarcely Buddhicised, unlike in case of later Bönpo rituals.
2It was while researching this ritual with Samten Karmay that I became aware of a fundamental confusion within western Tibetology: the mdos are indeed well-known in Tibetological literature but, more often than not, they are confused with one of their component parts, namely, the famous ‘thread-crosses’, called nam mkha’ (space) by Tibetans. This confusion is not, however, pervasive, as we shall see. The most obvious example of it is to be found in the work of René de Nebesky-Wojkowitz: in a chapter titled ‘Thread-crosses and thread-cross ceremonies’, he speaks of ‘the so-called mdos or “thread-crosses”’, stating further on that an ‘alternative term for mdos is nam mkha’’ (Nebesky-Wojkowitz 1975 [1956]: 369). This confusion renders the available information ambiguous and hence difficult to work with.
- 1 The entry for mdos in Jäschke’s dictionary has: ‘a cross formed of two small sticks, the ends of w (...)
- 2 It is, moreover, impossible to determine if the extrapolations in Das’s dictionary from Sandberg’s (...)
3In the first footnote of his chapter on ‘thread-crosses’, Nebesky-Wojkowitz offers a fairly ample bibliography, which may be supplemented by the more recent works cited below. However, if these publications are examined in chronological order, it becomes clear that the authors tend to rely unquestioningly on their predecessors – for their definitions of mdos, the confounding of mdos and nam mkha’, and for some of their explanations. A demonstration of this would be too long to include here. I shall, accordingly, limit myself to a brief sketch of the sources of the error. This, I believe, lies in H Jäschke’s dictionary (1881)1 and, more crucially, in S C Das’s dictionary (1902), which itself draws on Graham Sandberg’s Handbook of Colloquial Tibetan (1894).2
4In comparing the explanations offered by the two books, it is useful to unpack the process by which errors that are copied get amplified. If we begin with Das’s dictionary, the entry under mdos reads:
‘a tall, thin pared stick or wand provided with small cross-pieces, and with strings of various colours fastened to the ends of the transverse pieces and stretched thence to the main upright stick; the whole evidently intended to be a rude imitation of the mast of a vessel with yard-arms and rigging. Several of these structures, each 3 to 5 feet in height, are planted in the ground on the hill-side just above the house where some sick person is lying; and various magical ceremonies having been performed over the sick person, the btsan or demons which were afflicting him, and which are reputed to have a great passion for the rigging of a vessel, are presumed to be expelled and to take refuge in the toy masts outside the cottage. The sticks with their coloured strings are presently cast away as glud [ransom] down the ravine (Snd. Hbk.). There seem to be various names for the masts such as: — … skyel ba’i mdos; … lha mo’i bskang mdos; … sgrol ma’i g.yul mdos; … rgyal mdos; … btsan mdos (these names, obviously, are types or names of mdos rituals).’ (1902: 676–677, emphases added and Tibetan terms transliterated using the Wylie system).
5The source of these picturesque details is here mentioned within parentheses: Snd. Hbk., that is, Sandberg’s Handbook. Das’s dictionary also has an entry under bskang mdos, which adds nothing to the definition cited above.
6Referring to Sandberg’s Handbook, we observe that Das has, in fact, extrapolated from it quite singularly. The Handbook states:
- 3 We see that, although the Tibetan spelling given is incorrect (ldo), the pronunciation indicated d (...)
‘Lü’ (གླུད་) or ལྡོ་ Dö3 (in Sikkim): reeds or very thin straight twigs about 2 feet in height, strung with yarn of various colours and bits of stick, in close likeness to the yards and rigging on a ship’s mast [emphases added]. When anyone is sick, these are made and planted beside the path nearest to the house. As they are supposed to resemble the natural dwelling of demons which cause sickness, the demons are presumed to be enticed therein out of the sick man’s body.’ (Sandberg 1894: 196)
- 4 This photograph was taken in 1982 by Katia Buffetrille. I thank her for her permission to use it h (...)
7The phrases emphasised above in the two citations leave one utterly perplexed with regard to Das’s interpretation of Sandberg’s comparison. However, the astonishing photograph of the nam mkha’ on a rooftop in Darjeeling (fig 1) proves that, in some cases at least, Sandberg’s comparison is correct.4
Fig 1: Remains of a mdos deposited on a roof in Darjeeling
Photo: K Buffetrille
8Taking these texts as their point of departure, various authors have added their personal notes, though without taking up the question as a whole. We thus find ourselves confronted with a series of contradictory or unfounded statements:
9– The rituals of the mdos, also known as ‘thread-crosses’, have been categorized as a practice of the ‘folk religion’ (Tucci 1973: chapter 6, 226ff [1988: 176ff]), as ‘sorcery’ (Waddell 1934 [1894]), and more generally as a type of ‘magical rite’.
10Observing the existence of such thread-crosses in numerous tribal populations in Asia and across the world, F D Lessing compares them, no longer to a ship’s masts, but to a spider’s web, proposing that they function as ‘traps for demons’ (1942: 148–149, 173n29). While the idea did gain some currency, I have not yet found any trace of it in either the Tibetological literature or in elicitations from Tibetan informants.
11In the Religions of Tibet (1973 [1988]), Tucci – who offers in this work, as in his Tibetan Painted Scrolls (1949), the most interesting descriptions and insights regarding the symbolism of the mdos – arbitrarily opposes the ‘folk’ mdos ritual to the gto, a ‘powerful exorcistic ritual (drag po)’. He specifies that the one who ‘is to carry it [ie the gto] out must be expert in the meditation of one of his guardian deities (yi dam) who is here invoked in his fierce and terrifying aspect’ (Tucci 1973: 228–229 [1988: 177–180], see also 1949: 725, and Nebesky-Wojkowitz 1975 [1956]: 380).
12Tucci’s dichotomy is contradicted, firstly, on typological grounds: the mdos are instances of cho ga, that is, they involve the various phases of any ritual performed by the clergy (see Blondeau and Karmay 1988). Secondly, it is contradicted by the ritual texts themselves, as we shall see below. Thirdly, it is contradicted by Stephan Beyer’s account of the performance of the sGrol ma’i g.yul mdos by the monks of Khams sprul Rinpoche’s monastery in Dalhousie: Beyer was advised to have this ritual performed on his behalf by yogins because ‘it requires highly developed contemplative powers to create an entire world, and indeed the ritual itself presupposes a trained ability in instantaneous generation [of the deity]’ (Beyer 1973: 330).
- 5 Of course, certain errors or obscurities — very surprising in a work by Tucci — could well be due (...)
- 6 The term yas is one that is essential to understanding this category of rituals. Yet, attempts at (...)
13Furthermore, the The’u rang mdos ma, published by H Eimer and Pema Tsering (1973), offers an easily verifiable example of how the gto cannot be opposed to the mdos or considered radically different from the latter: folio 3b contains two perfectly parallel sentences. In them, one finds the expressions gto glud and mdos glud in corresponding positions. Only the second term – glud (ransom) – in each of these expressions is really clear. Nevertheless, the two parallel expressions prove a quasi-equivalence of the terms gto and mdos within this ritual. (And, as we shall see below, several collections of mdos bear the title of gTo ’bum, meaning ‘Collection of gto’.) Additionally, the The’u rang mdos ma allows us to nuance another of Tucci’s statements, according to which the yas is a ceremony ‘less important’ than the mdos, and which is performed by the lay donor himself (Tucci 1973: 227–228 [1988: 176–177]).5 By contrast, on folio 4b of the The’u rang mdos ma, the the’u rang are invited to take the yas glud, just as they were invited earlier in the text to take the gto glud, the mdos glud. It is clear that, for the author of this ritual text at least, the three terms are interchangeable.6
14– For all these authors, the mdos have pre-Buddhist and Bönpo origins. In this, it must be acknowledged, they follow the opinion expressed by certain Tibetan authors. While the pre-Buddhist or non-Buddhist origins of this ritual seem likely, they have not been established. The only exception is the presence of the term nam mkha’ in the royal funeral ritual, to be discussed below.
- 7 The same formulation is taken up by dPa’ bo gTsug lag phreng ba (see Haarh 1969: 103).
15As for the mdos rites of the Bönpo proper, only Tucci mentions them. However, he does so according to the erroneous classification in the rGyal rabs gsal ba’i me long which is, as we know, a Buddhist text. Tucci writes of the ’phrul gshen, ‘who wear coloured threads of wool … throw mdos and yas, erect temples (srid [sic]) and sacred symbols (rten) and expel calamities, present and future’ (Tucci 1949: 715).7 When one considers the Bönpo classification itself, as it appears in the gZi brjid (see Snellgrove 1967: 76–87), one sees that the mdos belong to the sNang gshen theg pa, and that their symbolism in that fifteenth-century text is identical to the symbolism of the Buddhists.
- 8 See the bibliography furnished by Nebesky-Wojkowitz (1975 [1956], chapter 19), particularly the te (...)
16– A number of mdos rituals have been cited and described summarily in various Tibetological studies.8 It is in these that the most flagrant contradictions emerge: the authors draw generalisations from the specific ritual they describe; the mdos, which equate with the nam mkha’ for these authors, are sometimes presented as the palace of a deity, and sometimes as the prison of a demon; they are variously placed on top of the house, thrown away at a crossroads, abandoned on a mountain or even burnt. Even the highly relevant analysis offered by Tucci, based on the ritual texts he read, suffers from the same contradictions and generalisations (1949: 725, 740n33): he first remarks, quite rightly, that ‘mdos can be of various kinds … and have different names according to the deities under whose protection they are placed or the kind of demons which they are supposed to expel or the purpose for which they are meant’ (ibid: 740n33). He then goes on to state:
‘As a rule the mdos is placed upon four square … basements … then on the upper basement a mast is placed which is crossed by a transverse bar; threads of various colours are then interlaced between the arms of the cross so formed… This is the zhal yas khang, the palace, where the gods invoked are supposed to reside’. (ibid)
17This is followed by a description of other preparations, and penetrating remarks on the symbolism of the mdos (on Sumeru, the construction of the cosmos) and the parallel to be established with the symbolism of the stūpa. However, what Tucci terms a general ‘rule’ is not to be found in all the rituals, and comes from a text that he names further down in a note. He also states: ‘It is therefore a cosmos which is here built and upon which the exorcist makes the gods or demons to dwell so that he might control them’ (ibid). One can only wish that Tucci had examined the ambivalent nature of the mdos adduced in this proposition, it being simultaneously the palace of the deity invited to preside over the cosmos, and the abode of the demon one wants to expel.
- 9 The brief definition of mdos proffered by Macdonald at the outset indicates that she does not conf (...)
18– Finally, Tibetologists have focused on the ‘thread-crosses’ to the veritable exclusion of the problem of the relationship between the mdos and the glud (‘ransom’) rituals. Yet Tucci, once again, has rightly noted that behind ‘every mdos lies implicitly the concept of ransom (glud). The ritual accordingly requires an image of the person (ngar mi, ngar glud) on whose behalf the ransom is to be effected’ (1988: 177). Likewise, Ariane Macdonald explicitly establishes that the mdos of Pehar that she describes ‘is a “ransom for life” ritual, srog glud’ (1982: 1026).9
- 10 Snellgrove too distinguishes between mdos and nam mkha’. He renders mdos as ‘ritual device’ and ‘q (...)
19It is this same notion of ransom that is highlighted by the Lopön Tenzin Namdak in his commentary to Snellgrove:10
- 11 My translation. Snellgrove’s translation, no doubt through a lapse of attention, is the opposite o (...)
mdos ni lha ’dre la dga’ ba’i rdzas / mdos la glud zer thub / glud la mdos mi zer = ‘mDos refers to items which are pleasing to the lha ’dre [or the gods and demons]; a mdos may be referred to as a glud [ransom, but] a glud [ransom] is not referred to as a mdos’ (1967: 300).11
20At the same time, this commentary confirms the problem.
- 12 Dwags po Rinpoche, personal communication.
21As a first hypothesis, it may be considered that the difference between the mdos and the glud rituals is based on the presence of ‘thread-crosses’ (nam mkha’) in the former, these being absent in the glud. This hypothesis seems to be supported by an opinion expressed, albeit with hesitation, by Dwags po Rinpoche. For him, the term mdos evokes a glud with nam mkha’ (which he prefers to term mdos ris, a quasi-synonym).12 However, Yontan Gyatso and Samten Karmay deny the association of these two ideas. Furthermore, there are ritual texts that describe mdos without nam mkha’, and glud with nam mkha’. Examples of both types of texts are to be found in Nebesky-Wojkowitz (1975 [1956]), in chapter 19 on ‘Thread-crosses and thread-cross ceremonies’, and in chapter 26 on ‘Protection against evil’. The question of distinguishing between the characteristic features of mdos and ransom rituals (glud) is certainly crucial. I am, however, unable to furnish a response for the present.
22Of course, this critique of previous studies does not mean that they are devoid of interest or that their speculations – particularly Tucci’s – do not prove relevant. However, it seems to me that only a close study of Tibetan sources will allow us to clarify the contradictions noted above, and arrive at a definition of the nature of the mdos and an understanding of their origins.
23As far as the nature of the mdos is concerned, the problem could be resolved in part if a satisfactory meaning and etymology for the term were to be found. However, neither the indigenous Tibetan dictionaries nor the rituals nor contemporary Tibetan informants have been able to furnish these. Considering only definitions from the most recent and most complete dictionaries, we find the following under the entry for mdos:
- 13 This dictionary adds that these practices pertain to the domain of superstition! The Tibetan reads (...)
- 14 The Tibetan is: mchod sbyin gyi tshab bam glud rdzas ’di la gnyis dbye rgyu yod pa’i yas stags ni (...)
– ‘A ritual to expel the lha ’dre [the gods and demons of this world], or the name for materials used to send away the lha ’dre’ (Dag yig gsar bsgrigs, p394).13
– ‘Substitutes for offerings and gifts [mchod sbyin], or materials for a ransom [glud], these being of two types: the yas stagsare offerings to be made to the gods [lha]; the mas stags are gifts to be presented to the demons [gdon bgegs]’ (Bod rGya tshig mdzod chen mo, p1387).14
- 15 See Mahāvyutpatti, nos 5061-5063.
- 16 The question of a possible shift from a real sacrifice to its substitute has been invoked in Blond (...)
24This last definition requires further explanation. First, consider the compound term mchod sbyin. The explanations that follow, and the numerous ritual contexts in which it is used, indicate that it should be understood – in its contemporary sense – as a combination of offerings made to gods, and gifts presented to demons. This is a terminological distinction important to Buddhists, so much so that one of their main grievances against the Bönpo is the fact that the latter ‘make offerings’ (mchod pa) to demons. Yet, we cannot neglect the fact that the same compound term was used by the first Buddhist translators to render, among others, the Sanskrit word yajña, meaning ‘sacrifice’ in the Vedic context.15 It is not a matter of extrapolating by decreeing that the mdos was originally ‘a substitute for sacrifices’. Nevertheless, it should be stressed that the term’s ambiguity entails that it must be interpreted with caution.16
- 17 The Chos grags dictionary, moreover, gives the alternative terms ya stag and yas bsags (accumulati (...)
25Consider next the expressions yas stags and mas stags. These are equally enigmatic even if we recognise, at first glance, a dichotomy that plays on the up/down opposition significant in case of offerings (offered above, to the gods) as opposed to gifts (given below, to the demons). We have encountered the term yas several times above, interpreted variously by different authors: Nebesky-Wojkowitz translates it as ‘treasures’ (1975 [1956]: 378), since these are all objects, simulacra and substitutes that are to be offered. Snellgrove translates yas stags as ‘ritual items’, a gloss that reveals little about the term’s meaning (1967: 14). The term yas is already present in the Dunhuang manuscripts. Rolf A Stein offers an interpretation to which I shall return below. He remarks in a note, citing the dPal khang (1960) and the Chos grags (dGe bshes Chos kyi grags pa 1997) dictionaries: ‘in modern times, yas is more or less synonymous with mdos, the framework with strings stretched over a cross that serves as a ransom (glud) paid to the divinities or demons’. He mentions the spellings yas thag and yas rtags for an expression that he considers is formed of ‘two independent but related terms’, sometimes replaced by rngan (reward, payment) and thags (Stein 1970: 182n36).17
- 18 For yas stags, the entry reads: bon lugs las bshad pa’i lha ’dre la brngan pa’i glud gtor cha zhig(...)
26The Dunhuang manuscripts must be studied to find the contexts in which this term appears, so as to try to define it more precisely. Here, I merely wish to demonstrate that it is scarcely clearer to contemporary Tibetans: the Bod rGya tshig mdzod chen mo, with whose definition we began, yields the following explanations for yas stags and mas stags:18
- 19 For brngan pa, the same dictionary gives the following two meanings: (i) bslu ba’i ched du rgyu dn (...)
yas stags: according to explanations from the Bön tradition, a sort of ransom gtor ma (glud gtor) given as a ransom (brngan pa) to the lha ’dre. (The word brngan pa here is thus understood in the sense of its first meaning given in this dictionary, namely that of ransom. The second meaning mentioned is that of ‘to give a gift’.19)
mas stags: according to explanations from the Bön tradition, one manner of giving a glud gtor to the lha ’dre.
27As was stated before, Buddhist authors commonly refer to the Bön tradition, but such tautological definitions yield little information. Furthermore, no dictionary contains an entry for stags, and the variants mentioned above only complicate the problem by opening up other semantic fields.
28In the end, the only thing that emerges clearly from the diverse Tibetan sources examined to date is that the mdos includes the ritual construction, and the assembled objects and materials as a whole, possibly including as well nam mkha’ (fig 2).
Fig 2: mkha klong gsang mdos, bonpo monastery in Dolanji, Himachal Pradesh.
Photo: Y Imaeda
- 20 For an analysis of the manuscript, see Haarh (1969: 357–379, especially 373–375 on the funeral pro (...)
- 21 Lalou (1952: 351n4), however, cites the description of the fabrication of a ngar mi — an effigy su (...)
29Apparently, the term mdos does not figure in the Dunhuang manuscripts. The ‘ransom’ rituals (glud), by contrast, are mentioned quite frequently in the non-Buddhist texts, and may therefore be considered to have been one of the main apotropaic rituals. We mentioned above that the only objective argument advanced by Tibetologists to affirm the ‘pre-Buddhist and Bönpo’ origins of the mdos seemed to be the presence of the term nam mkha’ in the royal funeral ritual. However, when we refer to the text itself, we see that, although the term does appear there – spelt as nam ka’ – the context is obscure, such that neither Lalou (1952) nor Haarh (1969) ventured to translate it. The term is found three times, in the enumeration of personages and objects that constitute the funeral procession.20 The order of this procession is identical in two instances, and slightly modified in a third instance. But, in each case, the nam mkha’ (one or several of them) are placed in between zo rig and smra zhal. The meaning of the first term is completely unknown, while that of the second is conjectural. It certainly is difficult to conceive of nam mkha’ as designating an object other than the ‘thread-crosses’ of the later rituals.21 Nevertheless, the doubt persists in the absence of other occurrences, yet to be found, in Dunhuang manuscripts, occurrences which would serve as a parallel, thereby enabling a better understanding.
- 22 This is cited in Hoffman (1950: 405). The very same passage is found in the Myang chos ’byung, exc (...)
- 23 My thanks to Samten Karmay for having kindly shared this information with me. (Note also that the (...)
30Besides, as we know, the rituals do not enter into the field of Tibetan theoretical reflections. The only general remarks that I have encountered thus far have pertained to questions of the origins of mdos (of whether these are Buddhist or not), and of the authenticity of claims attributing the creation of specific mdos to specific masters. Thus, it is through the rituals themselves and their comparison that we must seek to arrive at a clearer understanding of their nature and their signification, and at better hypotheses regarding their origins and transformations over time. There is a very large number of these rituals, belonging to all schools. Apart from Kong sprul’s series in the Rin chen gter mdzod – which has been known for quite a while, and which was partially used by Tucci and Nebesky-Wojkowitz – several collections of mdos have been published recently in India, some of them overlapping. The sheer abundance of these rituals, dating from at least the 11th-12th centuries to the present, raises yet another question, regarding their enduring popularity. Of this, the following serve as examples: Kong sprul, in his gTer ston brgya rtsa’i rnam thar, indicates that the thirteenth-century Guru chos dbang, in his childhood, studied a hundred and one treaties on the principal teachings of the mdos; Tāranātha, in the Myang chung,22 speaks of one Ra shag bon po, proclaimed an expert on mdos in a prophecy mentioned in a mdos thang yig, a work that unfortunately remains unknown to us. (We must, no doubt, recognise in this personage the Ra shag gter ston whose mdos rituals are included in the Rin chen gter mdzod, and who was condemned by Guru chos dbang’s father for having exclusively practiced mdos rituals. In case of the Bön tradition, following the Legs bshad mdzod, this name designates two gter ston.) Furthermore, Samten Karmay recently learnt in Lhasa that there could well be a mdos gzhung manuscript by the Fifth Dalai Lama among the texts preserved at the Potala.23 Here again, a systematic examination of the accessible sources will perhaps help us go beyond the sole reason expressly advanced by Tibetan authors to explain the popularity of these rituals, viz their great usefulness.
31Before we get to the comparison of two very different mdos rituals, I present here three general remarks. These deal with certain constants, revealed through a reading of thirty-odd ritual texts:
321. These rituals can have more than one function: far from being performed solely in case of illness, as the Western literature on the subject would have it, these are multifunctional apotropaic rituals to remove all obstacles created by hostile supernatural forces. One thus finds long, repetitive and stereotypic lists that enumerate all the cases in which their performance is beneficial to material and spiritual life. As an example one may cite the list established by gTer bdag gling pa for the Ritual of Four Hundreds (brGya bzhi’i cho ga), to be discussed shortly:
- 24 See brGya bzhi’i cho ga ’dod don lhun grub ces bya ba (1976: 117–118). See also a comparable remar (...)
‘This ritual is immeasurably useful against calumnies by the bandhe [Buddhist monks] and the Bönpo, “ensorcellments” [gtad], a number of enemies and thieves, people’s words of insult, imminent death, illness and the loss of livestock when these occur, epidemics in the land. In particular, [the ritual is useful] against possession by the rgyal ’gong and the yi dwags, death under an inauspicious astral combination, diminution of vitality and prosperity [dbang thang], the increase in enemies that harm one’s ‘luck’ [klung rta, sic], a number of calendrical troubles [keg] and catastrophes [chags], illnesses of babies and incurable illnesses, etc. In short, it is useful in eliminating accidental adverse causes [mi mthun pa’i ’gal rkyen sel ba].’24
- 25 On this classification, see Blondeau (1986: 101–102).
33In cases noted in Ladakh by Brauen (1982: 322) and amongst the Sherpas by Ortner (1978: 95–98), the mdos, and particularly the Ritual of Four Hundreds, are performed during funeral ceremonies. It is perhaps due to this multifunctionality that Kong sprul, in the Rin chen gter mdzod, links them with the Mahāyoga but does not classify them within one of the categories, viz. protection, exorcism, etc. Nor does he classify them as one of the four ritual actions (las bzhi). Instead, he places them in a special category, namely that of ‘various ritual actions’ (las sna tshogs).25
- 26 bKa’ ’khor lo bar pa’i yang bcud shes rab snying po’i mdo la brten pa’i bdud zlog bar chad kun sel (...)
342. We have also seen that each author privileges a particular place for the final placement of the mdos – crossroads, mountains, etc. It is true that many rituals treat this point very briefly, which can yield the impression that there are no alternatives. Nevertheless, when the texts elaborate on it, one sees that the final placement of the mdos depends, in fact, on its goal, and on the enemy to be vanquished. A good example is to be found in the Sher snying bdud zlog, a ritual to dispel the bdud demons in reliance on the Prajñā-pāramitā-hṛdaya-sūtra. The text gives the following instructions:26
- 27 The Tibetan here reads: phul phyogs ni. This passage does not refer to places where the ritual is (...)
‘As for the direction in which one should offer [that is, abandon]27 this mdos,
if it was made for a sick person, offer it in a large graveyard;
if it was made to restore the spar kha, offer it in the direction of the enemy [demon] who harmed;
if it was made for well-being, offer it at a crossroads or on any elevated path;
if it was made as a form of counter-magic, offer it in the direction of the one who placed the curse;
if it was made to cast a spell, offer it towards the enemy;
if it was made because the rgyal po were tormenting someone, offer it near a temple or a stūpa;
if it was made because the ’dre mo were tormenting someone, offer it at the bottom of the village or where three roads intersect;
if it was made to bring rain or because the klu were tormenting someone, offer it in a lake or a spring;
if it was made because the yul lha was tormenting someone, offer it in his direction.
In short, escort it [bskyal ba] in any direction in which that which causes harm is to be found.’
353. With regard to the Buddhist mdos, not only is their structure identical, but the stylistic formulae, and the mantra and the mudrā, are repeated and are, in turn, identical to those of the rituals that one may term ‘classical’.
36Thus it is that the gTo ’bum dgos ’dod sna tshogs re skong bkod pa (1978) places at the start of the collection a text describing the six mantra and six mudrā (sngags drug phyag drug) because, it states, the blessing (byin rlabs) by these six mantra and mudrā is the preliminary step of all the mdos rituals. These are standard mantra and mudrā that accompany the ritual offerings – of the two waters, perfume, incense, etc – and transmute these into noumenal offerings (see Beyer 1973 for a detailed discussion of this).
37I now turn to a comparison of two rituals: the wholly Buddhicised Ritual of Four Hundreds (brGya bzhi’i cho ga); and the scarcely Buddhicised Deer with Large Antlers (Sha ba ru rgyas) mentioned at the outset, which is a Bönpo ritual.
- 28 See gTer bdag gling pa (1976), brGya bzhi’i cho ga ’dod don lhun grub ces bya ba.
- 29 Yontan Gyatso, personal communication. My thanks to him for having very kindly elicited informatio (...)
38The titles of the two already serve to illustrate one of the difficulties involved in defining a ritual as a mdos ritual: neither of the two titles contains the term mdos (see note 27 above on a similar case). Yet, from the very first line of the Deer with Large Antlers, we find instructions for the preparation ‘of the mdos’. And it is the version of the Ritual of Four Hundreds written by gTer bdag gling pa that Kong sprul places at the start of his compilation of mdos in the Rin chen gter mdzod.28 Furthermore, this ritual is definitely regarded as a mdos by the dGe lugs pa.29 It derives its name from the items placed in four concentric circles around the central area:
– one hundred little stūpa or tsha tsha;
– one hundred lamps;
– one hundred gtor ma or ngar glud;
- 30 On these sorts of gtor ma and objects made of dough, see Nebesky-Wojkowitz (1975 [1956], especiall (...)
– one hundred ’phyang bu or bshos bu.30
- 31 Yontan Gyatso, personal communication.
39This ritual is performed very frequently, including on the margins of the Tibetan plateau, as we have seen. The dGe lugs pa too perform it frequently at the end of other rituals.31 Nebesky-Wojkowitz briefly describes five variants (whose texts are, for the most part, accessible today). However, it is significant, in light of our initial remarks, that he does not place this description in his chapter focused on ‘thread-crosses’ (that he also terms mdos), but rather in a much later chapter focused on ‘Protection against evil’ (1975 [1956]: 503ff).
- 32 See brGya bzhi’i cho ga ’dod don lhun grub ces bya ba (1976: 117).
- 33 Several tantra with this name are to be found in the bKa’ ’gyur and the rNying ma’i rgyud ’bum. Th (...)
40The Ritual of Four Hundreds is one of several mdos that the Buddhists explicitly link to the Indian tradition. gTer bdag gling pa explains that there are two versions of this ritual32: one according to the sūtra that he states is followed therein, and the other according to the tantra, which is condensed (bsdus pa) from the tantra of gShin rje nag po.33 The second version was ‘discovered’ in Bodhgaya and exists in the form of separate texts considered to have been transmitted by Padmasambhava. Although gTer bdag gling pa does not state it clearly, one may infer that this tantra-based version belongs to the gter ma tradition. Indeed, while I have not encountered texts purporting to be tantra-based, this very ritual is to be found in a gter ma text which lacks a colophon but is probably attributable to the same gTer bdag gling pa. Be that as it may, the gter ma version is very close to the ‘sūtra-based’ version, the chief difference lying in their respective modes of transmission: via Vairocana in case of the gter ma, and via Nāgārjuna and Atiśa (with a variant replacing Atiśa with gNubs Sangs rgyas ye shes) in case of the sūtra-based versions. We know also of this same ritual elsewhere: one text by the Second Dalai Lama, linked to the transmission by Padmasambhava; one by the first Paṇ chen, linked to the transmission by Nāgārjuna and Atiśa; several later versions, both abbreviated and developed, that confirm the great popularity of this ritual amongst the Buddhists.
- 34 See ’Phrin las brGya bzhi’i cho ga’i dbus phyogs zhugs pa, in gTo phran (1973: 273–376), preceded (...)
41However, we also have a Bönpo version, undatable for the present as the author cannot be identified.34 This version, of course, manifests certain differences from the Buddhist versions – with regard to the deities invited into the maṇḍala and the four demons specifically targeted, the ritual’s arrangements, its originator (gShen rab), and so forth. Another crucial difference is that the mythical precedent of the Ritual of Four Hundreds is not mentioned. Precedents are given, however, for several ritual actions inserted within the main ritual. Indeed, this Bönpo ritual is far more detailed than its Buddhist parallels. It begins, for example, with the rite of ‘taking the earth’ (sa blangs pa), which is a sort of ransom to the earth goddess (Sa’i lha mo) for the terrain on which the Four Hundreds is to be performed. The text also indicates variants depending on the goal that one wishes to attain or the gravity of the case. Yet, despite all these differences, the ritual structure is the same as that which will be described below in the Buddhist rituals. That said, this ritual is problematic on too many grounds – linguistic and others – for it to be discussed in detail here. I will, therefore, limit myself to an analysis of the Buddhist rituals.
- 35 See for example Das (1902: 666); and the entry under bdud bzhi in the Bod rGya tshig mdzod chen mo(...)
- 36 See the illustration of this arrangement in Ortner (1978: 97).
42The Buddhist versions have several variants, but they all essentially boil down to the following: at the centre of a maṇḍala formed of an eight-petalled lotus, one draws the right hand of the Buddha Śākyamuni in the gift-bestowing gesture. On the wrist, one places a statue of the Buddha facing the fingers. At the centre of the palm, in a vessel filled with grain and other things, one places an effigy of the sponsor for whom the ritual is being performed. This effigy is richly adorned and includes an item belonging to the sponsor. It is adorned with nam mkha’, rgyang bu, arrow, spindle, etc. On the four fingers of the Buddha, excluding the thumb, are placed the effigies of the four bDud (Māra) of traditional Buddhism: phung po’i bdud, ’chi bdag gi bdud, nyon mongs pa’i bdud, lha’i bu’i bdud,35 each with its own colour and attributes. (In variants, these are to be visualized.) All around, on the lotus petals, one places the four concentric circles of tsha tsha, etc.36
43The origins of this ritual are repeatedly mentioned at several stages of the performance: once, Indra was struck by illness. Nothing could cure him. So, he called out to the Buddha. Now the Buddha knew that the illness had been provoked by the four Māra. So, he summoned them and gave them the four hundred sets of objects that are present in the ritual to this day. Thus did the Buddha cure Indra. (The gter ma version gives more details: the Buddha summoned the four Māra on his hand …)
- 37 For a detailed description of the Deer with Large Antlers ritual, see Blondeau and Karmay (1988).
44When comparing this ritual with that of the Deer with Large Antlers,37 a number of common features emerge:
- 38 To the best of my knowledge, none of the Tibetologists cited in this article offer a translation o (...)
- 39 Clearly, what is missing here is direct observation and information from officiants who are mdos e (...)
45– Around a central effigy (the lamb-sized deer made of dough in the Sha ba ru rgyas, the statue of the Buddha in the Four Hundreds) are placed the objects and items that constitute the mdos. Among these are the main effigy, the substitute for the person for whom the ritual is performed, and ‘secondary’ effigies whose role is not specified. For the present, we may hypothesize that the latter represent substitutes for the main sponsor’s family circle. Indeed, the officiant’s injunctions, ordering the hostile powers to desist from misdeeds, club together the sponsor, the sponsor’s family circle, and the officiant himself. The main and secondary effigies are indiscriminately referred to as ngar glud and ngar mi.38 Nam mkha’, ‘thread-crosses’, are also mentioned in the list of preparations but not subsequently. This fact is constant in all the mdos rituals: going strictly by what the ritual texts state, the nam mkha’ do not seem to have a specific function.39
- 40 In the Rin chen gter mdzod. See Nebesky-Wojkowitz for a summary of the ritual (1975 [1956]: 377–38 (...)
- 41 In gTo ’bum dgos ’dod sna tshogs re skong bkod pa (1978: 493–500). The word ’phyongs is not to be (...)
46As for symbolic explanations of the nam mkha’, I have only found two types to date. The first symbolic explanation has it that each object of the mdos is placed in turn in symbolic correspondence with, on the one hand, a body part of the gods to whom the mdos is offered, and, on the other, the corresponding body part of the sponsor (body, speech, thought, blood, etc.). The objects are then offered in exchange for the human homologue. One finds this symbolism expressed in an almost identical manner in two rituals, Nyang ral Nyi ma ’od zer’s gter ma, the rGyal mdos dkar po drug mdos,40 and Guru chos dbang’s gter ma, the mDos kyi bsngo ba’i ’phyongs zhes bya ba.41 (It is surely not a matter of indifference that these two gter ston are part of the same transmission lineage.) The Guru chos dbang ritual establishes the following correspondence for the nam mkha’:
- 42 nam mkha’i ’phyongs ni : bhyo : nam zhes bya ba yangs pa la : mkha’ zhes bya ba cir yang snang : n (...)
Bhyo! Nam signifies vast, and mkha’ signifies the complete absence of phenomena. Space [nam mkha’] diffuses the luminous rays of the rainbow. It is non-dual with the mind of the Haughty Ones. I, the yogin, offer it as a ransom for my mind-vijñāna [rnam shes sems kyi glud], and that of my entourage. It will fulfil your cruel wishes’.42
- 43 Srid pa mkha’ ’gro’i gling bzhi spyi mdos, cited in Namkhai Norbu (1981: 59): ye shes lnga ldan na (...)
47It is also in a series of correspondences of the same type that one finds the following in the eighteenth-century sTag sham Nus ldan rdo rje’s gter ma text: ‘This nam mkha’ endowed with the Five Wisdoms [to be understood as: ‘materialised in the threads of five colours’] is non-dual with the face of the ḍākiṇī’.43 It remains to be verified whether this strongly Buddhicised symbolic system is to be found elsewhere and can be established as a constant in the Buddhist mdos.
- 44 This handwritten text has been studied in an unpublished thesis by E dell’Angelo. See also his disc (...)
48The second type of symbolic explanation comes from an apparently very ancient Bönpo ritual, the Srid pa’i spyi mdos,44 as cited by Namkhai Norbu (1981: 46–47). The ritual features gShen rab mi bo who, having subdued the harmful forces, btsan and others, substitutes the bloody offerings on which the latter fed, with grains, beer, etc. as well as the following: ‘The ransom for life, pay it by means of space [srog glud la nam mkhas brngan]; that for breath, by means of wind; that for heat, by means of butter-lamps [mar me]; that for blood, by means of water; that for flesh, by means of clay [kham pa]’. The play on meaning in the term nam mkha’ – meaning ‘space’ and ‘thread-cross’ – that is found in the Guru chos dbang ritual, is here complete, to the point that one wonders if there are indeed concrete ‘thread-crosses’ in this symbolic explanation based on the elements.
49A final remark about the two types of symbolic explanation: they are related, the first being perhaps a development of the second. In any case, we are very far from ‘traps for demons’.
50– One ‘invites’ (spyan ’dren) the deer and the Buddha in identical terms, asking each to descend into the ‘support’, that is, each one’s statue.
- 45 The expression in Tibetan used here is rgyu sbyor yon bdag, a stereotypic epithet used in all the (...)
51– The officiant self-generates as the deity, who is the first to have performed the ritual – gShen rab mi bo in case of the Deer with Large Antlers ritual, and the Buddha in the Four Hundreds. He then recites the mythical precedent and orders the evil powers assembled to submit as they once did, to accept the objects and the ‘ransoms’ given them. In exchange, these powers are ordered to cease tormenting the ‘donor of all these riches’,45 and to return, each to his own domain, on pain of punishment.
52– At the end of the ritual, all the objects and ritual elements are gathered together and carried far away. (The brGya bzhi’i cho ga of the Rin chen gter mdzod includes a short appendix, indicating with greater precision where these are to be placed.)
53Thus, the structure of the two rituals seems very similar. Nevertheless, there are notable differences – both with respect to each other and with respect to other ritual texts. And many points remain to be clarified:
54– The central role of the deer in the Sha ba ru rgyas is not made explicit. Samten Karmay and I (1988) have discussed several elements that suggest the mythical importance of this animal in ancient Tibet. However, the ritual itself offers no explanation of its pre-eminence. According to the mythical precedent, it serves as a ‘ransom’ for the minister’s soul; together with a bird for the king’s soul; and with a tree for the subject’s soul. On a side note: a bird and tree branches are among the objects placed around the deer’s effigy, but this seems to be the case in many mdos rituals. In the mythical precedent, a discussion of which takes up the major part of the text, the bird and the tree have the same importance as the deer.
55– In the Four Hundreds, as in all the mdos rituals wherein the central figure is a yi dam, a two-part structure emerges very clearly:
56(a) The first part is that of the standard sādhana with its stereotypic ritual sequences and formulae: the officiant mentally places the representation, the support of the deity, at the summit of Sumeru or at the centre of the maṇḍala, within a divine palace (gzhal yas khang). He presents the ritual offerings, transmuted through his meditation, and the various mantra and mudrā, to the deity in its noumenal aspect (ye shes pa). He self generates as the deity (bdag bskyed) and then causes the deity to enter its material support (mdun bskyed).
57(b) The presence of the deity within the ritual space being thus ensured, and the officiant invested with the deity’s power, the second step then involves summoning the evil worldly powers as guests (mgron po). He orders them to take all that is offered as a ransom, in particular, the effigy that substitutes for the individual they are tormenting. These riches should appease their thirst for vengeance on account of past offenses (lan chags).
- 46 Buddhist influences are not manifest here except in one part of the symbolic description of the an (...)
58Of course, this two-part structure does not exist in case of the Deer with Large Antlers ritual, which scarcely manifests any Buddhicisation.46 But what becomes of this structure in the Buddhist rituals themselves when the central power is itself a bdud, a btsan, etc? This question, I think, brings us to the heart of one of the ambiguities to be found in the Tibetological literature. I do not claim to resolve this ambiguity, only perhaps to pose the question more clearly.
- 47 mdos la byon …. glud la byon.
59For example, if one considers the ritual named bTsan mdos in the gTo ’bum mdos chog nyer mkho’i rigs phyogs gcig tu bsdebs pa dGos ’dod kun ’byung zhes bya ba collection (1978: 229–232), one finds a simple initial formula of homage to gShin rje gshed. In the mTsho sman rgyal mo’i dgu mdos (ibid: 106–112), we find a formula of homage to Vajrapāṇi. In both cases, the effigy of a btsan or of mTsho sman rgyal mo is placed at the centre of the ritual space, imagined as a maṇḍala or as Sumeru, and is surrounded by all the objects, the gtor ma, the nam mkha’, the animal effigies, etc. Once the preparations are completed, one passes immediately to the invitation (spyan ’dren) of the btsan or the four sMan gyi rgyal mo (Medicine Queens). In the bTsan mdos, the different categories of btsan (or more probably their leaders) are in turn described and invited with the same formula. For example, ‘Great Klu btsan, come to the mdos! 47 Lord of the mdos, come to the mdos! Great lord of the ransom, come to the ransom!’
60This series of invitations occupies the entire text, and the ritual ends abruptly with a few sentences: ‘This good ransom is far more wonderful than the man … Take the ransom instead of the sponsor. I, the vidyādhara, I send it to you … Henceforth, do not harm him. Carry out the tasks assigned to you!’ And as a note: ‘After that, send [the mdos] in the direction of the harm.’
- 48 This has been described summarily by Lessing (1942: 148–149), and in detail by Beyer (1973: 321–35 (...)
61Thus, when the Buddhist mdos does not openly, as it were, bring into play the deities of the noumenal plane, we are confronted with a ritual structure that is very similar to that of the Deer with Large Antlers. Nonetheless, we know to what extent the ritual texts are mere aides-memoire. When these texts are as short as the ones discussed as examples above, one may consider that they present only variants of a ritual that is considered standard and that is transmitted elsewhere, in part orally. However, if this were the case, one would not place the btsan, mTsho sman rgyal mo, etc. as central entities of the mdos, this place being reserved for the yi dam. Should we conclude that this is the basic structure, and that the sādhana part is a Buddhist superimposition? To the extent that a pre-Buddhist origin of the mdos is recognised, the response is clearly in the affirmative. And the fact that in certain rituals, such as the sGrol ma’i g.yul bzlog,48 the yi dam is placed apart from the mdos, as a sort of doublet, reinforces the impression of a Buddhist veneer. As things stand, however, this interpretation cannot be proven.
62– On the other hand, as Tucci has emphasized, it is clear that the goal of the ritual, whatever its degree of Buddhicisation, is to satisfy the hostile worldly powers, by symbolically offering them the universe and all its riches, with the effigies – substitutes for living beings – presented to them as a lure. Taking the lure, these powers would then cease their tormenting. All this is always given to them as a ‘ransom’ (glud), as attested in the examples cited above. This brings us back to the question raised earlier: what is the difference between the mdos and the glud rituals? To what has already been said above, I would only add that the terminology is the same: the Tibetan expression used is mdos/glud btang ba (literally, ‘to send a mdos/a ransom’), meaning ‘to perform a mdos or glud ritual’. In the rituals too the two terms are juxtaposed all the time, as we have seen. An additional example is that of the injunction to the worldly powers to take what is presented to them, which is a veritable leitmotif: mdos ’di bzhes / glud ’di bzhes (‘Take this mdos! Take this ransom!’).
63– The statement of the mythical precedent, thought to be a compulsory element of rituals of this type since the studies by Stein and Karmay, is fairly frequently absent: it is not to be found in the bTsan mdos, nor in the mTsho sman rgyal mo’i dgu mdos cited above. Nor is it to be found in other, sometimes longer, rituals. Here too, I can only note this fact for the moment.
64– Finally, the origins of the mdos rituals remain a matter of hypothesis. In the Rin chen gter mdzod, Kong sprul says that the foundational texts for the mdos (mdos kyi gzhung) are the tantra lHa ’dre stong gi dra ba’i rgyud and Ma mo’i rgyud. One cannot neglect this reference. Yet, a rapid glance at these tantra texts in the rNying ma’i rgyud ’bum – many of them appearing under the title of Ma mo’i rgyud – yields little hope of finding anything but another series of mdos rituals. These texts should, nonetheless, be examined more closely, as should the tantra cited as their sources in a certain number of rituals (as we saw above in case of the Ritual of Four Hundreds). But the multiplicity of canonical sources cited (including the Prajñā-pāramitā-hṛdaya-sūtra!), the affirmations that specific mdos were performed for the first time by the Buddha and transmitted by Nāgārjuna, provoke skepticism, seeming like so much subterfuge aimed at linking these rituals with the Indian Buddhist tradition.
- 49 The Srid pa mkha’ ’gro’i gling bzhi spyi mdos, see note 44.
65However, one cannot rely to any greater degree on Bönpo beliefs regarding the origins of the mdos. Namkhai Norbu, accepting these beliefs, holds that gShen rab mi bo was the initiator of these rituals. Equally dismissive of Buddhists who are contemptuous of Bön and of Tibetologists who speak of shamanism, Norbu explains the origins of the mdos thus: gShen rab mi bo sought to replace the bloody sacrifices of the native Bönpo who worshipped demons (gdon bon and bdud bon). He, therefore, instituted these rituals, stipulating that live victims had to be replaced by effigies (Namkhai Norbu 1988: 43–47, also 1982: 21 where he presents the same thesis in summary). Norbu’s position is interesting as, with it, we finally have another statement, in different terms, of the gap recognized by Tibetologists between the ancient Bönpo and organized Bön. Nevertheless, his position merely takes up anew the thesis presented in a Bönpo mdos without any particular critique of sources49. All one can do, therefore, is highlight the parallel between this version of the mythical precedent and the Buddhist version mentioned at the outset of this article, noting that neither the one nor the other seems to be founded in the Dunhuang manuscripts.
- 50 Regarding the Dunhuang manuscripts, see Lalou (1952), Haarh (1969), Stein (1970). For more recent (...)
66I have no alternative solution for the present regarding the origins of the mdos. Yet, if we accept their probable pre-Buddhist or non-Buddhist origins, several avenues of research open up. (Although the lack of proof of such origins was emphasized earlier, this is, nevertheless, the hypothesis that seems the most plausible to me currently.) These avenues all converge towards a comparison with the ancient funeral rituals – with what the few remaining Dunhuang manuscripts tell us about these rituals as well as with their vestiges in contemporary funeral rituals. To demonstrate it would first require a systematic and exhaustive comparison of all available material on funeral rituals in Tibet and Tibetan cultural areas, then their comparison with the ancient documents. Here, I can only indicate briefly what seem to me the most striking points in studies that have already been published.50
67We have seen that, while the term mdos is not to be found in the Dunhuang manuscripts (an absence that is not truly conclusive given the fragmentary state of the collection), other terms such as yas, glud (and gto/lto) are well documented in the collection, including in the manuscripts dealing with funeral rituals. We must now return to Stein’s comments, cited in part above, since these offer the most interesting food for thought (Stein 1970, 1971). I shall, however, set aside his definition of yas (‘scaffolding of threads stretched over a cross …’) as it seems to suffer from the confusion between mdos and nam mkha’, and to have been driven by a series of associations of ideas rather than by the text of the manuscripts. On the other hand, the following traits may be retained, which survive in funeral rituals across Tibetan cultural areas, with the minimum of local variation, traits that are strikingly similar to those that characterise the mdos:
68– These funeral rituals involve making an effigy of the deceased. This is personalized using symbols representing the latter (man or woman): the effigy is dressed in the person’s clothes and ornaments (whether really or symbolically), and bears an inscription of his or her name. Like the ngar mi in the mdos, this effigy is a substitute – here, for the deceased, though it is treated as living as it is his or her ‘soul’ that is to be attracted there, spoken to, etc.
- 51 Gorer (1967) and Snellgrove (1957) underscore the incompatibility of these practices with Buddhist (...)
69– The effigy is fed, and all the food and wealth it may need in the other world is placed before it. The food and wealth may be presented to the effigy painted on tsag li as in the Bönpo ritual described by Kvaerne (1985) or created mentally as in the ritual described by Snellgrove (1957).51 The location at which the effigy has been placed is also described to it as a richly adorned ‘palace’ (zhal yas khang). This image of a palace stems from pre-Buddhist beliefs. One may infer this from the insistence with which the Buddhist transposition of the funeral ritual presents it, formulating wishes that the deceased be reborn ‘in a miraculous palace’ (’phrul gyi zhal yas khang, Stein 1970: 160). Recall that, in the mdos, the ‘palace’ is the abode of either the deity or the hostile power. Here, the parallel between the mdos and the funeral rituals no longer holds in any rigorous sense. Nevertheless, one may re-establish this parallel by appealing to conceptions of death that anthropologists are eliciting amongst Tibetan populations. These conceptions involve, for instance, appeasing the deceased so that he leaves for the afterlife, barring his route of return towards the living – that is, just as is done in case of demons in the mdos.
- 52 The Karmapa died in Chicago on 5 November 1981 at 11:30 pm. local time. Holmes presented this mate (...)
- 53 Holmes did not know the term used to designate this structure and was unable to obtain precise inf (...)
- 54 Or khung pur. This is a term that I do not quite understand yet. It may be better to understand it (...)
70– The ‘soul tent’ (thugs gur) and the ‘body tent’ (ring gur), mentioned in the royal funeral ritual and the other Dunhuang manuscripts dealing with funeral rituals, have been discussed at length by Stein (1970: 178–179n22). He lists all related terms, including the following: thugs khang and ring khang, translated as ‘soul house (?)’ and ‘body house’; also thugs spur, translated as ‘soul and corpse’ or ‘honourable corpse’. Stein suggests that thugs gur and ring gur could designate the same thing, viz. an image of the deceased, the one being the honorific form of the other. Lalou (1952), by contrast, distinguishes them and proposes that ring gur be translated as ‘canopy’ (Fr dais). Stein cites two contemporary references in support of his view, in which the effigy is apparently termed gur among the Sherpas and the Tamangs. I have no new information to bring to bear on this point. Nevertheless, it should be noted that in Ladakh, and elsewhere too, the corpse is placed within a structure named spur khang, ‘corpse house’ (Brauen 1982). In the royal funeral rituals, it is a matter of ‘two tents’ (gur gnyis), which would seem to indicate two distinct objects or ‘constructions’ – one for the corpse, and one for the mind or ‘soul’, which could be represented by an effigy (Lalou 1952: line 32). Finally – and this brings us back to the mdos – in her richly detailed oral exposition on the Sixteenth Karmapa’s funeral ceremonies that took place at Rumtek over forty-nine days (from 6 November to 24 December 1981, the day of the cremation), Katia Holmes offered the following information, with photographs to illustrate:52 from 10 November, the day on which the body was brought to Rumtek, and until the cremation, the body of the Karmapa (referred to in the honorific register as sku gdung and sku pur) was kept in the chapel situated on the terrace while rituals were performed in the temples below. The body, in a posture of meditation, was placed at the centre of a maṇḍala drawn on a platform, inside a ‘residence’ (Fr demeure)53 a cuboid construction formed of a scaffolding draped in fabrics that hid it from view, and around which were heaped lamps, offerings, etc. We recognise here an arrangement that is very similar to that described in the mdos rituals, even as it calls to mind in some ways the ‘body tent’ (ring gur) of the Dunhuang manuscripts. A final point to retain is that, for the cremation, a ‘cremation chamber’ was constructed, named gdung khang, honorific for spur khang, ‘corpse house’.54 On the exterior, this ‘cremation chamber’ was shaped like a stūpa, adorned all along the perimeter with cloth garlands that reproduced the drwa ba dra phyed ornamentation so typical of iconography and architectural decoration. This ornamentation, in Holmes’ words, ‘symbolises that the gdung khang has become a palace’ (my emphases). And, to complete the symbolic reconstruction, before the Karmapa’s body was carried into the structure, drawings of maṇḍala were introduced within the base of the structure.
- 55 In gTo ’bum mdos chog nyer mkho’i rigs phyogs gcig tu bsdebs pa dGos ’dod kun ’byung zhes bya ba ( (...)
71As a last resort, one may argue that all these correspondences between mdos, stūpa, maṇḍala, Sumeru and a divine palace, and their counterparts in the funeral rituals are, all in all, logical because they stem from the same symbolism, as expressed by Tucci (1973 [1988]). There is, however, a final element underlying the unifying Buddhist interpretation that seems to link the mdos to the pre-Buddhist beliefs. This is the indication, encountered in several rituals, that one must ‘pitch a tent’ (gur) atop the mdos. For example, in the mTsho sman rgyal mo’i dgu mdos cited above, whose description of preparations is very brief, it is stated: ‘Atop a pure base, place a ’brang rgyas [a round gtor ma] made of the five grains, mark white points on the four directions and the centre, place [literally, pitch] a tent of white silk’.55
72– One of the essential aims of a funeral ritual, be it contemporary or of the Dunhuang collection, is to ‘show the path’ (lam bstan pa) to the deceased, that the latter attain unhindered the Land of Joy (dga’ yul) in case of the Dunhuang manuscripts, and liberation or a good rebirth in case of Buddhist rituals. The expression ‘show the path’ is also used to designate a ritual action included in the conclusion of the mdos rites: the officiant, in meditation, ‘shows the path’ to the beings who were previously summoned and are now ordered to leave the place and return forever to their own domains. This translates materially into the simple displacement of the figurines, in the direction in which these beings must leave. Alternatively, the figurines may be placed temporarily along a pathway leading to the door, to be abandoned subsequently at a specific location.
73– Stein has highlighted the similarity between concepts of illness and death, and thus in rituals relating to them, in ancient Tibet. It is, therefore, out of the question to assert that this bundle of concordances between the mdos and the funeral rituals proves that the former are derived from the latter. I will add only two points:
- 56 Note, however, that an annual mdos ritual for the entire community is also performed among the Lep (...)
74Firstly, we have already noted that mdos may be performed during funeral ceremonies. Yet this is hardly compelling; we know very well that the choice of rituals to be performed depends on astrological prescription, the sponsor’s wealth, and so forth. On the other hand, among the Lepchas (see Gorer 1967: 350–353), whose conceptions of the afterlife seem very similar to those of ancient Tibet, the mdos are integral to funeral rituals, performed to cast away the death-causing demon.56
- 57 We may, nevertheless, recall the role of factors such as the frequent transformation of ancient wo (...)
75The second point is rather fraught, and I advance it with the greatest circumspection. The impossibility of finding a specific meaning or etymology for the term mdos has been underscored. Recall as well that, in Tibetan, to ‘perform a mdos ritual’ is termed mdos btang ba, literally, to ‘send a mdos’. Now, in case of the manuscript of the royal funeral ritual, as is known, the greater part is devoted to the description of mdad or mdad chen po, a term that Lalou has translated as ‘funeral’ (1952). Bacot and Thomas, for their part, have translated it as ‘burial site’ or ‘cremation site’ (1940: 15, 16, 19, translation: 35, 41). Haarh, however, having compiled all occurrences of the term in the Dunhuang Annals and linked it with the attested form ’dad, has demonstrated convincingly that, on the one hand, this ceremony was performed three years (reckoned in the Tibetan style) after the death of the sovereign (1969: 360ff). On the other hand, the ceremony corresponded to a funeral feast (in the narrowest sense of the term) offered to the deceased, or rather to the latter’s effigy, a part of the feast serving as a ‘ransom’ on the path. It should be added that the verb used in case of this ceremony is ‘send’ (mdad btang ba). It is, I think, important to mention this terminological and conceptual parallelism, though I refrain from any hypothesis.57
- 58 See Stein (1970: 160, 170, 171 for the text). Stein has discussed these terms amply, and I shall n (...)
76A final avenue remains to be invoked: that of China, and its funeral and Taoist rituals. I venture along this path with even greater reserve, given that I cannot pursue the research myself, lacking the requisite training in the domain of Chinese studies. The idea is not new. Lessing, describing the mdos of Tārā cited above, compares the embroidered cloth pavilion placed before the effigy of the goddess to Chinese funeral pavilions (1942: 148). And I wonder if it is not this comparison that one finds already, at the start of the manuscript that Stein studied wherein we have: to transfer the merit of the ‘body tent’ (ring gur bsngo ba), one begins by making a smrang dar (?), ‘taking as a model a house of silk (?), a Chinese house’.58
- 59 See Stein (1979: 551ff).
- 60 We know that the avalambana, which takes place over the course of a month, is aimed at feeding the (...)
77Contemporary studies are gradually revealing the Chinese influences – Buddhist and non-Buddhist – in the Tibetan manuscripts of Dunhuang. This includes the Chronicle (Bacot et al 1940: 93–170), which was once considered a pure product of the indigenous literature. Stein, in particular, has supported the idea of such Chinese influences, and has shown that even Taoism and its rituals were not unknown in Tibet.59 In his typically cautious style, he compares similar traits within the two cultures but refuses to offer a definitive opinion on the question of a possible filiation. I shall imitate him closely and end this excursion outside the field of Tibetology with two references. The first reference is to the avalambana ceremonies, as these have been described by De Groot (1886: 403–435).60 On the last day of these ceremonies, enormous platforms, decorated with multicoloured flags and banderoles, are erected before the temples. An effigy of Yama is placed on these platforms (or sometimes off them). The faithful cover the platforms with food, heaped and hung on pyramid-shaped bamboo structures, as well as with paper representations of ‘wealth’ offered to the dead. The second reference is to two more recent articles, by Schipper (1985a, 1985b), who describes Taoist rituals in which one finds the main elements of mdos and glud rituals, including the importance attributed to astrology – nag rtsis, the putatively Chinese astrology – in the mdos texts (several of which are said to have been translated from Chinese). I add to this the following information that Schipper kindly communicated to me: in case of the funeral ritual, the officiant has before him a ‘thread-cross’ of variable size that seems to play no role, but whose presence is necessary.
78As we approach the end of this sometimes-venturesome demonstration, it remains premature to draw conclusions. The hypotheses are numerous, and it is not certain that the material available will permit the verification of each one of them. Nevertheless, what seems obvious to me is that, to achieve such an end, what is first needed is a patient re-examination of the ritual terminology, towards clarifying the meanings of these terms in the non-Buddhist manuscripts of Dunhuang, and the meaning given to them in the Buddhist literature that has partially preserved it. Moreover, since no research of this type has yet been conducted in Tibet, the importance of ethnological research on ‘marginal’ populations cannot be overemphasized. These populations, on the whole, hold out evidence of religious conceptions and practices that, even if only by analogy, shed light on the conceptual world from which the non-Buddhist manuscripts of Dunhuang originate, a world whose many traces may be found in Tibetan Buddhism to this day.
Flux RSS
