The Dead and Their Stories: Preliminary Remarks on the Place of Narrative in Tibetan Religion

An explicit focus on an “intermediate period” calls one’s attention to continuities between the “earlier” and the “later” periods.1 Paradoxically, it also forces one to approach individual problems from more of a longue durée perspective in order to define where, exactly, the differences between “early,” “intermediate,” and “later” lie. This is one reason that the conceit of the bar dar recommends itself, alongside other productive and justifiable periodizations of Tibet’s history, as a heuristic device for exploring historical issues. This article approaches one such issue – that

of the place of narrative in Tibetan religion. As Cathy Cantwell and Rob Mayer (2008) have pointed out in their discussion of very similar issues in connection with Dunhuang materials concerning Padmasambhava, the narrative element that appears to be so crucial in early Tibetan non-Buddhist ritual texts is far less important – though by no means absent – in the Indian Buddhist rituals that Tibet assimilated. The general picture that emerges is one of a process by which the narrative element decreased in importance as Buddhism came to dominate Tibetan ritual practices. Cantwell and Mayer have described one aspect of this process that was particularly relevant to the cultural transformation and social upheaval of the “intermediate

period,” namely, the indigenization of Buddhism by, among other things, infusing Buddhist ritual with narratives resembling those that often accompany non-Buddhist Tibetan rituals. These latter narratives often take the form of tales, which are then related to a concurrent or subsequent ritual performance. It is for this reason that I refer to them as “ritual antecedent tales” (Dotson 2008). Ritual antecedent tales enjoyed a central role in a pivotal arena for ritual competition in early Tibet: funeral rites. Through funeral rites and the 1 I take this opportunity to express my gratitude to the conference organizers for arranging such a successful conference and for kindly hosting the participants. I gratefully acknowledge the support of the Alexander von Humboldt Foundation and the German Federal Ministry of Education and Research, who support my current research project, “Kingship and Religion in Tibet.” I am grateful to Henk Blezer, Lewis Doney, George FitzHerbert, and Rob Mayer for their helpful comments to drafts of this chapter.

polemics surrounding them we see that the lines of debate are drawn not only on the key issue of animal sacrifice, but also on the role of narrative. To fully understand the narrative forms at work in this discussion, this article begins by introducing two specific ritual narrative forms: the ritual antecedent tale and the catalogue of ritual antecedent tales. The topoi and formulae of these narrative forms are detachable, and are shared by texts belonging to separate, performative settings, such as the “Account of the Minor Kings” which forms a part of Tibet’s chronicle-epic, the Old Tibetan Chronicle. The article then examines Buddhist funeral rites, and their different modes of interaction with traditional Tibetan funeral rites.

The transformation of one traditional rite by the Buddhist “Substitution Text,” for example, contrasts with the strategy of the “Story of the Cycle of Birth and Death” (Skye shi’i lo rgyus), an apocryphal Tibetan Buddhist text that draws on the Gaṇḍavyūha sūtra to provide a charter myth for the performance of Buddhist funeral rites. Both of these texts come from Cave 17 at Dunhuang, and are generally (though not authoritatively) dated to the ninth century. The third and final part of the article analyzes the rhetorical strategy of a later text, the “Account of the Food Provisioning [for the Dead]” (Zas gtad kyi lo rgyus). This is also a charter myth for the performance of Buddhist funerals, but one that appears to intentionally

disfigure and subvert traditional non-Buddhist ritual narrative forms. This latter polemic appears to be in dialogue with, or a rejoinder to anti-Buddhist polemics, such as those found in the Bsgrags pa gling grags, which accuse the Buddhists of plagiarizing Bon po works and stealing proprietary ritual narrative technology. The examination of the changing modes of interaction between Buddhists and their competitors through an analysis of the place of narrative in funeral rites reveals a changed context that maps an ideological shift taking place from the fall of the empire to the self-conscious avowal of new Tibetan identities such as Rnying ma and Bon in the centuries that followed. In the end, the Buddhist assimilation and subversion of these ritual narratives also underpinned the composition of new Buddhist master narratives that would refigure Tibet’s history from this time onward.

I. Tibetan Ritual Antecedent Tales and Non-Buddhist Funerary Texts Among the Dunhuang manuscripts one finds over a dozen interrelated funerary texts. Together with texts concerning healing, ransom rites, and divination, and alongside a vast corpus of Buddhist materials, these are some of our most important sources for the study of early Tibetan religion. The Dunhuang ritual texts are primarily focused on providing a mythical antecedent for a given ritual or for some feature of a ritual. The narratives stage such antecedents in a heroic, mythical setting that can involve kings,

queens, nefarious beings (e.g., srin, bdud), gods, and ritual specialists. They explain the origins of a rite, its necessity, and the proper methods for its performance. These tales, similarly to the invocation of a lineage tree in the context of a Buddhist teaching, serve to empower the ritual and the officiant by associating him or her with the previous successes of illustrious predecessors. Examining the corpus of our earliest extant ritual antecedent tales preserved in Cave 17 in Dunhuang, we can demonstrate the structure of the narratives and can begin to understand how they

were used. Most of these tales, almost all of which concern funeral rites, follow a basic plot of crisis and resolution. They also incorporate certain recurring topoi, motifs, and formulae. These elements can be expanded or contracted in the delivery, and they can also circulate beyond the confines of the ritual genre. As such, they constitute some of the basic units of narrative in Tibet, and we encounter the same formulae and topoi in later Bon and Buddhist ritual texts, in historical writings, and in folk tales. A ritual antecedent tale can be called smrang, rabs, or lo rgyus. The

former term is generally a genre designation, and unlike the latter two terms, it seems not to be found in the titles of texts or tales. Lo rgyus is the common word for “history,” but it is also used for charter myths and ritual antecedent tales.2 The same is true of rabs, which, besides designating ritual antecedent tales, can refer to non-ritual “accounts” of events.3 Vexing from the perspective of genre designations that aspire to precise distinctions, rabs refers to narratives that are mythological, and which often place themselves in a time of origins. This is true, for example, of the “Tale of the Separation of Horse and Kiang” (ITJ 731), an antecedent tale for the use of the horse as a psychopomp guide to the land

2 As noted by Helmut Eimer, the correct etymology of lo rgyus is a “continuum of tidings” or “series of reports”; a “succession of years” is a secondary meaning (Eimer 1979: 101–103). 3 Describing the polyvalence of the term rabs in an article that remains the locus classicus for the study of early Tibetan ritual narratives, Rolf Stein demonstrated that the term has two general meanings: 1) “succession” or “series”; and 2) “type” or “race” (Stein 1971: 537–38). The etymology of rabs as a “tale” or “account” might derive from the former, in which case a “succession [of words]” would be very similar to the etymology of lo rgyus. Stein also relates rabs to popular Indian and Chinese narrative genres by pointing out that in the Mahāvyutpatti, sngon gyi rabs translates the Sanskrit purāṇam, meaning “tales of past [events],” and that rabs seems to be the equivalent of the Chinese bian 變, as in the bianwen or “transformation text” genre of Tang popular literature. Stein makes this latter point based not on direct translation between Chinese and Tibetan, but on respective Tibetan and Chinese translations of a passage from the Mūlasarvastivadin Kṣudrakavastu.

of the dead.4 But rabs is also used in the titles of narratives like the “Account of Those who Served as Chief Councilors” (blon che bgyis pa ’I rabs; PT 1287, l. 63), which forms the second chapter of the Old Tibetan Chronicle, and constitutes a series of short vignettes of the careers of Tibet’s chief councilors that has been and continues to be used as a historical source.5 There are several characteristics that set ritual antecedent tales apart from other narrative forms. One is their use of a specific, formalized register of ritual language. It can be distinguished

from normal prose by its specialized lexicon and by its placing of synonymous terms or phrases in apposition. This, and other forms of parallelism, is a key feature of ritual antecedent tales, but one that we shall not explore in detail here. Rather, we shall briefly introduce the broad outlines of ritual antecedent tales for funerals, and in particular the “Tale of Lhe’u Yang ka rje” (Lhe’u yang ka rje’i rabs), which will serve as a point of departure for examining Buddhist funerary texts. The tale begins with a formulaic statement of temporal setting that sets its action in ancient times. The spatial setting of the tale is introduced alongside the dramatis personae:

Up above the sky, up atop the heavens, a certain ’Gon tsun phyva had no wife or bride. He searched for a wife and bride, and took a certain Tang nga Bdud mo tang, daughter of Bdud rje (“Lord of bdud demons”) Btsan tog skyold, as his wife and bride. The offspring they begat and conceived was born as a son, was birthed as a scion. They gave the boy a name, they gave the scion an appellation, Lhe’u Yang ka rje.

gnam gyi ya’ bla na / dgung gI ya stengs na’ / ’gon tsun phyva zhig khab dang / dbyal ma mcis kab dang dbyal btsald pa / bdud rje btsan tog skyold gyi sras mo tang nga bdud mo tang zhIg kab dang / dbyal du blangs te bshos tang /

4 See the summary and analysis in Stein 1971: 485–91 and that in Bellezza 2008: 529–37. The Tibetan Dunhuang manuscripts are cited according to their abbreviated shelfmarks. “PT” abbreviates Pelliot tibétain, and these documents are held at the Bibliothèque nationale de France in Paris. “IT” abbreviates “India Office Library, Tibetan collection,” and these documents are housed in the British Library in London. The combination of letter and number following “IOL Tib” constitutes the full shelfmark, e.g. “ITJ 734.” Where possible, I have included the title given in the text. In some cases, however, a single scroll will contain more than one named tale, and in other cases the tale will have no name. 5 Note that the “blon rabs” – actually a blon che bgyis pa ’I rabs – in the Old Tibetan Chronicle is not a “genealogy of councilors,” a fact that also suggests that while a rgyal rabs may be a “royal genealogy” – as it has been traditionally translated and generically designated – it is also a “story about kings”; see van der Kuijp 1996: 43.

nams / gyi sras / bu po ru / byung / sras dral du bltam bu’i mying btags / sras gI mtsan btags pa / lhe’u yang ka rje; PT 1134, ll. 68–71; cf. Stein 1971: 492.

Typically, the tales begin with a mother and a father, who conceive one or more children. This familial setting is an important characteristic of these ritual antecedent tales. The ubiquity of this introductory topos also lends itself to formalization, and to the use of noun pairs and appositional synonym phrases, as we see above with the noun pairs khab and dbyal, and the paired verbs in the phrase “the offspring they begat and conceived” (bshos dang nams gyi sras).6 Often the manner in which the characters are introduced is enough to set off an inexorable series of events

that leads to the illness or death of one of the principal characters. The most common and most adaptable trope for death in these tales is the “matrimonial narrative trope” (Dotson 2008: 47; Dotson 2013c). This, like the other tropes for death in ritual antecedent tales, is a trope in the literal sense of the word: it turns away from the denotative sense of the unfolding of events, and points towards death. This is due not necessarily to any belief in an inherent link between marriage and danger (though such a link cannot be dismissed), but rather to the fact that these tales must

“kill off” a protagonist so that his or her funeral may be performed. The “Tale of Lhe’u Yang ka rje” begins with a marriage, and then produces the death of the title character through the introduction of another trope for death, the “competition trope,” when Lhe’u Yang ka rje challenges his maternal relatives, the bdud demons, to a contest and to a race that, inevitably, result in his death (PT 1134, ll. 71–73; Stein 1971: 492–93). The death itself can either be simply stated or attended by a number of motifs or formulae. In the case of Lhe’u Yang ka rje we find the following common formula for a protagonist’s death: “the bdud demons came from the sky, the sri demons rose up from the earth, and they killed him and he was no more” (gnam las / ni bdud bcad sa las ni sri langste grongs gyis / myed; PT 1134, l. 74; Stein 1971: 493). There are other, more detailed images, including the poetic evocation of the decomposing corpse (ITJ 731r68; Stein 1971: 487), the motif of

6 Here the meaning of bshos, “to copulate or beget,” gives us the meaning of nams, which stands in relation to it as an appositional synonym. The latter term is otherwise obscure. These forms of expression obviously pose some problems for the translator. At the same time, they are a great boon for the lexicographer: if one knows the meaning of one term in such an appositional construction, this reveals the general meaning of the parallel term, should it be unknown.

the crumbled turquoise, and the motif of sullied beauty and lost virility (ITJ 731r122–24; Stein 1971: 490).7 Lhe’u Yang ka rje’s death is followed by motifs of sorrow and denial, and then by failed attempts to resurrect the deceased (PT 1134, ll. 74–89; Stein 1971: 493). While these topoi form a sizable part of this particular tale, they are absent in many others, and are essentially collapsible, optional narrative elements. Other narrative elements have a similar effect of prolonging the inevitable denouement. The favored priest, for example, may simply be enlisted to perform the funeral, or he may come only after the failure of others to do so (the “false priests motif”; cf. PT 1285r40; Lalou 1958: 162; Dotson 2008: 48). In

some cases, he must be approached via a go-between, and one go-between might fail before another succeeds in enlisting the favored priest (PT 1136, ll. 52–55). There are further narrative topoi that can postpone the resolution. Usually the resolution is the funeral itself, but prior to the funeral it may be necessary to perform preliminary rites. One may also need to secure psychopomp horses, yak-ox hybrids (mdzo), and sheep. For each of these there are expandable antecedent tales that describe their origins and prescribe their use in a funerary context. In the “Tale of Lhe’u Yang ka rje,” the title character’s father, Mgon tshun phyva, enlists the favored priest Bon gshin gshen drag, and plans to use two horses for Lhe’u Yang ka rje’s funeral. The horses escape, however, and have to be chased across a series of springs before a certain Dmu rje Rgya bdun catches them and returns them to Mgon tshun phyva (PT 1134, ll. 98–113; Stein 1971: 494–95). A series of standard formulae describe how one treats such horses, e.g.,

they are hobbled and placed in an enclosure where one feeds them green shoots and sweet barley flour and pours for them molassessweetened water. Then they are fitted with “bird horns.”8 In this case, it is expanded, and we are told that the horses accept Mgon tshun phyva’s request that they act as his son’s pyschopomp guides by “making the great reply with their mouths, and drinking the great poison with their lips” (lan chen / kha'is / blangs / dug chen mchus 'tungs; PT 1134, l. 116; Stein 1971: 495). The funeral itself is usually described in stereotyped phrases that concern the construction of the tomb and other elements of the ceremony, e.g., “they built a rgyal with eight cords on the bas, and erected a four-cornered se in the valley” (rgyal thag brgyad bas la bchas se gru bzhi lung du brtsi[g]ste; PT 1136, l. 59). Here too one finds formulae describing the psychopomp horses’ execution of 7 On the turquoise, see See Karmay 1998: 320. On the “hunting trope” as a further trope for death, see Dotson 2013b. I discuss these motifs, including the tropes for death, in The Victory Banquet: The Old Tibetan Chronicle and the Rise of Tibetan Historical Narrative (Habilitationsschrift, LMU München, 2013). 8 For a detailed discussion, see Bellezza 2008: 452–59; 517–

their task, namely, “using their courage to cross the passes and using their width to carry [the deceased] across the fords” (chab gang la ru bgyis / la yang ba rab du sbogs; ITJ 731r101–02). The “Tale of Lhe’u Yang ka’ rje” also includes instructions to the psychopomp sheep (skyibs lug mar ba), who guides the horses and the deceased to the land of the dead. This is considerably elaborated, and the tale describes their journey over fords and passes to the realm of the gods atop the ninth stage of heaven. There, it is emphasized, the animals and the deceased man are established equally as

ancestors, and enjoy the same status (PT 1134, ll. 124–33; Stein 1971: 496). The tale closes with a formula that relates it to the present as an antecedent for success. In its simplest form, this statement of relevance declares, “what was beneficial in ancient times shall be beneficial now; what was successful in ancient times shall be successful now” (gna’ phan da yang phan gna’ bsod da yang bsod; PT 1136, l. 60; Stein 1971: 504; Dotson 2008: 45). This ends the tale and enacts a passage out of narrative time and into the present situation to which this antecedent applies. It is here that we come up against the inherent limitations of our textual material. The ritual narratives appear to support a mimetic ritual to be performed during or after their recitation. Stein (1971: 482) came to a similar conclusion when he remarked of ritual antecedent tales that they give the origin of rites and ensure their efficacy, but also that they are mythical precedents that one imitates during the performance of the rite. The

formulaic statement of relevance, as well as the “performance notes” in these manuscripts (cf. below), imply that these narratives serve as mimetic templates for ritual practice, in which present-day priests and patients are invited to identify with their heroic, mythical antecedents. The nature and extent of this identification is not a point that one can press too far, given that the exact relationship of the ritual antecedent tale to the ritual act, e.g., preliminary or simultaneous; mimetic or divergent, is not something that may be established by textual artefacts alone. Nonetheless, the tales themselves make it clear that they are central to ritual practice, and suggest that they underpin mimetic ritual acts. This

brief description of ritual antecedent tales, spotlighting in particular the funerary narrative of the “Tale of Lhe’u Yang ka rje,” yields a fairly detailed compositional profile. Those versed in this compositional profile could skillfully manipulate the basic outline of a tale and “shuffle” relevant tropes, motifs, and formulae according to the occasion, the duration of the rite, etc., expanding some topoi and collapsing others. In addition, these topoi and story-types were detachable and portable. That is, they circulated not only within the ritual narrative tradition, but beyond it. We now turn to one such example, where a sub-genre of the ritual antecedent tale overlaps with the Old Tibetan Chronicle’s “Account of the Minor Kings.”

Redactional Outlines and “Catalogues” of Principalities From Marcelle Lalou’s 1965 article, “Catalogues de principautés du Tibet ancien” onward, scholars have referred to sequential arrangements of ritual antecedent tales, and similar non-narrative lists containing the names of several kingdoms, as “catalogues.” This term is descriptively accurate in those cases where we find only a list that includes, for example, the names of kingdoms, kings, and ritual specialists. Indeed, Lalou collated the data from such ostensibly non-narrative forms in order to present them side-by-side in a manner that she hoped would support future historical research.9 Rolf Stein treated some of this material in his magisterial study of early Tibetan ritual literature, and clarified the significance of the “catalogue” form:

Ne pouvant ou ne voulant pas toujours réciter tous les récits in extenso, ils ont été amenés à en abréger un certain nombre. Ce faisant ils sont arrivés à les réduire à un schéma invariable qui leur est commun. Leur résumés se bornent alors à étoffer ou à illustrer ce schéma d'un bref rappel des précédents ou des légendes d'origine. Cela finit par ressembler à une liste. (Stein 1971: 482.)

One can add to that while the term “catalogue” can describe the appearance of ritual antecedent tales when they are arranged in shorthand, this may not be the most apt term for these abbreviated tales, since they are in fact narratives, reduced by contraction to “redactional outlines.” This is the term John Charlot employs in his analysis of the combinations of stories and motifs in traditional Hawai’ian narratives. The redactional outline is the bare minimum from which the ritual specialist or bard can fill in a tale. In one of Charlot’s examples, a list of a hero’s opponents, for instance, doubles as a list of narratives, since each named opponent comes with a tale that the storyteller can choose to tell, or not, in the delivery (Charlot 1977: 491). This closely resembles the way in which “catalogues” of ritual antecedent tales can, though need not always be, expanded in their delivery. The redactional outline is the starkest form of the antecedent tale, which reduces it to nothing more than setting, names of dramatis personae, a briefly stated crisis and resolution, and a formula relating the antecedent to the present. Besides being an aide memoire, the redactional outline is also

9 It was Lalou’s stated wish that the scholars to whom she dedicated the article, Ariane Macdonald and Géza Uray, should use the materials to aid historical research (Lalou 1965: 189). On the problems posed by employing these documents as sources for Tibetan history, and particularly for the association of particular clans with particular territories, see Dotson 2012. See also Hazod 2009.

instrumentalized to create a sub-genre of ritual narratives, the “catalogue of ritual antecedent tales.” In its most straightforward typology, such a catalogue operates in the same way as a single antecedent tale, but simply adds one tale after another, and in the end applies the force of all of them to the present ritual with the same formulaic statement of relevance that is used in a single tale. The sequential arrangement of these tales allows them to be taken as a whole, rather than simply as a string of tales. This wholeness is expressed by the spatial organization of the settings of each tale, where the succession from one setting to the next typically follows the course of the Gtsang po (Brahmāputra) River. The coherence of

the catalogue’s tales as a single unit is often clear as well from the stated number of settings, which is usually nine or thirteen, both of which signify totality (Dotson 2012: 169–70). The movement in stages from one setting to another, upstream or, as is usually the case, downstream along the Gtsang po, also gives the impression of motion. In addition, the direction of travel can express intentionality, with upstream movements typically used for rites that recover something (e.g., a soul), and downstream movements used for expelling rituals (Gaenszle 2002: 127; Dotson 2008: 56–64). The inclusion of all of these settings within a single liturgy that reduces the tales to a spare redactional outline also conveys a sense of totality

or exhaustiveness, and of the catalogue “covering” the entire Tibetan ritualspatial universe.10 It is clear from the liturgies themselves how the tales expand and contract in the context of these “catalogues.” At the start of one of the catalogues in PT 1285, for example, the outlines of the narrative in the first two settings, Rtsang ro and Skyi ro, is given in full. We have details of the symptoms that ail the gods Rtsang la Bye’u and Skyi bla Bya rmang, and we are told of how for three days and three nights the respective healers Rtsang shen Snya lngag and Skyi gshen Rgyan ngar successfully healed them.11 Before the text goes on to the remaining eight narrative settings, there is a note to the liturgist, set off by eye-catching punctuation, which states, “one should proceed like this until [the tale set in the land of] Rkong yul bre snar” (/:/:/ rkong yul bre snar gyi bar du

10 In some cases, the settings may also pertain to diagnosis: if a patient’s affliction is not cured by the tale (and the accompanying rite) set in one place, then the officiant moves on to another tale (and presumably another rite) set in another place; PT 1285r95–97. 11 [±1] rtsang ro dbyes kar na’/ rtsang la bye’u zhig/ dbu snyung spyan du snyung/ gnya’ na ltag du tsa na’/ / rtsang shen snya lngag gI [±2] rkang (?) dmar la/ nub sum/ mngad du mngad/ / nang sum rbu ru rbu/ rtsang brang rub du bor/ / dbu snyung/ spyan snyung/ gnya’ na ltag [±2] $ / / skyI ro ljang sngon na skyI bla/ bya rmang dbu snyung/ spyan snyung ste/ skyi gshen rgyan ngar gyIs/ chu bya gnya’ rIngs la/ nub sum/ rbu ru rbu nang sum bshan du bshan/ skyI ’phrang rub du bo/ skyi bla bya rmangs/ dbu snyung/ spyan snyung yang bshos (ll. r171–74).

[xx] gyI bar du ’dI dang ’dra bar bgyI ’o/:/:/; r174–75). After this, the successive antecedent tales are presented in shorthand, catalogue format or redactional outline, e.g., “In ’Ol phu Dga’ dang, ’Ol lha Sha bzan was struck ill. ’Ol gshen ’Jang tsa mon yug healed him” (’ol phu dga’ da[ng] ’ol lha sha bzan / snyung ’ol gshen ’jang tsa mon yug gis / shug bya thi ba la rbu /; r177).12 Each of the eight remaining tales, including that set in Rkong yul bre snar, is given in this same sort of shorthand that the liturgist may fill out in his or her tellings. Similar notes to the liturgist

offer an insight into how these liturgies were conceptualized and how these documents were apparently used. Further on in PT 1285, for example, we find the phrase, “one should continue to go around in this manner from here down to Rkong yul” (’dI man chad rkong yul yan chad ’di bzhin bskor ro; l. r188). This verb, “to go around” or “to wander,” is found in a similar context in two other ritual antecedent tales, which both include the phrase nan ’di bzhin bskor ro, meaning something like, “one should proceed as you have in this telling” (ITJ 734, 3r117 and PT 1060, l. 73). One of these also instructs the liturgist to “proceed as [in the] above [tale]” with regard to other tales (gong ma bzhIn skor cig; PT 1060, l. 48).13 In a text

containing several ransom ritual antecedents we find many instances of the expression “proceed as with the previous tale” (nan snga ma bzhin du skor; ITJ 734, 4r159, 5r175, 5r184, 7r305), and “proceed as in the first tale” (nan thog ma bzhin du skor; 4r153). The use of the terms “recitation” (nan) and “go around” (skor or sometimes ’gor) clarify the terminology for the “catalogue of antecedent tales” form. A nan is the recitation of an antecedent tale. To “go around” or “wander” is to tell one tale after another or to link tales together, and this is called a nan skor. Like the spatial organization of the antecedent tales’ settings in the catalogue form, the verb skor imbues these tellings with a sense of movement. One should note here that a single antecedent tale can include different settings, and name different places, without being a catalogue of antecedent

tales. Such is the case with the mention of multiple suitors and their kingdoms in an antecedent tale in PT 1285. There, the list of the failed suitors arranges them one after another, westwards, against the eastward course of the Gtsang po 12 I paraphrase, since the translation of shug bya thi ba la rbu is uncertain. The verb rbu is synonymous with the more prevalent verb in this text, mngad, and they appear profusely in this text but little, if at all, elsewhere. Other synonymous verbs in the text are bslen and bshan. They are used variously, but in the above context the priest is the agent, and an animal – here a type of bird – is marked in the dative. A working hypothesis is that the priest expels the patient’s malady into the animal, in which case rbu/dbu is a variant of ’bud, “to drive out, expel” (Hill 2010: 203). 13 The imperative form shows clearly that these are performance notes or instructions. Where there is instead the future form, e.g., bskor, this has a comitative meaning: one “should” go around as in the above tale.

(Dotson 2008: 49–53). But no suitor plays an otherwise significant role in the tale, and it is only with the successful suitor and the marriage that we come to the crisis, its resolution, and its application to the present. It is for this reason that one can correctly refer to such a ritual antecedent tale with multiple such settings as a “narrative,” and not a “catalogue” (Dotson 2008: 45). More specifically, one can refer to a “single antecedent tale with multiple settings,” which accurately describes this example of the failed suitors from PT 1285 and, for example, the “Tale of the Rgyal byin” in PT 1040. In the latter, we have repeated antecedents of failure and death that are not individually resolved. What appear to be items of cursed dowry are implicated in the death of a woman’s successive husbands, and the deaths of successive kings who come to possess these

items. Unfortunately, the scroll breaks off before a resolution is given, so we cannot know the exact purpose of the tale. It is a fair guess that in an intact telling, the ritual specialist would have been called upon to deactivate or destroy the cursed objects, and that this would then be related to the present-day deactivation or exorcism of similarly malefic objects.14 One clear difference between a single antecedent tale with multiple settings and a catalogue of antecedent tales is the site of the resolution. In a catalogue, there is a resolution in each individual setting, when, for example, each god or king is healed. These healing antecedents can be related to the present ritual after each setting, and/or after all of the tales in the “catalogue” have been told. In a narrative with multiple settings there is not a resolution in each setting. Rather, each setting is the site of a crisis, and this crisis is not resolved until the final setting. As a narrative form, the catalogue of (ritual) antecedent tales is not limited to genre of ritual texts. Rather, we also find it – or something closely resembling it – in Tibet’s chronicle-epic, where many of the same

themes (totality, covering Tibet’s ritual-spatial universe, and instantiating present success through precedent) are brought to bear on an ostensibly historical genre. We observe this in the “Account of the Minor Kings” in the Old Tibetan Chronicle. There, in the Royal Genealogy (PT 1286), which forms the first part of the Old

14 See Stein 1971: 545, where he writes of this text, “Il s’agit plutôt de précédents d’une certaine méthode rituelle concernant le rgyal et le byin (puissance, bénédiction?).” The text requires further study, but it appears that Stein misunderstood rgyal byin. Two alternative understandings present themselves: 1) rgyal byin is short for “[the dowry that] Rgyal [de ched po] gave (byin) [his daughter]”; or 2) a rgyal byin is a nefarious force or demon. The latter explanation is supported by the term’s appearance in an inauspicious divination prognosis in the dice divination text ITJ 738: “this prognosis… led by hatred. Sri demons and rgyal byin and ’gong po demons will flourish from the pho sa” (mo ’dI chIg shIn [---] la bab ste / / mkhon kyIs sna drangs te / / pho sa nas srI dang rgyal byIn dang ’gong po che ste /; 2v14–15).

Tibetan Chronicle (PT 1286 and PT 1287), we find a list of kingdoms and the names of the kings and councilors who resided in each of them, proceeding, as in PT 1060 and other ritual texts, from west to east, that is, with the course the Gtsang po. It is introduced as a rabs: “The account of those who, acting as minor kings, resided in their respective castles in their respective minor kingdoms, and of those who served as councilors of the minor kings” (rgyal pran yul yul na / mkhar bu re re na gnaste / rgyal pran bgyId pa dang / rgyal pran gyI blon po bgyid pa ’I [rabs]; PT 1286, l. 6; Bacot, et al. 1940–1946: 83). This is the most non-narrative “catalogue” we have, and one after another it lists kings, kingdoms, and councilors. After giving its title, it begins with the first two kingdoms:

The overlord of Dar pa in Zhang zhung was Lig snya shur, and the two councilors were Khyung po Ra sangs rje and Stong lom Ma tse. In Phyed kar in Myang ro the lord was Rtsang rje ’i Thod kar and the two councilors were Su du and Gnang.

zhang zhung dar pa ’I rjo bo lig snya shur / blon po khyung po ra sangs rje dang / stong lom ma tse gnyIs / / [myang] ro ’i pyed kar na / rje rtsang rje ’i thod kar / blon po su du dang gnang gnyIs /; PT 1286, ll. 7–8.

It goes on to list a total of seventeen such entries, each with a territory, ruler, and two councilors. The directional orientation of the list, and its claim, at the end, to have listed thirteen minor kingdoms, indicates that it participates in the same traditions that inform the organization of space in ritual antecedent tales and in catalogues thereof. Here, as in catalogues of antecedents such as PT 1060, we find a clear symptom of adhering to such a formula: the stated number of kingdoms does not agree with the number of narrative settings that are actually listed (cf. Dotson

2012: 169–71). Similarly, the kings named in the Chronicle’s account of the minor kingdoms are not real people, but the same mythical kings found in ritual narratives, whose names are usually composed from vague royal ethnonyms.15 The purpose of the “Account of the Minor Kings” is not apparent until the end, when we are told that all of these kings vied with one another and that none could conquer the first royal ancestor, ’O lde spu rgyal, who subjugated them all. The account ends there, without any statement relating itself to present realities. 15 There may be exceptions to this, and we can point to Dgu(g) grI ’i Zing po rje, who is named as king of Ngas po, as someone who appears as a “historical” character in chapters three and four of the Old Tibetan Chronicle; see Bacot, et al. 1940–1946: 132–34 and Dotson 2013a. The Chronicle’s value as a source for Tibetan history is a point that I address in my Habilitationsschrift.

In considering the Chronicle’s “Account of the Minor Kings,” we can assess the extent to which it adheres to the typologies of ritual antecedent tales and catalogues of ritual antecedent tales outlined above. Formally, it has only the bare bones of our typology: dramatis personae, crisis, and resolution. If we view each setting as a redactional outline to be filled out, then we might assume that of each kingdom one might say, “king so-and-so and his councilors soand-so vied with the other kings and councilors, but could not defeat ’O lde spu rgyal, and was conquered.” Each individual setting, and the catalogue as a whole – if it be a catalogue – lacks the final element relating the tale(s) to the present. From this perspective, one can neither say that the “Account of the Minor Kings” is a single antecedent tale with multiple settings, nor that it is a catalogue of such

tales. It is clear, however, that the “Account of the Minor Kings” performs the same function as the spatial arrangement of redactional outlines does in a catalogue of antecedent tales: it presents a formulaic model of a specific, bounded ritual-spatial universe that would have been recognized as such by its audience. In this case, each setting is not the site of an antecedent tale; the point is rather the whole itself, and the Tibetan king’s assertion of dominion over this schematic representation of the world. This overlapping of form, content, and of genre designations leads to a productive line of enquiry on the relationship between ritual literature and historiography. We shall return to this theme after considering the variety of Buddhist attitudes towards Tibetan ritual antecedent tales and redactional outlines, particularly in the context of funeral rites. II. Tibetan Buddhist Appropriations of Ritual Antecedent Tales Animal sacrifice, or ritual killings, appears to lie at the center of early Tibetan funeral rites. The horses, sheep, and other animals that guided the deceased to the land of the dead were presumably only constituted as pyschopomps when they were sacrificed to join the man or woman in death. In the “Tale of the Yak-Ox Hybrid” (mdzo rabs) for example, the text describes the butchering of the yak-ox hybrid and the placement of its parts in the tomb.

They killed the mdzo mo and cast the four sections of flesh as bya za khra thog (food for the birds). They/he carried the four sections as the brother’s snying lan. They installed the four sections in the sister’s tomb, and installed the head and the internal organs as the sister’s mchun (pillow?).

mdzo mo bsade zha16 lhu bzhi go scogs na bya za khra thog du boro / / lhu bzhi go scogsni mying po snying lan du khyer / lhu bzhi go stsogs ni sring mo’i ngur du bcug mgo brang smad lnga ni sring mo’i / mchun du bcug /; PT 1068, ll. 54–57.

In the passage immediately following, however, we have a hint that an effigy might be used:

The loving brother gshen sought and sought; the eternal companion, the lifelong wealth/livestock. The blood, vermillion blood. The bones, conch bones. The hooves, iron hooves, the horns, golden horns, the eyes, the spug gem eyes.

mying po mnyes gshen gyis btsal btsal ba gzha gstsang gyi rogs tshe rabs gyi pyugso / khrag ni mtshal gi khrago / rus ni dung gi ruso / rmyig pa ni lcags / gi rmyig pa rva gser gi rva / / myig ni spugi myig/; PT 1068, ll. 57–59.

This is an interesting point, and it reminds us of the dynamic environment that produced these texts, none of which – in their extant Dunhuang exemplars – predate the introduction of Buddhism to Tibet. It may be that this passage relates to the ritual transformation of objects, so widespread in Tibetan ritual. Alongside this, we might consider the possibility that the Buddhist aversion to animal sacrifice, and perhaps a growing sentiment against it, may have led bon and gshen priests to substitute effigies for live animals in their rites. Such adjusted but still ostensibly “non-

Buddhist” rites may have also been used by Buddhists. This does not explode the general picture of a movement from widespread animal sacrifice to its piecemeal prohibition across the Tibetan cultural area, but rather lends some nuance to the models of interactions between Buddhists and sacrificers. As is well known, the use of scores of horses in funeral rites is clearly demonstrated from the excavation of tombs at Dulan, and the sacrifice of horses at a funeral is also referred to in the Old Tibetan Chronicle. It is also clear from the numerous, often fragmentary, Buddhist polemics against animal sacrifice in funeral rites that this practice was as contentious in early Tibet as it has been ever since. In a fragment of a catechism/polemic, we find the following passage, which could just as easily be authored by a lama or a gter ston in the present-day Himalayas:

According to the divine religion, “One should not kill living beings.” If one removes the roots of both fratricide/incest pollution and corpse pollution, then the gods shall be happy. “Renounce fighting in retribution for hurt and anger. Rely on dhāraṇī/spiritual sustenance for extensive and numerous joys.” So it says. Untie the knot of enmity! If one removes the root of faults, the gods shall be manifestly pleased. If they look upon such behavior, they shall not be made to be displeased and upset. They shall remain unruffled. Some say, “As for making the gods upset, it is none other than this: there are different ways to worship the gods. [You] worship Buddha with ska ma chos.17 We worship our sgo lha and yul lha with live animals. The rituals performed are not in accord, and so [the gods] become displeased and upset.” Are[n’t] there those who say this? But the Buddha…

lha chos las sems can no cog gI srog myi gcad do zhes ’byung ste / / smye bag dang ro bag gnyIs ka’I drung phyung zhing mchis na / / lha dgyes pa’I rigs / / gnod khro glen ba’I ’thab rtsod ni spangs / / dga’ dgu yangs pa’I gzungs ni bsten ces ’byungste / /mkhon gyI mdud pa ni bshig / /bkyon bar gyi drung ni phyung na / /lha dgyes par mngon te / / tshul ’dI rnams gzigs na yang / / myi dgyes shing ’thur ba’I gcugs ma mchis te / / myi ’thur bar yang gda’/ / kha cig na re / / lha ’thur ba’I gcugs lta zhig / / gzhan ma mchis kyang lha brjed pa la tha dad de / / sangs rgyas ni ska ma chos su gsol/ / bdag cag gI sgo lha dang / / yul lha ni srog chags kyIs gsol te / brjod pa’I cho ga myi ’thun bas / / myi dgyes shing ’thur bar ’gyur ro zhes mchi ba dag kyang mchis grang ste / / yong sangs rgyas; ITJ 990, ll. 6–13.18

Another fragmentary polemic, also about funerals, simply states, “it is inappropriate to kill” (bsad pa'I yang myI rigs; ITJ 562, 1r2). A further fragment from the same document begins to offer a strategy for avoiding sacrifice:

If it be inappropriate to violate the orders of the bon po and the gsas, then, since it is also inappropriate that such great suffering should result from doing as they say, and since, when funerals have been performed up until now and funeral rites have been performed according to the “Tale of the Psychopomp Sheep,” where its hooves

17 The religion of karma? 18 See also the brief treament of this text in van Schaik 2013: 234.

rend cliffs on which there are no paths and its snout sucks up [the water of] lakes that have no fords, but nowadays…

brgya zhIg la bon po / dang / gsas gyi bka’ gcag du myi rung ste / / de bzhin du mchi’ bar gyur na ni / de tsam las sdug bsngal cher yang / myI rung bas / yang skyIbs lug gyi rabs de bzhIn du mdad ni / sngan cad shId gtang ngo / ’tshal kyis / brag lam myed / do ’tshal ni / rmyig pas dral / mtsho rab myed do / ’tshal ni / sngur pas rngubs pas / deng sang du / phyogs su /; ITJ 489, 1v1– 4.

Unfortunately, the pothī-format leaf ends at this point, and its continuation has not been found.19 It would seem to be a charter for the transformation of the traditional funeral rite involving the psychopomp sheep. This is all the more likely for the fact that we have a text which details the “substitution” or “transformation” (bsngo ba) of traditional Tibetan funerals to accord with Buddhist ideology. Stein (1970) studied and translated this text, PT 239, along with a related fragment, ITJ 504.20 Imaeda (2007: 166–69) refers to it as a “substitution” text, and believes that it forms a suite of interrelated Buddhist funerary texts, together with the “Story of the Cycle of Birth and Death” (Skye shi’i lo rgyus) and

the “Guide to the Land of the Gods” (Lha yul du lam btsan pa).21 The “Substitution Text,” for example, seems to take it as given that one must use the frame of the non-Buddhist funeral. This is precisely the situation described in the above fragment, ITJ 489, which addresses a situation in which a Buddhist cannot abandon the traditional rite as dictated by the bon po, and must follow the customs given in the (non-extant, possibly fictitious) “*Tale of the Psychopomp Sheep.” The “Substitution Text” goes through the various elements of the “black funeral,” and transforms them so as to avoid animal sacrifice and emphasize the core Buddhist doctrines of karma and rebirth. While it is accommodationist in the sense that it is bound by the

structure of a non-Buddhist rite, the transformation of each stage of this rite is nevertheless also an occasion for direct confrontation. In the part of the rite that calls for the sacrifice of a sheep, for example, the text states, According to treatises of the black men, the custom of the black funerals, the origin narratives (smrang) for which the bon (priest) wants 19 See the discussion in van Schaik 2013: 249–51. 20 Since then another related fragment, ITJ 493, has been identified. 21 On the latter, see Lalou 1949. While “transformation” may be more descriptively adequate, it is inadvisable to refer to PT 239 as a “transformation text,” given that this famous genre of Chinese text (bianwen) was present in Tibet in the imperial period. I therefore follow Yoshiro Imaeda’s custom of calling PT 239 and related documents “substitution texts.”

a fee, and the tales (rabs) for which the demons want burnt offerings, the sheep is wiser than man, and the sheep is also more powerful than man. Though [the stories may say] that, each and every being is guided by his own karma, and this means that a sheep has no wish to lead you on the path, and has no wish to rend cliffs. Neither is a sheep able to serve as a guide, nor is it able to make plans, nor can it use its limbs to shoot an arrow. Trusting in what is real, trusting in the treatises of the white, divine religion, the ways of the white men, and the white funerals, we rely

on the white, divine religion. We do not insert hands and cold iron into your insides. We do not draw out your warm blood. We do not claw out your heart and organs. We do not hang your flayed skin from your shoulders. We do not grind your white bones in a mortar. We do not cook your red flesh in a pot. According to the way of men of the upper realms, we do not do the work of ’dre demons; we do not do the work of srin demons. The eyes, living eyes, sparkling; the ears, living ears, flopping; the horns, living horns, curling. You are left to graze on the grasslands, on the fields, pastures, and meadows. Fumigate with the purifying divine incense. By the virtue of offering powerful, divine mantras, may this deceased, whose name is xxx, wherever s/he may be [re]born, be free from the suffering of weapons and such things, and may s/he obtain an eternal body, free of the sufferings of birth and old age. Having well loaded the psychopomp sheep, may it be good and auspicious for the remaining relatives.

myI nag po ’i gzhung/ /shId nag po ’i lugs/ /bon yas ’dod smrang/ /’dre gsur ’dod gyI rabs las/ /myI bas nI lug ’dzangs la/ /myI bas kyang lug mthu che bar ’byung ba yang/ /sems can thams gyang sa so ’I las kyIs khrId pas/ /lug gyIs lam drang yang myI dgos/ / lug gyIs brag dral yang myI dgos / lug gyIs lam mkhan byed kyang myI nus / / lug gyIs blo byed kyang myI nus / lag dum gyIs mda' 'phen yang myI nus par / / ngon don22 la yid ches pas / / lha chos dkar po 'i gzhung / / myI dkar po 'i lugs shId dkar po 'i ches / / lha chos dkar po la rden23 nas / / lcags lag grang mo ni khong du ma bcug / / khong khrag dron po nI phyir ma phyung / / don snyIng smad lnga nI spar gyIs ma bdab / / g.yang bzhI nI phrag la ma gzar / / rus pa dkar po nI gdun24 la ma rdungs / / sha dmar po ni bzangs su ma btsos / / mtho res25 myI 'I lug kyIs 'dre 'I lam26 ma 22 Read nges don. 23 Read rten. 24 Read gtun. 25 Read ris.

byad27 / / srIn gI las nI ma byas ste / / dmIg gson myIg nI rIg rIg / / rna gson rna nI dab dab / / rus gson ru nI kyIl kyIl / / zhIng spang snar po la nI spang rtsI yan du za zhIng bzhag pa ste / / lha spos gtsang mas nI bsangs / / lha sngags gnyen pos pos bdab pa'I yon gyIs / / myIng 'dI zhes bya ba / / gar skyes gar skyes gyang / / mtshon cha las bstsogs pa'I sdug sngal thams cad las thar ba dang / / skye rgan ba'I sdug sngal myed pa'I g.yung drung gyI lus thob par shog shIg / / skyib lug legs par stad pas / / gnyen dun slad ma rnams la / / rjes bzang zhIng bkra shIs par gyur cIg /; PT 239, 7r4–10r3; Stein 1970: 162–64; Karmay 1998: 160.

The project of the text is clear: a site of sacrifice has now become an occasion for discoursing on karma. In an early expression of liberating lives, the animals that would have been sacrificed are now set free. In the section of the text that corresponds to and confronts the sacrifice of a horse, the Buddhists draw on their own narrative repertoire and tell the famous tale of Bālāha, Avalokiteśvara’s flying-horse incarnation, saving the sea-faring merchants in Singhala as related in the Kāraṇḍavyūha sūtra. After this tale we find a truncated version of the formulaic statement of relevance so integral to the antecedent tale form: “today, looking to tomorrow, this son of man’s karma is exhausted…” (dI rIng sang ltar myI bu las

zad cing/; PT 239, 12r4; Stein 1970: 164).28 Whereas this may have marked the end of a traditional ritual antecedent tale, here it is a transition into a parallel that is drawn between the deceased’s own helplessness and that of the merchants beset by demons on their voyage to Singhala. Then the man’s psychopomp horse is likened to Bālāha, in whose guise he is able to act as a savior for the deceased. In this way the Buddhists can retain the horse in a central role while transforming the reasons for doing so. It is a familiar strategy in the history of Buddhist interactions with competing

traditions. There is no doubt that the Buddhists vilify the non-Buddhist funerals. They are “black funerals,” performed by “black men,” in contrast to the “white funerals” of the “white, divine religion.” They attack their opponents’ fundamental texts or traditions (gzhung), and their ritual narratives (smrang and rabs). Besides this, they depict the bon po as being greedily concerned with their ritual fee (yas), and associate them, and their acts of animal sacrifice, with ’dre demons and srin demons. Such charges would continue to

26 Read las. 27 Read byas. 28 Here myi bu is a faint echo of myi rma bu mching rgyal, the traditional name used to refer to the patient in ritual antecedent tales in the same fashion that the “Substitution Text” refers elsewhere to “the deceased, whose name is xxx” (gshing mying ’di zhes bya ba; PT 239r2, ll. 4–5ff) (cf. Dotson 2008: 45, n. 23).

characterize Buddhist polemics against their bon po (and Bon po) opponents.29 Perhaps the most interesting point in the Buddhist attack is their targeting of named ritual antecedent tales. This brings us to a second mode of interaction and appropriation: the Buddhist construction of rival ritual antecedent tales. The “Story of the Cycle of Birth and Death,” several copies of which were found in Dunhuang, adapts well-known Buddhist narratives, like the Gaṇḍavyūha sūtra and the Sarvadurgatipariśodhana tantra, to narrate the divine prince Rin chen’s search for a remedy against death. This “history” or “account” (lo rgyus) is a long narrative that constitutes a charter myth for the performance of Buddhist funerals. In his study of the “Story of the Cycle of Birth and Death” (Skye shi’i lo rgyus), Yoshiro Imaeda describes the impetus for its composition:

The Buddhists, in fact, found themselves face to face with a population submerged in the indigenous religion. It was impossible to impose new ideas on them in an abstract way and so it was therefore necessary to employ an appropriate ‘pedagogical method’. It is this that led to the creation of the story of Rin chen that, situated in a time qualified by the use of ‘formerly’ (p. 137), could be used as a ‘story of origin’, an example of the efficacy of Buddhism in the search for a remedy against death. (Imaeda 2007: 170)

The “Story of the Cycle of Birth and Death” includes the key elements of the ritual antecedent tale form introduced above. We are introduced to the characters in heaven, and immediately to the crisis, namely the death of Rin chen’s father, the king of heaven, and Rin chen’s anguished attempt to come to terms with the dawning awareness of mortality. The main part of the narrative consists of Rin chen’s encounters with different teachers in

different settings. In each setting the protagonist, Rin chen, fails to find the remedy for birth and death that he seeks, and is sent to a further teacher. This is similar to several of the topoi in traditional ritual antecedent tales, among them the “failed suitors motif,” “false priests motif,” and “summoning the favored priest motif.” Unlike the “false priests motif” found in some ritual antecedent tales, this element is not used here for polemic ends. Some of the antecedent tales in PT 1285, for example, contrast the success of the favored ritual specialist with the failures of those outside of the tradition – in this case one hundred male gshen from the

29 Following a convention employed by Stein, I use the lower case, italicized bon to indicate the “bon po ritual tradition,” and the upper case Bon to indicate the selfconscious emergence of Bon as a religion against the backdrop of the rise of Buddhism as Tibet’s dominant religion; see Kapstein and Dotson 2007: ix, n. 4.

white, sunny mountain and one hundred female gshen from the black, shady mountain (Dotson 2008: 48). In the “Story of the Cycle of Birth and Death,” by contrast, each failed resolution or “false teacher” is rather a stepping-stone on a long narrative journey to meet the Buddha. All of these topoi have the effect of postponing the resolution and emphasizing the distance between the starting point and the state to be actualized. In a ritual context, one might say that the length of the journey and its many stages emphasize the distance between human and divine. It is only in the final narrative setting, in Magadha, that Rin chen meets the Buddha, who resolves the crisis. At the end, we also learn the relevance of the tale. This is not given with a formula that explicitly relates Rin chen’s journey to the present, but rather it takes the form of the Buddha giving general

instructions to tantric priests for how to perform funeral rites according to the Uṣṇīṣavijayā dhāraṇī. Like the story of Bālāha in the Buddhist transformation of the sacrifice of the psychopomp horse, the “Story of the Cycle of Birth and Death” also draws on Buddhism’s rich narrative repertoire. Its two main sources are the Gaṇḍavyūha sūtra and Sarvadurgatipariśodhana tantra. Unlike the story of Bālāha, however, the tale of Rin chen’s search for a remedy against death is less encumbered by its competitors’ ritual narrative technologies. The narrative does not underwrite a concurrent or mimetic ritual in the same way as a ritual antecedent tale appears to do. For example, despite the language used in each setting, where Rin chen requests a “remedy against death,” the Skye shi’i lo rgyus is more a bereaved son’s existential search for meaning than it is an antecedent for a ritual that constitutes a remedy against death (Imaeda 2007: 119). In the end the Buddha does not perform the funeral rite for Rin chen’s

father, nor does he embody a paradigm for subsequent ritual practitioners to emulate, but rather gives instructions to tantrikas. This is a subtle, but important diminishing of the role of narrative in ritual, but it belies a fundamental difference in orientation concerning the utility of narrative. These are not antecedent tales, but charter myths, and they present themselves as background preliminaries rather than as ritual centerpieces. The “Story of the Cycle of Birth and Death” is less a transformation of the traditional Tibetan ritual antecedent tale than an alternative ritual narrative technology. While the “Substitution Text,” enunciates a ritual that both mirrors and undermines the “black funeral” based on non-Buddhist ritual antecedent tales, the “Story of the Cycle of Birth and Death” dispenses with such mimetic opposition. This effectively resolves the apparent unease that Buddhists felt about a situation in which they were beholden to the traditional funerals and had to work within the outlines of rites that called for animal sacrifice. With the “Story of the Cycle of Birth and Death,” Buddhists could

compete more or less on their own terms, offering an alternative to the “black funerals.” Read in isolation, the “Substitution Text” and the “Story of the Cycle of Birth and Death” could easily be mistaken for diachronic developments in the Tibetan assimilation of Buddhism. The “Substitution Text” could be viewed as reflecting a time when Buddhists were forced to work within the fold of the dominant ritual tradition and its funerals, which their aversion to sacrifice compelled them to transform. The “Story of the Cycle of Birth and Death” would be taken to pertain to a later stage, when Buddhism was able to compete in the funeral ritual arena on its own terms. This is the type of error that one can easily make by studying

texts only for their contents without analyzing them as physical objects. As mentioned above, Yoshiro Imaeda sees the “Substitution Text” and the “Story of the Cycle of Birth and Death” as part of a suite of three texts together with the “Guide to the Land of the Gods” (Lha yul du lam btsan pa). Imaeda’s argument is based not only on the close relationship of their contents, but on the fact that the latter, whose sensibilities are very similar to those of the “Story of the Cycle of Birth and Death,” is found on the verso of the concertina (PT 239) that contains the “Substitution Text” on the recto. Moreover, the two texts are written in the same hand. It is clear, therefore, that the transformation of content and the direct polemical confrontation characteristic of the “Substitution Text” is a mode of interaction alongside other concurrent modes, such as the construction of a rival, Buddhist narrative in the “Story of the Cycle of Birth and Death.” III. Irony and Polemic in the “Account of the Food-Provisioning [for the Dead]” (Zas gtad kyi lo rgyus) The final text that we shall examine also has a complicated and unsettled transmission history. The “Account of the Food-Provisioning [for the Dead]” (Zas gtad kyi lo rgyus) is appended to the

Dba’ bzhed, one of the most famous early Buddhist histories of the introduction of Buddhism to Tibet. The Zas gtad kyi lo rgyus, or Zas gtad, is a polemical charter myth justifying the performance of Buddhist funeral rites. It has not been reliably dated, but some of its contents hint at its milieu. For example, it partakes heavily of the hagiographic tradition surrounding Vairocana, it belittles Bon, and it is aware of gter (ma).30 30 The Zas gtad kyi lo rgyus polemicist laments near the end of his text that “it is said that the foolish followers of the Bon tradition hid many items of wealth and enjoyment as gter” (bon lugs glen pa dag nor long spyod mang po gter du sbed pa yod skad; Wangdu and Diemberger 2000: 105; translation mine). In the context of the surrounding passage, which gives the textual basis for Buddhist funerals, this reference to gter most likely indicates Bon texts. It is also interesting in terms of “treasure texts” that the Zas gtad states that several Buddhist texts were translated from Sanskrit under Emperor Mu ne

Its treatment of early Tibetan ritual and historical traditions, even in refuting and lampooning them, also displays misunderstandings that are probably indicative of its temporal remove from the imperial period. To name only a few, it treats the title “warlord” (zing po rje) as if it were a proper name; it mistakes the name of this ruler’s stronghold; and it uses Mchims Dwags po as a compound toponym despite the fact that Mchims and D(w)ags po were separate, albeit neighboring, kingdoms. More egregiously, the entire setting of the Zas gtad kyi lo rgyus is a historical anachronism: it has Mu ne brtsan (po), presiding over the debate concerning the performance of his father Khri Srong lde brtsan’s funeral, despite

the fact that Mu ne brtsan (po) predeceased his father (Dotson 2007: 13, n. 48). The Zas gtad appears after the end of the Dba’ bzhed, and is not found in other extant versions of the Sba bzhed. For these reasons, it cannot be dated on the same bases that one might use to date the Sba bzhed/ Dba’ bzhed itself, whose core narrative traditions ultimately go back to the founding of Bsam yas Monastery (Sørensen 1994: 10–14). At the earliest, the Zas gtad dates to the end of the “intermediate period.” Both temporally, and, as we shall see, thematically, it forms a fascinating contrast with the earlier materials just reviewed. The Zas gtad evinces an extremely complex relationship with ritual narrative forms. It offers an account of a legendary debate between a Bon po, Mchims Btsan bzher legs gzigs, and a Buddhist, Vairocana, about how to perform the funeral of Emperor Khri Srong lde brtsan (742—c.800). The Buddhists win the debate, the funeral is performed in a Buddhist way, and this is presented as the charter for performing Buddhist funerals. One element of the prescribed funerals is food-offerings to the dead (zas gtad), hence the name of the text. As in the case of the “Story of the Cycle of Birth and Death,” the whole of the Zas gtad kyi lo rgyus is a charter myth that bears some resemblance to traditional Tibetan ritual narrative forms. It introduces the time, setting, and dramatis personae, presents the crisis and its resolution, and states the account’s relevance to the present. Like the “Story of the Cycle of Birth and Death” and the “Guide to the Land of the Gods,” it champions the use of the Sarvadurgatipariśodhana tantra and the Uṣṇīṣavijayā dhāraṇī, among others, for

btsan po, and that some of these “were hidden in a black leather box (bse sgrom nag po) in the dBu rtse zangs khang (the copper palace in bSam yas)” (Wangdu and Diemberger 2000: 104). This could be taken to imply that the funeral rites for which the Zas gtad kyi lo rgyus forms the charter myth were imagined to have been unearthed from this black box, though this is not stated explicitly. There is a voluminous literature on “revealed treasures,” “treasure texts,” and gter ma that I will not summarize here. Suffice it to say that the earliest attested textual revelations probably date to the early eleventh century; see Davidson 2005: 225.

funeral rites. As with these other Buddhist narratives, this is not a template for a mimetic ritual, but a charter myth that explains the triumph of these rites over “black Bon funerals.” The Zas gtad differs, however, in a few very important respects. First, whereas the “Substitution Text” transforms traditional funerals and undermines them “from within,” and the “Story of the Cycle of Birth and Death” and the “Guide to the Land of the Gods” offer competing alternatives, the Zas gtad is entirely dismissive of its non-Buddhist competitors. Second, while these earlier Buddhist narratives, like ritual antecedent tales, narrated the exploits of mythological figures, the Zas gtad polemicist chose historical figures as his

dramatis personae. We shall return to this latter point after examining in detail the rhetorical strategies by which the Zas gtad dismisses and degrades its (imagined) opponents. The Zas gtad kyi lo rgyus doubles as a charter myth and a polemic. The latter aspect is expressed in the familiar form of a staged debate. This allows the polemicist to put his opponent’s argument in the mouth of a villainous straw man, and to roundly defeat it through the words of his champion. In such a set-up one expects a tour de force, and the Zas gtad does not disappoint. Mchims Btsan bzher legs zigs

argues in favor of performing Bon funerals, worshipping royal tombs, and supplicating mountain deities – essentially a succinct statement of what the Zas gtad polemicist imagines to have been, or erects as, the traditional status quo against which the Buddhists had to struggle in imperial Tibet’s ritual economy. It is particularly interesting that Mchims Btsan bzher evokes Tibet’s traditional constitutional mythology in terms similar to that found in the Old Tibetan Chronicle and the eulogies of imperial-period pillar inscriptions.31 He evokes the descent of the first king in a verse that overlaps with one found in the Old Tibetan Chronicle.32 He then recounts the origin of funeral rites, but, in a display that reveals the polemicist’s ignorance of his opponents’ beliefs, claims that these funerals came about not during the time of Dri gum btsan po and Spu de gung rgyal, but during that of Lha tho tho ri

31 I borrow the term “constitutional mythology” from Martin Mills, who uses it to describe how traditional Tibetan “religious histories” such as the Rgyal rabs gsal ba’i me long, and particularly the myth of Srong rtsan sgam po (c.605–649), serve as “a means for people in general to constitute their understanding of legitimate governance and history” (Mills 2009; italics in original). 32 The Old Tibetan Chronicle (in the Royal Genealogy) states, “Even the trunks of the trees strained to the utmost. Even the springs rippled on their banks. Even the boulders and so forth saluted elegantly and thoughtfully” (shing sdong po yang bang thang thang / chab lu ma yang dngo sil sIl / gor pha bong la stsogs pa yang mnyed khrung khrung gis pyag ’tshal lo; PT 1286, ll. 33–34; Bacot, et al. 1940-1946: 86). The Zas gtad reads, “The trees used to bow their bodies, the solid boulders used to jump” (shing sdong po ni sku dud dud / gor pha bong ni ’phar thang thang; 27a, l. 7; Wangdu and Diemberger 2000: 97–98, n. 387).

snyan shal, an ancestor promoted by Buddhists during the time of the polemicist’s milieu.33 Mchims Btsan bzher continues by valorizing the worship of the royal tombs and the royal divinities such as the mountain god Yar lha sham po. He then recalls the Yarlung Kingdom’s conquest of the minor kingdoms, its rapid expansion, its “lofty realm” (srid mtho) and “traditional art of governance” (gtsug lag).34 Allowing Buddhist monks to perform the royal funeral, argues Mchims Btsan bzher, would be to risk endangering the traditional order of things, and would disrupt the realm (chab srid) (Wangdu and Diemberger 2000: 97–99). The allusions to, and near quotations of, traditions found in the Old Tibetan Chronicle, are also found in

Vairocana’s response. Here the tenor changes to one of mockery, however, particularly when Vairocana references traditional ritual narrative forms. Vairocana’s counter-argument unfavorably contrasts tombs, palaces, and the royal mountain god Yar lha sham po with Nālanda Monastery in India, Buddhist pure lands such as Bde ba can and Padma can, and the Protectors of the Three Families (rigs gsum mgon po), including Vajrapāṇi. It is the most succinct statement one could imagine for how one replaces imperial cosmology and non-Buddhist royal palladia with those of Buddhism. The rhetorical heart of Vairocana’s rebuttal is a mock “catalogue of ritual antecedent tales” that forms a contrast with the Old Tibetan Chronicle’s “Account of the Minor Kings.” As in a traditional redactional outline, the tale

33 That Lha tho tho ri has been substituted for Dri gum btsan po and Spu de gung rgyal is clear from the transmission and corruption of an already corrupt tradition concerning Spu de gung rgyal. The latter, according to the Old Tibetan Chronicle, was known as Grang mo gnam gser brtsig when he died: “The son of Dri gum btsan po was Spu de gung rgyal, [who was between] the seven heavenly Dri [and] the six earthly Legs. When living, [he was called] Spu de gung rgyal. When he died, [he was called] Grang mo gnam gser brtsig” (dri gum btsan po ’i sras / / spu de gung rgyal gnam la dri bdun / sa le legs drug bshos na / spu de gung rgyal grongs na / / grang mo gnam gser brtsig / gser brtsig gI sras; PT 1286, ll. 48–49; see also Zeisler

2011: 194– 200). The verb bshos, which means “to procreate” in every other appearance in the genealogy, accounts for past mistranslations of this passage, and also for the Zas gtad’s transformation of a posthumous name into a bride: “Lha tho tho ri snyan shal united with Grang lung gyi Gung sman, and Grang lung gyi gung sman performed his funeral” (lha tho tho ri snyan shal dang grang lung gi gung sman du bshos te/ gtong ’dad gtong ba yang grang lung gi gung sman gyis gtong ngo //; Dba’ bzhed, 27v2; Wangdu and Diemberger 2000: 98). The promotion of Lha tho tho ri as Buddhist ancestor is but one example of a long process by which the genealogy adapted to evolving political realities. 34 Gtsug lag is a polyvalent term that refers, in a constitutional context, to traditional governance and the traditional order of things. It can also mean divination, and it can refer to fundamental texts; see Stein 2010: 182–87 and Hahn 1997: 353–54.

set in the first setting is given in full, and those that follow are compressed. It begins as follows:

To say that gtsug lag and the rites and diagnoses of Bon are good is also a falsehood. Khri ’phang gsum, king of Zing po rje, worshipped the compassionless god Thang lha yar lha. The two Ag gshen [[[Wikipedia:priests|priests]]] of ’Phan yul killed many animals, such as yaks, sheep, and horses. ’Drir ’grin and Gco mi and others, the ’dre demons that want burnt offerings and the bon that want iron – many gathered together.35 Because he added subsequent ones to his previous misdeeds, and practiced a false religion, his bondservants Nyang, Sbas, and Gnon – the three –, with Tshe spong the messenger making four, along with his stronghold, Khung lung rgyab bu snang, were gathered under the dominion of Spu rgyal Tibet.

bon gyi lto dpyad dang gtsug lag legs bgyi ba de’ang sho pe ba lags te / zing po’i rgyal po khri ’phang gsuM gyis lha snying rje med pa’i thang lha yar lha gsol/ ’phan yul gyi ^ag gshen rnaM 2 kyis gnags lug dang rta lasvo pa dud ’gro mang po ni bsad / ’drir ’grin dang gco mi lasvo pa ’dre gsur ’dod dang / bon lcags ’dod mang po bsags/ sdig lnga ma’i steng du phyi ma mnan ste ’khrul pa’i chos la spyad pas kho’i bran nyang sbas gnon dang 3 / tshe spongs ’phrin dang bzhi mkhar khung lung rgyab bu snang dang chas nas spu rgyal bod kyi mnga’ ris su ’dus pa lags te de lto che zhing zhal bsod pa lags saM; Dba’ bzhed, 29v2– 6.

Vairocana’s redactional outline then briefly repeats the same basic narrative in the settings of the kingdoms of Zhang zhung, ’A zha, Mchims Dwags po, and Snubs. Like ’Phan yul, each fell to Tibet / Spu rgyal as a result of worshipping Bon po gods, sacrificing animals, and performing “black Bon funerals” (Wangdu and Diemberger 2000: 101–103). After finishing with the final setting, Vairocana gives a short statement of relevance: “Because of such major sins it is improper to perform funerals according to the Bon system” (Wangdu and Diemberger 2000: 103). Mchims Btsan bzher retorts in disbelief with rhetorical questions that ask if Vairocana would have it that the monks will now control all secular and military affairs. To this Vairocana triumphantly announces,

35 This appears to be a corruption of a passage similar to that found in the “Substitution Text”: “According to the treatises of the black men, the custom of the black funerals, the origin narratives (smrang) for which the bon (priest) wants a fee, and the tales (rabs) for which the demons want burnt offerings…” (myI nag po ’i gzhung/ /shId nag po ’i lugs/ /bon yas ’dod smrang/ /’dre gsur ’dod gyI rabs las/; cf. above).

“We monks can do it!” The Buddhists have won the debate, and Vairocana and others perform Khri Srong lde brtsan’s funeral in a Buddhist fashion. Vairocana’s mock account of the minor kingdoms is a polemic within a polemic. From its garbled, yet recognizable quotation of similar anti-bon polemics of the type found in the “Substitution Text,” it clearly belongs to a tradition of polemics that target “black Bon funerals.” It is also clearly familiar with the Tibetan constitutional mythology that informs the Old Tibetan Chronicle. Vairocana refers to how the subjects of the ruler Khri ’phang gsum, “Nyang, Sbas, and Gnon – the three –, with Tshe spongs the messenger making four,” fell under the power of Spu rgyal Bod. This is

the standard phrase used to describe these four clans in the context of their formative role in the creation of the Tibetan Empire when they defected from Zing po rje Khri pangs sum, narrated in the third and fourth chapters of the Old Tibetan Chronicle (Bacot, et al. 1940-1946: 134–39).36 The Zas gtad’s allusions to the Old Tibetan Chronicle or to closely related narratives brings Vairocana’s “mock catalogue” into conversation with the “Account of the Minor Kings” discussed above. The latter uses the redactional outline form to evoke an image of totality over which the Tibetan emperor asserts dominion. It overlaps closely with “catalogues of ritual antecedents” that set a crisis and its resolution, e.g., healing the sick, in each setting. In that case, one cannot safely claim that the Chronicle borrowed the form from ritual texts or vice-versa, since they may each be

drawing on a common, shared topos. Here, however, Vairocana’s relationship to the catalogue of ritual antecedents form is clear: it is the object of his parody. Despite using the same settings and the same redactional outline form as in the catalogues of ritual antecedents, Vairocana’s catalogue gives antecedents for failure rather than antecedents for success. Kingdom after kingdom failed because its king supported “black Bon funerals,” and these antecedents, taken together, inform us that such a funeral performed for Emperor Khri Srong lde brtsan can only have the same malefic effect. Vairocana has self-consciously appropriated a ritual narrative form associated with his opponents, and in so doing he has also disfigured it. Rather than arraying antecedents for ritual success in a sequential narration, he presents antecedents of failure. In one sense, Vairocana has beaten the Bon po at their own game, like an atheist quoting scripture to Christians by way of rebutting Christian views. On the other hand, his “mock catalogue” inverts the logic of the catalogue of ritual antecedent tales, and stands in relation to them as the Black Mass does to the Catholic Mass.

36 Uray (1967: 500) takes the appearance of this same phrase in the Mkhas pa’i dga’ ston to be “a paraphrase of the sentence quoted from the Ziṅ-po-rǰe Narrative” in the Old Tibetan Chronicle.

The rhetorical strategy of disfiguration and irony is also evident in Vairocana’s pun at the end of his account of the conquest of ’Phan yul in the first narrative setting. He asks, “lto che zhing zhal bsod pa lags saM?” This use of the term lto che by Buddhists to belittle Bon po opponents refers to the Bon po use of the homophonous gto rites and dpyad rites as a key element of the rituals that they perform in order to earn their livelihoods (Karmay 2010: 54). It is through gto rites and dpyad rites, so the pun goes, that a Bon po becomes wealthy and acquires a big belly (lto

che, a homophone with gto dpyad).37 Such humorous wordplay is a familiar rhetorical strategy. Considering examples such as “the Rabbinic deformation of euangelion, in the sense of the Christian Gospels, into ‘aven gilayon’ or ‘avon gilayon’, a kind of nonsense phrase but suggesting something like ‘falsehood/perversion of blank parchment’,” Philip Alexander writes of deformation or parody that it “is a witty put-down that not only expresses disapproval and devalues what the opponent holds dear but invites the hearer to disassociate himself from the butt of this humour” (Alexander 2008: 86). These conclusions certainly apply to Vairocana’s argument and to the rhetoric of the Zas gtad.

Bon po Rhetoric on Buddhist Appropriation This sort of mockery is far removed from the adaptation of the traditional funeral rites in the “Substitution Text.” Within the mode of ironic subversion espoused by the Zas gtad one can hardly imagine Buddhist willingness to make use of Bon ritual and narrative technologies for anything other than derision. Considering their generally ambivalent approach to non-Buddhist ritual

narratives, and the Buddhist backgrounding of even their own rival charter myths such as the “Story of the Cycle of Birth and Death,” one can read the Zas gtad in the context of polemics not only about funerals, but also about the place of narrative in Tibetan ritual. In this context, the Zas gtad might be in dialogue with criticisms that Buddhists were appropriating bon / Bon ritual narrative forms. We find exactly this charge in the

Bsgrags pa gling grags, a sort of expanded “Bon anti-Sba bzhed” whose action also takes place mostly during the reign of Khri Srong lde brtsan. In its account of a Buddhist versus Bon po conflict set during this emperor’s reign, Bon pos first defeat Padmasambhava in a contest involving magical killing and resurrection, and show him to be an underhanded charlatan. Then they defeat the Buddhists in a debate in which Dran pa Nam mkha’ criticizes Buddhists for their lack of a narrative element in their religious system:

37 A similar pun seems to work with zhal bsod pa: it literally means “to please the face,” and the near-homophone zhal gso ba can mean “to repair,” but also, and perhaps more to the point here, “to feed your face.”

“Instead of merely fixing your mind on nothing but nonconceptualization, do you, like Bon, have antecedent tales (dpe) and stories (lo rgyus) to tell in order to benefit beings? … If you have no story of the origin, then you have no method of performing funeral rituals for the dead.”

khyod mi rtog pa tsaM la blo gtad pa tsaM las / ’gro don byed pa la dpe dang lo rgyus bon gzhin bshad rgyu yodaM /… byung khungs lo rgyus med pas shi ba ’dur thabs med; G.yung drung bon gyi sgra bsgrags pa rin po che’i gling grags (from the Bon Bka’ ’gyur), folio 37b, ll. 3–4, or “page” 74 in the Arabic numbering. 38

In his summation, Dran pa Nam mkha’ accuses the Buddhists of appropriating Bon ritual technologies: “As there are no discourses on funeral rites in the discourses of Dharma, you have appropriated the model of Eternal Bon, and now, without any shame, you lead beings” (chos kyi lung las ’dur lung ma byung pas / gyung drung gi bon la dpe blang nas / skye bo khrid phyigs ngo tsha med /; G.yung drung bon gyi sgra bsgrags pa rin po che’i gling

grags (from the Bon Bka’ ’gyur), folio 38a, l. 4). These are charges to which the Bon pos obviously believed the Buddhists were vulnerable. One need not even read between the lines to suppose that Dran pa nam mkha’ is taking aim at something like the “Substitution Text” or the “Story of the Cycle of Birth and Death”: in a subsequent passage in the Bsgrags pa gling grags we find a list of Buddhist texts supposedly plagiarized from Bon, including the *Rta g.yag gi gtad yar (“The Ritual Loading of the Horse and Yak”), the *Lha bu rin chen gyi shi rabs (“The Tale of the Death of the

Divine Son Rin chen”), and the *’Dod khams su lam btsan pa (“Guide to the Desire Realm”). Buddhist monks, we are told, used Bon po texts as models, and transformed them by giving them Buddhist names (bon rnams chos su ming stags la bsgyur ro).39 These three titles correspond rather well to the “Substitution Text,” the “Story of the Cycle of Birth and Death,” and the “Guide to the Land of the Gods,” even if

38 The same passage begins at folio 65a (63a in the Arabic numbering), l. 3 of the “Oslo version,” the Bon chos dar nub gi lo rgyus grags pa rin chen gling grags ces bya ba/ dmongs pa blo’i gsal byed. The latter is virtually identical with the account given in the dbu can Bon chos dar nub kyi lo rgyus bsgrags pa rin chen gling grags zhes bya ba dmongs pa’i blo’i gsal byed, whose account starts at folio 60a, l. 4. 39 Bon chos dar nub gi lo rgyus grags pa rin chen gling grags, 72a5–b2 (70a in the Arabic numbering), and Bsgrags pa rin po che’i gling grags (from the Bon Bka’ ’gyur), 41b (p. 82), ll. 2–4.

the corresponding title for the latter is probably satirical.40 The name of the “Tale of the Death of the Divine Son Rin chen” obviously refers to the protagonist of the “Story of the Cycle of Birth and Death,” but it evidently takes it that he, and not his father, is the deceased. It is also interesting that this charter myth should be targeted for charges of plagiarism, since it is essentially a Tibetan apocryphal quasi-sutra based on the Gaṇḍavyūha sūtra and the Sarvadurgatipariśodhana tantra (Imaeda 2007: 171). If it is not the content that is at issue, perhaps it is the use of narrative as an important element of ritual that the Buddhists are accused of appropriating. What may lie behind these charges, therefore, is an

unease not only about Buddhist plagiarism of the contents of Bon po rites, but about Buddhist appropriation of ritual narrative forms. Juxtaposing the polemics of Vairocana and Dran pa nam mkha’ – or rather those of the Zas gtad kyi lo rgyus and the Bsgrags pa gling grags – they almost appear as if they are in dialogue. Vairocana’s polemic would seem to overturn Dran pa nam mkha’s two main points: he dismisses “antecedent tales and stories” by disfiguring them, and argues that it is the Bon funerals, and not the Buddhist ones, which are false rites. Vairocana’s dismissal of the ritual antecedent tale form also inoculates the Zas gtad against Dran pa nam mkha’s charge of appropriating Bon po ritual technologies, a point to which its predecessors such as the “Substitution Text” and the “Story of the Cycle of Birth and Death” may have been seen to be vulnerable. Whether these competing Bon and Buddhist accounts of conflict over funerary practices are in fact in dialogue remains to be determined through text-critical studies of the Sba bzhed / Dba’ bzhed, the Bsgrags pa gling grags, and related texts.

Ritual Antecedent Tales and Charter Myths in Relation to Histories Despite its parody of ritual antecedent narrative forms, the Zas gtad is itself a charter myth. Like the “Story of the Cycle of Birth and Death,” it uses narrative to provide a justification for the use of specific Buddhist funeral rituals. Indeed, when we consider the basic plot outlines, the narrative’s contours superficially resemble those of a traditional ritual antecedent tale: Khri Srong lde brtsan’s funeral is to be performed in Brag dmar; false priests arrive and fail (or are overcome); the exemplary priest Vairocana performs the funeral; and funerals are now to be performed in a Buddhist manner. Rather than deploying the mythical kings and priests who populate traditional ritual antecedent tales, however, the Zas gtad takes historical figures and casts them 40 The first title recalls a line in the “Substitution Text” from the long passage translated above: “…having well loaded the psychopomp sheep,…” (skyib lug legs par gtad pas).

as characters in a tale whose form and outcome is predetermined. This is a very important shift, and one that is relevant not only to such “accounts” (lo rgyus) and charter myths, but to developments in Tibetan approaches to historical narrative in, for example, the chos ’byung and rgyal rabs genres. Of course the strategies present in the Zas gtad are the result of developments over time. When the Old Tibetan Chronicle was composed, the narrative topoi found in ritual antecedent tales were a central part of Tibetan cultural production. The Chronicle’s composers naturally drew on these in their narrations, as in the “Account of the Minor Kings” and the episode of Princess Sad mar kar, which is framed within a matrimonial narrative trope (Macdonald 1971: 263– 65; Uray 1972: 35–36; Dotson 2013c). By the time of the Zas gtad, such narrative forms were no longer part of a

common cultural heritage, but were associated with an increasingly marginalized group, and Buddhists drew on other, separate narrative models. This ascendant Buddhist epistème was then read back anachronistically into the imperial period, and imperial-period ideologies, e.g. “traditional art of governance” (gtsug lag), were conflated with Bon, and demonized. The relationship of the Zas gtad kyi lo rgyus to history brings us back to the physicality of the document, and specifically to its position at the end of the Dba’ bzhed. Given its statement of relevance to funeral rites, we might expect the Zas gtad to appear near the beginning of, or as an addendum to a manual for funeral rites. So what is it doing here, at the end of the “historical” narrative of the coming of Buddhism to Tibet and the foundation of Bsam yas Monastery? To brush aside the question, we might say that the Zas gtad is simply out of place, and is essentially there by accident. Indeed there are several other cases of such juxtapositions of

separate texts in a single manuscript, where it is inadvisable to make too much of their links or relationships. On the other hand, if we take the Zas gtad’s placement at the end of the Dba’ bzhed as meaningful, then we must consider that the Zas gtad may be doing something else besides underwriting the practice of Buddhist funerals. As we have seen, it is not only a polemic about funerals, but also a polemic about narrative. The allusions to and near-quotations of the Old Tibetan Chronicle place Vairocana’s mock account of minor kingdoms in dialogue with that of the Chronicle or closely related narrative traditions. It is as if the two “catalogues” form a symmetry of the type preferred by Tibetan historians, where Vairocana’s ironic disfiguration of the catalogue of antecedents form effectively closes a chapter of Tibetan narrative historiography that began with the type of “innocent” use of shared narrative forms that we find in the Old Tibetan Chronicle. This is not the same sort of narrative symmetry whereby the empire, created from the chaos of the minor kingdoms, devolves into the chaos of the regional principalities; it is not the

unraveling of the silken knot of the empire’s religious laws and the breaking of the golden yoke of its royal laws;41 it is not that irony has replaced innocence in a benighted time. Rather, the Zas gtad refutes one vision of imperial power and harmony and confidently replaces it with another. Vairocana’s “ironic catalogue” is in this sense a leitmotif for the deactivation of those putative imperial-period ideologies that do not accord with emerging post-imperial religio-political identities. One of its strategies, as we have seen, is to label as “Bon po” those remembered or imagined elements of earlier imperial ideologies that have now become objectionable. It is in this sense a charter myth for a new sort of narrative of the imperial period, very much of a piece with that found in the Dba’ bzhed that precedes it.

Conclusions Tibetan polemics about funerals are also polemics about the place of narrative in ritual. Examining early Tibetan ritual antecedent tales (rabs) and redactional outlines, we can almost discern the script of the rituals that they purport to accompany and underpin. The narrative topoi can expand and contract, suspending the inevitable resolution in the same way that the manifold elements of a ritual can postpone the denouement of its climax. Briefly surveying such tales, we have demonstrated how they are constructed, and how a sub-genre, the “catalogue of ritual antecedent tales,” reduces the tales to redactional outlines in order to array them in meaningful ways to convey directionality and/or totality. Faced with these narratives, and their central role in funeral rituals in particular, Buddhists adopted a number of strategies. When it was necessary, perhaps due to the expectations of clients, to use the broad outlines of the traditional rite, Buddhists transformed its individual elements in order to avoid

animal sacrifice. They also drew on their own narrative repertoire, as in the case of the insertion of the story of Bālāha in the “Substitution Text.” Another strategy was to perform Buddhist funeral rites according to the Sarvadurgatipariśodhana tantra and other canonical texts. In this case, however, the Buddhists also composed narratives that gave the background for these rituals and effectively functioned as charter myths. We see this in the case of the “Story of the Cycle of Birth and Death.” Moving forward in time, we find a further rhetorical strategy of parody and dismissal. In the “Account of the Food Provisioning [for the Dead]” (Zas gtad kyi lo rgyus), the traditional “catalogue of ritual antecedent tales” form is transformed into a catalogue of antecedents of failure that disprove, rather than underwrite, the 41 For these instances of symmetry and for others, mostly drawn from the chos ’byungs of Lde’u and Jo sras Lde’u, see Vitali 2004: 109–110, nn. 5–6. On the symmetry of the minor kingdoms and the “regional principalities,” see Dotson 2012: 171.

efficacy of Bon funerals. While using the catalogue form to disprove his opponent’s argument, the polemicist simultaneously mocks the relevance of this form by debasing it. This strategy of disfiguring traditional ritual narrative forms, which were by then associated with the Bon po, also addressed charges that the Buddhists were stealing proprietary ritual narrative technology. With a few nods to Tibetan narrative forms, the “Story of the Cycle of Birth and Death” draws largely on Indian Buddhist narratives. The outlines of its two main sources, the Gaṇḍavyūha sūtra and Sarvadurgatipariśodhana tantra, already include setting, crisis, resolution, and a statement of relevance. The same can be said for any number of Indian Buddhist sutras and tantras, not to mention all sorts of other narratives (Cantwell and Mayer 2008: 293–94). Viewed in the context of Indo-

Tibetan Buddhism, these, like the Tibetan enthusiasm for the myth of Rudra, are examples of how Tibet selectively assimilated and subtly transformed certain elements of Buddhism. From this perspective, Buddhist narrative moved towards the indigenous model, in which narrative was more than just a preliminary to ritual, but constitutive of it. The “Story of the Cycle of Birth and Death” is a product of this movement. By the time of the Zas gtad, such accommodations created a sense of unease among Buddhists, and gave rise to Bon po charges that Buddhists plagiarized their use of ritual antecedent tales. Looked at from the perspective of Tibetan ritual and religion in an inclusive sense, and beginning from our earliest sources, h

owever, texts like the “Story of the Cycle of Birth and Death” and the Zas gtad are periodic signposts of a long process by which the role of narrative in ritual was diminished. While these may be charter myths, and models of exemplary ritual practice to the extent that the Bsgrags pa gling grags targeted them as instances of Buddhist appropriation of Bon po ritual narrative traditions, they do not claim to be templates or models for mimetic ritual in the same was as the ritual antecedent tales do. The general trend, then, appears to be one in which Buddhism’s move from the periphery to the center of Tibetan religious life was accompanied by a diminution of the role of narrative in Tibetan ritual. Like the Tibetan assimilation of Buddhism itself, this was and is a long and incomplete process.

Works Cited ITJ 489 Fragmentary anti-bon polemic (from same text as ITJ 562) ITJ 493 Fragment of “Substitution Text” ITJ 504 Fragment of “Substitution Text” ITJ 562 Fragmentary anti-bon polemic ITJ 731 “Tale of the Separation of Horse and Kiang” ITJ 734 Long scroll containing ransom (glud) rites ITJ 738 Dice divination text ITJ 990 Buddhist catechism/ anti-bon polemic PT 126 Small scroll containing ’Phrul kyi byig shus phyi ma la bstan pa’i mdo and “The envoy of Phyva to Dmu” PT 239 “Substitution Text” PT 1040 “Tale of the Rgyal byin” PT 1060 A ritual text involving horses PT 1068 “Tale of the Yak-Ox Hybrid” (mdzo rabs) PT 1134 Funerary texts, including the “Tale of Lhe’u Yang ka’ rje” PT 1136 Scroll including the “Tale of Blood Brotherhood” and the “Tale of Lho rgyal Byang mo tsun” PT 1285 A long scroll containing healing rituals PT 1286 The “Royal Genealogy,” properly a part of the Old Tibetan Chronicle PT 1287 The Old Tibetan Chronicle

Alexander, Philip S. 2008. Insider/outsider labelling and the struggle for power in early Judaism. In Green, Nile and Mary Searle-Chatterjee, eds, Religion, Language and Power. London & New York: Routledge, 83–100. Bacot, Jacques, Frederick W. Thomas, and Charles-Gustave Toussaint. 1940– 1946. Documents de Touen-houang relatifs a l’histoire du Tibet. Paris: Libraire Orientaliste Paul Geuthner. Bellezza, John Vincent. 2008. Zhang zhung:

Foundations of Civilization in Tibet. Wien: Verlag der Österreichischen Akademie der Wissenschaften. Cantwell, Cathy and Rob Mayer. 2008. Enduring myths: smrang, rabs and ritual in the Dunhuang texts on Padmasambhava. In Pommaret, Françoise and Jean-Luc Achard, eds, Tibetan Studies in Honor of Samten Karmay. Part II – Buddhist and Bon po Studies. (Special Issue, Revue d’Etudes Tibétaines 15), 277–314. Charlot, John. 1977. The application of form and redaction criticism to Hawaiian literature. The Journal of the Polynesian Society 86.4: 479–502. Davidson, Ronald. 2005. Tibetan Renaissance. Tantric Buddhism in the Rebirth of Tibetan Culture. New York: Columbia University Press.

Dotson, Brandon. 2007. “Emperor” Mu-rug-brtsan and the 'Phang thang ma Catalogue. Journal of the International Association for Tibetan Studies 3: 1–25. ——. 2008. Complementarity and opposition in early Tibetan ritual. Journal of the American Oriental Society 128.1: 41–67. ——. 2012. At the behest of the mountain: gods, clans and political topography in post-imperial Tibet. In Scherrer-Schaub, Cristina Anna, ed., 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. Leiden: Brill, 159–204. ——. 2013a. Warlords versus true kings. In Schaeffer, Kurtis, Matthew T. Kapstein, and Gray Tuttle, eds, Sources of Tibetan Tradition. New York: Columbia University Press, 41–43. ——. 2013b. The princess and the yak: the hunt as narrative trope and historical

reality in early Tibet. In Dotson, Brandon, Kazushi Iwao, and Tsuguhito Takeuchi, eds, Scribes, Texts, and Rituals in Early Tibet and Dunhuang. Proceedings of the 12th Seminar of the International Association for Tibetan Studies, Vancouver, 2010. Wiesbaden: Reichert Verlag, 61–85. ——. 2013c. The unhappy bride and her lament. Journal of the International Association for Bon Research 1: 199–225. Eimer, Helmut. 1979. rNam thar rgyas pa. Materialien zu einer Biographie des Atiśa (Dipaṃkaraśrījñāna). 2 Teile, AF 67. Wiesbaden: Otto Harrassowitz. Gaenszle, Martin. 2002. Ancestral

Voices: Oral Ritual Texts and their Social Contexts Among the Mewahang Rai of Eastern Nepal. Münster: LIT Verlag. Hahn, Michael. 1997. A propos the term gtsug lag. In Krasser, Helmut, et al., eds, Tibetan Studies. Proceedings of the 7th Seminar of the International Association for Tibetan Studies, Graz, 1995. Volume 1. Wien: Verlag der Österreichischen Akademie der Wissenschaften, 347–54. Hazod, Guntram. 2009. Imperial Central Tibet: an annotated cartographical survey of its territorial divisions and key political sites. In Brandon Dotson, The Old Tibetan Annals: An Annotated

Translation of Tibet’s First History. With an Annotated Cartographical Documentation by Guntram Hazod. Wien: Verlag der Österreichischen Akademie der Wissenschaften, 161–231. Hill, Nathan. 2010. A Lexicon of Tibetan Verb Stems as Reported by the Grammatical Tradition. München: Verlag der Bayerischen Akademie der Wissenschaften. Imaeda, Yoshiro. 2007. The History of the Cycle of Birth and Death: a Tibetan narrative from Dunhuang. In Kapstein, Matthew T. and Brandon Dotson, eds, Contributions to the Cultural History of Early Tibet. Leiden: Brill, 105– 81.

Kapstein, Matthew T. and Brandon Dotson. 2007. Preface. In Kapstein, Matthew T. and Brandon Dotson, eds, Contributions to the Cultural History of Early Tibet. Leiden: Brill, vi–xii. Karmay, Samten Gyaltsen. 1998 (1983). Early evidence for the existence of Bon as a religion in the royal period. I

n Samten Gyaltsen Karmay, The Arrow and the Spindle: Studies in History, Myths, Rituals and Beliefs in Tibet. Kathmandu: Mandala Book Point, 157–68. ——. 1998 (1991). The soul and the turquoise: a ritual for recalling the bla. In Samten Gyaltsen Karmay, The Arrow and the Spindle: Studies in History, Myths, Rituals and Beliefs in Tibet. Kathmandu: Mandala Book Point, 310– 38. ——. 2010. Tibetan indigenous myths and rituals with reference to ancient Bön texts: the Nyenbum (Gnyan ’bum). In Cabezón, José Ignacio, ed., Tibetan Ritual. Oxford: Oxford University Press, 53–68. van der Kuijp, Leonard. 1996. Tibetan historiography. In Cabezón, José Ignacio and Roger Jackson, eds, Tibetan Literature: Studies in Genre. Ithaca: Snow Lion, 39–

56. Lalou, Marcelle. 1949. Les chemins du mort dans les croyances de Haute-Asie. Revue de l’histoire des religions 135.1: 42–48. ——. 1958. Fiefs, poisons et guerisseurs. Journal Asiatique 246: 157–201. ——. 1965. Catalogues des principautés du Tibet ancien. Journal Asiatique 253: 189–215. Macdonald, Ariane. 1971. Une lecture des Pelliot Tibétain 1286, 1287, 1038, et 1290: essai sur la formation et l’emploi des mythes politiques dans la religion royale de Sroṅ bcan Sgam po. In Études Tibétaines dédiées à la mémoire de Marcelle Lalou. Paris: Adrien Maisonneuve, 190–391. Mills, Martin.

2009. The many bodies of King Songtsen Gampo: constitutional mythology in Tibetan history. Presented at the International Conference: Exploring Tibet’s History and Culture, Delhi, November 19th, 2009. Typescript. van Schaik, Sam. 2013. The naming of Tibetan religion: bon and chos in the imperial period. Journal of the International Association for Bon Research 1: 227–57. Sørensen, Per K. 1994. Tibetan Buddhist Historiography: the Mirror Illuminating the Royal Genealogies. Wiesbaden: Harrassowitz Verlag. Stein, Rolf A. 1970. Un document ancien relatif aux rites funéraires des Bon-po tibétains. Journal Asiatique 258: 155–85. ——. 1971. Du récit au rituel dans les manuscrits Tibétains de Touen-houang. In Études Tibétaines dédiées à la mémoire de Marcelle Lalou. Paris: Adrien Maisonneuve, 479–547.

——. 2010. Rolf Stein’s Tibetica Antiqua, with Additional Materials, trans. Arthur McKeown. Leiden: Brill. Uray, Géza. 1967. Traces of a narrative of the Old Tibetan Chronicle in the Mkhaspa’i dga’-ston. Monumenta Serica 26: 498–505. ——. 1972. Queen Sad-mar-kar’s songs in the Old Tibetan Chronicle. Acta Orientalia Academiae Scientiarum Hungaricae 23.1-3: 5–38. Vitali, R. 2004. The role of clan power in the establishment of religion (from the kheng log of the 9th-10th century to the instances of the dByil of La stod and gNyos of Kha rag). In Cüppers, Christoph, ed., The Relationship

Between Religion and State (chos srid zung ’brel) in Traditional Tibet. Proceedings of a Seminar held in Lumbini, Nepal, March 2000. Lumbini: Lumbini International Research Institute, 105–157. Wangdu, Pasang and Hildegard Diemberger. 2000. Dba' Bzhed: The Royal Narrative Concerning the Bringing of the Buddha’s Doctrine to Tibet. (PhilosophischHistorische Klasse, Denkschriften, Band 291. Beiträge zur Kultur- und Geistesgeschichte Asiens Nr. 37.)

Wien: Verlag der Österreichischen Akademie der Wissenschaften. Zeisler, Bettina. 2011. For love of the word: a new translation of Pelliot tibétain 1287, the Old Tibetan Chronicle, chapter I. In Imaeda, Yoshiro, Matthew T. Kapstein, and Tsuguhito Takeuchi, eds, New Studies of the Old Tibetan Documents: Philology, History and Religion. (Old Tibetan Documents Online Monograph Series, Vol. III.) Tokyo: Institute for the Languages and Cultures of Africa and Asia, 97–213.