Printer Friendly

Complementarity and opposition in early Tibetan ritual.

The Bon religion of Tibet is sometimes referred to as a form of heterodox Buddhism, or as the "fifth sect" of Tibetan Buddhism. As an institutionalized religious sect, it has its origins in the tenth and eleventh centuries. (1) In its outward appearance, its beliefs, rituals, and iconography, Bon, to an outside observer, differs little from Tibetan Buddhism. One of the most visible differences in this respect, and one that has been noted many times before, is that whereas Buddhists circumambulate sacred sites in a clockwise direction, the Bon-po do so in a counterclockwise fashion. Perhaps the greatest difference between Bon and Buddhism lies in their competing approaches to Tibetan history. (2) The issue at stake in this contest is the Bon religion's claim to antiquity, and, by extension, primacy of place in Tibet's religious landscape. Bon-po historiographers claim that the Bon religion was more or less fully formed when Buddhism took root in Tibet. Further underlining their aspirations to antiquity, the Bon-po claim the Buddha Sakyamuni as a Bon-po teacher within the lineage of the founder of Bon, Gshen-rab Mi-bo. Although Buddhist historiographers, arguing from the vantage point of reigning orthodoxy, generally concede that Bon is a "pre-Buddhist" religion, they customarily assert that Bon developed in phases, and that the modern Bon sect took form through plagiarism and mimicry of Buddhism. Reciprocally, Bon historiography lays the same charge at the door of the Buddhists. (3)

Modern scholars have tended to follow the Buddhists' lead in this argument by focusing mainly on the legitimacy of Bon-po claims to antiquity. Most have fallen victim to unexamined assumptions about "pre-Buddhist Bon," "the royal Bon religion," or even "pre-Buddhist Bon shamanism." In this way they create a sort of catch-all category for all "pre Buddhist," non-Buddhist, and by implication anti-Buddhist, Tibetan ritual practices, and tend to associate these with the cult of divine kingship prior to and during the period of the Tibetan Empire (c. 600-c. 850). (4) Those who have studied Old Tibetan ritual texts among the Dunhuang manuscripts, in particular R. A. Stein and Samten Karmay, indeed uncover a good deal of continuity between Old Tibetan ritual practices and later Bon ritual practices. (5) These Old Tibetan ritual practices are generally carried out by ritual specialists referred to as bon, bon-po, and gshen. As a result we have a curious situation whereby early Tibetan ritual specialists, bon-po, are referred to by the same term as adherents of the later systematized Bon religion. (6) This has perhaps exacerbated the insidious tendency to construct "pre-Buddhist Bon" as a hazy foil against which the introduction of Buddhism to Tibet is set. This category gained a semblance of nuance, however, when the French Tibetologist Marcelle Lalou argued that a plurality existed in early Tibet's ritual landscape whereby one class of ritual specialists, the bon, opposed another, the gshen.

In a pioneering article Lalou translated the Dunhuang document PT 1285, in which the bon are repeatedly glorified at the expense of the gshen, who fail in their healing remedies where the bon succeed (7) Other scholars took up this claim regarding the opposition between bon and gshen in their own work. In particular. R. A. Stein also examined the opposition between gshen and bon on numerous occasions in his own treatment of PT 1285 and other Dunhuang ritual texts. (8) While Stein revealed some further dynamics of the relationship between these two types of ritual specialists, pointing out, for example, that their ritual technologies overlapped and that they often worked in tandem, his most influential writings reified the model of simple opposition. (9) In my rereading of this document, I argue that the model of opposition between bon and gshen is based on a false dichotomy, and that these two classes of priests were identical, or nearly identical, ritual specialists involved mostly in healing, divination, and funerary rites. They belonged to a single ritual complex in which bon and gshen were interchangeable. Further, in my reading of PT 1285 I have uncovered a hitherto unnoticed feature of the text: it contains two specific types of liturgies. The first, which is described here as a "narrative" liturgy, constitutes the majority of the text, while the second type, referred to here as a "catalogue of healing antecedents," represents a minor, but significant part of the text. The most intriguing aspect of this revelation is that while the recited oral journey in the narrative liturgy moves upstream along the Gtsang-po River (Brahmaputra) from east to west, the "catalogues of healing antecedents" move in the opposite direction. I consider the significance of this divergence, first as possible evidence in support of Lalou and Stein's mode of bon I gshen opposition, and then, after dismissing this, I suggest that this directional opposition represents two different types of practices within a single ritual complex. In this way I argue that the model of opposition between bon and gshen as two distinct classes of ritual specialists must be replaced by a model of complementarity according to which these two interchangeable, and perhaps identical, classes of priests deployed diverging but complementary ritual technologies within a single tradition.


The ritual jurisdiction of ban and gshen has been mapped out through analyses and translations of numerous Dunhuang ritual texts. (10) In all known cases the bon and gshen are ritual specialists involved mainly in healing rites and funeral rites, and their ritual functions overlap a great deal. In the Old Tibetan document under consideration, PT 1285, bon and gshen are deployed as healers to either gods or kings in legendary episodes that serve as ritual antecedents in a healing liturgy. In this context the bon and gshen healers employ the same ritual technologies: they perform mo and phya divinations (mo btab phya klags). These are similar types of prognoses regarding an individual's future. (11)

Elsewhere a bon-po, Ltam-bon Dmus-long, performs gto and dpyad rites. (12) A word on these two terms: Karmay explains that the gto consists of two parts, a narrative followed by a rite. The narrative recounts the mythical antecedents of the rite and thus guarantees its effectiveness. (13) This is, of course, a fundamental principle in Tibetan ritual, and is not unique to gto rituals. As a ritual category, the term "gto rituals" in fact covers a wide range of Tibetan ritual practices, and is something of an umbrella category. (14) The term dpyad, which often appears together with gto, indicates diagnosis, and is often used in a medical context. (15) This is, of course, perfectly consonant with the Old Tibetan healing liturgies in which the term appears. Considering gto rituals and diagnoses (dpyad) together, Karmay asserts that the former is "probably quite effective at least on the psychological side whereas the dpyad deals purely with the physical side of the patient." (16)

Other Old Tibetan documents reveal that gto and dpyad rituals were not peculiar to bon, but were performed by gshen as well. In fact, one funerary document analyzed by Stein, PT 1068. includes an episode where "Father (Pha) Gshen-rab Mi-bo" arrives with two gshen for a woman's funeral. Here Gshen-rab states that the bon will perform the gto rite and the gshen will perform the dpyad. (17) Coming from the lips of a central, recurring figure in Old Tibetan ritual literature, and indeed one who would be recognized (or reimagined) as the founder of the Bon religion in the tenth and eleventh centuries, there could hardly be a clearer archetype for ritual cooperation and complementarity. Another funerary document, PT 1042. has bon and gshen working side by side at a funeral. (18) In the funerary context the bon and gshen also perform the funeral and burial rites known as shid, rmang, and mdad. (19)

As noted by Stein, another Old Tibetan ritual text, ITJ 734, names various types of bon-po--lha-bon, g.yang-bon, and phya-bon--who perform rites such as calling in good fortune / life essence (g.yang du glan) (ITJ 734, 2 recto, I. 61; 3 recto, I. 88) and, above all, ransom rites (glud), which are the theme of the text (see infra). (20) They are also involved in creating soul-houses (rla-khyim, rla-gang). soul-strongholds (rla [rlan]-mkhar), and soul-paths (ral-lam) for the soul (sku-rla) of the deceased (ITJ 734, 2 recto, 11. 68-71; 3 recto. II. 104-5). In addition, they perform lesser-known rites involving various strings and cords, such as the dmu-dag, lha-dag, and gsas-dag (ITJ 734, 6 recto, 1. 224).

In summary, bon [po] and gshen were ritual specialists operating during the period of the Tibetan Empire (c. 600-c. 850). They were important figures in Tibetan society, and performed divination (mo, phya), gto rituals, diagnoses (dpyad), ransom rites (glad), funeral rites (skid, rmang, mdad), and other rites, mainly in the context of healing rituals and funeral rites. Having briefly described the ritual functions of bon [po] and gshen and demonstrated their important role in Tibetan society during the imperial period, I will now discuss in greater detail the relationship between these two types of ritual specialists, as it appears in PT 1285.


Here we are concerned mainly with healing rites. Best exemplified by ITJ 734 and PT 1285, these texts are written liturgies meant to be performed by the healer and differing depending on the source of the malady and the nature of the patient (e.g, human or divine). The liturgy proceeds by citing a ritual antecedent for the patient's illness and recovery. There are two distinct types of liturgies found in PT 1285, which I have designated as "narrative liturgies" and "catalogues of healing antecedents." We shall see examples of each below. The first type of liturgy is exemplified by the "matrimonial narrative," in which a couple from one region wishes to marry their daughter to the ruler of another region. In some cases there are several suitors, while in other cases there are none. The marriage is generally unhappy, and ends in illness or death, usually by poisoning or as the result of an ill-advised hunting trip. (21) The bulk of PT 1285, sections I, V, VI, and VII, consists of just such narratives. Related narratives, involving possession by a demon (gdon) and poisoning, are found in sections II, III, and VIII. After the illness, or death, the patient is cured by a ritual specialist, almost always described as a bon-po. After the end of the narrative, the officiant, that is, the person who actually uses the liturgy as part of a rite, relates these antecedents of healing to the present case, bringing their collective weight and healing power to bear on the actual patient: "Today, if one looks to tomorrow, this benefits Rma-bu Mehing-rgyal and brings him good fortune" (de ring sang Ita na myi rma bu mchlng rgyal I 'dl la plan te bsod) (PT 1285, 11. 115-16). Myi Rma-bu literally means 'man, son of man', and Stein has demonstrated the equivalence of the terms mi, rma, and smra as all indicating 'man'. (22) He further considers Rma-bu Mching-rgyal to be a formulaic epithet. In this case I am inclined to read Rma-bu Mching-rgyal as 'everyman', and assume that the officiant would here insert the name of his or her patient. (23)

While most of PT 1285 consists of these narrative rites, the text also contains a number of "catalogues of healing antecedents." These are often reduced to a shorthand where only the dramatis personae and their settings are listed. Such passages have the appearance of a little more than catalogues. (24) The successive stories in each "catalogue" cannot really be regarded as episodes in a narrative, since the plot is nearly always the same. Within the text each of the "catalogues" passes through various ancient kingdoms of Tibet, evoking them as sites for antecedent healing vignettes. As with the narrative rites, after the final variant of the story the officiant relates these antecedents of healing to his actual patient with the same formula: "Today, if one looks to tomorrow, this benefits Rma-bu Mching rgyal and brings him good fortune." In this way the healers empower their remedies by appealing to mythical antecedents in which the rulers or gods of various regions were healed of their maladies through the agency of individual ban or gsken healers, and by piling antecedent upon antecedent during the course of an oral journey across the land of Tibet. The liturgist identifies with his heroic predecessors and gives the patient a royal template for recovery. (The extent and nature of this "identification" is an important matter both for assessing the [in]applicability of shamanic categories to early Tibetan ritual and for considering the Tibetan assimilation of generation and completion practices associated with the Vajrayana. It is likely, however, to remain an open question.) The appeal to precedent or antecedent (dpe) is an essential part of Tibetan ritual, and I would suggest that in this case it is not merely a preliminary to treatment, but an essential part of the remedy itself.

As healing liturgies specifically designed to give the patient a mythical antecedent for his or her own recovery, these narratives and "catalogues" of ritual antecedents constitute models of and for an ordered world. As such, their geographical reach is a reconnaissance of the cultural universe. It is notable, therefore, that this universe extends along the course of the Gtsang-po River, and includes hardly any far-flung territories. This geographical limitation could either point to an early date for the text's initial composition, or, more likely, designate the boundaries of a particular liturgical setting as defined by the ritual specialists who employed the text. Notably, all other known catalogues of principalities and catalogues of healing antecedents travel along the Gtsang-po River, with few exceptions. (25) It seems therefore, that the river is the organizing principle here. These named places, apart from a few mythical territories like the "land of gods" and the "land of demons (srin)" are well-known areas, and are centered on central Tibet.

Similar evocations of territory are found in chanted journeys practiced all over the Himalayan region. This process, by which quotidian nomina, as settings for mythical antecedents, explicitly become healing numina, is described in some detail by Gaenszle in the context of "shamanic journeys" of the Mewahang Rai of Eastern Nepal.
 This partly explains why the image of the journey is so prevalent: it
 evokes the 'history' of the ancestors and expresses an emotional
 attachment to the territory inhabited by them. Even though many
 persons have never travelled along the path of the shaman's chant it
 is a journey through real and still existent villages and localities
 which the audience generally knows, if not from first.
 hand experience, from stories and events linked with them. But in any
 case, as has been pointed out by Desjarlais (1989 [p. 291]), the
 landscape traversed in the journey is both symbolical and physical,
 'for a symbolic matrix has been graphed on the physical landscape',
 thus forming a 'healing geography'. (26)

While the applicability of the term "shaman" to ancient Tibetan ritual specialists such as ban and gshen remains unsettled. Gaenszle's observations concerning the Mewahang Rai recitations certainly hold true of those found in PT 1285. (27) In these types of oral journeys there is a double movement at work, whereby the chant traverses known geographical sites in a step-by-step manner, thus emphasizing distance between man and divinity in this rite, but paradoxically makes the distant near and the ancient present by applying the ritual antecedents to the actual patient. (28)

PT 1285 has been transcribed, translated, and paraphrased by Lalou, who observed that the text is characterized by an unstable orthography and numerous spelling variants / errors. I can do little to improve on her work, but here paraphrase one of the ritual antecedent tales in the liturgy in order to offer a taste of the text itself before moving on to more structural concerns. The following is a paraphrase of section V of PT 1285, which narrates 'Ol-rje Zin-brang's efforts to marry off his daughter, 'Ol-za Lhani-bu.
 In the land of 'Ol-phu Dga'-dang, 'Ol-rje Zin-brang and his wife.
 Brag-za Slebs-sma, had a beautiful daughter named 'Ol-za Lham-bu.
 Skyl-rje Rmang, from Ba, in Skyi-ro Lchang-sngon, courted her, but
 there were bad portents. Brag-rgyal Thang-po, from Nya-lung Brag-'or,
 then courted her as well, but there were also bad portents. 'Ol-za
 Lham-bu's father, 'Ol-rje Zin-brang, thought for three days and for
 three nights, and decided that it was better to have a neighbor a
 short way off than to draw near one who is a long way off. (29) He
 therefore ordered his daughter to marry Brag-rgyal Thang-po.

 'Ol-za Lham-bu was unhappy with this, and despised Brag-rgyal Thang
 po's realm. She missed her home in 'OI-bu Dga'-dang, and she and
 Brag-rgyal Thang-po lived together without having any sexual union
 whatever. (30)

 'Ol-za Lham-bu then poisoned her father. 'Ol-rje Zin-brang. Though
 one hundred male and female gshen convened, they did not find the
 nature of the illness. Then 'Ol-bon Ljang-tsa gshen gyi Mon-yug found
 the nature of the illness, and cured it. O1-rje Zin-brang was even
 better than before. Looking now at the man Rma-bu Mching-rgyal, this
 benefits him and brings him good fortune. (31)

This is a good example of a "matrimonial narrative." It is rather short, however, and a longer narrative, such as that of section I, would include more intermediary suitors.


In her introduction to PT 1285 Lalou notes the project of the text: it compares the healing techniques of the bon with those of the gshen and constantly glorifies the former while derogating the latter. It must be pointed out, however, that while this is true of roughly the first half of the document, the second half mentions the healing exploits of the gshen with hardly any mention of the ban. Roughly speaking, the first part of the document (sections I, II, III, V, VI, VII, and VIII), constituting the bulk of the text, contains "narratives" in which it is mostly the bon who acts as priests, while in the second, smaller part of the document (sections IV, IX, X, and XI), the "catalogues of healing antecedents," it is predominantly the gshen who acts as priests. In some cases, however, the two types of healers appear to be interchangeable.

In her analysis Lalou notes the formulaic nature of the gshen who fail where the ban succeed. (32) They are always one hundred male white-turbaned (thod-dkar) gshen from Dagsri Dkar-po ("the white, sunny mountain"), who sometimes chant the words shar or sha-ra-ra (PT 1285. 11. 39-40, 66-67, 86-87, 110, 136-37, 151-52, 165-66), and one hundred (in one case nine hundred) female gshen from Sribs-ri Nag mo ("the black, shady mountain"), who are sometimes qualified by the obscure term zhu-bub (PT 1285, II. 40-41, 67-68, 87, 110, 136-37, 152, 166-67). (33) While Lalou noted the formulaic nature of these inept gshen, she still took at face value the opposition between "bon and gshen" in the text. In fact, however, the opposition is not between bon and gshen as such, but between bon-po--a term which, as we shall see, appears to qualify both bon and gshen--and a specific class of mythical healers who hail not from locales along the Gtsang-po, as the other healers do, but from the "white, sunny mountain" and the "black, shady mountain," which are not toponyms, but formulaic indicators of the male and female principles yang [TEXT NOT REPRODUCIBLE IN ASCII.] and yin [TEXT NOT REPRODUCIBLE IN ASCII.]. (34) In either case, it is a matter of one priest within the tradition of the healing liturgy--either a bon or a gshen--juxtaposed with one hundred priests and one hundred priestesses from outside this tradition. Taken together with the interchangeable nature of bon and gshen in parts of the text, and with the fact that the text appears to be balanced between those liturgies that glorify the bon and those that glorify the gshen--and in one case, as we will see, a gshen triumphs over the hundreds of male and female "formulaic gshen]"--it is evident that we must set aside the model of opposition between bon and gshen put forward by Lalou and repeated by Stein and others.

Moving on to the individual liturgies in PT 1285, and examining them structurally, what crystallizes before us is the most intriguing aspect of this text, its ordering of territory. The first liturgy in PT 1285 is found in section T of the text, which appears to concern the wooing of Thang-nga Bria-ma, daughter of Smra"i-dang Ding-dings kyi rje and his wife (dbyal) Gun-gun Ma-btsun, in the land of Smra-yul Thag-brgyad/Smar yul Thag-brgyad. (35) The names of her suitors are listed along with their territories until a match is found with Thang-ba G.yu-thang of G.yu-ro Lung-sum. Thang-ba G.yu-thang is then struck ill after a hunting trip, and a single bon healer, Smra-bon Zing-ba, succeeds in healing him, in contrast to the inept remedies of one hundred male white-turbaned gshen (pho-gshen thod-dkar) from Dags-ri Dkar-po ("the white, sunny mountain") and the nine hundred female gshen (mo-gshen) from Sribs-ri Nag-mo ("the black, shady mountain"). (36)

As our primary concern is not with the precise contents, but with the broad outline of these liturgies, the list of territories is presented below in tabular form in the order in which they appear. The table only includes the name of each territory and its ruler, leaving out the names of the bride, the healer, and so on. Note that the first territory is that of the bride's father, while the last is that of her husband. All intervening territories are those of unsuccessful suitors. In other words it is not the case that each territory is the site of a healing antecedent.
Table 1. Territories Listed in a Matrimonial Healing Narrative, PT
1285, Section I (ll. 16-58)

 Territory Ruler

 1 Smra-yul Thag-brgyad Smra-dang Ding-ding kyi rje
 2 Rkong-yul Bre-sna Rkong-de Dkar-po
 3 Myang-dang Thags-sum Myang-tshun Rgyal-po
 4 Mchims-yul Dgu-sul Mchims-rje Gnyi-sla Che'u
 5 Dags-shul Shing-nag Drang-rje Rnol-nam
 6 Rngegs-yul Gru-bzhi Rngegs-rje Gling-brang Che'u
 7 Phu 'Ol-phu Dga'-dang 'Ol-rje Zin-brang
 8 Dbye-mo Yul-drug Dbye-rje Mkhar-pa
 9 Yar-lung Sogs-kar 'O-lde Spu-rgyal
10 Skyi-ro Lchang-sngon Skyi-rje Rmang-po
11 G.yu-ro Lung-sum Thang-ba G.yu-thang

Sections II (II. 58-79) and III (II. 79-90) of PT 1285 are narrative liturgies similar to that in section I: they relate the healing of an illness by a ban after unsuccessful remedies attempted by the hundred formulaic male and female gshen. In section II, Thang-ba Dmu-thang (37) of Dmu-yul is healed by Dmu-bon Ye'u-than Rmang-ba after being possessed by a demon (lit., "a demon enters his mind"; thugs su gdon bchug) (1. 62). Part III is a similar healing narrative concerning Ga-gar Ltangs kyi rje of Yar-lung Sogs-dkar, who resides in the stronghold (mkhar) Bar-pa Zo-brang, and is also possessed by a demon (11. 81-82). After the inept remedies of the hundred formulaic male and female gshen, he is cured by Lde-gshen Rmun-bu.

Section IV differs from the first three sections in that it is not a narrative, but a "catalogue of healing antecedents." This section is brief, and more similar in type to an actual catalogue of principalities than to the healing narratives that precede it.
Table 2. "Catalogue of Healing Antecedents," PT 1285, Section IV
(ll. 90-97)

 Territory Ruler Healer

1 Rngegs-smo Gru-bzhi Rngegs-rje Glum-'brang Rngegs-shen
 Tshe'u Sto-bo So-ngan

2 Dags-shul Shing-nag Dags-rgyal Spro-zin Dags gyi Rgyal-
 [nad] gshenHo-da

3 Mchims-shul] Mchims-rje Gnyi-zla Tshe'u Mchims-gshen Lo
 rab 'bring snar

4 Zhong-du Dam-dru Zhong-rje Drum-po Zhong-shin Lon-
 shin Thang po

5 Rkong-yul Bre-sna Rkong-rje Dkar-po Rkong-shen Dog-
 po Dog-rnying

We return to narrative liturgies in sections V, VI, and VII, each beginning with a matrimonial narrative similar to that in section I. Section V (11. 97-116), paraphrased above, begins in 'Ol-phu Dga'-dang with 'Ol-rje Zin-brang's attempts to marry off his daughter, 'Ol-za Lham-bu. One unsuccessful suitor, Skyl-rje Rmang, from Ba, in Skyl-ro Lchang-sngon. is followed by the successful candidate, Brag rgyal Thang-po, from Nya-lung Brag-'or. The unhappy bride then poisons her father, who is healed by 'Ol-bon Ljang-tsa gshen gyi Mon-yug, following the inept remedies of the hundred formulaic male and female gshen.

Section VI (ll. 117-43) contains matrimonial narratives involving mostly non-human actors. The ruler of Rgya-yul Gtan-bzangs, Rgya-rje Mying-mtshan, wishes to marry off his beautiful daughter, Thang-nga Rgya-mo-thang. The unsuccessful suitors hail from the land of the gods, the land of the serpent spirits (klu), and the land of men, and the successful candidate is from the land of demons (srin). In a perfect parallel with section V, the bride's father is poisoned, and then healed by Leg-tang Rmang-ba. (38)

Section VII (II. 144-60) presents the same type of narrative consisting of an unhappy marriage and a poisoned father. In this case Bal-rje La-nam of Bal-yul Leng-tang marries off his beautiful daughter. Thang-nga Bal-mo-thang, to Nags-rje Khri-ba of Nags-yul Deng-ba. Once the father is poisoned, following the useless remedies of the hundred formulaic male and female gshen, they call a bon-po, Bal-bon Rum, who succeeds in healing him. The narrative names no intervening suitors.

Section VIII (11. 161-70) is an aborted healing narrative set in Ltam-shul Gung-dang. It seems not to be a matrimonial narrative, but the ruler of Ltam-shul, Ltam-rje Ya-bo, is poisoned. Here a ban healer, Ltam-bon Dmus-long, succeeds where the hundred formulaic male and female gshen fail. After a damaged portion at the end of line 170, however, the narrative ends midstream, and a "catalogue of healing antecedents" begins on line 171, resembling section IV in that it presents more of a catalogue of principalities, and is not devoted to the glorification of the ban, but instead focuses on the activities of the gshen. This new catalogue, section IX, differs from section IV, however, because instead of healing rulers or their kin, in section IX the gshen heal divinities. This is a profound difference and demonstrates a fundamental tenet of the relationship between man and divinity in ancient Tibet: not only men, but gods as well, must be purified. The catalogue begins in Rtsang-ro, and ends with the statement "as for the others, proceed as above" (gzhan nl gong ma bzhln bgyl 'o).
Table 3. "Catalogue of Healing Antecedents," PT 1285, Section IX (11.

 Territory God Healer

 1 Rtsang-ro Dbyc-kar Rtsang-la Bye'u Rtsang-shen Snya-lngag

 2 Skyi-ro Ljang-sngon Skyi-bla Bya-rmang Skyi-gshen Rgyan-ngar

 3 Yar-lung Stsogs-kar Yar-la Sham-po Lde-gshen Rmun-bu

 4 Dbye-mo Yul-drug Dbye-bla Spyil- Dbye'-gshen Khar-bu Ljon-
 gangs phyug

 5 'Ol-phu Dga'-dang 'Ol-lha Sha-bzan 'Ol-gshen 'Jang-tsa Mon-

 6 Rngegs-shul Gru- Rngegs-lha Bya- Rngegs-gshen So-nang
 bzhi rmang Sto-bo

 7 Dags-shul Shing- Dags-lha Sgam-po Dags-gshen Rbam-

 8 Mchims-shul Dgu- Mchims-lha Than- Mchims-gshen Lo-rab
 sul 'tsho 'Bring-rab Snar-skyol

 9 Zhon-yul Dam- Zhong-[lha] Thang- Zhong-gshen Lon-shin
 drug mo Thang-po

10 Rkong-yul Bre-snar Rkong-lha Des-legs Rkong-gshen [+ -2]

Sections X and XI, both "catalogues of healing antecedents," are similar in structure to section IX. There appears to be no formal break between sections IX and X, nor between X and XI, and the latter two should probably be considered as a single section. They differ qualitatively from section IX, however, in that they name men and not gods as the patients of the gshen. Section X announces at the end of its second stanza concerning Skyi-ro Ljang-sngon that it will continue in this manner down to Rkong-yul (1. 188). Section XI then begins with Yar-lung Sogs-kar. Section XI stops at the end of the recto, and the first seven or eight lines of the verso are left blank, followed by two lines that are upside down in relation to the rest of the verso. This suggests that the second half of the sections X and XI, consisting of approximately seven lines of text, is missing. As with sections IV and IX, these are laconic "catalogues" that mention only illnesses, and do not include the narrative behind these maladies.
Table 4. "Catalogues of Healing Antecedents," PT 1285 Sections X
(11. 184-88) and XI (11. 188-93)

 Territory Ruler Healer

1 Rtsang-pho Phyed- Rtsang-rje Po-bla Rlsang-gshen Snya-ngag
 kar /Risang-rje Phwa

2 Skyi-ro Ljang- Skyi-rje Rmang- Skyi-gshen Rgyan-ngar
 sngon po

3 Yar-lung Sogs-kar 'Ol-lde Spu-rgyal Lde-gshen Rmun-bu

4 Dbye-mo Yul-drug Dbye-rje Khar-pa Dbye'-gshen Khar-bu Ljon-

5 'Ol-phu Dga'-dang 'Ol-rje Zing-brang 'Ol-bon 'Jang-tsa Mon-yug

The names of the healers correspond exactly with those in section IX, excepting 'Ol-gshen 'Jang-tsa Mon-yug's transformation into 'Ol-bon 'Jang-tsa Mon-yug, which illustrates the interchangeable nature of these two types of ritual specialists. Looking back to section IV, which resembles sections IX-XI, it can now be observed that section IV contains the second half of the list missing from sections X and XI. It is likely that the interpolation of the "catalogue" in section IV was an error on the part of the scribe.

Combining the data from sections IV, IX, X, and XI, which all represent one type of "catalogue" yields the following table. Where the fields for territory and healer overlap but differ, the data from section IX appear first.
Table 5. "Catalogue of Healing Antecedents," PT 1285 (Sections IV, IX,
X, and XI)

 Territory Ruler God Healer

 1 Rtsang-ro Rtsang-rje Rtsang-la Bye'u Rtsang-shen
 Dbye-kar/ Po-bla/ Snya-lngag/
 Rtsang-pho Rtsang-rjc Rtsang-gshen
 Phyed-kar Phwa Snya-ngag

 2 Skyi-ro Skyi-rje Skyi-bla Skyi-gshen
 Ljang-sngon Rmang-po Bya-rmang Rgyan-ngar

 3 Yar-lung 'Ol-lde Yar-la Lde-gshen Rmun-bu
 Stsogs-kar/ Spu-rgyal Sham-po

 4 Dbye-mo Dbye-rji: Dbye-bla Dbye'-gsben
 Yul-drug Khar-pa Spyil-gangs Khar-bu

 5 "Ol-phu 'Ol-rje "Ol-lha 'Ol-gshen
 Dga'-dang Zing-brang Sha-bzan 'Jang-tsa Mon-yug
 'Jang-tsa Mon-yug

 6 Rngegs-shul Rngegs-rje Rngegs-Iha Rngegs-gshen
 Gru-bzhi/ Glum-'brang Bya-rmang So-nang Sto-bo/
 Rngegs-smo Tshe'u Rngegs-shen Sto-bo
 Gru-bzhi So-nang

 7 Dags-shul Dags-rgyal Dags-lha Dags-gshen
 Shing-nag/ Spro-zin Sgam-po Rbam-dang-tsha/
 Dags-shu Dags gyi
 Shing-nad Rgyal-gshen Ho-da

 8 Mchims-shul Mchims-rje Mchims-lha Mchims-gshen
 Dgu-sul Gnyi-zla Che'u Than-'tsho Lo-rab 'Bring-rab

 9 Zhon-yul Zhong-rje Zhong-[lha] Zhong-gshen
 Dam-drug/ Drum-po Thang-mo Lon-shin Thang-po
 Zhong-du /Zbong-shin
 Dam-dru Lon-shin Thang-po

10 Rkong-yul Rkong-rje Rkong-lha Rkong-gshen [+ -2]
 Bre-snar Dkar-po Des-legs /Rkong-shen
 Dog-po Dog-rnying

As noted above, these laconic "catalogues" represent a small but divergent portion of the text. Unlike the matrimonial narratives exemplified by section I, no suitors or other figures separate the protagonists: each territory listed is the site of a healing antecedent that is applied to the patient. This contrasts with the matrimonial "narratives," where intermediary suitors and their realms separate the liturgies' protagonists.

Looking now to the ordering of space in these two types of liturgies, the matrimonial and poisoning narratives that constitute the bulk of the document, sections II, III, VII, and VIII, are narratives set in a single territory, and cannot therefore be easily compared with the ordering of territory in the "catalogues." Section V, with but one unsuccessful suitor in its matrimonial narrative, lists three territories. Section VI, while a bit longer in this respect, is not geographical in nature, as it concerns gods, serpent spirits (klu), men, and demons (srin). Only section I names enough territories to be employed for a fruitful structural comparison with the "catalogues" in terms of its ritual ordering of space.

As the concern here is the liturgical ordering of space, only the territories, and not the rulers, gods, or healers, are included in the table below. For ease of comparison, the territories invoked in the "narrative" from section I are presented here alongside those listed in the "catalogue" from section IX. The third column lists the territories from the first column in reverse order for ease of comparison, and to illustrate the inverse correspondence of named territories.
Table 6. The Inverse Ordering of Territory in "Narrative Liturgies" and
"Catalogues of Healing Antecedents"

 Territory (Section I; Territory (Section IX; Territory (Section
 "Narrative") "Catalogue") I "Narrative"--in

 1 Smra-yul Thag-brgyad Rtsang-or Dbye-kar G.yu-ro Lung-sum

 2 Rkong-yul Bre-sna Ski-ro Ljang-sngon Skyi-ro Lchang-

 3 Myang-dang Thags- Yar-lung Stsogs-kar Yar-lung Sogs-kar

 4 Mchims-yul Dgu-sul Dbye-mo Yul-drug Dbye-mo Yul-drug

 5 Dags-shul Shing-nag 'Ol-phu Dga'-dang Phu 'Ol-phu Dga'

 6 Rngegs-yul Gru-bzhi Rngegs-shul Gru-bzhi Rngegs-shul Gru-

 7 Phu 'Ol-phu Dag'-dang Dags-shul Shing-nag Dags-shul Shing-

 8 Dbye-mo Yul-drug Mchims-shul Dgu-sul Mchims-yul Dgu-sul

 9 Yar-lung Sogs-kar Zhon-yul Dam-drug Myang-dang Thags-

10 Skyi-or Lchang-sngon Rkong-yul Bre-snar Rkong-yul Bre-sna

11 G.yu-ro Lung-sum Smra-yul Thag-

As is evident from this table, the "narrative liturgy" in section I follows almost the exact opposite order of that followed by the "catalogues" exemplified by section IX. The former moves from east to west up the Gtsang-po River, while the latter moves in the opposite direction, downstream, from west to east. The only significant differences are that the "narrative" does not include Rtsang-ro Dbye-dkar, but names G.yu-ro Lung-sum in the corresponding place, and has Myang-dang Thags-sum in the place corresponding to the catalogue's Zhon-yul Dam-drug. Myang-dang Thags-sum presumably corresponds to the territory of the minor kingdom (rgyal-phran) Myang-yul, usually referred to as Myang-yul Rta-gsum or Myang-yul Shing-nag. This corresponds generally to Rgya-mda' County in Kong-po. (39) Finally, the "narrative" begins with Smra-yul Thag-brgyad, which is not found in the "catalogues." As noted already, smra is a synonym for mi, meaning 'man'. The term 'eight cords' (thag-brgyad) is often used to describe a structure in funeral rites, which seems to be some sort of tent, or soul-house that acts as a microcosm and houses the deceased or his effigy. (40) Smra-yul Thag-brgyad may therefore not refer to an actual place, but simply serve as a formulaic point of departure.

The places invoked are part of an evocation of the ritual or liturgical universe, but they simultaneously represent known territories. While the location of Zhon-yul is uncertain, and that of Rngegs-yul is tentative, the territories invoked may be sketched on the map below. (For a detailed discussion of the ancient principalities or "minor kingdoms," their locations, and a brief survey of the tombs that they appear to have left behind, see Hazod's survey in Dotson, The Old Tibetan Annals.)

It may be significant that in all other known Dunhuang liturgies that include an oral circuit of territory, the movement is downstream, from west to east along the Gtsang-po River, i.e., the opposite of that of the "narrative liturgy" of section I in PT 1285. The best known non-liturgical catalogue of territory, which appears in the Royal Genealogy (PT 1286), also moves in the direction of the river. As this document also contains a genealogy of the Tibetan royal line and appears to have enjoyed official sanction, one could argue that this west-to-east movement down the Gtsang-po River represented an orthodox ordering of territory. Another semi-official document--or, rather, a muddled copy of an official or semi-official document--PT 1290, which apparently records the enthronement of Khri Gtsug-lde-brtsan (ruled 815-841), follows this same west-to-east, downstream movement in its catalogues of principalities. This same movement is found in the catalogues of territory in later histories such as the mid-to-late-thirteenth-century Rgya bod kyi chos 'byung rgyas pa of Mkhas-pa Lde'u (p. 225) and the mid-sixteenth-century Mkhas pa'i dga' ston by Dpa'-bo Gtsug-lag Phreng-ba (pp. 155-56). In this case the ordering of territory up the river from east to west in the "narrative" rite would perhaps represent a heterodox tradition.


This suggestion is particularly intriguing when we observe that the healers in the "narrative liturgies" are predominantly ban, and the healers in the "catalogues" are almost always gshen. It is tempting, therefore, to regard the "catalogues" as representative of a "gshen tradition," as opposed to the matrimonial and poisoning narratives, which might be regarded as representative of a "ban tradition." Given the latter's apparently heterodox position in terms of its ordering of territory, this could be taken as an enticing precursor to the anti-nomian identity of the later institutionalized Bon religion in relation to Buddhism, and in particular to the practice of counterclockwise circumambulation. Unfortunately, there are other indicators within the liturgies that suggest that this ban / gshen opposition and the suggested orthodox versus heterodox liturgical ordering of space is little more than a false dichotomy. As mentioned already, while section IX names a healer called 'Ol-shen 'Jang-tsa Mon-yug (1. 177), section XI refers to him as 'Ol-bon 'Jang-tsa Mon-yug (1. 191). Both of these occur in gshen-dominated "catalogues." Furthermore, this same figure is found in the bon-dominated "narrative" in section V as 'Ol-bo[n] Ljang-tsa Gshen gyi Mon-yug (1. 110). Were the bon I gshen dichotomy a hard-and-fast reality, one would not find bon priests in the "gshen catalogues" and gshen priests in ""bon narratives." Likewise, these terms would not be used interchangeably, to the extent that one ritual specialist, 'Ol-bo[n] Ljang-tsa Gshen gyi Mon-yug, could be both a bon and a gshen at one and the same time. Likewise, the success of Lde-gshen Rmun-bu in healing Ga-gar Ltangs kyi rje in section III, after the inept remedies of the hundred male and female gshen (1. 88), would be unlikely were this indeed a healing narrative belonging to a tradition that supported only the bon at the expense of the gshen.

The interchangeability of bon and gshen, and the presence of bon priests in the "catalogues" and gshen in the "narratives," are not, however, the only, or indeed the most persuasive, pieces of evidence against reading a bon / gshen dichotomy in PT 1285. In section I, following the useless attempts of the hundred formulaic male and female gshen, we find the phrase, "Go [to] the bon-po Lam-ba! Go [to] Smra-bon Lam-ba" (ban po lam ba' [la] mchls / smra bon lam ba [la] mchis) (1. 41). In section II, in exactly the same context, we find the phrase, "Go to the bon-po!" (bon po lha (41) mchi shig) (1. 68). This instruction is followed by the appearance of the successful healer, Dmu-bon Ye'u-than Rmang-ba, who heals Thang-ba Dmu-thang. The same phrase is found in exactly the same context in section III, and is followed by the appearance of the healer Lde-gshen Rmun-bu (1. 88). This repetition is pertinent because the phrase explicitly qualifies a gshen, Lde-gshen Rmun-bu, as a bon-po. While section IV consists of a laconic catalogue and does not amount to a narrative, the passage on the ruler of Rkong-yul is slightly more elaborated than the other sections, and, after a description of the ruler's suffering, we again find the phrase "Go to the bon-po" (bon-po la mchl shig) (1. 94). This is followed by the arrival of Rkong-shen Dog-po Dog-rnying, who succeeds in healing the ruler. Not only is this gshen qualified as a bon-po, but this phrase, "Go to the bon-po!" appears here in the context of a gshen-dominated "catalogue." This emphasizes the interchangeability--and perhaps even the identity--of bon and gshen with vivid clarity. This same phrase "Go to the ban-pol" follows the inept remedies of the hundred formulaic male and female gshen in sections VI (1. 137), VII (11. 152-53), and VIII (1. 167).

We can draw from these considerations two significant conclusions in PT 1285 and other Old Tibetan ritual texts: bon and gshen are interchangeable and more or less identical ritual specialists, and the opposition between bon and gshen that Lalou and Stein read in the liturgies of PT 1285 is a false dichotomy. Considering this interchangeability alongside their overlapping ritual jurisdictions, we arrive at a significant revelation regarding early Tibetan ritual traditions: bon and gshen are both qualified as bon-po, and are part of the same ritual tradition, which, at the risk of extrapolating too much on the basis of the present text, might be fairly described as the "bon-po ritual tradition." This is not to collapse a pluralist notion of early Tibetan religion within this ritual tradition, for such fundamental features of Tibetan religious life as the cult of mountain deities may have remained outside its sphere of influence. In addition, the inclusion of gshen within the "bon-po ritual tradition" again underlines the continuity between early bon-po concepts and practices and those of the institutionalized Bon religion. At the same time, this is a picture gleaned from liturgical descriptions of bon and gshen, and does not describe relationships obtaining between actual bon and gshen in a given time and place. It is, rather, an exemplar for members of the tradition to follow. One must also bear in mind that PT 1285 and other Dunhuang ritual texts are not dated, and one cannot safely assume that they date to the imperial period (c. 600-c. 850) or to the Tibetan occupation of Dunhuang (786-848). In a similar vein, Blezer ("sTon pa gShen rab," 423-24, n. 6) has recently suggested that these texts are rather late (ninth and tenth centuries), and stand at or near the beginnings of a self-conscious bon-po ivention of tradition that would feed into the emergence of the Bon religion in the tenth and eleventh centuries, a contention that is examined in further detail below.

While these conclusions arising from a re-reading of the Old Tibetan ritual text PT 1285 are significant in and of themselves, one difficult question remains unanswered: if the neatly diverging ordering of territory noted above does not indicate opposition and/or complementarity between bon and gshen, then what does it signify?


Oral evocations of territory and chanted journeys similar to those contained in the Old Tibetan document PT 1285 are found in many parts of the Himalayan region. Within Himalayan anthropology the ritual organization of space is a prominent topic, and work in this field provides a clear interpretive framework for conceptualizing the inverse ordering of territory in PT 1285. In what follows I will consider this inverse ordering of territory in light of this interpretive framework with an eye to outlining and defining the problem.

As noted above, Gaenszle analyzes similar liturgical evocations of territory among the Mewahang Rai of eastern Nepal. Gaenszle distinguishes two types of oral journeys based on their "intentionality."
 Basically, one can distinguish two types: either one travels toward a
 distant place, and this usually implies taking something along,
 namely the offerings; or, one returns from a distant location, and
 this usually implies bringing something back, generally the "souls."

In the Mewahang Rai context, as in the early Tibetan case under consideration, these journeys are oriented along a river, and the ritual specialist's chant travels upstream to recover items such as lost souls, and downstream, towards the south, to discard items. In Gaenszle's example, this conceptual orientation along the river corresponds not only to a north-south axis, but also includes the categories of up and down and life and death. (43)

Let us now apply this discursive framework to the directional orientations in early Tibetan liturgies. The territories named in the matrimonial narrative in section I of PT 1285 begin in the east and proceed westward, upstream along the Gtsang-po River. In the "catalogues," by contrast, the movement is downstream from west to east. Were these liturgies typologically cognate with those described by Gaenszle, we would expect the upstream movement in the narrative of section I to be associated with recovering something, and the downstream movement of the "catalogues" to be concerned with abandoning something.

In order to consider this model fairly, it is necessary first to refine our understanding of the rites described in PT 1285. In the first place we should state unequivocally that although these oral journeys are represented as only going in one direction, they necessarily imply a return. These rituals describe a communication with the supernatural, and there is always an exchange or result, which returns to the client or group he represents. This is part of the double movement described above, by which evoking the distance between man and the supernatural paradoxically brings them near and facilitates their ritual exchange. (44)

The details of the "catalogues" in PT 1285 are particularly sparse: a king or god is struck ill and is healed by a priest--usually a gshen, but sometimes a bon--and that is all. Their laconic nature prevents any detailed analysis of the patients' conditions, although we do find "headaches" (dbu-snyung) and "eye aches" (spyan-snyung) (ll. 94, 171, 173, 184-85, 186-87, 189). Elsewhere, the mind (thugs) (ll. 185, 187, 189) seems to be affected. In most cases, however, it is only stated that the patient is "struck ill" (snyung).

Further clues concerning the class of rites to which the "catalogues" pertain may be gathered from a somewhat more elaborated "catalogue" in the Old Tibetan ritual text ITJ 734 that forms part of the same tradition. (45) As with the "catalogues" in PT 1285, these move downstream along the Gtsang-po River from west to east, and gshen predominate as healers. These catalogues contain short antecedent healing scenarios that all follow the same theme. A ruler is captured by a demon (srin), and then a gshen arrives to perform a "ransom" (glud) ritual. To do so, he calls the local deity into an effigy (lha la nyan du btsugs), and exchanges this with the demon, thus liberating the ruler. At the end of each of these antecedents, the officiant relates this success to the present case, and calls all the gods from the antecedent tales into an effigy (nyan), which is cast to all the demons. (46) As with the catalogues in PT 1285, the healers are predominantly referred to as gshen, although there is one, Rgya-bon Leg-dang Rmang-ba (8 recto, 1. 350), who is referred to as a bon. Nowhere do we find the phrase "Go to a bon-po."

Typologically then the "catalogues" in ITJ 734, which appear to be closely related to those in PT 1285, are "ransom" (glud) rites in which an effigy is cast away. The movement downstream, coupled with the casting away of the effigy, clearly corresponds to Gaenszle's typology of oral journeys that travel downstream, removing something unwanted to a distance from the actors. But do the narratives in section I moving in the opposite direction, upstream, constitute a corollary rite involving the recovery of something?

While the narratives may be more detailed than the "catalogues," they are not necessarily any clearer as a result. Only section I of PT 1285 clearly expresses the inverse ordering of territory with respect to the "catalogues." Furthermore, while each individual territory in a "catalogue" contains a healing antecedent, each narrative, though it may name several territories, contains only one healing antecedent. In the case of the "matrimonial narrative" in section I, for instance, the eleven named territories emerge from the list of suitors intervening between the protagonist's home and that of her eventual husband. One could object that the ordering of territory is due simply to the locations of the respective homes of the bride and groom. One might question, therefore, whether a structural comparison with the "catalogues of healing antecedents" is even possible. In my view, however, the ordering of territory in the narrative of section I is too obvious and too structured to be insignificant, and must therefore be read as integral to the ritual's design.

The typology of these narratives always involves catastrophic illness or death stemming from hunting trips, demonic possession, or poisoning. In section I it is as a result of a hunting trip to the north. In sections II and III, it is due to possession by a demon (thugs su gdon bchug). In sections V, VI, and VII, brides poison their fathers as revenge for their unhappy marriages. The remedies, however, are all similar and generic "healings," and there is no evidence, for example, that the maladies are due to a lost or wandering soul that the healer must recover. This being the case, the narratives in PT 1285 do not appear to correspond to Gaenszle's paradigm of upstream ritual journeys to recover something that needs to be brought back.

Another possible interpretation presents itself. Gaenszle notes that while the downstream movement is associated with sending something away, with the plains to the south, and with death, the upstream direction is associated with the north and with life. The directional association of life and death is also taken up by Oppitz in his discussion of the directional orientation of rites practiced by the Northern Magar in western Nepal. Oppitz emphasizes that there are several directional orientations among the Magar--east to west, north to south, clockwise and counterclockwise, circular and semi-circular--and that these cannot be reduced to a single dimension. (47) Each of these orientations, some of which overlap, is aligned with a set of concepts appropriate to the rituals in which it is invoked. Among the Northern Magar the east-to-west orientation roughly corresponds to the north-to-south orientation of the Mewahang Rai: east represents life, and is marked by a mountain pass that constitutes a watershed, and west represents death. (48) This of course also corresponds to the path of the sun, and clockwise ("sunwise" in the northern hemisphere) movement, generally employed in dance, likewise corresponds to life, while counterclockwise movement signifies death. (49)

Among the Northern Magar those oral journeys to recover lost souls proceed upstream to the east, to the watershed of Jaljala Pass, which marks the border between the living and the Beyond. (50) This upstream movement corresponds perfectly with Gaenszle's typology of those oral journeys for the recovery of souls. This west-to-east movement of the lost or wandering soul mimics the movement of the deceased soul, which, escorted by a psycho-pomp sheep, also proceeds in stages to Jaljala Pass, each place along the way being invoked by the shaman in his oral journey. (51) On the face of it, this guided movement of deceased souls to the east, the direction of life, might seem contrary to Gaenszle's typology, as one might expect the deceased to travel to the west, the direction of death. Discussing the funeral rites of the Mewahang Rai, however, Gaenszle himself, drawing on the work of Bloch and Parry, notes that such funerary rites often abound with symbols of fertility and the regeneration of life.
 The forces of decay, mortality and impermanence are contrasted with
 those of productiveness, vitality and eternal stability, and by thus
 creating a transcendental sphere the funerary rituals can 'dramatise
 the victory of order over biology.' (52)

This serves to empower the funerary rite as a victory of eternity and order, and a denial of, or even victory over, death. With this in mind, the eastward movement of deceased Magar souls upstream in the direction of life is not at all counterintuitive. Thus while the malignant spirits of the deceased (asan, masan) are associated with the west, the direction of death, the resting place of the Magar soul in the Beyond is located to the east, the direction of life, upstream past the watershed of Jaljala Pass.

This model, according to which the upstream movement (eastward in the Magar example), associated with life, predominates in funeral rites, might be profitably applied to the early Tibetan liturgies under consideration, if it should be the case that the narrative rites in PT 1285 may deal not with healing, but with death. On the face of it, this is an unlikely proposition, but the similarity of healing and funerary rites is such that this possibility must be considered.

Transposing this conceptual framework onto Tibetan soil, the easy correspondence among the Northern Magar of east, life, and upstream as opposite to west, death, and downstream is complicated by the fact that the Gtsang-po River flows from west to east, that is, against the path of the sun. There appears to be good reason, however, to regard the river as primary and the cardinal directions as secondary in the Old Tibetan liturgies. An Old Tibetan ritual text concerning horses, PT 1060, follows the course of the river from west to east through thirteen kingdoms; the first and last kingdoms mentioned are called "the upper head of the river" (chab gyi ya bgo) and "the lower tail of the river" (chab gyi ma gshug) (PT 1060, 11. 63, 92). This same pair of terms is employed in the same way in the funerary text PT 1134, 11. 39-40. (53) This generic pair of terms could describe the course of any river, but are here applied to the Gtsang-po. Therefore the "upper head of the river" corresponds to the area near the source of the Gtsang-po, which is given sometimes as Zhang-zhung and sometimes as Rtsang. (54) This is an important point, as it demonstrates that the interpretive framework employed here to approach the directional orientation in Old Tibetan ritual liturgies is not imposing foreign categories, but adopting those recognized within the tradition under consideration. It also demonstrates that the river, not the path of the sun, serves as the ordering principle for these rites.




As explained above, the directional orientation of the Magar shaman's chanted journey for recovering lost souls and that for guiding souls to the Beyond are exactly the same. This identity once again underlines the close relationship between healing rites and funeral rites, and emphasizes the difficulty of distinguishing between them. In the Old Tibetan ritual texts under consideration, the difficulty is due not only to the parallel ritual methodologies apparent in these rites, but also to the obscurity of the language. Obviously, the difficulty of distinguishing between healing and funerary rites is exacerbated in the present case by the circumstance that all the available data are necessarily textual; we cannot simply observe the patient to see whether he is living or dead.

Still, it is possible to compare the liturgies of PT 1285 with similar Tibetan ritual texts to assess whether or not the narrative in section 1 (and the other narratives in PT 1285) might be concerned with a funeral. The paradigms of illness in the narrative liturgies of PT 1285--ill-fated marriages, hunting accidents, possession, and poisoning--are not uncommon in Old Tibetan ritual texts, and are also found in the Rnel dri 'dul ba'i thabs as early institutionalized Bon texts such as the Klu 'bum. In some, though not all, cases, they serve as paradigms of death and precede funeral ceremonies. Stein notes two examples from Old Tibetan funerary texts PT 1134 and PT 1136 where hunting accidents precipitate the protagonists' deaths. (55) In another case, in the Old Tibetan ritual text PT 1068, though a hunting trip passes without incident, it may be responsible for the death of the hunter's sister, who remained at home. (56) Typically these hunts entail trips to the north, to a country referred to as Byang-ka Snambrgyad (above, n. 25). Indeed it is to this place that Tshangs-chen Rab-'byor travels for a hunting trip in the first "episode" in the Klu 'bum nag po, a popular recitational text of the later Bon religion. There he befriends a srin demon, and they hunt together until they are attacked by a group of fearsome bdud demons, Rol-po Kya-bdun, who kill Tshangs-chen Rab'byor's srin companion and wound Tshangs-chen Rab-'byor with an arrow. (57) The medical interventions that follow are unsuccessful, and Tshangs-chen Rab-'byor dies, his legs filled with pus (Klu 'bum nag po, 4a4). His death is followed by funerary rites involving such illustrious figures as Gshen-rab Mi-bo and his son, Mu-cho Ldem-drug. At one point there is talk of healing / resurrecting (sos) Tshangs-chen Rab-'byor, but this does not succeed (Klu 'bum nag po, 15a3-4, 16b5-6). This practice of "reviving" the dead is also found in the Old Tibetan funerary text PT 1068, mentioned above in the context of bon and gshen complementarity and partnership. Here a bon performs the gto rite and the gshen performs the dpyad rite for a deceased woman, who is then "cured of her illness and no longer dead." (58) They then perform her funeral rite (shid), so it is evident that the meaning of "no longer dead" must not be taken in a modern medical sense. Stein relates this ritual "resurrection" to later Bon practices found in the Klu 'bum and also to Mosso funeral rites. (59) He further notes the semantic relationship between the words 'to nourish', 'to heal', 'to be alive', and 'life'. (60) In the Mosso context, Stein states that the deceased is said to be "healed" during the course of the funeral rite, and that "this healing seems to guarantee the life or the fertility of his descendants" (translated from the French). (61)

We find similar themes in the Rnel dri 'dul ba'i thabs, which contains several matrimonial narratives that are strikingly similar to those in PT 1285. This document, recently unearthed from a stupa in southern Tibet along with several other non-Buddhist, Buddhist, and medical texts, cannot be dated with certainty. The editor of the volume in which these documents are transliterated and reproduced, Pasang Wangdu, suggests that most of them date to the eleventh century and after, but that some may be earlier and relate to the imperial period. He also rightly notes the close relationship of these texts with the Dunhuang ritual documents under consideration here, though without going into any detail. (62)

The Rnel dri 'dul ba'i thabs concerns the removal of various types of obscurations, most of which are embedded in a matrimonial narrative framework of ritual antecedent tales. In one such antecedent tale, following her marriage to Zangs-' gar Se'u-pa, 'Gos-za Phyam-'darma rides off to visit her parents, but is killed by a mischevious shepherd who hits her in the head with a stone, knocking her from her horse. A dri-bon then performs various rites, including her entombment. (63) These rites seem to be for the benefit of the surviving husband, who is pacified by them, rather than explicitly for the deceased. Similarly, there are other matrimonial narratives in this text where, following a marriage, the bride and her child die, apparently in childbirth, due to the work of a srin demon that enters into the pregnant woman.
 In the ninth month 'Go-ra, the demon of Skyi, went into the mother,
 into the rgu, went into the depths of the mother. It caused water
 sickness to rise up from the depths, and the blood from the depths
 ame out the door. The mother's head was facing upwards. The child's
 head was facing downwards. (64)

The rites that follow, however, are to purify the situation for the surviving husband and father. (65)

Another rite follows an accidental death that occurs as two families are arranging a marriage. And while there seems to be a funeral involved, performed by the illustrious Dur-gshen Rma-da' and by Pha Ya-ngal Gyim-kyong, another intriguing figure, the point of the rite is not the funeral itself, but the aversion of reprisals (sha glan) from the dead man's relatives. (66)

While the narratives of the Rnel dri 'dul ba'i thabs are cognate with those of PT 1285, we should take care when using the latter, and those of the Klu 'bum or the Mu cho'i khrom 'dur, to try to illuminate the former. Although this method might supply some insight, it also risks anachronism and can lead to misunderstandings. Still, we can see from the narratives of the Rnel dri 'dul ba'i thabs and from Dunhuang ritual texts that while the trope of the matrimonial narrative, like the hunting expedition, signals that one of the protagonists will die, it does not mean that death (or funerary rebirth) is the focus of the rite for which the narrative is an antecedent tale. This could just as easily be for the purification of the survivor's relationship with the supernatural world, or with the human world, as in the case of averting reprisals.

Like the matrimonial narrative frame, the hunting expedition to Byang-ka Snam-brgyad, followed closely by calamity, is a formulaic narrative trope within these rites, and, as a model for imbalance within the natural world that requires correction, provides a stereotyped diagnosis by the presiding ritual specialist. By putting death, illness, or another form of imbalance into narrative form, they render it both explicable and treatable. It is by providing the patient (or the deceased) with a story, however stereotyped, that these ritual specialists persuade him of the efficacy of their rites, and thereby ensure them.

In summary, the matrimonial narrative and the hunting narrative are tropes that precede the death of one of the principal characters in the antecedent tales, but this is only a prelude to the main ritual, which can concern healing, purification, funeral rites, and probably other types of rites as well, including even those concerned with expelling or counteracting impurities. In other words, the narrative framing of a ritual antecedent tale seems to tell us very little about the sort of rite it precedes, and one cannot say that matrimonial narratives relate necessarily to funeral rites. Still, this does not address the matter of the ordering of space in an upstream trajectory in section I of PT 1285 and its possible correspondence to models of afferent or "life-oriented" rituals in the Himalayas. Unfortunately, we have not yet turned up another upstream ordering of territory cognate with that of section I of PT 1285, but one should not necessarily assume that the afferent model does not apply or that upstream narratives were not in use. One can also add that while the narrative in section I does not appear to be concerned with the recovery or retrieval of something lost, it is--insofar as it is ostensibly a healing rite--"life-oriented."

The supposition that the "patient" in the above narrative is deceased is not, however, integral to the analysis. The main point is that the inverse ordering of space in the liturgies of PT 1285 represents ritual technologies within a single tradition, and that this phenomenon of ordering space in either an upstream or a downstream movement should be read in the context of such ritual dichotomies as expelling and recovering, efferent and afferent, and death-oriented and life-oriented. The downstream movement is employed in the "catalogues" of ITJ 734 to cast away an effigy in order to make an exchange with demons that plague a patient. By contrast, the upstream movement of the "narrative" liturgy in section I of PT 1285 should be "life-oriented" or concerned with recovering something. Further, these two types of directionally oriented liturgies do not necessarily exclude one another. They may well have been employed as complementary technologies appropriate to successive phases within a larger rite. Indeed this is the case in the rites described in PT 1042, the most famous Old Tibetan funerary text. Here a ransom ritual is performed before the deceased is guided to the land of the dead. (67) In other words, the downstream movement of his effigy occupies the demons and throws them off his scent, giving the deceased a "clean slate" for his journey to the land of the dead.

It is tempting to view upstream and downstream evocations of territory within these liturgies as signifying two sides of the same coin as rites that can heal, and employ a life-oriented, upstream trajectory, or those that can avert misfortune, and proceed in a death-oriented / expelling movement downstream. This would probably be to overly schematize a situation that, as we have seen in an investigation of Old Tibetan ritual texts alongside similar, slightly later sources, is in practice far more complex.

Although the liturgies themselves only represent a one-way movement, in each case a return is implied as part of the ritual exchange. The "return" in the "ransom" rites, for example, is the renewal and health of the patients for whom these liturgies are performed. The movement of these liturgies is therefore circular, and forms a logical parallel with circumambulation. Here, too, we find issues of complementarity and opposition. While it is common to speak of Buddhist clockwise circumambulation in contrast to Hindu counterclockwise circumambulation of Mt. Kailash, or Buddhist clockwise circumambulation in contrast to Bon counterclockwise circumambulation, there are in fact pilgrimages where Bon-po circumambulate clockwise, and those where Buddhists go counterclockwise. (68) Often this is due to ritual prerogatives, but it is just as often bound to issues of identity and self-definition. (69) It would be shortsighted to assume that similar matters of identity politics did not play a role in the composition of Old Tibetan ritual texts, and indeed in the ordered structure of the liturgies analyzed above.


While we have seen that bon and gshen work together in numerous healing and funerary rites, it must also be remembered that they do so only as actors in ritual narratives. We know the names of the bon and gshen in these heroic precedents, and the names of the patients and their tormentors, but nowhere does the text reveal (in a colophon, for example) who employed these liturgies. Although it is highly likely that those who employed such texts as liturgies identified themselves as bon-po, the texts may well have been composed in a time and/or milieu where the distinction between bon and gshen was forgotten or poorly understood. (70)

Just as the identity of the officiant is not altogether certain, the identity of the intended patient presents an enigma. The rituals described among the Mewahang Rai and the Northern Magar, like many throughout the Tibetan cultural area, are rooted in local places. Here the operative word is local, for the land, and those deities that inhabit it, play a crucial role in a person's health and well-being. In many cases, these are the deities of the local mountain that serves as a watershed for the settlements below. These gods of place are so important to Tibetans that they often take them along when migrating to a new area. In this way, sometimes local gods (yul-lha) undergo a transformation by which they become the fictive paternal gods (pho-lha) of an emigrant group. (71) More literalistic coping strategies include traveling with a bag of earth from one's homeland and mixing this into food in order to remain healthy. The point in all these cases is the connection with the local gods of one's immediate village and region. But this begs the question: whose gods are the gods in the liturgies of PT 1285 and similar Old Tibetan ritual texts? These liturgies are not rooted in a single place, but incorporate places and their attendant deities from all along the course of the Gtsang-po. (72) In the liturgies described above, however, one would tend to expect local divinities appropriate to the patient's village. As such, the liturgies seem slightly artificial. Further, this geographic sweep suggests that the bon-po ritual tradition conceived of itself as having a pan-Tibetan jurisdiction.

I can offer a few hypotheses to explain the existence of such seemingly artificial pan-Tibetan liturgies. The first is to assume that they stand as witnesses to a pan-Tibetan ritual tradition, in which the priests have invented an imperial tradition from local traditions by substituting the genius locii of, say, a village for those of an entire cultural area comprising the course of the Gtsang-po and its side valleys. In this way they lay claim to an expanded ritual jurisdiction, and do so by employing a vocabulary understood at the most local levels. This hypothesis rests on the assumption that local rituals, such as ransom rites, preceded and informed their elaboration on a larger scale. The opposite--that local traditions observed today derive from imperial traditions like those described in PT 1285--might also be claimed, though my own inclination is to view this as secondary, and to see in many aspects of Tibetan imperial religion a sort of "local religion plus," where regional and local ritual traditions and territorial cults are expanded and adapted for imperial use, and only then filter back down to a local level.

While the non-local nature of the territories evoked in early Tibetan ritual texts certainly demonstrates a claim to extensive ritual jurisdiction, it is unclear whether or not this transpired as a result of the need to invent imperial traditions with the rise of the Tibetan Empire, or whether this is testimony to a later invention of tradition by an increasingly self-conscious group of bon-po in the tenth century. On the other hand, this is not an "either/or" question, and it may well be that such territorial claims grew out of early imperial traditions and were reimagined and redeployed with the emergence of a more self-conscious invention of bon/ Bon tradition in the tenth and eleventh centuries.

For an alternative hypothesis, we may turn now to the patient. Were these rites truly performed for an "everyman," they would then stand as a testament to a cultural and political sense of pan-Tibetan identity the likes of which has not been observed since. This seems unlikely, since, as desribed above, similar evocations of territory recorded in Tibetan and Himalayan ethnographies are almost always local in nature. Who, then, embodies the course of the Gtsang-po? One answer would be that these and similar rituals, with their incorporation of Tibet's varied telluric deities, were designed for the emperor himself. This is not to say that the liturgies found in PT 1285, ITJ 734, and related texts are themselves royal rituals, but rather they would also seem to demonstrate a process inverse to that of their proposed creation, namely, the filtering down of imperial (or royal) ritual practices to a more common level. Again we are left with a circular movement, though this time it is between royal and popular ritual, a phenomenon that is well documented and which may be observed, for example, in the various ritual uses in the Tibetan cultural area of the dmu cord, the technology by which Tibet's ancient kings travelled between heaven and earth. (73)


The above analysis has demonstrated that the model of opposition between ban and gshen, two types of ritual specialists found in Old Tibetan sources, is based on a false dichotomy. From Old Tibetan ritual texts it is evident that ban and gshen were identical, or nearly identical, ritual specialists involved mostly in healing, divination, and funerary rites. Within their ritual complex, it appears to be all but insignificant whether a given priest is referred to as a ban or a gshen; these terms are interchangeable and refer to specialists who perform identical rites and often work in tandem. The only exception or caveat is that the gshen tend to predominate as officiants in the ritual antecedents for "ransom" (glud) rites.

The ritual liturgies under consideration here, and, specifically, their inverse ordering of space, according to which the "matrimonial narrative" of section I of PT 1285 moves upstream along the Gtsang-po River from east to west, while the laconic "catalogues of healing antecedents" move downstream from west to east, could easily be read to bolster the model of bon I gshen opposition. Such an interpretation, however, is highly unlikely. This directional opposition represents not a conflict between two traditions, but two different types of practices within a single ritual complex. As such, it underlines the diverging functions of these two types of liturgies.

Parallel liturgies from Himalayan ethnography provide a clear interpretive framework for some general conclusions regarding issues of directionality in the Old Tibetan liturgies under consideration. Gaenszle's typology of oral journeys, with two basic types, those that escort something away (downstream) and discard it, and those that search for something (upstream) to bring back, are profitably applied to the "catalogues of healing antecedents" in PT 1285 and the similar "catalogue" in a closely related Old Tibetan ritual text, ITJ 734. The latter--and almost certainly the former as well--is concerned with "ransom" (glud) rites in which an effigy is cast away. The movement is downstream, towards the direction associated with such categories as lower, death, and female. Reciprocally, the upstream movement moves towards such conceptual categories as life, upper, and male. This upstream liturgy corresponds to the "narrative" liturgy in section I of PT 1285, which ostensibly concerns healing. As characteristics of complementary ritual technologies, the upstream and downstream orientations of these liturgies might also be employed in the context of a single, larger rite, such as a funeral. Most importantly, their directional orientation characterizes specific types of liturgies within the bon-po ritual tradition, and should not be read as evidence for opposition, but as an example of complementarity and totality within a single tradition.

It should perhaps be noted that the foregoing discussion does not construct the "bon-po ritual tradition" as a royal or state religion. In previous scholarship, there is a strong tendency to create "royal religion" or "Bon royal religion" as a sort of catchall category for those practices in early Tibet that appear to be non-Buddhist, and this must be resisted. Likewise, "royal religion" or "state religion" is not necessarily a title or prize to be awarded to whichever tradition's priests controlled central rites such as royal funerals and rites of renewal like the ransom rite, for there are strong indications that the ritual practices that buoyed the Tibetan divine kingship and the state were pragmatic in nature, and not limited to a single religious tradition.

The work of Himalayan ethnographers is employed here only to provide a conceptual framework for reading these Old Tibetan ritual texts, and I have not posited any causative link between earlier and later ritual traditions. At the same time, there are many striking similarities between the ritual tradition under consideration and those operative among Himalayan groups, and this has led some to cautiously posit just such a link. (74) Whether in the realm of language or culture, the proposition that the center's antiquity is preserved on the present-day periphery is problematical, and can lead to circular arguments. The suggestion that many of the beliefs and practices found among ethnically Tibetan groups in the Himalayas first originated in Tibet, however, is far less problematic, and the present contribution, whether it adds weight to this proposition or not, at least demonstrates the conceptual proximity of early Tibetan ritual practices to similar practices found in the Himalayas.



I should like to record my deep debt of gratitude to Henk Blezer for a stimulating email discussion over several months and for insightful comments to earlier drafts of this article that significantly reshaped my analysis. Thanks are also due to Charles Ramble, with whom I read passages of PT 1285, and to Cristina Scherrer-Schaub and to Remi Chaix at the Bibliotheque nationale de France for enabling me to consult

the manuscript itself. I am also grateful to Nick Allen and to Guntram Hazod for their comments on an earlier draft, to the anonymous readers for their constructive feedback, and to Stephanie Jamison for her thorough editing. Some of this research was carried out under the auspices of a Fulbright Fellowship in China and Tibet.

(1). The Bon religious sect appears to have taken shape in approximately the early eleventh century through the work of Gshen-chen Klu-dga' and his disciples; see D. Martin. Unearthing Bon Treasures: Life and Contested Legacy of a Tibetan Scripture Revealer, with a General Bibliography of Bon (Leiden: Brill. 2001), 92.

(2). This point is argued persuasively and in detail in Z. Bjerken, "The Mirrorwork of Tibetan Religious Historians: A Comparison of Buddhist and Bon Historiography" (PhD diss., Univ. of Michigan, 2001).

(3). While Tibetan Buddhist monk historiographers have certainly enjoyed more influence in this field. Bon historiography is best regarded not as a heterodox tradition, but rather as a competing orthodoxy. For an excellent overview of Bon historiography and its strategies, see H. Blezer. "The Bon of Bon: Forever Old," in Proceedings of the Tenth Seminar of the International Association for Tibetan Studies, Oxford 2003 (Leiden: Brill, forthcoming).

(4.) The locus classicus for such assumptions in Tibetan studies regarding sacred kingship is Giuseppe Tucci's article, "The Secret Characters of the Kings of Ancient Tibet," East and West 6 (1955): 197-205. Here Tucci speculated along Frazerian lines that were later taken up in detail by Erik Haarh in his book. The Yarlung Dynasty (Copenhagen: Gad. 1969). By contrast with Tucci. Ariane Macdonald. working almost exclusively from Old Tibetan texts, uncovered many of the principles behind Tibetan kingship and early Tibetan religion in her work, "Une lecture des Pelliot tibetain 1286. 1287. 1038, et 1290: Essai sur la formation et I' emploi des mythes politiques dans la religion royale de Sron bean Sgam po." in Etudes libetaines dediees a la memoire de Marcelle Lalou (Paris: Adrien Maisonneuve, 1971), 190-391.

(5.) Stein's and Karmay's interpretations of these instances of continuity differ, however. Karmay tends to see this continuity as evidence for a system of beliefs and practices constituting a Tibetan religion. Stein, on the other hand, sees the Old Tibetan bon and gshen as ritual specialists, but not necessarily as adherents of a religious sect. The debate is outlined succinctly in R. A. Stein. "The Indigenous Religion and the Bon-po in the Dunhuang Manuscripts" (orig. pub. 1988), in The History of Tibet, vol. Led. Alex McKay (London: Routledge Curzon. 2003). 592, 595. Bjerken also outlines Karmay's position and his imperatives in some detail, and summarizes Stein's position in Z. Bjerken, "The Mirrorwork" (above, n. 2). 206-11. 218-19. See most recently H. Blezer. "sTon pa gShen rab: Six Marriages and Many More Funerals," in Tibetan Studies in Honour of Samten Karmay. part 2: Buddhist and Bon po Studies, ed. Francoise Pommaret and Jean-Luc Achard (special issue of Revue d'Etudes Tibetains 15 [2008]): 426-29.

(6.) Here I follow Stein's custom of employing the capitalized term "Bon" to refer to the institutionalized religion, which took shape in approximately the early eleventh century, while referring to the ancient Tibetan priests with the italicized, lower-case "bon" or "bon-po." These conventions are not intended, however, to prefigure the outcome of the debate concerning the continuity between the earlier and later traditions; M. T. Kapstein and B. Dotson, "Preface," in Contributions to the Cultural History of Early Tibet, ed. Kapstein and Dotson (Leiden: Brill, 2007), ix. n. 4.

(7.) M. Lalou, "Fiefs, poisons et guerisseurs," Journal asiatiaue 246 (1958): 157-201. "PT" abbreviates "Pelliottibetain," which refers to the Tibetan texts in the Pelliot collection of the Bibliotheque nationale do France. The number following "PT" indicates a document's press mark. Similarly, the abbreviation "ITJ." or, to be precise, the "IT" in "ITJ." indicates that the text is an Old Tibetan document from the India Office Library, kept in the British Library in London. The final letter (e.g.. "J") and numbers following indicate a document's press mark.

(8.) See R. A. Stein. "Du recit au rituel dans les manuscrits tibetains de Touen-Houang," in Etudes Lalou (above, n. 4). 479-547: idem. Tibetan Civilization (London: Faber, 1972), 236-38: and idem. "Indigenous Religion" (above, n, 5), 584-614.

(9.) R. A. Stein. Tibetan Civilization. 236-38.

(10). The roles of bon and gshen in early Tibetan ritual are expertly summarized in G. Orosz, ""Folk Religion in the Ritual Manuscripts of Ancient Tibet." in Demons and Protectors. Folk Religion in Tibetan and Mongolian Buddhism, ed. Beta Kelenyi (Budapest: Ferenc Hopp Museum of Eastern Asiatic Art, 2003), 19-26.

(11) N. Norbu. Drung, Deu and Bon: Narrations, Symbolic Languages and the Bon Tradition of Ancient Tibet (Dharmasala: Library of Tibetan Works and Archives. 1995, repr. 1997), 63. Aside from being a term for a prognosis, phywa I phya is similar to g-yang in the sense that it can mean 'good fortune' or "luck', though in a rather more visceral or substantive sense than it carries in English. Phywa also indicates a class of sky gods intimately associated with the royal line as the paternal ancestor.- of the first Tibetan sovereign. See S. G. Karmay, "The Appearance of the Little Black-Headed Man" (orig. pub. 1986), in The Arrow and the Spindle: Studies in History, Myths, Rituals and Beliefs in Tibet, ed. Samten Gyaltsen Karmay (Kathmandu: Mandala Book Point, 1998). 250-51.

(12.) PT 1285,11. 167-68; Lalou, "Fiefs," 172, 186.

(13.) S. G. Karmay, "A General Introduction to the History and Doctrines of Bon" (orig. pub. 1975), in The Arrow and the Spindle, 141. See also Stein, "Du recit au rituel." 482; and C. Cantwell and R. Mayer, "Enduring Myths: sMrang, Robs and Ritual in the Dunhuang Texts on Padmasambhava," in Studies Karmay (above, n. 5): 292-93.

(14.) In his study of Mi-pham's treatment of gto rituals, Lin Shen-yu underlines their diversity: they are used for healing disease, expelling evil spirits such as gdon, averting misfortune, and assuring victory and prosperity. Many of the rituals are firmly rooted in the practicalities of village life, with some concerned with "enabling pregnancy or solving problems of the production of yoghurt and beer, even gaining the love of a wife who does not appreciate her husband or changing the sex of an embryo." Emphasizing their significance to Tibetan folk religious practices, and the principles of sympathetic magic behind them. Lin Shen-yu distils the function of gto rituals down to two basic intentions: "to avoid and eliminate disaster, and to bring luck and happiness." Lin Shen-yu, "Tibetan Magic for Daily Life: Mi pham's Texts on Gto-Rituals," Cahiers d'Extreme-Asie 15 (2005): 110-11,119.

(15). Karmay, "General Introduction," 141. Along with divination (mo) and calculation (rtsis), gto and diagnosis (dpyad) form the basis of the "vehicle of the gshen of divination" (phyagshen theg-pa), the first of the nine vehicles in the institutionalized Bon religion: D. L. Snellgrove. The Nine Ways of Bon: Excerpts from gZi brjid (London: Oxford Univ. Press. 1967), 9. The first four vehicles are known as the "vehicles of causation." and include many of the rites performed by bon and gshen priests in Dunhuang ritual documents. These are also known as "black water" (chab nag) practices, and they constitute the first of the "four gates of Bon" (bon sgo bzhi).

(16.) Karmay. "General Introduction," 141. Concerning the physical aspect of their healing practices, there is one instance in the text where the bon healer Smra-bon Zing-ba washes the patient's mouth sores in snow and his hand sores in a lake:smara bon zing ba mchis I lam ba mchis smra I bon zing ba vis I kha strive gangs la bgrus I lag smye misho las bkrus (PT 1285. 11. 41-42; Lalou, "Fiefs." 166. 183).

(17.) PT 1068. 11. 91-92: Stein. "Du recit au rituel." 522.

(18.) Stein, Tibetan Civilization, 238.

(19.) Stein, "Indigenous Religion," 597. The terms skid and mdad are often found together, and may be interchangeable. The latter term is used for all royal funerals recorded in the Old Tibetan Annals; see B. Dotson, The Old Tibetan Annals: An Annotated Translation of Tibet's First History: With an Annotated Cartographical Survey by Cunt ram Hazod (Vienna: Verlag der Osterreichischen Akademie der Wissenschaften. forthcoming), index. The term rmang, on the other hand, is both less frequent (in a funerary context) and less straightforward.

(20.) Stein, "Du recit au rituel," 507 and idem. "Indigenous Religion." 596-97. The line numbers employed by Stein and by Thomas, who edited and translated this text (F. W. Thomas. Ancient Folk Literature from Northeastern Tibet [Berlin: Akademie Verlag, 1957]), differ from those found in the transliteration at Old Tibetan Documents Online (OTDO). The OTDO transliteration takes note of the fact that the scroll is cut into eight sheets, and this numbering is followed here. The translideration is found online at and in the first volume of the OTDO Monograph Series (Yoshiro Imaeda and Tsuguhito Takeuchi et al., eds.. Tibetan Documents from Dunhuang Kept at the Bibliotheque Nationals de France and the British Library. Old Tibetan Documents Online Monograph Series, vol. 1 [Tokyo: Research Institute for Languages and Cultures of Asia and Africa. 2007]. 276-89).

(21.) As a result of the ritual technologies and beliefs that operate here, it is difficult to know whether a rite of "healing" pertains to a patient who is living or one who is deceased. This point will be taken up in detail below.

(22.) Stein, "Indigenous Religion." 602.

(23.) For other examples where healing paradigms arc applied to this Tibetan everyman in a similar Old Tibetan ritual text, ITJ 734, see Stein, "Du recit au rituel," 503, 504. This interpretation of Rma-bu Mching-rgyal is bolstered by the appearance of a similar formula in the Rnel dri 'dul ba'i thabs, a comparable bon-po liturgical text recently unearthed from the Dga'-thang 'Bum-pa-che stupa in Gtam-shul in southern Tibet. Here we find matrimonial narratives strikingly similar to those in PT 1285, and the liturgies typically end with a successful healing followed by the phrase, "Thus in ancient times it has been beneficial and auspicious. Now it will be beneficial and auspicious for whomever you chant it" (gna' de Itar plum de bsod do// da su la gyer ha la phan de bsod do//); Pa-tshab Pa-sangs Dbang-'dus and Glang-ru Nor-bu Tshe-ring. eds., Gtam shut dga' thang 'bum pa che ruts gsar du rnyed pa'i bon gyi gna' dpe bdams bsgrigs (Lhasa: Bod Ijongs bod yig dpe rnying dpe skrun khang, 2007), 42, 45, 146, II. 9-10. 151, 1. 3. In another liturgy of the same text, the formula is altered such that "whomever" (su) is replaced by "this man"'(mi di); ibid. 50, 162,1.8.

(24.) Stein. "Du recit au rituel," 482-83.

(25.) See B. Dotson, "At the Behest of the Mountain: Gods, Clans and Political Topography in Post-Imperial Tibet," in Old Tibetan Studies Dedicated to the Memory of Professor Ronald E. Emmerick (1937-2001): Proceedings of the 10th Seminar of the International Association for Tibetan Studies, Oxford 2003, ed. Cristina Anna Scherrer-Schaub (Leiden: Brill, forthcoming), and M. Lalou, "Catalogues des principautes du Tibet ancien," Journal asiatique 253 (1965): 189-215. The mention of Byang-ka Rnam-brgyad in PT 1286 and other catalogues, generally taken as an alternate name for the Northern Plateau (Byang-thang), is a notable exception. According to Stein, however, it indicates a Turkic country (Stein, "Indigenous Religion," (602).

(26.). M. Gaenszle, "Journey to the Origin: A Root Metaphor in a Mewahang Rai Healing Ritual," in The Anthropology of Nepal: Peoples, Problems, Processes, ed. Michael Allen (Kathmandu: Mandala Book Point, 1994), 258.

(27.) In a recent article Bjerken reviewed in detail the manner in which "shamans" and "shamanism" have been employed in relation to Tibetan religion, particularly Bon. but he did not deal with early Tibetan ritual specialists such as bon and gshen: Z. Bjerken, "Exorcising the Illusion of Bon 'Shamans': A Critical Genealogy of Shamanism in Tibetan Religions," Revue d 'Eludes Tibetaines 6 (2004): 4-59.

(28.) A. Hofer, "Nomen est Numeir. Notes on the Verbal Journey in Some Western Tamang Oral Ritual Texts," in Himalayan Space: Cultural Horizons and Practices, ed. Balthasar Bickel and M. Gaenszle (Zurich: Volkerkundermuseum Zurich, 1999), 226-27.

(29.) Alternatively, lam rings nye dit bos may mean "to take as a relative one from afar." In this sense, the phrase is reminiscent of "to establish marriage relations" (gnyen lam 'dzugs).

(30.) It is worthy of note here that the prospective bride's mother is a lady of Brag (Brag-za), and possibly related to the suitor, Brag-rgyal Thang-po. The marriage could therefore reflect either a system of (delayed) direct exchange, or patrilateral cross cousin marriage. This is usually proscribed in Tibetan societies, and could be a cause of the couple's unhappy marriage.

(31.) PT 1285.II. 97-116: Lalou. "Fiefs." 169-70, 184-85.

(32.) Ibid.. 162,

(33.) Lalou believed this term possibly referred to a particular type of headdress or head ornamentation (ibid., 201). These same formulaic hundred male and female gshen appear in a liturgy in ITJ 734, 2 recto II. 48-49 (Stein, "Du recit au rituel," 506: Imaeda and Takeuchi el al., Tibetan Documents, 277). The text is damaged and fragmen tary, but there they do not seem to be opposed with any other type of ritual specialist, and it appears that their remedies are successful. On these formulaic gshen, who also make an appearance in the Rnel dri did ba'i thabs (Gtam shut, 52, 165. II. 3-4), see Blezer, "sTon pa gShen rab" (above, n. 5), 430-31 and n. 19.

(34.) Stein. "Du recit au rituel." 510-11.

(35.) Lalou ("Fiefs." 159) notes that the subscribed ra btags may be an abbreviation for Smar-yul.

(36.) Ibid., 183. Dags should read gdags: its pairing with sribs indicates that it refers not to the place name of Dags-po / Dags-yul, but, as Stein ("Du recit au rituel," 510-11) pointed out, to the sunny, or male principle (Chinese [TEXT NOT REPRODUCIBLE IN ASCII.] yang), as opposed to the female [TEXT NOT REPRODUCIBLE IN ASCII.] yin. These terms are sometimes also translated according to the gender connotations, as husband (stangs) and wife (dbyal) respectively (R. A. Stein, "Tibetica antiqua I: Les deux vocabularies des traductions indo-tibetaine et sino-tibetaine dans les manuscrits de Touen-Houang," Bulletin de l'Ecole francaise d'Extreme-Orient 72 [1983]: 194). On gdags and sribs in this and similar contexts, see also Blezer, "sTon pa gShen rab," 430, n. 19.

(37.) This, like Thang-ba G.yu-thang, is a formulaic name found in other Old Tibetan ritual texts and in later Bon texts (Stein, "Indigenous Religion," 601).

(38.) This figure, often called Leg-tang Rmang-po, is found in institutionalized Bon sources such as the Legs bshad mdzod as a famous scholar and translator from China. See S. G. Karmay, The Treasury of Good Sayings. A Tibetan History of Bon (1972; rpt. Delhi: Motilal Banarsidass, 2001), 16, 17.

(39.) This location is discussed in detail in G. Hazod's annotated cartographical survey in Dotson, The Old Tibetan Annals.

(40.) Stein. "Du recit au rituel," 494, n. 44.

(42.) M. Gaenszle, Ancestral Voices: Oral Ritual Texts and Their Social Contexts among the Mewahang Rai of Eastern Nepal (Munster: LIT Verlag, 2002), 127. See also idem, "Traveling Up--Traveling Down: The Vertical Dimension in Mewahang Rai Ritual Journeys," in Himalayan Space (above, n. 28), 137. Gaenszle is in general agreement with Hofer, who, working in Nepal among the Western Tamang, refers to these two types of journeys as "efferent" and "afferent" respectively (Hofer, "Nomen est Numen," 212).

(43.) M. Gaenszle, 'Travelling Up--Travelling Down," 147-49; idem, Ancestral Voices. 128-29.

(44.) Hofer, "Nomen est Numen," 226-27.

(45.) Karmay, in his analysis of glud rites, also notes the close relationship between these two Old Tibetan ritual texts. Further, he compares the liturgies with a clear parallel from the institutionalized Bon religion called a gludrabs (S. G. Karmay, "The Man and the Ox: A Ritual for Offering the glud" (orig. pub. 1991), in The Arrow and the Spindle, 342-46.

(46.) For a translation of the relevant section of this "catalogue," see S. G. Karmay, "Concepts of Territorial Organization and Their Transformation into Buddhist Sacred Sites," in The Arrow and the Spindle, vol. 2, ed. Samten Gyaltsen Karmay (Kathmandu: Mandala Book Point, 2005), 36-46. According to Stein, "Du recit au rituel," 503, n. 70, the nyan / nyen is closely related to the ransom (glud) rite, and may indicate a ritual object, perhaps an effigy of the deceased. Orosz ("Folk Religion" [above, n. 10], 25) tentatively glosses nyan with brnyan 'form', and translates it with 'duplicate' or 'likeness'. The present analysis confirms that the nyan is the physical aspect of the ransom rite, the effigy itself. As such, it appears to be similar to the effigy or 'figurine' (ngar / ngar-glud) employed in the ransom rites of the institutionalized Bon; see Karmay, "The Man and the Ox," 340-41. Karmay ("Concepts of Territorial Organization," 46) seems to have missed this point in his own discussion of the phrase nyan du btsugs in this passage, which he renders as 'to invoke'.

(47.) M. Oppitz, "Cardinal Directions in Magar Mythology," in Himalayan Space, 189.

(48.) Ibid., 186-89.

(49.) Ibid., 185-87.

(50.) Ibid., 188-89.

(51.) Ibid., 189; idem, "Death and Kin amongst the Northern Magar," Kailash 9.4 (1982): 388.

(52.) M. Gaenszle, "The Making of Good Ancestors. Separation, Transformation and Exchange in Mewahang Rai Funerary Rites," in Ways of Dying: Death and Its Meanings in South Asia, ed. Elisabeth Schombucher and C. P. Zoller (New Delhi: Manohar, 1999), 50.

(53.) Imaeda and Takeuchi et al., Tibetan Documents, 146, Chab gyi ya-bgo / Chab kyi ya-mgo is found on its own in the funerary text PT 1136. ll. 30. 46 (Tibetan Documents, 156, 157), and also in PT 1289, a funerary text concerned with animal sacrifice (PT 1289, 2 recto, 1. 12; Tibetan Documents, 246). Stein ("Du recit au rituel," 492, n. 37) also notes here the up/down opposition and the attendant male/female gender opposition at work in the pair Chab gyi ya-mgo and Chab gyi ma gshug.

(54.) See also Blezer, "sTon pa gShen rab," 451, n. 74.

(55.) Stein, "Du recit au rituel," 493, 501.

(56.) Ibid., 522.

(57.) Klu 'bum nag po: Gtsang ma klu 'bum chen mo; A Reproduction of a Manuscript Copy Based upon the Taranatha Tradition of the Famed Bonpo Recitational Classic, vol. IV, Klu 'bum nag po. Rtag-brtan Phun-tshogs gling-based edition (Dolanji: Tibetan Bonpo Monastic Centre, 1977), 2a4-2b6. This entire episode is paraphrased J. V. Bellezza, Zhang Zhung: Foundations of Civilization in Tibet: A Historical and Ethnoarchaeological Study of the Monuments, Rock Art, Texts, and Oral Tradition of the Ancient Tibetan Upland (Vienna: Verlag der Osterreichischen Akademie der Wissenschaften, 2008), 482-92. For further information on the Rol-po Skya-bdun, see J. V. Bellezza, Spirit-Mediums, Sacred Mountains and Related Bon Textual Traditions in Upper Tibet: Calling Down the Gods (Leiden: Brill, 2005), 287-302.

(58.) na sde sos ba ni mchis / shisde ba ni ma mchis gyis (PT 1068, 1. 92: Stein, "Du recit au rituel," 522).

(59.) Ibid., 484. See also a similar episode in the Mu cho'i khrom 'dur, a later cycle of Bon-po funeral texts, where a corpse is restored, and then can walk and speak (Bellezza, Zhang Zhung, 399).

(60.) Stein later developed this point in an article, "Un ensemble semantique tibetain: Creer et procreer, etre et devenir, vivre, nourrir et guerir," Bulletin of the School of Oriental and African Studies 36.2 (1973): 412-23.

(61.) Stein, "Du recit au rituel," 484.

(62.) Gtam shul (above, n. 23),I. Blezer ("sTon pa gShen rab," 430-31, n. 19) for his part suspects that a date "no earlier than the eleventh century" may be more prudent.

(63.) Gtam shul, 42, 146, 1. 9. This narrative comprises 40-42 and from 143, 1. 1 to 146, 1. 10.

(64.) zla ba ngo rgu na' / skyi srin 'go ra yis // mar zhugs rgu ru zhugs / ma'i khong du zhugs / chu nad khong nas bslang / khong thrag sgo ru phyung / ma mgo gyen du bstan / u 'go thur du bstan /; Gtam shul, 45, 151, 1. 11 to 152, 1. 3. This would appear to ritualize and place within a comprehensive story the circumstances of complications with pregnancy or death in childbirth--all too common occurrences in the Tibetan cultural area. The same theme is taken up in the marriage of the lord of Rtsang at pp. 47-48, the marriage of the lord of Gnubs at p. 48, and those of the lord of Skyi at pp. 48-49, of the lord of Yar-lungs at pp. 49-50, of the lord of Dbye-mo at p. 50, and of the lord of Mchims at pp. 50-51. One notes here also a downstream movement from Rtsang to Mchims in the setting of these related narratives, which could be interpreted as expelling, in that all of the rites in this text seek to 'subdue' ('dul) or 'remove' (phyung) various types of impurities (rnel, dri [ma]).

(65.) This is evident at p. 50, 162, 1. 8, where the rites benefit Dbye-rje Khar-pa. Interestingly, in the previous liturgy, set in Yar-lungs, this is done for the benefit of a god, Yar-la Sham-po (Gtam shul, 50, 161, 1. 6).

(66.) Gtam shul, 51-52, 164, 1. 1 to 167, 1. 3. The Ya-ngal are often placed alongside Mtshe-mi and Gco as the personal priests of the first Tibetan king, and they are famous as one of the key priestly clans in Lubra, in northern Nepal; see Snellgrove, The Nine Ways of Bon, 4-5 (above, n. 15).

(67.) mdad chen po 'i tshe yang // cho ga gzhan ni stor ma / chen po dang 'dra 'o // thugs glud lu gu dkar po gcig / (PT 1042, ll. 97-98); Lalou, "Rituel Bon-po des funerailles royales," Journal asiatique 240 (1952): 346, 356. The glud rite is also found within a larger funeral ritual in the Mu cho'i khrom 'dur; Bellezza, Zhang Zhung, 398.

(68.) T. Huber, The Cult of Pure Crystal Mountain: Popular Pilgrimage and Visionary Landscape in Southeast Tibet (Oxford: Oxford Univ. Press, 1999). 13. It is worth noting here the existence of Indian precedents for employing clockwise or counterclockwise circumambulation according to the type of ritual involved. The clockwise circumambulation is known as pradaksina, as opposed to the counterclockwise prasavya movement. This latter practice is associated with rites for the ancestors, or pitr, see C. Sen, A Dictionary of Vedic Rituals Based on the Srauta and Grhya Sutras (Delhi: Concept Publishing Company, 1978), 85-86. 151.

(69.) Sec K. Cech, "The Social and Religious Identity of the Tibetan Bonpos," in Anthropology in Tibet and the Himalaya, ed. Charles Ramble and M. Brauen (Zurich: Volkerkundesmuseum der Universitat Zurich, 1993), 42-44. See also Huber, Cult, 65, 253, n. 57. In the case of Tsa-ri. it should be noted that women circumambulate counterclockwise as a copying strategy, since they are forbidden to follow the clockwise path (ibid., 121 ff.).

(70.) Blezer, "sTon pa gShen rab," 422.

(71.) This illustrates the mutability of descent and residence in Tibet, where households and lineages often go hand in hand. See, for example, M. Kind. "Abducting the Divine Bride. Reflections on Territory and Identity among the Bonpo Community in Phoksumdo, Dolpo," in Territory and Identity in Tibet and the Himalayas, ed. Katia Buffetrille and H. Diemberger (Leiden: Brill, 2000), 279-80.

(72.) Elsewhere, in the context of divination texts, I have referred to such a concatenation of gods as constituting and "imperial pantheon," brought together by conquest and administrative centralization; B. Dotson, "Divination and Law in the Tibetan Empire: The Role of Dice in the Legislation of Loans, Interest, Marital Law and Troop Conscription," in Contributions to the Cultural History of Early Tibet, 59.

(73.) On the derivation of royal rituals from existing ritual vocabularies us an integral component of their effectiveness, see M. Bloch, "The Ritual of the Royal Bath in Madagascar: The Dissolution of Death. Birth and Fertility into Authority," in Rituals of Royalty: Power and Ceremonial in Traditional Societies, ed. David Cannadine and S. Price (Cambridge: Cambridge Univ. Press, 1987) 271-97. On the dmu cord or dmu ladder, which links the Tibetan kings to their celestial origins, but which also links man and bride in a Tibetan wedding ceremony, and is used in local territorial cults to bind men to their headman and thence to their local deity, see Karmay. "General Introduction" (above, n. 13), 150-51; idem, "A Comparative Study of the yul lha Cult in Two Bonpo Areas and Its Cosmological Aspects" (orig. pub. 2000), in The Arrow and the Spindle, vol. 2, ed. Samten Gyaltsen Karmay (Kathmandu: Mandala Book Point, 2005), 56-57.

(74.) See, for example, Hofer, "Nomen est Numen" 230-31.
COPYRIGHT 2008 American Oriental Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2008 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Dotson, Brandon
Publication:The Journal of the American Oriental Society
Article Type:Report
Geographic Code:9CHIN
Date:Jan 1, 2008
Previous Article:The first Islamic legal theory: Ibn al-Muqaffa' on interpretation, authority, and the structure of the law.
Next Article:Southern girls or Tibetan knights: a Liang (502-557) court performance.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |