Hostname: page-component-745bb68f8f-f46jp Total loading time: 0 Render date: 2025-02-11T07:05:36.843Z Has data issue: false hasContentIssue false

The evolution of the Buddhist rakṣā genre in the light of new evidence from Gandhāra: The *Manasvi-nāgarāja-sūtra from the Bajaur Collection of Kharoṣṭhī Manuscripts

Published online by Cambridge University Press:  15 May 2014

Ingo Strauch*
Affiliation:
Université de Lausanne, Switzerland
Rights & Permissions [Opens in a new window]

Abstract

This article examines the *Manasvi-nāgarāja-sūtra, a unique text in the rakṣā genre of Buddhist literature dating to the early centuries of the common era. In addition to exploring special features of the vocabulary and meaning of this text, the article places the *Manasvi-nāgarāja-sūtra in the wider context of Buddhist textual and ritual practice in the first, second and third centuries ce.

Type
Articles
Copyright
Copyright © SOAS, University of London 2014 

The discovery of a great number of Gāndhārī manuscripts in the 1990s and early 2000s changed our picture of early Buddhism in many ways. Texts from numerous genres of Buddhist literature are represented in the corpus in the form of very early manuscripts that can be dated between the first century bce and the second or – in the case of the Bamiyan fragments – the third century ce.Footnote 1 There is no need to stress the importance of these early attestations for our knowledge of rather early forms of Buddhism, standing as they do on the boundary between oral tradition and written culture, and in a crucial period of Buddhist history when it was expanding dramatically from South Asia to Central and East Asia.

Among these new manuscript collections (the Bajaur Collection of Kharoṣṭhī manuscripts, see Strauch Reference Strauch2008), is one text (Fragment 3) that can be attributed to the class of rakṣā or protective texts. As far as we know, this fragment represents the earliest preserved remains of such a text. This fact alone promises new insights into the early history of this genre. Before I describe this new text and its impact on our knowledge of early rakṣā literature, I would like to highlight briefly some of the known features of this genre in order to set the frame against which the new Gāndhārī text has to be interpreted.

Early representatives of Buddhist rakṣā literature

In a ground-breaking article Peter Skilling (Reference Skilling1992) described a special class of Buddhist texts for which he introduced the term “rakṣā literature”.Footnote 2 This type of literature consists of texts which are meant to bestow magical power and to protect the persons who command them from a series of calamities. Some of these texts were composed long after the Buddha's death, while others seem to belong to the earliest strata of Buddhist literature. The Theravāda Buddhists subsume these texts under the label paritta, a Pali word based on Old Indian pari-√trā “to protect”. The synonymous term rakṣā (Pali rakkhā) is found both in Theravādin and non-Theravādin traditions and is, according to Skilling, preferable for the designation of this group of texts. In the tradition of the Mūlasarvāstivādins these and some other texts are assembled under the term mahāsūtra or “Great sūtra” (cf. Skilling Reference Skilling1992: 125–9; Reference Skilling1994; Reference Skilling1997).

The idea behind these texts can be connected with one of the most important features of the Buddha – his ability to grant security and protection. One text from the Pali Dīghanikāya, the Soṇadaṇḍa-sutta, expresses this characteristic feature with the words:

Samaṇo khalu bho gotamo yasmiṃ gāme vā nigame vā paṭivasati, na tasmiṃ gāme vā nigame vā amanussā manusse viheṭhenti.

(DN I 116.14)

In whatever town or village the samaṇa Gotama stays, non-humans do not harm the people of that town or village.

(Skilling Reference Skilling1992: 111).

In the rakṣā literature this specific capacity of the Buddha was now extended to an entire class of texts. One of the most prominent of these early protective texts was the Āṭānāṭika-sūtra, which can be described as “a charm to protect the monk meditating in the forest from unsympathetic demons (yakkha/yakṣa)” (Gethin Reference Gethin1998: 168). Indian versions of this are preserved in the Pali canon of Theravāda Buddhism (DN III 194–206) and in the manuscript remains of the Sanskrit canons of the Sarvāstivāda and Mūlasarvāstivāda schools discovered in Central Asia and in Gilgit in Northern Pakistan (Hoffmann Reference Hoffmann1939, repr. 1987).Footnote 3 In the Āṭānāṭika-sūtra the mahārāja Vaiśravaṇa hands the Buddha a magic spell with the following words, which subsume the main purpose and contents of the sūtra as follows:

santi bhadanta bhagavataḥ śrāvakā bhikṣavo bhikṣuṇya upāsakā upāsikā ye ’raṇyavanaprasthāni prāntāni śayanāsanāny adhyāvasanti | santi cātra vyāḍā yakṣā amanuṣyā naivāsikā ye bhagavatpravacane ’bhiprasannā anabhiprasannāś ca | (…) | sādhu bhadanta bhagavataḥ śrāvakā bhikṣavo bhikṣuṇya upāsakā upāsikā idam evātānāṭikāṃ sūtraṃ vi(dyāṃ rakṣāṃ pādavandanīṃ vistareṇ)odgṛhṇīyur (dhārayeyur ātmano guptaye rakṣāyai sukhasparśavihāratāyā anabhiprasannānāṃ vyāḍānāṃ yakṣā-)ṇām amanuṣyānāṃ naivāsikānām abhiprasā(dāyābhi)prasann(ā)nāṃ ca [bhū](yobhāvāya |)

(Hoffmann Reference Hoffmann1939: 35–7, repr. 1987: 51–3).

Es gibt, o Ehrwürdiger, Jünger des Erhabenen, Mönche und Nonnen, Laienanhänger und Laienanhängerinnen, die an Waldwildnissen gelegenen, einsamen Lagerstätten sich aufhalten. Und dort befinden sich böse Yakṣa-Ortsdämonen, die der Verkündigung des Erhabenen ergeben, und solche, die ihr nicht ergeben sind. (…) Gut mögen, o Ehrwürdiger, des Erhabenen Jünger, Mönche und Nonnen, Laienanhänger und Laienanhängerinnen diese Āṭānāṭika-Sūtra, (Zauberformel, -Schutzzauber, -Fußverehrung in vollem Umfang) lernen (und behalten zu ihrem Schutze, zur Bewachung, zum Wohlsein und um die nicht ergebenen bösen) Yakṣa-Ortsdämonen zur Ergebenheit zu bringen wie um die ergebenen zu vermehren

(Hoffmann Reference Hoffmann1939: 80 f., repr. 1987: 96 f.).

With regard to the purpose and intended audience of the magic formulas, the text makes three points clear: a) The reason why these magic formulas are introduced are because of monastics living in a remote and potentially dangerous environment. They are especially associated with ascetic practices. b) The protective quality of the text is achieved by magic power, which has to be activated through memorization. c) Access to this power is extended to the whole Buddhist community, consisting of four groups: monks, nuns, lay men, and lay women.

These characteristics are shared by all extant versions of the text and it can therefore be suggested that they belong to its ancient core. But if we compare the Pali with the Sanskrit versions, we observe an important difference. Some of the Sanskrit versions insert in several places a textual passage which is not found in the Pali text. It is introduced by the phrase tadyathā and followed by isolated words. The passage is concluded by the word svāhā, a Sanskrit term meaning “hail” or “success”:

tadyathā | bilimāha | balimele | purā | pure | ghori | gandhārī | cori | caṇḍāli | sopakke | sthulasāpati | padumāpati svāhā |

(Hoffmann Reference Hoffmann1939: 33–5, repr. 1987: 49–51).Footnote 4

In the terminology of the rakṣā literature these little textual components are called mantra or mantrapada, in some cases also vidyā, all terms meaning “magic spell”. These mantras consist of various syllables, some of which represent intelligible lexemes, while others are not intelligible at all (Skilling Reference Skilling1992: 153–6). Other texts – generally Mahāyāna – introduce the lexeme dhāraṇī into the terminology, resulting in terms like mantradhāraṇī (used by Asaṅga) or dhāraṇīmantrapada (found in the Megha-sūtra) (cf. Skilling Reference Skilling1992: 150–52). A whole collection of mantras and dhāraṇīs is found in the Saddharmapuṇḍarīka-sūtra, in the chapter called dhāraṇīparivartaḥ (Vaidya Reference Vaidya1960b: 233–5). This text uses the terms mantrapadadhāraṇī, dhāraṇīmantrapada and dhāraṇīpada side by side, combining the inherited terminology (mantra, mantrapada) with the new one and leaving no doubt that they have to be perceived as synonyms.Footnote 5 According to Asaṅga's classification of dhāraṇīs, the mantradhāraṇī is one of four dhāraṇī categories which are governed by a Bodhisattva, consisting of dharmadhāraṇī, arthadhāraṇī, mantradhāraṇī and bodhisattvakṣāntilābhāya dhāraṇī. The mantradhāraṇī is defined as follows:

tatra mantradhāraṇī katamā | iha bodhisattvas tadrūpāṃ samādhivaśitāṃ pratilabhate yayā yāni mantrapadānītisaṃśamanāya sattvānām adhitiṣṭhati tāni siddhāni bhavanti paramasiddhāny amoghāny anekavidhānām ītīnāṃ saṃśamanāya | iyam ucyate bodhisattvasya mantradhāraṇī

(cited after Lamotte Reference Lamotte1949–80, IV: 1857).

Qu'est-ce que la dhāraṇī de mantra? Le Bodhisattva gagne une telle maîtrise en concentration que par elle il consacre les syllabes magiques destinées à apaiser les fléaux de tous les êtres, et ces syllabes deviennent ainsi effectives, suprêmement effectives et infaillibles pour apaiser de multiples fléaux. Telle est, chez le Bodhisattva, la dhāraṇī de mantra

(tr. Lamotte Reference Lamotte1949–80, IV: 1858).Footnote 6

Skilling convincingly calls this type of mantra “protective”. This is the category found in Śrāvakayāna texts and early Mahāyāna sūtras, and has to be distinguished from the “spiritual mantras” which are linked with spiritual goals like enlightenment and were introduced not later than the second century ce (Skilling Reference Skilling1992: 151).

As he pointed out, the earliest texts that used these elements were composed in the early centuries of our era in North India.Footnote 7 In many cases the texts were not complete new compositions, but used the protective texts of traditional Buddhism, which became enriched by magic spells. The ways in which this new material was incorporated into existing textual units are different. In the case of the Āṭānāṭika-sūtra the mantras were inserted into the main textual body. However, the manuscripts show that this approach was followed only by some recensions, partially independently of each other.Footnote 8

Another method is represented by the text tradition of the Nagaropama-sūtra (Bongard-Levin et al. Reference Bongard-Levin, Takamichi and Wille1996). Here an entire text called vyākaraṇa is added as an appendix to a traditional sūtra. But supplementing the old rakṣā texts is only one way to incorporate this new material into the Buddhist literary tradition. As Peter Skilling (Reference Skilling1992: 113) has shown, whole new classes of texts emerge. Probably the most important of these are the so-called Pañcarakṣā texts or “Five Protections”, a collection of sūtras which also frequently use an older canonical core, but develop out of this completely new texts. As Skilling stresses: “these texts were extremely popular in Northern India, Nepal, and Tibet, as may be seen from the numerous manuscripts kept in libraries around the world” (Reference Skilling1992: 138). In general, it is very difficult to ascertain exactly when the composition of these new texts took place. The manuscript evidence is rather late and seems to be separated from the actual events by several centuries. But there is good evidence that at least some of the texts which were later assembled under the label Pañcarakṣā circulated in an earlier form in the first centuries ce.Footnote 9 Thus the earliest Chinese translations of the Mahāmāyūrīvidyārājñī (henceforth: Mahāmāyūrī) belong to the fourth century ce (T. 986).Footnote 10 An early Sanskrit version of this text is part of the so-called Bower manuscript found at Kucha in Central Asia and which can be dated to the early sixth century ce (Sander Reference Sander, Yaldiz and Lobo1987).

The Gāndhārī *Manasvi-nāgarāja-sūtra (BajC 3)

Thanks to the new discoveries from Gandhāra, we are now in a position to present the earliest manuscript of such a newly composed rakṣā text. It is part of the Bajaur Collection and is written in a variety of the Kharoṣṭhī script of the Indian north west and composed in the language of that region, the so-called Gāndhārī. Both script and language were in use up to the third or early fourth century ce (cf. Strauch Reference Strauch, de Voogt and Quack2011). According to their philological and palaeographical features, the texts from the Bajaur Collection can be dated to the first two centuries of our era. Given the date suggested by Peter Skilling for the introduction of the use of mantras in Buddhist rakṣā literature (“the beginning of the Common Era”) the evidence for this manuscript would suggest it falls at the very beginning of the development just described.

The text of the Bajaur Fragment 3 (BajC 3) is written on both sides of a birch bark scroll measuring about 17 × 39 cm. It is composed as a short sūtra describing how the nāgarāja Manasvin approached the Buddha and handed over to him a special mantra, which could be used to bear all kind of calamities. The structure of the sūtra can be summarized as follows:

  1. 1. Introduction (nidāna): Śrāvastī.

  2. 2. Manasvin approaches the Buddha.

  3. 3. Manasvin's concern: “The monks are not anymore devoted to the discipline of wakefulness” (na bhūyo jāgarikāyogam anuyuktā viharanti).

  4. 4. Buddha's answer: Reason = fear.

  5. 5. Manasvin utters the mantra and its effects.

    1. 5.1. The effects of the mantra: fearlessness

    2. 5.2. The mantra

    3. 5.3. The spheres of use of the mantra: way, forest, king's palace, enemies, fire, water, etc.

  6. 6. Buddha repeats the mantra and its effects.

    1. 6.1. The effects of the mantra: fearlessness

    2. 6.2. The mantra

    3. 6.3. The spheres of use of the mantra: way, forest, king's palace, enemies, fire, water, etc.

  7. 7. The nāga king instructs the Buddha to teach the mantra.

  8. 8. The Buddha instructs Ānanda to teach the mantra.

  9. 9. The conclusion.

Although the nāga king Manasvin is not found in old or south Indian Buddhist texts, he seems to have been quite popular in north Indian Buddhism. His name:

is exclusively known from Northern Buddhist texts like the Lalitavistara (ed. Lefmann Reference Lefmann1902: 204, lines 9–10; 219, line 9), the Saddharmapuṇḍarīkasūtra (ed. Vaidya Reference Vaidya1960b: 2.17), and the Mahāvyutpatti (chapter 168, Sakaki Reference Sakaki1965: 227–31). His special association with magical practices is indicated by his mentioning in protective texts like the Mahāmāyūrī, (ed. Takubo Reference Takubo1972: 5, 41, cf. ed. Ol'denburg 1899: 221, 247) and the so-called Āṭānātikahṛdaya, a Central Asian appendix to the respective sūtra (ed. Hoffmann Reference Hoffmann1939: 104 (repr. 1987: 120), v. 14, cf. Sander Reference Sander, Yaldiz and Lobo1987: 207 f.) (Strauch Reference Strauch2008: 120).

Manasvin's name is also found in other protective texts, like in the mantra of the Mahādaṇḍadhāraṇī-Śītavatī (Hidas forthcoming b) and in the Megha-sūtra (Bendall Reference Bendall1880: 306 f.).Footnote 11 According to this evidence the Gāndhārī sūtra most probably belongs to a northern Buddhist tradition.

The following discussion aims to determine the position of this newly discovered Gāndhārī protective sūtra within the already known Buddhist rakṣā traditions discussed above. For this purpose we will concentrate on a few selected passages.Footnote 12

The effects of the mantra

The first part describes the efficiency of the spell. As can be seen from the summary of the contents, it is repeated twice (5.1 and 6.1). It is therefore possible to reconstruct this passage to a considerable degree.

Quotation 1 (BajC 3, lines 14–18)

... ṇa teṇa maṇuśaṇa bhayidave [ṇa a](maṇuśaṇa ṇa vyaḍe)aṇa bhayidave ṇa drigheaṇa bhaidave ṇa jat[u]ṇa [bhayidav](e) [ṇa] +  +  +  +  +  + ṇa yakṣaṇa bhayidave ṇa praduṭhacitaṇa bhayidave ṇa śa[str]eṇa [ka](le kaṭave) ṇa agiṇa dajidave ṇa udeami kale kaṭave ‧ ṇa viṣeṇa kale (kaṭave ...)

Translation

... he has not to be afraid of human beings, (of non-human beings, of wild beasts,) of snakes, of *demons,Footnote 13 of ?, of yakṣas (i.e. demons), of evil-minded, he cannot die from a weapon, he is not to be burnt by fire, he cannot die in the water, he cannot die from poison.

The major topic of this list is fearlessness (na bhayidave, Skt. na bhetavyam), supplemented by different kinds of immunities against all kinds of calamities. The given list largely corresponds to a conventional inventory which generally agrees with the so-called ten antarāyas, “hindrances, obstacles”, which are already known to the Vinaya (cf. CPD s.v.), among them the king, thieves, fire, water, human beings, non-human beings, wild beasts, snakes (rāja, cora, aggi, udaka, manussa, amanussa, vāḷa, siriṃsapa).

There are more parallels that point to the canonical background of these ideas. The positive counterparts to these calamities are part of another conventional list that contains the so-called advantages (P. anisaṃsa, Skt. anuśaṃsa) that one obtains from various religious practices. One such list is found in the Mettānisaṃsa-sutta (AN V. 342.1–13), a text belonging to the protective literature of the Pali tradition. The means by which these advantages can be attained is defined as mettā “friendliness”. As Schmithausen, in his work Maitrī and Magic (Reference Schmithausen1997) shows, maitrī (P. mettā) as a specific Buddhist concept has much in common with otherwise magic or supernatural capacities. Only by his maitrī did the Buddha subdue an elephant, by maitrī is he able to achieve the most unbelievable things. It is therefore quite natural that the concept of maitrī becomes an adequate starting point for the introduction of magic practices. Schmithausen (Reference Schmithausen1997: 67) describes this process thus:

Though, on the one hand, friendliness tends to become regarded as an autonomous means for protection ... effective with regard to insentient forms of nature like water, fire or poison, there is, on the other hand, a tendency to have some doubt about its protective effectivity. At any rate, it is often supplemented or even replaced by other protective devices like commemoration of the Buddha or the Three Jewels, or appealing to their (or other powerful beings') protective power, or “tapping” this power by means of truth magic. Increasingly, magical formulas (mantra, dhāraṇī) come to be employed ... .

The text in the Mettānisaṃsa-sutta (AN V 342.1–13) runs as follows:

sukhaṃ supati, sukhaṃ paṭibujjhati, na pāpakaṃ supinaṃ passati. Manussānaṃ piyo hoti, amanussānaṃ piyo hoti, devatā rakkhanti, nāssa aggī vā visaṃ vā satthaṃ vā kamati, tuvaṭaṃ cittaṃ samādhiyati, mukhavaṇṇo vippasīdati, asammūḷho kālaṃ karoti, uttariṃ appaṭivijjhanto brahmalokūpago hoti

(AN V 342.5–10).

(1) One sleeps well; (2) one awakens happily; (3) one does not have bad dreams; (4) one is pleasing to human beings; (5) one is pleasing to spirits; (6) deities protect one; (7) fire, poison, and weapons do not injure one; (8) one's mind quickly becomes concentrated; (9) one's facial complexion is serene; (10) one dies unconfused; and (11) if one does not penetrate further, one fares on to the brahmā world.

(tr. Bodhi Reference Bodhi2012: 1573).

Similar, but not entirely identical lists of obstacles/calamities and advantages are found throughout Buddhist literature of different genres and periods (cf. for some of them Strauch Reference Strauch2007/08: 41–6). And they become a typical and characteristic feature of the newly composed protective texts, as for example the vyākaraṇa of the Nagaropama-sūtra:

yaḥ kaścin māriṣa idaṃ vyākaraṇaṃ ... manasikariṣyati saḥ ahinā na daṃkṣyati viṣaṃ kāye na tariṣyati śastraṃ na kramiṣyati nodakena kālaṃ kariṣyati agninā na dhakṣyati rājāno ’pi na prasahiṣyaṃti corā na muṣiṣyaṃti rājakulamadhyagato ’pi svastinottariṣyati gāḍhabandhanabaddho ’pi mokṣyati āsannāsamāgato ’pi abhyavakāśagato bhaviṣyati sarve ca kṛtyakākhordamantraveṭāḍa prativigamiṣyaṃti sarve ca bhūtagaṇā na viheṭhayiṣyaṃti anyatra pūrvakeṇa karmaṇā.

(Bongard-Levin et al. Reference Bongard-Levin, Takamichi and Wille1996: 85; II.11, cf. also II.25)

Whosoever, Good Sir, will ... reflect upon this exposition, he will not be bitten by snakes, nor will poison work in his body, nor will weapons have effect (against him), nor will he drown, nor will he be burned by fire, nor will kings suppress him, nor will thieves rob him; even when he is in the midst of a royal palace, he will escape (any misfortune) on account of his good luck; even when he is bound by tight fetters, he will be freed; even when he is encroached upon (by others), he will be unconstrained; and all sorceries, maledictions, curses, and demonic arts will be neutralized; and all the legions of ghosts will not harm him except as a result of his former karma.

(Bongard-Levin et al. Reference Bongard-Levin, Takamichi and Wille1996: 99)

Of special importance for the evaluation of the Gāndhārī text, however, is the parallel to the Mahāmāyūrī. Although the conventional list of calamities against which the spell is supposed to help is found here in an extended form and in a slightly different terminology, it also refers to “fearlessness” as the main aim of the mantra's use:Footnote 14

rājabhayāc caurabhayād agnibhayād udakabhayāt bandhakabhayāt pratyarthikabhayāt

(Takubo Reference Takubo1972: 42, cf. ed. Ol'denburg 1899: 248)

(shall protect) from the fear from kings, thieves, fire, water, allies and enemies.

This list is complemented by another, closely related, enumeration of situations in which the mantra should be memorized (manasi-√kṛ):

iyaṃ cānanda mahāmāyūrī vidyārājñī grāmagatena manasikartavyā. araṇyagatena manasikartavyā. pathigatena manasikartavyā. utpathagatena. rājakulamadhyagatena. cauramadhyagatena. agnimadhyagatena udakamadhyagatena. pratyarthikamadhyagatena. parṣanmadhyagatena. vivādamadhyagatena. ahidaṣṭena. viṣapītena. sarvabhayasannipātitena ca manasikartavyā.

(Takubo Reference Takubo1972: 9–10, cf. ed. Ol'denburg 1899: 224)

This queen among the spells, this Mahāmāyūrī, should be memorized by a person, who has got into a village, who has got inside a forest, who has got on the path or off the path, who has got inside the king's palace, who has got amidst thieves, who has got into fire, who has got into water, who has got amidst enemies, who has got into an assembly, who has got into a dispute, who was bitten by a snake, who has drunk poison, (to sum up) who is afflicted by all sorts of fear.

This compositional principle is also found in the Gāndhārī sūtra where we find an almost identical list.

Quotation 2 (BajC 3, lines 39–42)

aya khu maṇaspia vija ◦ pathagadeṇa maṇasikaṭave upasa[gadeṇa] (ma)[ṇa](si)kaṭave rañagadeṇa maṇasikaṭave rayaülamaǰagadeṇa maṇasikaṭave +  +  +  +  + midramaǰagadeṇa maṇasikaṭave amidramaǰagadeṇa maṇasikaṭave ‧ (śastrataragadeṇa) maṇasikaṭave ‧ agiataragadeṇa maṇasikaṭave udeaataraga(deṇa maṇasikaṭave) viṣapideṇa maṇasikaṭave

Translation

This is the spell of Manasvin. It should be memorized by a person who has got on the path, who has got off the path, who has got inside a forest, who has got inside the king's palace, (...) who has got amidst friends, who has got amidst enemies, (who has got in between weapons), who has got into fire, who has got into water, who has drunk poison.

If we compare the lists of both texts in a synoptic way, their partial overlapping and general parallelism as well as their relationship to the categories of antarāyas as described in the Pali Vinaya becomes reasonably obvious, as shown in Table 1.

Table 1. Comparative list of calamities (antarāya) and advantages (anuśaṃsā)

As can be seen in Table 1, the correspondence between BajC/list 2 and Mahāmāyūrī/list 2 is particularly high and seems to speak in favour of a common source of both traditions. While the first lists are much indebted to the canonical models of the antarāya and anuśaṃsa lists, these later lists seem to reflect a slightly different terminology which is closely connected with the time of the composition of these new texts. Parallels from the vyākaraṇa of the Nagaropama-sūtra Footnote 15 and from the Aṣṭasāhasrikāprajñāpāramitā Footnote 16 can support this evaluation.

The addressees: Forest monks

According to the above-cited passage from the Āṭāṇāṭika-sūtra the protective function of the sūtra was particularly associated with a special group of monks, those who are said to dwell in isolated places such as, for example, forests. The Gāndhārī sūtra ends with the sentence:

Quotation 3 (BajC 3, lines 46–51)

ughiṇahi aṇade maṇaspia vijaraya ‧ payauṇehi aṇad(e) ma[ṇaspia vija]raya arañiaṇa bhikhuṇa hidae bheśadi suhae caduṇa pariṣaṇa bhikhuṇa bhikhuṇiṇa uaseaṇa uasiaṇa arthae bheśadi hidae suhae bheśadi [pha](ṣa)viharadae bheśadi ‧ eva bhate [bhagava] [a]iṣpa aṇado ? ? + + ? [va ye] + + va aiṣpa aṇado maṇaspio ca ṇagarayo sadevamaṇuṣasaro ?(bhagavado) bhaṣido abhiṇadi

Translation

Ānanda, take up this king-spell of Manasvin. Ānanda, learn this king-spell of Manasvin. It shall be for the welfare and the joy of the monks dwelling in forests. It shall be for the benefit, for the welfare, for the joy, for the comfort of the four assemblies, (i.e.) the monks, the nuns, the laymen, the laywomen. Thus (may it be), Venerable Lord, (spoke the honourable Ānanda.) (...) Ānanda and the nāga king Manasvin, together with gods, humans and asuras, enjoyed the speech of the Venerable.

Again the text supports two points:

  1. 1) The magic spell is to be distributed among the four assemblies, i.e. among all followers of the Buddhist order, including the laymen and laywomen.

  2. 2) A special position, however, is given to the forest monks. They are labelled here with the distinctive term arañia (Skt. āraṇyaka, Pali āraññika). Forest living was already in early Buddhist monasticism one of the twelve or thirteen kinds of ascetic practices (dhutaguṇa, dhutaṅga) (cf. Gethin Reference Gethin1998: 98 f.). Āraṇyakas are monks devoted to ascetic practices in an isolated environment. That they are in special need of remedies against natural calamities such as fire, water and snakebites, calls for no special remark (cf. e.g. Schmithausen Reference Schmithausen1997: 35).

The magical formula (vidyārāja, mantra)

The Gāndhārī sūtra calls the magic formula vijaraya, Skt. vidyārāja, spell-king. This term has to be connected with feminine vidyārājñī “spell-queen” or “queen among spells” which is regularly used in the texts of the Pañcarakṣā tradition to designate the magical spell. There seems to be, however, some evidence that the feminine gender of this term was the result of a later development, possibly influenced by the growing role of the female element in Tantric practices. Thus Hidas noted that the oldest manuscripts of the Mahāpratisarā-Mahāvidyārājñī, one of the texts of the Pañcarakṣā tradition, use the masculine title Mahāpratisara-Mahāvidyārājarājā (2012: 21 f.). He explains this change in gender by the fact “that the roots of this tradition go back to Brahmanism, to texts as early as the Atharva-veda”. The Gāndhārī sūtra may add another example for the early masculine gender of the spell, although it can hardly be used to strengthen Hidas' argument about an Atharvavedic origin.

Beside the term vijaraya (Skt. vidyārāja) the Gāndhārī text uses the term mantra in its Gāndhārī equivalent matra. Like the other elements of the text, the mantra is also reproduced twice: once as received by the Buddha from the nāgarāja's mouth (= 5.2); and once as announced by the Buddha (6.2). Due to this repetition it is possible to reconstruct most of its contents.

Quotation 4 (BajC 3, lines 34–8)

saṃyasida ◦ paṃḍara ◦ kara a ◦ keyura ◦ [d]u[d]ura ◦ data ◦ davia ◦ bidumadi ◦ śirimadi ◦ teyavadi ◦ rasagapa[di] (◦ puruṣa ◦ k[a]ja[a]) vihatra ◦ taragatriṭhi ◦ iṣimuha ◦ hili cili ◦ khili cili ◦ kha ? + + + + (vihala) rasagatriṭhi ◦ oru a ◦ uru a ◦ dadura ◦ daha ◦ utama ◦ suut[ta](ma) + + + utamatama

Like many other mantras it consists of a multitude of isolated terms. Some of them are magical syllables, which can be found in different Indian magical traditions, like e.g. hili cili, khili cili.Footnote 17

The present paper is not the place to provide a comprehensive discussion of all the terms listed in the mantra/vidyārāja of the sūtra. Instead, I wish to stress a feature of this list that is indicated by the text itself. The mantra is concluded by the words:

Quotation 5 (BajC 3, lines 29–30)

evameva maṇaspia evameva maṇaspia (evameva ma)ṇaspi[a ‧ imeṣa] mahaviṣaṇa ṇamo jaṇea ‧ ṇa teṇa maṇuśaṇa bhayidave ...

Translation

Thus is (the spell) of Manasvin, thus is (the spell) of Manasvin, thus is (the spell) of Manasvin. (Who) would know the names of these great poisons has not to be afraid of human beings …

The text thus explicitly states that the terms contained in the mantra are the names of poisonous substances that are obviously subsumed here as a group called mahāviṣa “great poisons”. By knowing their names, they are subdued, and – moreover – convey general immunity against all sorts of calamities. Fortunately, there exists a list of such mahāviṣas in the Mahāmāyūrī (Takubo Reference Takubo1972: 55, cf. Ol'denburg 1899: 257–8):

udgṛhṇa tvam ānanda mahāviṣāṇāṃ nāmāni. tadyathā.

aṇḍarā, paṇḍarā, karaḍā, keyūrā, bhūtāṃgamā, bhūtapati, vindupati, śiripati, tejapati, tejograpati, yaśopati, araḍā, taraḍā, taradā, tarataraḍā, duttarā, dantājuhā, juhā, johā, jolā, melā, phurā, guhā, rucirā, danturā, ilikicikā, kirikirikā, kāmvā, śatanturā, vikuli, kirimi, taraṅgā, riṣṭā, āmravati, jambumati, manumati, kamale, vimale, kuṇḍale, ahituhi, duhi, vakve, vakvadūte, vatsamāte, mahāgāre, tulamve sulamve svāhā.

ity ete ānanda mahāviṣās te ’py anayā mahāmāyūryā vidyārājñyā svāter bhikṣor mama sarvasatvānāṃ ca rakṣāṃ kurvantu

Take up, Ānanda, the names of the Great Poisons, namely:

aṇḍarā, paṇḍarā, karaḍā, keyūrā, bhūtāṃgamā, bhūtapati, vindupati, śiripati, tejapati, tejograpati, yaśopati, araḍā, taraḍā, taradā, tarataraḍā, duttarā, dantājuhā, juhā, johā, jolā, melā, phurā, guhā, rucirā, danturā, ilikicikā, kirikirikā, kāmvā, śatanturā, vikuli, kirimi, taraṅgā, riṣṭā, āmravati, jambumati, manumati, kamale, vimale, kuṇḍale, ahituhi, duhi, vakve, vakvadūte, vatsamāte, mahāgāre, tulamve sulamve. Hail.

These are, Ānanda, the names of the Great Poisons. They too shall protect by the Great Peacock spell-queen the monk Svāti, me and all living beings.

The words tadyathā “namely” and svāha “hail” at the beginning and at the end of the enumeration make clear that this text passage represents a mantra. As can be seen from a comparison between both these mahāviṣa lists, they share a number of identical or closely related terms (Table 2, see also Table 3, columns A and B).

Table 2. Names of poisons from the mantra of BajC 3 and from the Mahāmāyūrī

Table 3. Comparative list of names of poisons from different Buddhist traditions and Āyurvedic literature

The close relationship between the Gāndhārī sūtra and the Mahāmāyūrī as an early representative of the class of Pañcarakṣā texts was already indicated by the common use of the term vidyārāja/vidyārājñī and the antarāya-based lists of calamities/spheres of use. It seems to be further corroborated by the partly identical lists of “Great Poisons”.

Especially interesting are the pairs bidumaṯi-vindupati, śirimaṯi-siripati and teyavadi-tejapati which seem to indicate that the list of the Mahāmāyūrī had been translated from a source composed in Gāndhārī where the feminine possessive suffix was given as vadi (teyavadi = Skt. tejovatī). This suffix was obviously misunderstood by the Mahāmāyūrī or its source as representative of Skt. pati, which is homonymous in Gāndhārī.

The tradition connected with these designations of poisonous substances/plants is not restricted to these two mantras. Thus there is another spell in the Mahāmāyūrī that also contains some of the terms belonging to the list of “Great Poisons” (indicated here in bold print, see Table 3, column E).

iyaṃ cānanda mahāmāyūrī vidyārājñī kāśyapena samyaksaṃbuddhena bhāṣitā cābhyanumoditā ca.

tadyathā. aṇḍare paṇḍare kaṇḍare maṇḍare khaṇḍare jaṃbu juṃbunadi jaṃbuvati, matte maṇḍitike, amare siddhi, hara hara hara hara, paśu paśu paśu paśu paśupati svāhā siddhi siddhi siddhi siddhi svāhā (Takubo Reference Takubo1972: 44, cf. Ol'denburg 1899: 223, 250).

And, Ānanda, this Great Peacock queen-spell was announced and approved by Kāśyapa, the Completely Enlightened, namely:

aṇḍare paṇḍare kaṇḍare maṇḍare khaṇḍare jaṃbu juṃbunadi jaṃbuvati, matte maṇḍitike, amare siddhi, hara hara hara hara, paśu paśu paśu paśu paśupati. Hail. siddhi siddhi siddhi siddhi. Hail.

Traces of this tradition are also found outside the Mahāmāyūrī. There are a few other rather short variants of mantras referring to some of the characteristic terms. In the Śārdūlakārṇāvadāna a formula called ṣaḍakṣarīvidyā is handed over by the Buddha to Ānanda with the following words (Mukhyopadhyaya 1954: 4–5, parallel terms indicated here in bold, see Table 3, column C).Footnote 18

udgṛhṇa tvam ānanda imāṃ ṣaḍakṣarīvidyām dhāraya vācaya paryavāpnuhi ātmano hitāya sukhāya bhikṣūṇāṃ bhikṣuṇīnām upāsakānām upāsikānāṃ hitāya sukhāya. Iyam ānanda ṣaḍakṣarīvidyā ṣaḍbhiḥ samyaksambuddhair bhāṣitā, caturbhiś ca mahārājaiḥ, śakreṇa devānām indreṇa, brahmaṇā ca sahāpatinā. mayā caitarhi śākyamuninā samyaksambuddhena bhāṣitā. Tvam apy etarhi ānanda tāṃ dhārasya vācaya paryavāpnuhi. yad uta tadyathā

aṇḍare pāṇḍare kāraṇḍe keyūre’rcihaste kharagrīve bandhumati vīramati dhara vidha cilimile viloḍaya viṣāṇi loke viṣa cala cala golamati gaṇḍavile cilimile sātiniṃne yathāsaṃvibhakte golamati gaṇḍavilāyai svāhā //

Take up, Ānanda this ṣaḍakṣarī spell, keep it, recite it, learn it for the welfare, the joy of yourself, for the welfare, the joy of the monks, nuns, laymen and laywomen. Ānanda, this ṣaḍakṣarī spell was spoken by the six Completely Enlightened, and by the Four Great Kings, and by Śakra, the king of the gods, and by Brahman, the Lord of the World. And now it is spoken by me, Śākyamuni, the Completely Enlightened. Now you too, Ānanda, keep it, recite it, learn it, which is as follows:

aṇḍare pāṇḍare kāraṇḍe keyūre ’rcihaste kharagrīve bandhumati vīramati dhara vidha cilimile viloḍaya viṣāṇi loke viṣa cala cala golamati gaṇḍavile cilimile sātiniṃne yathāsaṃvibhakte golamati gaṇḍavilāyai. Hail //.

The references to viṣa in the mantra probably indicate the specific character of the spell and its association with poisonous substances. In the Mahāsāhasrapramardinī (Iwamoto Reference Iwamoto1937: 33), another Pañcarakṣā text formed around the core of a canonical sūtra – the *Ratna-sūtra (Pali Ratana-sutta)Footnote 19 – we find the following passage:

vīryeṇa tejasā teṣāṃ viṣam astv aviṣaṃ sadā /

tatra mantrapadā bhonti nirviṣā viṣadūṣaṇāḥ //

syād yathedam /

harikeśi / nakile / rehile / amare aṇḍare paṇḍare / kaṭake / keyūre / hase hase hase / khase khase khase / kharaṅge / marugahaṇe svāhā // mumukṣa svāhā / hile svāhā / mile svāhā //

hatā gaṇḍāḥ kilāsāś ca vaisarpāś ca vicarcikāḥ /

piṭṭakā lohaliṅgāś ca kacchūr bhavati saptamī //

rāgo dveṣaś ca mohaś ca ete loke trayo viṣāḥ /

nirviṣo bhagavān buddho buddhatejohataṃ viṣam //

rāgo dveṣaś ca mohaś ca ete loke trayo viṣāḥ /

nirviṣo bhagavān dharmo dharmatejohataṃ viṣam //

rāgo dveṣaś ca mohaś ca ete loke trayo viṣāḥ /

nirviṣo bhagavān saṃghaṃ saṃghatejohataṃ viṣam //

viṣasya pṛthivī mātā viṣasya pṛthivī pitā /

etena satyavākyena viṣāḥ sarve syur nirviṣāḥ //

By the energy, by the heat of them (i.e. Buddhas, Pratyekabuddhas, Arhats, etc.) the poison shall always be non-poisonous. There are mantra words which are poisonless (nirviṣa), which make poison effectless (viṣadūṣaṇa), namely:

harikeśi. nakile. rehile. amare aṇḍare paṇḍare. kaṭake. keyūre. hase hase hase. hase khase khase. kharaṅge. marugahaṇe Hail. mumukṣa Hail. hile Hail. mile Hail.

Destroyed are goitres (gaṇḍa), cases of “white leprosy” (kilāsa), the effect of the visarpa decease, itch (vicarcikā), cases of tartar (piṭṭaka), bloody abscesses, and scab (kacchū) as the seventh.

Greed, hatred and delusion – these are the three poisons in the world.

Without poison is Lord Buddha, by the heat of the Buddha the poison is destroyed.

Greed, hatred and delusion – these are the three poisons in the world.

Without poison is Lord Dharma, by the heat of the Dharma the poison is destroyed.

Greed, hatred and delusion – these are the three poisons in the world.

Without poison is Lord Saṃgha, by the heat of the Saṃgha the poison is destroyed.

The earth is the mother of the poison, the earth is the father of the poison.

By this true speech all poisons may be non-poisonous.

This passage of the Mahāsāhasrapramardinī is especially informative. Although it reinterprets the term viṣa “poison” in terms of Buddhist dogmatics, its reference to diseases, which are supposed to be caused by the influence of poisonous substances, gives a clear indication of the close connection of this type of magic with Indian sciences, especially with medicine (āyurveda).

Beside these parallels in Sanskrit texts there exists a small Tibetan sūtra, which is also associated with the nāga king Manasvin. In their rgyud department (= Tantra) the Tibetan Kanjurs contain a little text with the title ’phags pa klu'i rgyal po gzi can gyis žus pa žes bya ba’i gzuṅs (see Eimer Reference Eimer1989: 106, Nr. 284). Its Sanskrit title is given in different spellings. While the Western Group of Kanjurs uses the title (ārya-)vira/vīra-nāgarāja-paripṛcchā-nāma-dhāraṇī (cf. e.g. Narthang [Csoma de Körös Reference Csoma de Körös1881: 328, no. 17] and Stog 615 [Skorupski Reference Skorupski1985: 280]), the Eastern Group transliterates the nāgarāja's name either with namasvi or with tapasvi. The correct name of this nāgarāja was obviously known to the Mahāvyutpatti (3285), which gives as equivalent of Tib. klu'i rgyal po gzi can Skt. manasvī-nāgarāja. The original Sanskrit title of the Tibetan dhāraṇī text should therefore be reconstructed as Manasvi-nāgarāja-paripṛcchā-nāma-dhāraṇī (cf. Strauch Reference Strauch2007/08: 45 f.).

The narrative frame of both sūtras seems to be rather different: in the Tibetan text the story is situated in the Indraśaila cave on the Vaidehaka mountain where snakes are tormenting the monks. The nāga king asks the Buddha what can be done to make an end to this disturbance. Following this request the Buddha announces the mantra.Footnote 20 Contrary to this, the Gāndhārī text treats the mantra as a gift from the side of the nāga. Another difference between the texts is the use of the designation dhāraṇī, which is avoided in the case of the Gāndhārī text, but is clearly found in the Tibetan parallel, both in its Sanskrit form in the transliteration of the title and in its Tibetan equivalent gzuṅs. That both texts – the Tibetan dhāraṇī and the Gāndhārī rakṣā sūtra – are at least related to each other can be shown by the mantras, which share a common inventory of terms. The mantra of the Tibetan text runs according to the Derge edition of the Kanjur (D 659) as follows (parallel terms indicated here in bold print, see Table 3, column F):

tadyathā paṇḍare. karaṭe. keyūre. bodhyaṃgamate. bindumate. araḍa. taraḍa. taruḍa. dahadahana. etadaha. jele jela. phale. śata śata. kule nakuli. ekarṣi taraṅgapriṣṭi.

While a number of words are parallel to the Gāndhārī text, some terms are only found in the other parallels (cf. Table 3). This shows that the two texts do not directly depend upon each other, but seem to go back to a common source that was also used in other texts, such as for example the Mahāmāyūrī.

The place where such a source should be looked for was already indicated by the Mahāsāhasrapramardinī that showed clear connections with Indian scientific traditions. It is probable that the list of “Great poisons”, that seems to form the origin of the terms of the mantra, is based on an inventory of poisonous plants or substances which was known not only in Buddhist circles or in magical contexts, but was formed in an early botanic or medical tradition of India. This evidence is further corroborated by the fact that the Tibetan text also explicitly refers to the “Great poisons and medicines” using the term dug daṅ sman chen po bcva brgyad po “eighteen great poisons and medicines”.

At present it is not possible to trace such a list in any of the preserved early Āyurvedic texts, but at least some of the terms can be cautiously identified or associated with healing or poisonous plants mentioned in later literature. Thus the term bidumaṯi of the Gāndhārī text can be connected with Skt. bindukinī or binduka, the latter cited by Hellwig (Reference Hellwig2009: 150) under the category “poison”. Other terms such as pāṇḍara, karaṇḍa and śrīmatī are attested in the dictionaries more generally as a “kind of plant” (see PW s.vv.). More specific is the connotation of taraṭa as a certain medical plant (PW s.v.) and of tejovatī, a kind of pepper (“Piper chaba W. Hunt”, PW s.v.). The last term is also found in Madanapāla's Nighaṇṭu. According to Dash (Reference Dash1991: 65) it designates Zanthoxylum alatum. Madanapāla also refers to keyūra as a designation of a medical plant: “yam (Dioscorea bulbifera)” (Dash Reference Dash1991: 360).

Further research is required here to identify these species and to clarify whether they have a specific connection with poisonous or healing substances. But there is good reason to assume that the lists reproduced in the Mahāmāyūrī and in the various magical formulae are based on an otherwise lost botanical or medical tradition. Table 3 lists the distinctive terms of the different traditions.

Summary

The Gāndhārī*Manasvi-nāgarāja-sūtra is a representative of early post-canonical rakṣā literature. Its compositional structure and the phraseology of its narrative part are based on that of canonical protective sūtras, which served as models for this newly emerging class of texts. Especially close is the relationship of the Gāndhārī sūtra to some texts of the Pañcarakṣā group and among them especially to the Mahāmāyūrī. Thus the Gāndhārī mantra contains, according to the text, the names of poisonous substances (mahāviṣa) which are closely related to the lists of mahāviṣas as preserved in the text of the Mahāmāyūrī. Some of the terms are also found in mantras that are part of other post-canonical protective texts. They are also supposed to be effective against poisons and show that the composition of the Gāndhārī *Manasvi-nāgarāja-sūtra was part of a broader development in north Indian Buddhism that is reflected in different types of texts.

Another, and probably later, representative of the Manasvi-dhāraṇī is found in the Tibetan Kanjur. Despite their diverging narrative frames, the mantra of this text shows parallels as well to the Gāndhārī as to the Mahāmāyūrī text. This shows that all three versions represent independent branches of a common tradition. This common tradition seems to be rooted in a list of mahāviṣas from a seemingly lost botanical or medical tradition. If this assumption is correct, such a scientific tradition could be identified as one of the possible sources for Buddhist rakṣā practices.

Footnotes

1 A comprehensive survey of the corpus of Gāndhārī literature will be available with Falk and Strauch Reference Falk, Strauch, Harrison and Hartmann2014.

2 See also Skilling's discussion of the rakṣā character of the mahāsūtras and their relation to the Theravāda parittas in the general introduction to his translation of these texts (1997: 63–88).

3 A new and complete version, which is part of the Dīrghāgama manuscript, studied by Jens-Uwe Hartmann (Reference Hartmann, Chojnacki, Hartmann and Tschannerl2000; Reference Hartmann2004) and his students and colleagues, has recently been edited by Lore Sander and Siglinde Dietz (Sander Reference Sander2007). It is supposed to originate from Gilgit.

4 Other – only partially preserved – mantras are found on the following pages: Hoffmann Reference Hoffmann1939: 39, 63, 69, 75, 77 and 1987: 55, 79, 85, 91, 93.

5 For a discussion of the terms mantrapada and mantradhāraṇīpada and numerous references cf. Pagel Reference Pagel2007: 57–61.

6 The specific meaning of dhāraṇī in Asaṅga's text is not undisputed. Skilling (Reference Skilling1992: 150, fn. 3) obviously perceives dhāraṇī in the basic sense “spell”, rejecting Braarvig's suggestion (Reference Braarvig1985: 19 f.) that the term designates the capacity “in retaining or remembering spells” rather than the spell itself. However, such an understanding seems to be shared by Lamotte who stresses: “Elle (la Dhāraṇī) n'est pas à proprement parler un mantra, une formule magique, comme on traduit généralement; elle est au premier chef la mémorisation des enseignements de tous les Buddha”. See also Pagel Reference Pagel2007: 59 f., fn. 50. Although the text of the Saddharmapuṇḍarīka-sūtra cited above clearly supports Skilling's interpretation of dhāraṇī as “spell”, it is probably advisable to bear in mind this broader meaning of dhāraṇī when applied in a Mahāyāna context. For dhāraṇī in the sense of “mnemonic device” cf. also Nattier Reference Nattier2003: 291 f., fn. 549. An extensive discussion of the semantic values of the term dhāraṇī is now provided in Davidson Reference Davidson2009.

7 According to Peter Skilling “the practice of rakṣā developed further with the use of mantras or vidyās, by the beginning of the Common Era at the very latest” (1992: 168, see also, the more detailed remarks on pp. 163 f.).

8 According to Helmut Hoffmann (Reference Hoffmann1939: 87, repr. 1987: 103), the mantras entered the text of the Āṭānāṭika-sūtra only in Turkestan. They are not found in the Chinese and Tibetan translations. However, according to Lore Sander, mantras are also part of the Dīrghāgama version, which probably originates from Gilgit. Their exact location and wording is usually different from that in the Central Asian version (Sander, email communication in November 2012).

9 Cf. for the Mahāpratisarāvidyārājñī, which “already existed in the fourth century ce or even in the third in some early form” (Hidas forthcoming a). Cf. also Hidas Reference Hidas2012: 24, fn. 16.

10 A comprehensive treatment of the complicated textual history of the Mahāmāyūrī in the Chinese traditions is now available in Sørensen Reference Sørensen2006.

11 These two references I owe to Gergely Hidas.

12 The discussion is based in part on my previous online publication (Strauch Reference Strauch2007/08: 40–47).

13 The meaning of jatu- is not completely clear. The suggested translation connects this term with Skt. yātu “Bez. einer Gattung von Dämonen, die in allerhand spukhaften Formen erscheinen” (PW s.v.). The word is already known in the Ṛgveda, but is especially found in Atharvavedic contexts. For the related Iranian term translated as “sorcerer” (Avestan yātu, Middle Persian jādūg) in different literary and historical contexts see Forrest Reference Forrest and Oktor Skjærvø2011: 63–70. The word is also discussed with references to previous treatments by Boyce (Reference Boyce1975: 85). A connection with Old Indian jantú (CDIAL 5110) cannot be completely ruled out. The parallel in the Niya document 565 jaṃdunaṃ could speak in favour of such an interpretation. For this word Bailey (Reference Bailey1948: 332) could establish the meaning “snake”. But in view of the preceding synonymous drighea- (Pali dīrgha-) “snake”, the translation “demon” is preferred here.

14 The list is found in two different forms (Takubo Reference Takubo1972: 28, 42).

15 ... pathagatā apy utpathagatā api ārāmagatā api śūnyāgāragatā (Bongard-Levin et al. Reference Bongard-Levin, Takamichi and Wille1996: 83; II.2) “... whether they are on the (right) path or have gone astray, whether they are in pleasure gardens or in deserted houses” (tr. Bongard-Levin et al. Reference Bongard-Levin, Takamichi and Wille1996: 96).

16 na ca khalu punar devaputrās tasya kulaputrasya vā kuladuhitur vā imāṃ prajñāpāramitām udgṛhṇato dhārayato vācayataḥ paryavāpnuvataḥ pravartayamānasya araṇyagatasya vā vṛkṣamūlagatasya vā śūnyāgāragatasya vā abhyavakāśagatasya vā pathi gatasya vā utpathagatasya vā aṭavīgatasya vā mahāsamudragatasya vā ... bhayaṃ vā bhaviṣyati, stambhitatvaṃ vā bhaviṣyati, utpatsyate vā (Vaidya Reference Vaidya1960a: 25.18–22) “There will be no fear, will be no paralysis, and they will not appear, Gods, for the noble man or woman who is learning, keeping, reciting, mastering and teaching this Perfection of Wisdom, not in a forest, not at the root of a tree, not in an empty house, not in an open place, not on the path, not off the path, not in the woods, not on the ocean” (cf. Strauch Reference Strauch2007/08: 45).

17 See e.g. Mahāpratisarā-Mahāvidyārājñī: hili 2 | mili 2 | kili 2 | cili 2 | sili 2 (Hidas Reference Hidas2012: 181). For non-Buddhist occurrences see e.g. the Agnipurāṇa: oṃ kili kili oṃ khili khili vili vili (Mitra Reference Mitra1870–79: 231) and the closely related spell in the Garuḍapurāṇa (1.38.7): kilikili khilikhili milimili cilicili (Kumar Reference Kumar and Kumar2006: 78).

18 Cf. for this passage and Tibetan extracts thereof, Skilling Reference Skilling1992: 157, fn. 1.

19 For a study of the Mahāsāhasrapramardinī see Hidas (forthcoming a).

20 A comparative study of both versions is currently being carried out with Cristina Scherrer-Schaub in the framework of a joint seminar at the University of Lausanne. The author thanks Cristina Scherrer-Schaub for the preliminary information.

References

Bailey, H.W. 1948. “Irano-Indica”, Bulletin of the School of Oriental and African Studies, 12: 319–32.Google Scholar
Bendall, Cecil. 1880. “The Megha-Sūtra”, Journal of the Royal Asiatic Society of Great Britain & Ireland (New Series), 12, 286311.Google Scholar
Bodhi, Bhikkhu. 2012. The Numerical Discourses of the Buddha. A Translation of the Anguttara Nikāya. Translated from the Pāli. (The Teachings of the Buddha Series.) Boston: Wisdom Publications.Google Scholar
Bongard-Levin, Gregory, Boucher, Daniel, Takamichi, Fukita and Wille, Klaus (eds). 1996. The Nagaropamasūtra: An Apotropaic Text from the Saṃyuktāgama. A Transliteration, Reconstruction, and Translation of the Central Asian Sanskrit Manuscripts. (Sanskrit-Texte aus dem buddhistischen Kanon: Neuentdeckungen und Neueditionen. Dritte Folge. Sanskrit-Wörterbuch der buddhistischen Texte aus den Turfan-Funden, Beiheft 6.) Göttingen, 7131.Google Scholar
Boyce, Mary. 1975. A History of Zoroastrianism: The Early Period. (Handbuch der Orientalistik.) Leiden.Google Scholar
Braarvig, Jens. 1985. “Dhāraṇī and Pratibhāna: memory and eloquence of the Bodhisattvas”, Journal of the International Association for Buddhist Studies, 8, 1729.Google Scholar
Csoma de Körös, Alexandre. 1881. “Analyse du Kandjour, recueil des livres sacrés au Tibet”, Annales du Musée Guimet 2, 131349.Google Scholar
Davidson, Ronald M. 2009. “Studies in Dhāraṇī literature I: revisiting the meaning of the term Dhāranī”, Journal of Indian Philosophy, 37, 97147.Google Scholar
Dash, Vaidya Bhagvan. 1991. Materia Medica Of Ayurveda. Based on Madanapāla's Nighaṇṭu. Delhi.Google Scholar
Eimer, Helmut. 1989. Der Tantra-Katalog des Bu ston im Vergleich mit der Abteilung Tantra des tibetischen Kanjur. Studie, Textausgabe, Konkordanzen und Indices. (Indica et Tibetica 17.) Bonn.Google Scholar
Falk, Harry and Strauch, Ingo. 2014. “The Bajaur and Split Collections of Kharoṣṭhī manuscripts within the context of Buddhist Gāndhārī literature”, in, Harrison, Paul and Hartmann, Jens-Uwe (eds), Recent Advances in Buddhist Manuscript Research (Beiträge zur Kultur- und Geistesgeschichte Asiens, 80; Denkschriften der philosophisch-historischen Klasse, 460). Vienna: Österreichische Akademie der Wissenschaftenm, 5178.Google Scholar
Forrest, Satnam K. Mendoz. 2011. Witches, Whores, and Sorcerers: The Concept of Evil in Early Iran. Foreword and other contributions by Oktor Skjærvø, Prods. Austin.Google Scholar
Gethin, Rupert. 1998. The Foundations of Buddhism. Oxford.Google Scholar
Hartmann, Jens-Uwe. 2000. “Zu einer neuen Handschrift des Dīrghāgama”, in Chojnacki, Christine, Hartmann, Jens-Uwe and Tschannerl, Volker M. (eds), Vividharatnakaraṇḍaka: Festgabe für Adelheid Mette. (Indica et Tibetica: Monographien zu den Sprachen und Literaturen des indo-tibetischen Kulturraumes, 37.) Swisttal-Odendorf, 359–67.Google Scholar
Hartmann, Jens-Uwe. 2004. “Contents and structure of the Dīrghāgama of the (Mūla-)Sarvāstivādins”, Annual Report of the International Research Institute for Advanced Buddhology at Soka University for the Academic Year 2003, 119–37. Hachioji.Google Scholar
Hidas, Gergely. 2012. Mahāpratisarā-Mahāvidyārājñī. The Great Amulet, Great Queen of Spells. Introduction, Critical Editions and Annotated Translation (Śata-Piṭaka Series, 636.) New Delhi.Google Scholar
Hidas, Gergely. Forthcoming a. “Rituals in the Mahāsāhasrapramardanasūtra” (Pre-typeset version). To appear in Mirnig, N. et al. (eds), Pushpika: Tracing Ancient India through Texts and Traditions. Contributions to Current Research in Indology. Volume I. Oxford.Google Scholar
Hidas, Gergely. Forthcoming b. Mahādaṇḍadhāraṇī-Śītavatī. An apotropaic scripture of the Buddhist Mantranaya. Introduction, critical edition and annotated translation.Google Scholar
Hoffmann, Helmut (ed.). 1939. Bruchstücke des Āṭānāṭikasūtra aus dem zentralasiatischen Sanskritkanon der Buddhisten. (Kleinere Sanskrit-Texte, 5.) Leipzig (repr. 1987. Monographien zur Indischen Archäologie, Kunstgeschichte und Philologie, 3. Stuttgart, 7–121).Google Scholar
Iwamoto, Yutaka (ed.). 1937. Mahāsāhasrapramardanī, Pañcarakṣā I. (Beiträge zur Indologie, 1.) Kyoto.Google Scholar
Hellwig, Oliver. 2009. Wörterbuch der mittelalterlichen indischen Alchemie. Groningen.Google Scholar
Kumar, Pushpendra (ed.). 2006. The Garuḍa Mahāpurāṇam (English translation by Kumar, PushpendraDelhi.Google Scholar
Lamotte, Etienne. 1949–80. Le Traité de la Grande Vertu de Sagesse de Nāgārjuna. Louvain.Google Scholar
Lefmann, S. (ed.). 1902. Lalita Vistara. Leben und Lehre des Çākya-Buddha. Textausgabe mit Varianten-, Metren- und Wörterverzeichnis. Halle.Google Scholar
Mitra, Rajendralal (ed.). 1870–79. Agni Purāṇa: A Collection of Hindu Mythology and Traditions. 3 volumes (Bibliotheca Indica, 65, 1–3). Calcutta.Google Scholar
Nattier, Jan. 2003. A Few Good Men. The Bodhisattva Path according to The Inquiry of Ugra (Ugraparipṛcchā). (Studies in the Buddhist Traditions.) Honolulu.Google Scholar
Ol'denburg, Sergej (ed.). 1899. Otryvki kašgarskich i sanskritskich rukopisej iz sobranija N.F. Petrovskago. Zapiski Voctočnago otdelenija Imperatorskago Russkago Archeologičeskago Obstchestva, 11 (1897–98). St. Petersburg, 207–64.Google Scholar
Pagel, Ulrich. 2007. Mapping the Path: Vajrapadas in Mahāyāna Literature. (Studia Philologica Buddhica. Monograph Series, XXI.) Tokyo.Google Scholar
Sakaki, Ryōzaburō (ed.). 1965. Bon-zō-kan-wa shih-yaku taiko hon-yaku meigi tai-shu. 2 vols. (Suzuki Research Foundation, Reprint Series, 1.) Tokyo (3rd ed.).Google Scholar
Sander, Lore. 1987. “Origin and date of the Bower Manuscript, a new approach”, in Yaldiz, Marianne and Lobo, Wibke (eds), Investigating the Indian Art. Berlin, 313–23.Google Scholar
Sander, Lore. 2007. “Preliminary remarks on two versions of the Āṭānātīya (Āṭānāṭika)-Sūtra in Sanskrit”, Journal of the International College for Postgraduate Buddhist Studies, 11, 159–96.Google Scholar
Schmithausen, Lambert. 1997. Maitrī and Magic: Aspects of the Buddhist Attitude toward the Dangerous in Nature. (Österreichische Akademie der Wissenschaften, Philosophisch- Historische Klasse, Sitzungsberichte, 652. Band; Veröffentlichungen zu den Sprachen und Kulturen Südasiens, Heft 30.) Vienna.Google Scholar
Skilling, Peter. 1992. “The Rakṣā literature of the Śrāvakayāna”, Journal of the Pali Text Society, 16, 109182.Google Scholar
Skilling, Peter. 1994. Mahāsūtras: Great Discourses of the Buddha. Volume I: Texts. Critical editions of the Tibetan Mahāsūtras with Pāli and Sanskrit Counterparts as Available. (Sacred Books of the Buddhists, 44.) Oxford.Google Scholar
Skilling, Peter. 1997. Mahāsūtras: Great Discourses of the Buddha. Volume II. Parts I and II. (Sacred Books of the Buddhists, 46.) Oxford.Google Scholar
Skorupski, Tadeusz. 1985. A Catalogue of the sTog Palace Kanjur. (Bibliographia Philologica Buddhica, Series Maior VI.) Tokyo.Google Scholar
Sørensen, Henrik H. 2006. “The spell of the Great, Golden Peacock Queen: the origin, practices, and lore of an early esoteric Buddhist tradition in China”, Pacific World Journal, Third Series, 8, 89123.Google Scholar
Strauch, Ingo. 2007/08. The Bajaur Collection: A New Collection of Kharoṣṭhī Manuscripts. A Preliminary Catalogue and Survey. Online version 1.1 (http://www.geschkult.fu-berlin.de/e/indologie/bajaur/publication/strauch_2008_1_1.pdf?1347696630).Google Scholar
Strauch, Ingo. 2008. “The Bajaur Collection of Kharoṣṭhī manuscripts – a preliminary survey”, Studien zur Indologie und Iranistik, 25, 103–36.Google Scholar
Strauch, Ingo. 2011. “The character of the Indian Kharoṣṭhī script and the ‘Sanskrit revolution’: a writing system between identity and assimilation”, in de Voogt, Alex and Quack, Joachim (eds), The Idea of Writing. Writing Across Borders. Leiden, 131–68.Google Scholar
Takubo, Shūyo (ed.). 1972. Ārya-Mahā-Māyūrī Vidyā-Rājñī. Tokyo.Google Scholar
Vaidya, P.L. (ed.). 1960a. Aṣṭasāhasrikā Prajñāpāramitā with Haribhadra's Commentary called Āloka. (Buddhist Sanskrit Texts, 4.) Darbhanga.Google Scholar
Vaidya, P.L. (ed.). 1960b. Saddharmapuṇḍarīkasūtra. (Buddhist Sanskrit Texts, 6.) Darbhanga.Google Scholar
Figure 0

Table 1. Comparative list of calamities (antarāya) and advantages (anuśaṃsā)

Figure 1

Table 2. Names of poisons from the mantra of BajC 3 and from the Mahāmāyūrī

Figure 2

Table 3. Comparative list of names of poisons from different Buddhist traditions and Āyurvedic literature