1. Introduction
The present paper focuses on one group of Old Tibetan manuscripts from Dunhuang: IOL Tib J (henceforth ITJ) 433 and 434, located in the Stein Collection in London, and Pelliot tibétain (henceforth PT) 179, 247, and 248, located in the Pelliot Collection in Paris. Due to a mix-up in the order of the folios, the manuscripts were separated and have then been stored in different locations. Here they will be reordered and reunited according to recently identified parallel texts in the Tibetan and Chinese Tripiṭaka.
Upon scrutinizing the Tibetan and Chinese Tripiṭaka, it becomes evident that several texts deal with the liturgies for the seven Tathāgatas including Bhaiṣajyaguru,Footnote 2 which encompass the liturgy of reciting (gzungs bklag cho ga) the Bhaiṣajyaguru-sūtra (Sūtra of Bhaiṣajyaguru, henceforth Bhg), and the liturgy of offering to the seven Tathāgatas including Bhaiṣajyaguru (previously known as the “Medicine Buddha”Footnote 3), all sharing a common structure in their rituals.
Bu ston records in his Chos 'byung that Śāntarakṣita authored three versions of the liturgies of the seven Tathāgatas, described as follows:
slob dpon zhi ba 'tshos mdzad pa'i de bzhin gshegs pa bdun gyi mchod pa'i cho ga rgyas 'bring bsdus gsum 'di mdo sde'i phyogs su gtogs par sems so || (Chos 'byung: 180b)
Ācārya Zhi ba 'tsho (i.e. Śāntarakṣita) made the Liturgy of the Venerating of the Seven Tathāgatas [in] three [versions], extensive, medium and brief, which I think should be included in the section of the sūtra.
It can be inferred that the three versions (extensive, medium, and brief) produced by Śāntarakṣita refer to the “original” versions of D 3132, D 3133, and D 3134, respectively.Footnote 4 These are the only existing works in the Tibetan Tripiṭaka that address the same subject matter and have been attributed to Śāntarakṣita.
Before commencing my investigation of the Old Tibetan manuscripts pertaining to the rituals of the Saptatathāgatas (seven Tathāgatas) and identifying the parallel texts in the Tibetan and Chinese Tripiṭaka, I had initially assumed that these liturgies might have been transmitted from Tibet to China. Consequently, the chain of their textual transmission appeared to be: Classical Tibetan → Chinese.
However, this raised questions about the period preceding the authorship of these texts as Classical Tibetan works and their compilation into the Tibetan Tripiṭaka. I became curious about the possible existence of a “pre-canonical” version of these texts. There appeared to be a question mark at the starting point of their textual transmission: ? → Classical Tibetan → Chinese.
To address this uncertainty, I sought to identify extant Old Tibetan manuscripts related to these liturgies that could be dated to the era of the so-called “bstan pa snga dar” (earlier spread of Buddhism). This would allow us to elucidate a “pre-canonical” form covered in these manuscripts, representing the earliest stage of the tradition. This would then enable a deeper understanding of how the worship of a single Bhaiṣajyaguru, as depicted in the Sanskrit text of Bhg, developed into the worship of the seven Tathāgatas including Bhaiṣajyaguru. Moreover, this investigation could shed light on the rituals in these liturgies, their significance, and the broader context of the early introduction of Buddhism into Tibet.
With these questions in mind, I embarked on my investigation of the Old Tibetan manuscripts. Through this research, I hope to have now filled the gap in the chain of textual transmission of these liturgies, and have also addressed several related questions.
For the purposes of this paper, the focus is primarily on the investigation of the Old Tibetan manuscripts concerning the liturgies for the seven Tathāgatas and their connection to canonical texts, particularly in the Tibetan and Chinese Tripiṭaka.
2. Previous studies of IOL Tib J 433 and other Old Tibetan manuscripts
In Louis de La Vallée Poussin's catalogue (de La Vallée Poussin Reference De La Vallée Poussin1962: 140–41), ITJ 433 is described as comprising folios 9–100, with dimensions of 6.1 × 24.5 cm, presented in the form of a poṭhī, and folios are scribed by two different hands. This manuscript contains the pūjā (veneration) to the Saptatathāgatas.
As for other Old Tibetan manuscripts concerning the same topic, Lalou (Reference Lalou1939: 68–9) identified PT 248 as a fragment of the abhiṣeka ritual of Tathāgata Bhaiṣajyaguruvaiḍūryaprabha. However, she did not provide specific information about the content of PT 247 that seemed to be related.
Later, Dalton and van Schaik (Reference Dalton and van Schaik2006: 177–8) observed the correlation between ITJ 433 and the canonical text Q 135 [= D 503 De bzhin gshegs pa bdun gyi sngon gyi smon lam gyi khyad par rgyas pa zhes bya ba theg pa chen po'i mdo, Sūtra of the Detailed Account of the Previous Aspirations of the Seven Tathāgatas], Q 3955 [= D 3134], ITJ 434 and Pelliot tibétain 179, which are texts about the former aspirations or vows of the seven Buddhas (De bzhin gshegs pa bdun gyi sngon gyi smon lam, Saptatathāgatapūrvapraṇidhāna). They also mentioned that an additional folio from ITJ 433 can be found in ITJ 434, but their attempt to insert ITJ 434 after “ITJ 433/1: 35v” (folio 35 of ITJ 433) led in the wrong direction, and they expressed uncertainty about the nature of the correspondence with the canonical version: “once again the nature of the correspondence with the canonical version is obscure” (Dalton and van Schaik 2006: 178).
In his study, Schopen (Reference Schopen1978: 73) stated that there are “only two small fragments of a Tibetan translation of this ‘text’ [i.e. *StP = D 503; my addition] from Tun-huang [i.e. Dunhuang]”, and he was unsure “whether these fragments come from Bhg (D 504 bCom ldan 'das sman gyi bla bai ḍūrya'i 'od kyi sngon gyi smon lam gyi khyad par rgyas pas zhes bya ba theg pa chen po'i mdo) or *StP [D 503]”.
While these previous studies have provided important information, the correspondences they have pointed out between the Old Tibetan manuscripts and the Tibetan canonical texts can now be clarified. The correspondence between ITJ 433 and the canonical texts Q 135 [= D 503] and Q 3955 [= D 3134] does not appear convincing, and the link between ITJ 433 and ITJ 434 is incorrect. Moreover, those studies have not explored the correspondences between the manuscripts and the Chinese canonical texts.
Here, I will rectify Lalou Reference Lalou1939 and Schopen Reference Schopen1978 by identifying PT 247 and PT 248. Furthermore, I will reunite other manuscripts related to the same topic that have been separated in the Stein Collection and Pelliot Collection. It should be noted that both of Schopen's presumptions (Bhg and *StP) are misleading, since the two fragments he mentioned (PT 247 and PT 248) belong together with ITJ 433, rather than being Bhg and *StP as presumed. And PT 247 is, in fact, just a missing folio of ITJ 433.
The field of studies on the sūtra and the worship of the seven Tathāgatas including Bhaiṣajyaguru is extensive and diverse. There are now many excellent studies which are not, however, directly relevant to the focus of this paper and therefore will not be discussed here in detail. For those interested in exploring further research on this subject, Schopen's Reference Schopen1978 work serves as an excellent starting point. It offers a comprehensive overview of the available sources and scholarly interpretations, shedding light on various aspects of this field of study.
3. Remarks after the comparison of parallel texts
3.1. Remarks on the manuscripts
After the correct parallel texts of the Old Tibetan manuscripts had been identified, it became evident that the sequence of the folios was in disorder, with some folio numbers evidently being wrong. So, with the help of the parallel texts, the “original” order of the folios of the manuscripts is reconstructed, just like putting a chaotic puzzle together piece-by-piece to get a full view of the picture.
Normally, the folio numbers offer reliable information about the order of a manuscript. However, in the case of ITJ 433, the folio numbers were written in a very cursive way, leading to mistakes in their sequence. For instance, the folio number 33 (so sum) was mistakenly written after 62 (ro nyis), most likely due to the similarity of the characters so in 33 (so sum) and ro in 63 (ro sum). Consequently, the folio order of ITJ 433 is as follows: [61 missing], 62, and then it continues with 33 after 62, i.e. [61 missing], 62, 33, 34. The folio numbers appear to have been added later and were incorrect.
Furthermore, it is clear now that PT 247 in Lalou's catalogue belongs to the same manuscript as ITJ 433, and it can even be identified as one of the missing folios of ITJ 433, i.e. folio 45. The majority of the folios of PT 179 could also be inserted into ITJ 433 as its missing folios.Footnote 5 Therefore, the “original” folio numbers of the whole text might be as follows: 3, [4–21],Footnote 6 22, [27], 28–30, [31], PT 179 (32), [33, 34], ITJ 433 (35), 36, [37], PT 179 (38), ITJ 433 (39), 40–44, [45], 46–49, 52r,Footnote 7 50v, [51], 52, [53], PT 179 (54), ITJ 433 (55), [56–8], 59, 60, [61], 62, 33 [sic], 34, [35, 36], PT 179 (37), [38–40], ITJ 433 (41), [42], 43, PT 247 (44), [45], 46–48, [49], 50–53,Footnote 8 55, [56], 57, [58, 59], 60, [61], 62, [63], 64–74, ITJ 434 (75), [76], ITJ 433 (77), (no number = 78), 79–83, [84], 85–91.Footnote 9
There is one potential problem with the pagination. According to the parallel canonical text, the folios 90 (dgu bcu thaṃ), 91 and 92 [missing] of PT 179 could be inserted at the very beginning of ITJ 433, followed by ITJ 433 folios 93‒99 and folio 3. It could also be possible that after folio 91 of ITJ 433, which is the end of the liturgy according to the canonical version, there is a further text dealing with the preparation of the liturgy, corresponding to folio PT 179 (folio 90), 91, [92], ITJ 433 (93), 94–99 and 3. These two assumptions obviously require further evidence and explanation.
If we put the Sanskrit sūtra, i.e. Bhg, into the history of the textual transmission of worshipping Bhaiṣajyaguru, a process of liturgicalization from sūtra to practice is evident. The relations between the Old Tibetan manuscripts, the canonical versions, and the “foundational text”, i.e. the Sanskrit text of Bhg, have been established as follows (see Table 1).
Overview of the structure and relationship between sūtra, liturgy, and practice
3.2. The result of the comparison
The comparison has yielded significant findings. It is now evident that the majority of the manuscripts deal with the liturgy of the seven Tathāgatas including Bhaiṣajyaguru, and one Tibetan parallel text in the Tibetan Tripiṭaka has also been found. ITJ 433 (or, more precisely, most of itFootnote 10) contains an Old Tibetan version of D 3132 De bzhin gshegs pa bdun gyi sngon gyi smon lam gyi khyad pa rgyas pa'i mdo sde'i man ngag (Personal Instructions on the Sūtra of the Detailed Account of the Previous Aspirations of the Seven Tathāgatas). Additionally, since D 3133 shares most of its passages with D 3132, we can consider a major part of D 3133 as derived from the Old Tibetan version represented by ITJ 433. This Old Tibetan version later “developed” into the Classical Tibetan version D 3132, as evident in the Tibetan Tripiṭaka, and was also copied and preserved in Dunhuang (see Figure 1).
Presumed process tree of D 3132 and ITJ 433
From the perspective of textual transmission, a series of texts containing the liturgical method (chog sgrigs, jin gui 經軌; or vidhi, cho ga, yi gui 儀軌 ) of the seven Tathāgatas can also be traced back to a common source, i.e. an Old Tibetan version containing the same topic as ITJ 433. These texts share almost the same structure and ritual procedures, leading us to presume that the tradition of worshipping the seven Tathāgatas originated from this Old Tibetan version found in ITJ 433 and other Old Tibetan manuscripts.Footnote 11 Notable texts in this tradition include Panchen IV bLo bzang Chos kyi rgyal mtshan's (1570–1662) bCom ldan 'das sman bla'i mdo chog snying po bsdus pa yid bzhin nor bu (Bojiafan yaoshi jinggui xinyao lüeshe ruyi moni 薄伽梵藥師經軌心要略摄⋅如意摩尼, Extracted Essential Liturgical Method of Bhaiṣajyaguru, the Wish-fulfilling Jewel),Footnote 12 Ye shes rgyal mtshan's (1713–93) sMan bla'i mdo mchog gi snying po'i chog sgrigs rin po che'i them skas (Yaoshi jinggui xinyao huibian baoti 藥師經軌心要彙編．寶梯, Compiled Essential Liturgical Method of Bhaiṣajyaguru, the Jewel-ladder), ʼjam mgon 'ju Mi pham rgya mtsho's (1846–1912) sMan bla'i mdo chog bsdus pa bdud rtsi'i bum bzang (Yaoshi jingwang luexiu yigui ganlu shengping 藥師經王略修儀軌 ⋅ 甘露圣瓶, Extracted Liturgical Method of Bhaiṣajyaguru, the Auspicious Vessel of Amṛta) and so forth.
3.3. The purpose of ITJ 433 / D 3132, 3133 and their affiliation
The purpose of authoring the text becomes evident from the colophon of D 3132:
dbang phyug dam pa'i mnga’ bdag dpal lha btsan po lha sras khri srong lde'u btsan gyi sku tshe bsring ba dang | dbu rmog btsan pa dang | chab srid mtho ba dang | las sgrib sbyang ba dang | tshogs gnyis spel ba'i ched du | de bzhin gshegs pa bdun gyi sngon gyi smon lam gyi khyad par rgyas pa'i mdo sde'i man ngag | slob dpon bo dhi satvas mdzad || (rgyud, pu 275a4–275a5.)
For the sake of the longevity of the Lord of the noble lords, the Auspicious Heavenly bTsan po, the Heavenly Son Khri Srong lde'u btsan (r. c. 754–97), the might of [his] helmet [= authority], the supremacy of the reign, the purification of [his] stained [= bad] karman, and the development of the two accumulations [of his gnosis and merit]… Ācārya Bodhisattva made the Personal Instructions on the Sūtra of the Detailed Account of the Previous Aspirations of the Seven Tathāgatas. (Slightly changed based on van der Kuijp Reference Van der Kuijp2004: 5.)
It is also worth mentioning that mKhas grub rje noted that:
[…] because the ācārya Śāntarakṣita has composed a rite [based on them] (Toh. 3133, no author listed).Footnote 20 He composed it consistent with the structure of Kriyā-Caryā rites, in that he has the preliminaries of observing the Sabbath (upoṣadha), and so on. […] (tr. Lessing and Wayman Reference Lessing and Wayman1978: 109).Footnote 21
Given the above information, we are informed that Ācārya Śāntarakṣita composed D 3133 following the structure of Kriyā-Caryā (bya spyod, action-tantra) rites, which implies that it might not be unreasonable to categorize D 3132 similarly, suggesting that D 3132 also belongs to Kriyā-Caryā (bya spyod) rites.
From the perspective of practitioners, there is a need to put the sūtra (in this case, the Sanskrit Bhg) into practice, which necessitates the creation of instructions that are more suitable and comprehensible to practitioners. Thus, various forms of elucidation have been developed, such as commentaries (vṛtti, 'grel ba), expositions (bhāṣya, bshad pa), perpetual commentaries (pañjikā, dka' 'grel), personal instructions (upadeśa, man ngag), manuals (vidhi, cho ga), liturgical method (chog sgrigs), ritual for sūtra (mdo chog) and so forth.
4. The liturgicalization of the texts
This section focuses particularly on the process of worshipping the seven Tathāgatas that is covered in ITJ 433 and other Old Tibetan manuscripts. It shows the difference between the sūtra itself and the rituals, i.e. how the Bhaiṣajyaguru-sūtra was transformed into a religious practice of worship through liturgy, which I would like to define as “liturgicalization”. That is to say, the sūtra itself is liturgicalized into religious practice.
As for the “ritual”, I follow the definition given by Tambiah:
Ritual is a culturally constructed system of symbolic communication. It is constituted of patterned and ordered sequences of words and acts, often expressed in multiple media, whose content and arrangement are characterized in varying degree by formality (conventionality), stereotypy (rigidity), condensation (fusion), and redundancy (repetition) (Tambiah Reference Tambiah1985: 128).
As we see later in the main structure of ITJ 433, rituals such as the four empowerments and the act of paying homage to and taking refuge in the seven Tathāgatas align well with Tambiah's definition.
In this paper, I try to distinguish between the terms “ritual” and “liturgy”. To be exact, the term “ritual” is used to designate performance-centred rituals, while “liturgy” designates the entire text as a manual for worship (pūjā). The liturgy includes a series of well-choreographed rituals and certain collective formulas for the conduct of worshippers. As for the term “practice”, I take it in a narrow sense of “religious practice”.
4.1. The main structure of the text as liturgy
In light of the previous explanation, the text of ITJ 433 (thus also D 3132) is considered to be the liturgy of the seven Tathāgatas, structured as follows:
I. The beginning/preparation (pūrvagama, sngon 'gro, qian xing 前行)
i. Taking a bath
ii. Keeping fasting
iii. Arranging the maṇḍala
iv. Generating the four immeasurables (catvāry apramāṇāni, tshad med pa bzhi, si wuliang 四無量).Footnote 22
II. The main action (spyod pa, zhengxing 正行)
i. Four empowerments (adhiṣṭhāna, byin gyis rlob pa, jiachi 加持)
a. Empowerment of the ground (sa gzhi byin gyis rlabs pa, jiachi dadi 加持大地)
b. Empowerment of the palace (khang bzang byin gyis rlabs pa, jiachi gongdian 加持宫殿)
c. Empowerment of the teaching thrones (gdan khri byin gyis rlabs pa, jiachi fazuo 加持法座)
d. Empowerment of the offerings (mchod pa byin gyis rlabs pa, jiachi gongpin 加持供品)
ii. Inviting the noble ones (spyan 'dren pa, yingqing shengzhong 迎請聖眾)
iii. Paying homage to and taking refuge in the “Three Jewels” and expressing aspirations
a. Paying homage to and taking refuge in the Buddhas, i.e. the seven Tathāgatas including Bhaiṣajyaguru and Śākyamuni
b. Paying homage to and taking refuge in the dharma
c. Paying homage to and taking refuge in the saṃgha (including Mañjuśrī, Trāṇamukta, Vajrapāṇi, the 12 Great Generals of Yakṣas)
III. The end (mjug, jiexing 結行)
i. Offering
ii. Confession
iii. Sending the noble ones back to their places
iv. Dedication
Some of the rituals above (II. iii and III. i, ii, also listed in Table 1) roughly correspond to the structure of the saptavidhānuttara-pūjā (the seven-branch offering, Tib. yan lag bdun pa, Chin. qi zhi gong 七支供).Footnote 23 The standard form consists of the following seven branches:
i. Obeisance (vandanā): Reciting the formula of the Three Refuges and praising the excellent qualities of the Buddha, Dharma, and Saṅgha.
ii. Offering (pūjana): It includes elaborate offerings to each of the senses and, in tantric rituals, so-called inner and secret offerings.
iii. Confession of wrongdoing (pāpadeśanā): Acknowledging and repenting any unvirtuous deeds.
iv. Admiration or rejoicing (anumodana): Taking delight in and admiration for the merits and virtues of others.
v. Requesting the Buddhas to turn the wheel of dharma (dharmacakrapravartanacodana): Beseeching the Buddhas to teach the Dharma for the benefit of all beings.
vi. Requesting the Buddhas not to pass into parinirvāṇa (aparinirvṛtādhyeṣaṇa): Beseeching the Buddhas to remain in the world and continue benefitting all sentient beings.
vii. Dedication of merit (pariṇāmanā): Mentally or ritually dedicating/directing the accumulated merit (puṇya) to the welfare and enlightenment of all beings (based on PDB 2014: 776, 52, etc.).
This sevenfold ritual form is “a common liturgical legacy that cut across the bounds of individual sectarian and cult interests and functioned as an integrative basis for Chinese Buddhist devotion as a whole” (Stevenson Reference Stevenson1987: 436). It is deeply rooted in the Indian tradition,Footnote 24 and later became a standard practice in many Mahāyāna traditions. It also became a common element in tantric pūjās.
The origins of the seven-branch offering are closely connected with the Buddhāvataṃsaka-mahāvaipulya-sūtra (The Sūtra of the Ornament of the Buddhas, Chin. T 293 Dafangguang fo huayan jing 大方廣佛華嚴經, Tib. Sangs rgyas phal po che zhes bya ba shin tu rgyas pa chen po'i mdo), particularly with the chapter known as “the chapter of Bodhisattva Samantabhadra” (Puxian pusa xingyuan pin 普賢菩薩行願品), which covers the Ten Aspirations of Bodhisattva Samantabhadra (Puxian(pusa) shida xingyuan 普賢(菩薩)十大行願):Footnote 25
(1) to worship the buddhas 禮敬諸佛, (2) to praise the tathāgatas 稱讚如來, (3) to make offerings 廣修供養, (4) to repent sin 懺悔業障, (5) to rejoice in the merits attained by others 隨喜功德, (6) to ask the buddhas to teach 請轉法輪, (7) to ask the buddhas to stay in the world 請佛住世, (8) to follow the buddhas for study 常隨佛學, (9) to be friends with all beings 恆順衆生, and (10) to devote one's merits to the salvation of others 普皆廻向 (DDB).
The following paragraphs will provide details of the rituals that are contained in ITJ 433 and other Old Tibetan manuscripts.
4.2. The popular ritual: prolonging life with banners and lamps
It is noteworthy that mchod pa'i sbyor ba ni (“the method of offering”) is added by the author of the text to specify the exact method conducting this ritual to achieve one of the prominent purposes in the mundane world, i.e. to prolong life. The sentences following mchod pa'i sbyor ba ni were named by the practitioners as “prolonging life with banners and lamps”, and achieved extreme popularity:Footnote 26
The same ritual, with identical details on how to offer to Bhaiṣajyaguru, is also present in another Old Tibetan manuscript fragment, PT 248:
PT 248: (a1) dge 'dun la zas dang skom dang | yo byad thams chad cI nus su mchod cIng g.yog byos shig | (a2) nyin gcig la lan gsuṃ bcom ldan 'das sman gyi bla be du rya'i 'od gyi rgyal po de bzhin gshegs pa'i mtshan (a3) yid la dran bar byos shig || mye mar bzhi 〇 bcu rtsa dgu bud pas la | gzugs re re la yang mye mar bdun (a4) bdun zhog ste | mye mar gcig gyang shing 〇 rta'i 'khor lo tsham du byos la || zhag bzhi bcu rtsa dgu'i (a5) bar du ơd ma zhil bar bya || tshon sna lnga la lha'I ba dan bzhi mtho bzhi bcu rtsa dgu pa byos shig ||
a2. nyin gcig la lan gsuṃ] nyin mtsan lan gsuṃ du ITJ 433, nyin lan gsum mtshan lan gsum du D.
4.3. Other rituals recorded in ITJ 433 and ITJ 434
4.3.1. Offering to the seven Tathāgatas
In addition to the rituals of offering banners and lamps, ITJ 433 and ITJ 434 contain several other rituals of worshipping the seven Tathāgatas. They involve the offering of perfumes, incense, oil lamps, garlands, ornaments, songs, musical instruments, drums, and chanting the dhāraṇī/mantra, among other practices.
Regarding the placement of ITJ 434, it has been clarified above that it is not to be inserted after ITJ 433/1:35v. Instead, it corresponds to one of the missing folios of ITJ 433, namely, to folio 75 (don lnga).
4.3.2. Empowerment of the palace
In the ritual of the empowerment of the palace, which is part of the preparation process, the comparison is made between the Old Tibetan manuscript ITJ 433 and the canonical version D 3132. It becomes evident that D 3132 contains more detailed information and is longer than ITJ 433. For instance, the term “sgo gsum” (three gates) in ITJ 433 is fully explained in the canonical version D 3132, where it is described as “stong pa nyid” (emptiness), “mtshan ma med pa” (signlessness), and “smon pa med pa” (wishlessness). The additional details in D 3132 provide a more comprehensive understanding of the ritual of the empowerment of the palace compared to ITJ 433:
4.4. Adding of mantras
Mantras hold significant importance and are utilized for a wide range of purposes in various cultural and religious contexts. They are considered sacred and potent in many spiritual traditions, including Buddhism, Hinduism, and other belief systems. Just as Burchett (Reference Burchett2008) pointed out that “[H]ealers chant mantras to cure illness; a worker may mutter a mantra as he lifts a heavy load; others may speak mantras in order to attain success in business or love, to protect themselves from evil spirits and enemies, or to ensure conception or safe childbirth”.
These mantras are integrated into the liturgy to bestow blessings, invoke the power of the Tathāgatas, and aid practitioners in their spiritual journey. The repetitive chanting of mantras during the ritual plays the role of creating a sacred and transformative experience for the worshippers, enhancing their connection with the divine and facilitating the fulfilment of their aspirations and wishes.
There are several mantras included in ITJ 433, D 3132, and D 3133. Here I present three of them. One of the mantras present in ITJ 433 is known as “Mantra of dispelling the karmic veil said by the Buddha Stainless Gold” (jinse baoguang miaoxing chengjiu rulai shuo chuyezhang shenzhou 金色寶光妙行成就如來說除業障神咒), but is also named “rDo rje'i ri 'joms pa” (Destroying the Diamond Mountain) in the Tibetan Tripiṭaka and po jinggangshan 破金剛山 in the Chinese Tripiṭaka, also recorded in the Tangut manuscript from Khara-khoto (hei shui cheng 黑水城, lit. the city of black water). This mantra has been traced back to various Chinese, Tibetan (both in Old Tibetan manuscripts from Dunhuang and Tibetan canonical texts), Tangut, and Sanskrit parallels, not just limited to those related to the seven Tathāgatas. It is found in different contexts and used for various purposes such as healing, protection, and success in different endeavours. The same mantra was also transcribed from Chinese into Tangut in Инв. №. 885: 
tsə̣1 dzji̱j2 nia̱2 ɣjiw2 bji1 śjạ1 tha1 ˑjij1 mer2 tjị1 tśhja2 ˑio̱w1 lwer2 lhejr2 phju2 ？ (*Chin. Yaoshi liuliguang qifo zhi benyuan gongde jingdian shangjuan 藥師琉璃光七佛之本願功德經典上卷, Sūtra of the Detailed Account of the Previous Aspirations of the Seven Buddhas including Bhaiṣajyaguruvaiḍūryaprabha, first fascicle (juan)).
The mantra can also be found in some other texts which are unconnected with the seven Tathāgatas, such as T 433 Wuqian wubai foming shenzhou chuzhang miezui jing 五千五百佛名神呪除障滅罪經 (Sūtra of Five Thousand Five Hundred Buddha Names and Dispelling Harm and Extinguishing Sin) and its Tibetan parallel D 262 Sangs rgyas kyi mtshan lnga stong bzhi brgya lnga bcu rtsa gsum pa (*Buddha-nāma-sahasrapañcaśatacaturtripañcadaśa); T 665 Jinguangming zuishengwang jing 金光明最勝王經 (Sūtra of Golden Light, or Sūtra of Supreme Golden Light, the Sovereign King) with its Tibetan parallel D 555 'Phags pa gser 'od dam pa mchog tu rnam par rgyal ba'i mdo sde'i rgyal po theg pa chen po'i mdo, D 556 'Phags pa gser 'od dam pa mdo sde'i dbang po'i rgyal po shes bya ba theg pa chen po'i mdo; T 982 Fomu dakongque mingwang jing 佛母大孔雀明王經 (Sūtra of the Great Peahen, Queen of Mantras), T 984 孔雀王呪經 (Sūtra of the Peacock King Mantra), T 985 Foshuo dakongque zhouwang jing 佛說大孔雀呪王經 (The Buddha Speaks of the Sūtra of the Great Peacock Mantra King) with their Tibetan parallel D 559 Rig sngags kyi rgyal mo rma bya chen mo and the Sanskrit parallel Mahāmāyūrīvidyārājñī.
The second mantra is the “Mantra of the beryl light producing the samādhi power of the Tathāgata” (De bzhin gshegs pa'i ting nge 'dzin gyi stobs bkyed pa'i bai ḍūrya'i 'od ces bgyi ba'i gzungs, rulai dingli liuliguang zhou 如來定力琉璃光咒), depicted in folio 58–9, then repeated in folio 63–4:
The third example is the “Mantra of remembrance spoken by Vajradhara, Īśvara, Indra and the Four Great Kings” (zhijinggang shifan sitianwang shuo yinian zhou 執金剛、釋梵四天王說憶念咒):
The presence of various versions and transcriptions of the same mantra across different texts and languages is not uncommon in religious and spiritual traditions. The translation and transcription of mantras from one language to another, especially when dealing with esoteric or sacred texts, can be challenging and may lead to variations in expression. In the case of the “Mantra of being free from all troubles and being fulfilled what is wishing for, spoken by the Bodhisattva Vajradhara” (zhijinggang pusa shuo ling wuwei zhongnao suoyuan manzu zhou 執金剛菩薩說令無眾惱所願滿足咒), it appears that different translators and scribes have approached its rendering differently. It is noticeable that Shaluoba 沙羅巴 has transcribed, rather than translated, one sentence of the Tibetan text D 3133 in T 925, which thus seems quite esoteric:
ITJ 433: yang dag par rdzogs pa'i sangs rgyas bdun po dag la phyag 'tsal lo || rdo rje 'dzin thams (73r3) chad la phyag 'tsal lo ||
Pay homage to seven fully awakened Buddhas! Pay homage to all Vajradharas!
This sentence in ITJ 433 is the same in D 3132 and D 3133, but when transcribed into Chinese in T 925, Shaluoba 沙羅巴 took a different approach and condensed the sentence into a mantra-like form: nanwu saduonan sanmiao sanputuonan nanwu saliva(erhe) variluo(erhe)tuonan “南無馺多喃 三藐三菩陀喃 南無薩哩嚩(二合)嚩日囉(二合)陀囉喃”, which seems to have been influenced by a Sanskrit sentence: *namaḥ saptānāṃ samyaksaṃbuddhānāṃ namaḥ sarvavrajadharāṇām.Footnote 29 This transcription may have been an attempt to encapsulate the essence of the original sentence in a mantra format suitable for recitation and meditation. Comparably, the translator of T 927 Yaoshi qifo gongyang yigui ruyiwang jing 藥師七佛供養儀軌如意王經 (Sūtra of the Ritual Procedure for Making Offerings to the seven Buddhas including Bhaiṣajyaguru, the Wish-Fulfilling King),Footnote 30 Gongbuzhabu 工布查布 (mGon po skyabs), translated the same sentence in the bDe gshegs bdun gyi mchod pa'i chog bsgrigs yid bzhin dbang rgyal (Ritual Procedure for Making Offerings to the Seven Sugatas, the Wish-Fulfilling King, henceforth DChYB) directly into normal Chinese: dingli qizun zhengdeng zhengjue dingli yiqie zhijingang “頂禮七尊正等正覺。頂禮一切執金剛”, which shows some similarities in sound and structure to the mantra-like form in T 925. Apparently Shaluoba 沙羅巴 took T 451 into consideration when he was translating T 925, since Yijing 義凈 did almost the same thing in T 451: nanmo saduonan sanmiao sanfotuonan nanmo sapobazheluo daluonan 南麼馺多喃三藐三佛陀喃南麼薩婆跋折囉達囉喃 (with only minor variations in sound change):
It is possible that Shaluoba 沙羅巴 and other translators faced difficulties in translating certain passages, especially those that contained mantras or ritual elements, which are often highly symbolic and context-specific. The use of transcription without translation may also serve to acknowledge the uniqueness and sanctity of the original language, allowing the text to retain its ritual power and significance. It may also be an invitation for those who are initiated or trained in the specific tradition to interpret and understand the text within its intended religious or ritual framework.
These variations in transcriptions and translations of mantras highlight not only the flexibility and adaptability, but also the susceptibility to errors in the transmission of sacred texts in different linguistic and cultural contexts. They also illustrate the importance of preserving the essence and meaning of these sacred verses while accommodating the nuances of different languages and scriptural traditions.
5. Concluding remarks
The findings of the two-dimensional textual analysis have provided valuable insights into the transmission of the liturgies of the seven Tathāgatas including Bhaiṣajyaguru. Based on this analysis, the following two preliminary conclusions have been drawn:
The Old Tibetan material discovered in the manuscripts ITJ 433, ITJ 434, PT 179, and PT 248 serves as a crucial and indispensable, but previously unnoticed, link in understanding the textual history of these liturgies, which fills a blank in the chain of the textual transmission of the liturgies of the seven Tathāgatas. It has revealed that the transmission chain of these liturgies is Old Tibetan → Classical Tibetan → Chinese, rather than the previously assumed Classical Tibetan → Chinese.
Moreover, the Old Tibetan language exhibits distinct features that are not present in the Classical Tibetan used in the Tibetan canonical version.Footnote 31 The Old Tibetan version appears to be an earlier and more condensed form of the liturgy compared to the longer and more elaborate canonical version. This suggests that the liturgy underwent a process of liturgicalization over time, with the addition of blessing procedures, reciting procedures, and various mantras to the foundational text of the sūtra. The Old Tibetan manuscripts shed light on the liturgicalization of the sūtra, indicating that the tradition of worshipping the seven Tathāgatas was already established in early Tibet. This might also explain that there are no witnesses of any Old Tibetan version of D 503 and D 504, the official Tibetan translations of BhgSkt and *StPSkt, since these two texts have already been transformed into liturgies such as D 3132 and so forth.
Roughly corresponding to the structure of the seven-branch offering (saptavidhānuttara-pūjā), the addition of the blessing procedure, the reciting procedure, and several mantras to the foundational text of the sūtra point to a liturgicalization of the sūtra. This liturgicalization, as seen in ITJ 433, D 3132 and other related texts, provided a practical and accessible way for practitioners to conduct their religious practices. So, the chain of transmission would be like this: sūtra + saptavidhānuttara-pūjā (or similar liturgical forms) → liturgy → practice. It allowed them to follow prescribed procedures and recite mantras, making the worship of the seven Tathāgatas more straightforward and meaningful. As a result, these liturgies became instrumental in the worship and practice of the seven Tathāgatas, facilitating the transition of the sūtra into the context of worship and devotion. Through liturgicalization, the Bhaiṣajyaguru-sūtra was transformed into worship and practice on its way from ancient India to Tibet.
This research has illuminated a dynamic and trans-regional transmission of the worship of the seven Tathāgatas, with the Old Tibetan material playing a significant role in understanding the early stages of this transmission. This study enriches our understanding of the religious and textual history of this specific worship and its development in the Tibetan context, and even in the broader context of Indo-Tibetan Buddhism. It also emphasizes the importance of considering diverse linguistic sources and their ritual implications in the study of religious texts and practices.
6. Abbreviations
- Bhg
Bhaiṣajyaguru-sūtra
- D
sDe dge edition of the Tibetan Tripiṭaka bKa’ 'gyur and bsTan 'gyur
- DChYB
bDe gshegs bdun gyi mchod pa'i chog bsgrigs yid bzhin dbang rgyal
- DDB
Digital Dictionary of Buddhism (via http://www.buddhism-dict.net.emedien.ub.uni-muenchen.de/ddb/)
- ITJ
IOL Tib J, Stein's Collection of Tibetan Dunhuang Manuscripts preserved at the British Library in London (formerly in the India Office Library (IOL), via IDP (International Dunhuang Project): http://idp.bl.uk)
- Инв.
Eluosi, KDYSF, Zhongguo, SKMY. ed. 2015. Heishuicheng Manuscripts Collected in the Institute of Oriental Manuscripts of the Russian Academy of Sciences 俄藏黑水城文獻 Ecang Heishuicheng Wenxian. Shanghai: Shanghai Chinese Classics Publishing House 上海古籍出版社.
- OTDO
Old Tibetan Documents Online (https://otdo.aa-ken.jp)
- P (Q)
Peking (Qian Long) edition of the Tibetan Tripiṭaka bKa’ 'gyur and bsTan 'gyur
- PDB
The Princeton Dictionary of Buddhism. Robert E. Buswell and Donald S. Lopez (eds). Princeton: Princeton University Press, 2013.
- PT
Pelliot tibétain, Pelliot's Collection of Tibetan Dunhuang Manuscripts preserved at the Bibliothèque Nationale in Paris (via https://gallica.bnf.fr)
- *StP
*Saptatathāgata-pūrvapraṇidhāna-viśeṣa-vistāra-sūtra
- T
Taishō shinshū daizōkyō (大正新修大藏經) (via CBETA: https://cbetaonline.dila.edu.tw/zh/)
- Chin.
Chinese
- Skt.
Sanskrit
- Tib.
Tibetan
- Tang.
Tangut
- tr.
translation
- add.
added (addit)
- om.
omitted (omittit)
7. Sigla
- *
reconstruction
- ṃ
anusvāra (rje su nga ro)
- $
page initial sign mgo yig
- I
gi gu rlog
- |
daṇḍa
- ||
double daṇḍa
- {{ }}
erased/deleted by the scribe
- { }
deleted by the editor
- << >>
added by the scribe
- < >
added by the editor
- 〇
string-hole
- []
supplements; illegible or disappeared, but supplied by the editor
Open access
