outcome

2012.1.8 Bill M. Mak (Kyoto University)

1

Ratnaketu-parivarta, Sūryagarbha-parivarta and Candragarbha-parivarta of

Mahāsaṃnipātasūtra (MSN) -

Indian Jyotiṣa through the lens of Chinese Buddhist Canon 1

One of the greatest difficulties working with Indian texts is their general lack of a sense of

history. Historical events are often conveniently placed in some mythological past, and

even if dates and numbers are given, they are so fantastic that only the diehard literalists

would take them seriously. This is particularly true for Indian jyotiṣa

(astronomical/astrological) texts as the astral science which was thought to be transmitted

to humankind from the divine beings since time immemorial. In the case of compilations,

old texts were simply collated together, as in the example of Varāhamihira’s

Pañcasiddhāntikā, leaving it to the readers to decide what the interrelation between its

components are. In such a manner, the highly-esteemed jyotiṣa knowledge of different

historical periods is confounded together and only with painstaking philological skills and

a certain amount of luck can one reconstruct a textual history from purely internal

evidences.

External evidences often shed new lights on such kind of texts, but at the same time they

introduce a whole new set of problems. In many cases they lead to more questions than

answers. These include parallel materials in works within or without the same textual

lineage, as well as translations in other languages. The Buddhist jyotiṣa tradition had all

along emulated its mainstream Vedic/Hindu counterparts and by and large it preserves the

older form of Indian jyotiṣa. From the 3rd – 9th century CE, a large amount of these Indian

Buddhist jyotiṣa materials, many of them no longer extant, were translated into Chinese

and were included in various Chinese Buddhist canons. Thus the Chinese Buddhist

translations serve as a time capsule of Indian jyotiṣa materials throughout a large part of

1 This research was supported by a grant from the Kyoto University's Kyoto Erasmus Program which is funded by 'The Institutional Program for Young Researcher Overseas Visits' of the Japan Society for the Promotion of Science. The paper was presented on Jan 8 2012 in the panel “Sanskrit and Science” at the 15th World Sanskrit Conference, New Delhi. A full version of the paper is planned be published in the Proceedings of WSC.


2

the first millennium of our era. The questions we ought to ask are: Firstly, what exactly do

these Chinese materials contain and how reliable are they? Secondly, what can they tell us

about the nature and historical development of Indian jyotiṣa at least from the Buddhist

angle? Lastly, how does this historical insight compare with our current knowledge of

Indian jyotiṣa.

In this paper, my focus will be a text called Mahāsaṃnipātasūtra (MSN), translated into

Chinese in various stages from late 4th to late 6th century CE. This text contains extensive

references to Indian jyotiṣa and appears to be extremely popular in Central Asia as

Buddhism was spread from its native India to the rest of Asia. Leaving aside the Central

Asian and Chinese influences the text might have picked up by the time it reached the

Chinese in its present form, the text is basically a translation of an Indic exemplar no

longer extent. As it turns out, the Chinese translation of the Mahāsaṃnipātasūtra has in

fact preserved the different stages of Indian jyotiṣa and its structure and composition could

only be justified within a historically informed framework.2

I. Classification of Jyotiṣa literature

Before we proceed to the Chinese materials, let us look at how the jyotiṣa tradition was

conceived. One of the approaches to this question is to examine how the jyotiṣa literature

is classified, both traditionally as well as from the contemporary perspectives.

Traditionally, jyotiṣa was viewed as one of the auxiliary disciplines associated with the

Vedas (vedāṅga). The prevalent classification follows the tripartite scheme given in

Varāhamihira’s (505-578 CE) Bṛhatsaṃhitā, namely gaṇita (mathematical astronomy),

horā (genethlical astrology or horoscopy) and saṃhitā (natural astronomy or

miscellaneous divinations). Varāhamihira produced three major works in each of these

categories, namely Pañcasiddhāntika, Bṛhajjātaka, and Bṛhatsaṃhitā, which in turned

became the authority in the respective domains up to the present day. Given jyotiṣa is a

2 For a comprehensive, philological study of the Chinese translation of MSN, see Mak 2012 [forthcoming]. The astronomical materials have been treated extensively in Zenba 1957. At the present, only few Sanskrit fragments of the relevant passages have been identified (Hoernle 1916, Saerji 2005).


3

practical science, it is but natural that the materials are classified according to the genre

which closely related to their functions and operations.

A more historically rigorous approach was attempted by Pingree, who classified Indian

jyotiṣa based on the places of origin of the materials, which fall largely into five historical

periods: i) Vedic (ca. 1000 B.C. - 400 B.C); ii) Babylonian (ca. 400 B.C. - 200 A.D.): e.g.,

Vedāṅgajyotiṣa (VJ); iii) Greco-Babylonian (ca. 200-400): e.g., Yavanajātaka (YJ); iv)

Greek (ca. 400-1600): e.g., Āryabhaṭīya, PS; and v) Islamic (ca. 1600-1800).3 Pingree’s

scheme is the fruit of a lifetime’s work dedicated to comparing astronomical materials of

different ancient cultures, including Babylonian, Egyptian, Greek, Indian and Arabic.

While many of Pingree’s arguments are logical and well-found, there are cases which have

shown how Pingree’s over-emphasis and faith in Babylonian elements in Indian jyotiṣa

had led the great scholar into making some grave mistakes.4

A less speculative and more scientific way of classifying the jyotiṣa texts is to examine the

position of vernal equinox described therein. Due to the precession of equinoxes (at about

1 degree in 72 years), the reference point in different jyotiṣa texts also varied: a) Kṛttikā

(2350 B.C.): Taittirīya-s, Atharvavedapariśiṣṭa, VJ, BS; b) Bharaṇī (1300 B.C.): VJ; c)

Aśvinī (300 A.D.): PS, BS. Although this phenomenon is well recognized by scholars of

Indian texts, its implication has not been examined seriously, possibly since the materials

of different stages are often mixed together within a single text. Nonetheless, looking at it

from another angle, this observation is in fact not only helpful toward disentangling the

confounded Indian materials, but turns out to be an important way of making sense of the

whole corpus of the Chinese materials as we shall see.

A summary of the above three types of categorization is as follows:

a. Based on genre (Varāhamihira, 6th century CE)

3 Pingree 1981. 4 The most striking example would be Pingree’s edition of the 79th chapter of Yavanajātaka, which Pingree amended heavily to agree with Babylonian data. As K.S. Shukla later pointed out, most of Pingree’s emendations are incorrect and unnecessary. The original readings in Pingree’s apparatus were in fact mostly correct!


4

i) Gaṇita: Pañcasiddhāntika (PS) ii) Horā: Bṛhajjātaka (BJ) iii) Saṃhitā: Bṛhatsaṃhitā (BS) b. Based on origin/materials (Pingree 1981) i) Vedic (ca. 1000 B.C. - 400 B.C) ii) Babylonian (ca. 400 B.C. - 200 A.D.): Vedāṅgajyotiṣa (VJ) iii) Greco-Babylonian (ca. 200-400): Yavanajātaka (YJ) iv) Greek (ca. 400-1600): Āryabhaṭīya, PS v) Islamic (ca. 1600-1800) c. Based on position of vernal equinox

a) Kṛttikā (2350 B.C.): Taittirīya-s, Atharvavedapariśiṣṭa, VJ, BS b) Bharaṇī (1300 B.C.): VJ c) Aśvinī (300 A.D.): PS, BS

II. Chinese translations of Jyotiṣa materials

One of the key advantages of working with the Chinese Buddhist translations is that most

of them are relatively clearly dated and their textual history (at least the translations, and in

some cases, also the original) are often carefully documented. The earliest text extant

containing extensive jyoitṣa materials is Zhi Qian’s支謙 translation of

Śārdūlakarṇāvadāna (ŚKA), titled Modengjia jing摩登伽經 (230 CE). In this work, the

28 nakṣatra-s or the Indian lunar mansions, the 9 luminaries (graha-s), as well as

astronomical measurements such as the so-called Metonic cycle and the gnomic

measurements throughout the year are given. Subsequently, for the next 600 years or so,

new jyotiṣa and more sophisticated materials were progressively introduced. By the 9th

century, most of the basic Indian jyotiṣa materials found in the works Varāhamihira and

Āryabhaṭṭa, such as the Hellenisitc weekday order, the horoscopes and the ephemerides,

and some basic calendrical techniques are reflected the Chinese translations. Curiously,

within this corpus, MSN represent a transitional stage between the old and the new, where

new materials are tentatively introduced without the necessary elaboration.

An overview of the key Chinese translations containing Indian jyotiṣa materials are given

as below.

230 C.E. Mātaṅgasūtra 摩登伽經(=ŚKA)

T1300 by Zhi Qian支謙

28 nakṣatras from Kṛttikā, 9 graha-s, Metonic cycle, monthly gnomic


5

measurement 265-316 Akālakālasūtra

時非時經 T794 by Ruo Luoyan若羅嚴

gnomic measurement by half-month

308 Śārdūlakarṇāvadāna (ŚKA)舍頭諫太子二十八宿經

T1301 by Dharmapāla竺法護

Descriptions of 28 nakṣatra-s, explanation of units

[396?]<586 MSN大集經 T397 by Dharmakṣema/ Narendrayaśas

12 zodiacal signs

401-413 Mahāprajñāpāramitopadeśa 大智度論

T1509 by Kumārajīva鳩摩羅什

Four types of month

550-567 Commentaries to Abhidharmakośa

T1559/1644 by Paramārtha真諦

Reference to geocentric theory in India (Āryabhaṭīya?)

645-664 Various works T1545/1558/1563/ 1579 by Xuanzang玄奘

comprehensive introduction of Indian cosmology, calendrics and planetary size

702 Mahāmayūrīdhāraṇīrāja孔雀咒王經

T985 by Yijing義淨

Nine grahas

710-727 Various tantric works T1304,1310,1311 by Yixing一行

Horoscopy

718 Navagrahakaraṇa九執曆 by Gautamasiddha瞿曇悉達

Siddhānta algorithms

742-764 Xiuyao jing宿曜經 T1299 by Amoghavajra不空

27 nakṣatra-s from Aśvinī, weekday astrology

806-866 Qiyao rangzai jue七曜攘災決

T1308 by Jinjutuo(?)金俱吒

Planetary ephemeris

III. The Transitional character of MSN

The Chinese translation of the MSN contain three chapters (parivarta-s) – Ratnaketu,

Sūryagarbha and Candragarbha, which contain materials relevant to the present study:.

While all of them deal with various aspects of Indian jyotiṣa, they each have a different

character, representing their different sources, often dislocated in place and time. The

differences among the three chapters will be discussed under the following headings: a)

Context; b) Representation of the lunar mansions; c) Number and order of the luminaries;

d) Zodiac; e) Astronomical measurements.


6

a) References5

The jyotiṣa materials in the three chapters of MSN were attributed to different sources

according to the texts themselves: i) Jyotīrasa (Ratnaketu); ii) Jyotīrasa/Karoṣṭha/Garga

(Sūryagarbha); iii) Buddha/Brahma (Candragarbha). Broadly speaking, we can see that

jyotiṣa materials are often incorporated into Buddhist texts somewhat apologetically in the

early phase. In the Mahāyāna case, jyotiṣa knowledge is often considered a form of

expedient (upāya), employed for the benefits for the sentient beings though they are not

considered genuine Buddhist teachings per se. In the late stage, in particular among the

so-called Tantric works, they are seen as authentic Buddhist teachings and are often

employed directly with no justification given or required. The different contexts of which

jyotiṣa materials are incorporated in the three chapters of MSN thus coincide with this

general development. I would suggest this however is by no means coincidental as we shall

see further evidences of the transitional character of MSN which captured Indian jyotiṣa in

its different stages.

b) Lunar mansions

In almost all cases when the lunar mansions are mentioned, Kṛttikā is given as the first

nakṣatra.6 In one case, Bharaṇī is incoherently mentioned first among the nakṣatra-s. As

mentioned already earlier, the changing order of the lunar mansions (from Kṛttikā of 2350

BC to Bharaṇī 1300 BC) reflects the shift of vernal equinox, i.e. precession. MSN thus

reflected the old jyotiṣa materials which took Kṛttikā as the first nakṣatra. In India, after

300 CE, Aśvinī (equivalent to Aries) is established as the first nakṣatra. This is in fact also

reflected in the order of the zodiacal signs which we will see later, as well as in almost all

Chinese translations made after MSN.

5 The question why jyotiṣa materials, often irrelevant, if not antithetical to the Buddhist teachings, were incorporated into the Buddhist texts is a curious one and will require separate treatment. The topic will be dealt in a forthcoming paper of mine, titled “Indian jyotiṣa (astronomical/astrological) materials in Chinese Buddhist Translations - Why were they there in the first place?” to be presented at the symposium “Cross-Cultural Transmission of Buddhist Texts: Theories and Practices of Translation”. University of Hamburg. Jul 23, 2012. 6 In Ratnaketu, kṛttikā was translated to Jiao instead of Mao which reflected the confusion with the Chinese system on the part of the translator (Zenba 1957, Mak 2002).


7

Chapter First mansion Contents (D=descriptive; P=predictive)

T397-9 Ratnaketu Jiao角 (=Kṛttikā) D: mole in body part P: On individual (character and longevity) based on birth

T402 (Ratnaketu parallel)

Mao昴 (=Kṛttikā)

T397-14 Sūryagarbha Fasc. 41

Mao昴 (=Kṛttikā) D: i) Presiding deity; ii) Name; iii) Number of stars; iv) Shape of asterism; v) Span in degrees.; vi) Objects for worship

T397-14 Sūryagarbha Fasc. 42

Wei胃 (=Bharaṇī) P: i) General Divination; ii) On sickness; iii) On individual based on birth; iv) On individual based on conception…

T397-15 Candragarbha Jiao角 (=Kṛttikā?) D: Kūrmavibhāga - two sets of correspondence between mansions and Asian kingdoms

c) Luminaries

Nothing much can be said about the luminaries of MSN since no details except their

enumeration is given. It is nonetheless noteworthy that 8 and 7 graha-s are represented in

Sūryagarbha and Candragarbha respectively, in contrast to the 9 graha-s which were

already prevalent in India as seen in the work of Varāhamihira’s works of roughly the

same period. The order of the luminaries appear haphazard which is characteristic of

Buddhist works.7 Moreover, it shows no Hellenistic influence which is found in later

Chinese works.8 In this respect, the luminaries represented in MSN are not different from

those of earlier Chinese translations, reflecting an earlier representation of the graha-s in

ancient India whose order is not significant.

7 See Yano 1986:27-28. In the case of Sūryagarbha, the order of the luminaries is typical of the Chinese. 8 The earliest Hellenstic weekday order (☉ ☽ ♂ ☿♃ ♀ ♄) evident in Sanskrit work extant is found in Yavanajātaka 79.52-54. c. 250 AD (?).


8

Sūryagarbha: Jupiter-Mars-Saturn-Venus-Mercury-Sun-Moon-Rāhu (8) Candragarbha: Sun-Moon-Mars-Jupiter-Saturn-Mercury-Venus (7)

d) Zodiac

The twelve zodiacal signs seen in the Chinese Sūryagarbha and Candragarbha of MSN is

the earliest representations of the Zodiac within the Chinese corpus. The names of the

zodiacal signs are translated by meaning and by the Sanskrit pronunciation in the two

respective chapters, which are spuriously assumed to be translated by the same person,

Naredrayaśas. Regardless of who the translators of these two passages actually were,

apparently the materials were fairly new to the translator(s) since the translation style was

not yet fixed at this stage. Furthermore, no definitions are given on the horoscopes in the

Sūryagarbha, despite the zodiacal signs themselves are carefully incorporated into the

lunar mansion-based astrology. At any rate, the enumeration of the twelve zodiacal signs

beginning with Aries (equivalent to Aśvinī) represents the latest stage of astronomical

observation if the precession of equinoxes is taken into consideration as we have discussed

earlier.

Zodiac in Chinese translations of MSN Sign English Sanskrit T397-14

Sūryagarbha T397-15 Candragarbha (十二辰)

♈ Aries meṣa 持羊之神彌沙 ♉ Taurus vṛṣa 持牛之神毘利沙 ♊ Gemini mithuna 雙鳥之神彌偷那 ♋ Cancer karkaṭa 蟹神羯迦吒 ♌ Leo siṃha 師子之神呵 ♍ Virgo kanyā 天女之神迦若 ♎ Libra tulā 秤量之神兜邏 ♏ Scorpio vṛścika 蝎神毘梨支迦 ♐ Sagittarius dhanvin 射神檀尼毘 ♑ Capricorn makara 磨竭之神摩伽羅 ♒ Aquarius kumbha 水器之神鳩槃 ♓ Pisces mīna 天魚之神彌那


9

e) Astronomical measurements

One of the most intriguing piece of information we can glean from the astronomical

passages of the Sūryagarbha of MSN concerns the gnomic measurements and the

day-night ratio. Although rather unfortunately the length of the gnomon itself is not given

in the text, the day/night ratio on the summer solstice described (18:12) is possible only

when observed at around 32°30’ N.9 This ratio has been noted in a number of other

Mahāyāna jyotiṣa texts, suggesting these works could have be composed or at least mainly

circulated near the Northwest frontier of present India, where it is now known as the

Gandhāra region.

Conclusion

To sum, the following points may be made:

- Chinese Buddhist translations from c. 250 – 1000 CE capture the evolution of Indian

jyotiṣa in three stages – old, transition and new.

- Up to 6th century CE, Buddhist texts preserved the oldest type of lunar astrology (28

nakṣatra-s) in India.

- Sūryagarbha and Candragarbha of the compilation “Mahāsaṃnipātasūtra” (<600 A.D.)

reflects a transitional stage where new jyotiṣa elements such as the Zodiac and new

coordinates were introduced.

- Parallel development is reflected in Varāhamihira’s Bṛhatsaṃhitā where the older lunar

astrology and the newer horoscopy coexisted.

- By 8th century CE, tantric astrological works reflect an overwhelming interests in horā

which required a more sophisticated form of calendrics (e.g. ahargaṇa) and computational

astronomy (e.g. karana).

Chinese translations

Astrology Astronomy Indian correspondences

Old > 300 A.D. ŚKA MSN-Ratnaketu

Lunar astrology/ 28 nakṣatras

VE=Kṛttikā (2350 B.C.), 5 years yuga

TS, AV-P, Gargasaṃhitā

9 Niu 1981; Mak 2002.


10

Transitional c. 600 A.D. MSN-Candragarbha MSN-Sūryagarbha

Lunar astrology/ Zodiac / 7-8 grahas

VE=Kṛttikā/ Bharaṇī (1300 B.C.)

VJ, BS

New c. 800 A.D. - Navagrahakaraṇa, Xiuyao jing, Qiyao rangzaijue

Horoscopy based on 9 grahas /27 nakṣatras

VE=Aśvinī (300 A.D.), siddhānta, ephemerides

YJ, BJ, PS


11

Abbreviations AVŚ Atharvaveda. Śaukana. Edited by R. Roth and WḌ. Whitney. 1856. Atharva Veda

Sanhita. Berlin. BS Bṛhatsaṃhitā by Varāhamihira. Edited by Dvivedī, Kṛṣṇacandra. 1997.

Bṛhatsaṁhitā by Śrī Varāhamihirācārya with the Commentary of Bhaṭṭotpala. Varanasi: Sampurnanand Sanskrit University.

VJ Vedāṅgajyotiṣa by Lagadha. Edited by Kuppanna Sastry & K V Sarma. 1985. Vedāṅga Jyotiṣa of Lagadha in its Rk and Yajus Recensions. New Delhi: Indian National Science Academy.

Bibliography Braarvig, Jens. 1993. Akṣayamatinirdeśasūtra - The Tradition of Imperishability in

Buddhist Thought. Oslo: Solum Forlag. Hasuzawa Shōjun 蓮沢成純 (trans.). 1930-1. Kokuyaku issaikyō - Daishūbu 1-3 国訳一切経-大集部1-3. 東京: 大東出版社.

Hatani Ryōtai 羽渓了諦. 1934. 「大集経と佉羅帝との関係」.『宗教研究』新11:731-744.

Hoernle, August Friedrich Rudolf. 1916. Manuscript Remains of Buddhist Literature Found in Eastern Turkestan : Facsimiles with Transcripts, Translations, and Notes. Oxford: Clarendon Press.

Kurumiya Yenshu. 1975. Bibliographical Notes of the Ratnaketuparivarta. Tokyo: Rissho Daigaku Hokekyo Bunka Kenkyujo].

———. 1978. Ratnaketuparivarta : Sanskrit Text. Kyoto: Heirakuji Shoten. ———. 1979. Ratnaketuparivarta : 'Dus Pa Chen Po Rin Po Che Tog Gi Gzuns 'Dus Pa

Chen Po Dkon Mchog Dbal Zes Bya Ba'i Gzuns, Being the Tibetan Translation of the Ratnaketuparivarta. Kyoto: Heirakuji Shoten.

Leumann. 1908. "Über die einheimischen Sprachen von Ostturkestan im frühen Mittealter." Zeitschrift Der Deutschen Morgenländischen Gesellschaft 62:83-110.

Lévi, Sylvain. 1902. "Notes chinoises sur l'Inde." Bulletin de l'école Française d'extrême-Orient 2:246-255.

———. 1905. "Notes chinoises sur l'Inde. V. Quelques documents sur le bouddhisme indien dans l'Asie centrale (première partie)." Bulletin de l'école Française d'extrême-Orient 5:253-305.

Mak, Bill M. 2012 [forthcoming]. “Silk Road Transmission of Astrological Lore to China - Indian, Chinese and Central Asian elements in Mahāsaṃnipātasūtra (T397). Silk Road: Interwoven History. Cambridge, MA: Association for Central Asian Civilizations & Silk Road Studies.

Matsuda Shinya, 松田慎也. 1997. Daishūbu 「大集部」. In Daijō kyōten kaisetu jiten『大乗経典解説事典』. 東京: 北辰堂. 171-194.

Niu Weixing, 钮卫星. 2004. Xiwang fantian《西望梵天》. 上海: 上海交通大学出版社. Pingree, David. 1981. "History of Mathematical Astronomy in India." In Dictionary of

scientific biography. New York: Charles Scribner's Sons. 533-633. ———. 1989. "Indian Planetary Images and the Tradition of Astral Magic." Journal of the

Warburg and Courtauld Institutes 52:1-13.


12

———. 1990. "The Purāṇas and Jyotiḥśāstra: Astronomy." Journal of the American Oriental Society 110 (2): 274-280.

Saerji萨尔吉. 2005. 《〈大方等大集经〉之研究》. Ph.D. dissertation. Peking University. Shinjō Shinzō 新城新藏. 1928. 「二十八宿の傳來」. In 『東洋天文學史研究』. 京都: 弘文堂.

Yabuki Keiki 矢吹慶輝. 1934. “大集月藏經題解”. Kokuyaku issaikyō - Daishūbu 4 国訳一切経-大集部4. 1-14. 東京: 大東出版社.

Yano Michio 矢野道雄. 1986.『密教占星術』. 東京: 東京美術. ———. 1992.『占星術師たちのインド』. 東京: 中央公論社. ———. 2004.『星占いの文化交流史』. 東京: 勁草書房. ———. 2011. 『インド数学の発想 IT大国の源流をたどる』. 東京: NHK出版部. Zenba Makoto, 善波周. 1956. 「仏典の天文暦法について」. 『印度学仏教学研究』

4 (1): 18-27. ———. 1957. 「大集経の天文記事」.『日本仏教学会年報』22: 101-116.

outcome

Documents