Chapter 5: Contact and Transformation: The Formation of Siddhānta Astronomy¶
5.1 The Problem of Origins¶
Between the five-year calendar of the Vedāṅga Jyotiṣa and the Āryabhaṭīya of 499 CE lies a transformation. The earlier astronomy had no planetary theory, no geometrical model of the heavens, no trigonometry and no method of predicting eclipses. The Āryabhaṭīya, and the systems summarised by Varāhamihira a few years later, have all of these: a spherical earth at the centre of a spherical cosmos, a zodiac of twelve signs divided into degrees and minutes, epicyclic models for the Sun, Moon and five planets, a table of sines, and complete procedures for eclipses. Almost nothing survives from the centuries in which this change took place, and historians have had to reconstruct it from fragments, from the later texts, and from comparison with Mesopotamian and Greek astronomy.
This chapter describes what is known about that transformation and the debate about it. Its conclusion, in brief, is that Greek astrology and elements of Hellenistic mathematical astronomy reached India in the early centuries CE and were decisive for the new science; and that what Indian astronomers made of them, from the sine function to the great integer cycles, was original and distinctive.
5.2 Mesopotamian Elements¶
Chapter 3 described the parallels between the Vedāṅga Jyotiṣa and the Mesopotamian compendium MUL.APIN (Pingree 1973). Some of the oldest layers of Indian omen literature also have Mesopotamian counterparts, and Pingree argued that a part of the omen material of the Gārgīya Jyotiṣa and of the Atharvavedic supplements descends from the great Mesopotamian omen series Enūma Anu Enlil, transmitted through the Achaemenid empire (Pingree 1997). The sexagesimal subdivision of the day into sixty nāḍikās, each of sixty vināḍikās, may belong to the same stream.
Later, a second wave of Babylonian mathematical astronomy reached India indirectly, through Greek intermediaries. The Vasiṣṭha system summarised by Varāhamihira (below) uses arithmetical methods for the Moon and planets that closely resemble the linear and step-function schemes of Seleucid-period Babylonian astronomy, which Greek astronomers had also adopted (Neugebauer and Pingree 1970–1971; Neugebauer 1975).
5.3 Greek Astrology in India¶
5.3.1 Routes of contact¶
After Alexander’s campaigns (327–325 BCE) Greek-speaking kingdoms ruled parts of Bactria and north-western India for two centuries, and trade between Roman Egypt and the ports of western India flourished in the first centuries CE. The Western Kṣatrapas, rulers of Gujarat and Mālava with their capital at Ujjayinī, were in close commercial contact with the Roman world. It was in their territory that Greek astrology was translated into Sanskrit, and Ujjayinī later became the meridian city of Indian astronomy.
5.3.2 The Yavanajātaka¶
The key document is the Yavanajātaka (“Horoscopy of the Greeks”), a Sanskrit verse treatise on Greek horoscopic astrology by Sphujidhvaja. In Pingree’s edition, the text states that it versifies a prose translation made by Yavaneśvara (“lord of the Greeks”) in a year Pingree computed as 149/150 CE, and that the verse version itself was completed in 269/270 CE (Pingree 1978b). The earlier date rests on an emended reading of a damaged verse, and Mak (2013), using newly found manuscripts, has questioned both the reading and Pingree’s account of the text’s history. The verse text belongs at any rate to the third century CE.
The Yavanajātaka introduced into Sanskrit the apparatus of Hellenistic astrology: the twelve signs of the zodiac and their attributes, the horoscopic “places” or houses counted from the ascendant, the planetary lords of signs, the aspects between planets, the decans, and the techniques for judging a nativity. Its final chapter contains rules for computing the positions of the Sun, Moon and planets, of a kind that Pingree identified as Greek adaptations of Babylonian methods. The work was followed by the Vṛddhayavanajātaka of Mīnarāja (perhaps fourth century) and by the astrological works of Satya and others that Varāhamihira used two centuries later.
5.3.3 Greek words in Sanskrit astronomy¶
The clearest evidence of the Greek contribution is linguistic. Varāhamihira, in the opening chapter of his Bṛhajjātaka, gives the Greek names of the signs alongside the Sanskrit ones: kriya (Greek krios, Aries), tāvuri (tauros), jituma (didymoi), kulīra, leya (leōn), pāthona (parthenos), jūka (zygon), kaurpya (skorpios), taukṣika (toxotēs), ākokera (aigokerōs), hṛdroga (hydrochoos) and ittha (ichthyes). Other loanwords became permanent parts of the technical vocabulary of both astrology and astronomy (Table 5.1).
| Sanskrit | Greek | Meaning in Sanskrit |
|---|---|---|
| horā | hōra | half of a sign; horoscopy as a whole |
| kendra | kentron | cardinal place of a horoscope; anomaly of a planet |
| liptā | lepton | minute of arc |
| drekkāṇa | dekanos | one third of a sign (decan) |
| jāmitra | diametron | the seventh place, opposite the ascendant |
| āpoklima | apoklima | the “cadent” places of a horoscope |
| paṇaphara | epanaphora | the “succedent” places |
Varāhamihira also gives a learned Sanskrit derivation of horā from ahorātra (“day and night”) by dropping the first and last syllables. This is a scholastic etymology; the word is Greek. The use of kendra, originally the Greek word for the “centre” of an epicycle, for the anomaly of a planet in Indian planetary theory shows that Greek geometrical astronomy, not only astrology, entered the Indian vocabulary.
The seven-day week, with days named after the seven planets in the order Sun, Moon, Mars, Mercury, Jupiter, Venus, Saturn, also came with Hellenistic astrology; its order derives from the Hellenistic scheme of “planetary hours”, explained in Chapter 13. The Yavanajātaka knows the week. Weekdays begin to appear in Indian inscriptions in the Gupta period.
Indian authors acknowledged their debt. Varāhamihira quotes the view that although the Yavanas are foreigners (mlecchas), this science is well established among them, and they are honoured like seers (Varāhamihira 1981–1982).
5.4 The Pañcasiddhāntikā¶
5.4.1 Varāhamihira’s compendium¶
The best view of Indian astronomy just before and around Āryabhaṭa comes from the Pañcasiddhāntikā (“Five Siddhāntas”) of Varāhamihira, whose epoch is 505 CE. Varāhamihira, who worked in the region of Ujjayinī and was active in the first half of the sixth century, summarised five astronomical systems then current, each named after a supposed author: Paitāmaha (“of the Grandfather”, Brahmā), Vāsiṣṭha (of the sage Vasiṣṭha), Romaka (“of the Romans”), Pauliśa (of Pauliśa, probably a Greek “Paulos”) and Saura (“of the Sun”). The text survives in only a few corrupt manuscripts. It was first edited by Thibaut and Sudhākara Dvivedī in 1889; the standard edition, with translation and a mathematical commentary, is that of Neugebauer and Pingree (1970–1971).
Varāhamihira also judged the five systems. He states that the Pauliśa is accurate, that the Romaka is close to it, that the Saura is more accurate still, and that the remaining two, the Paitāmaha and Vāsiṣṭha, are far from the truth (Neugebauer and Pingree 1970–1971).
5.4.2 The five systems¶
Paitāmaha.
The Paitāmaha system, summarised very briefly, is a version of the five-year calendar of the Vedāṅga Jyotiṣa, with a year of 366 days. Its inclusion shows that the old calendar was still known in the sixth century, and Varāhamihira’s verdict shows that it was obsolete.
Vāsiṣṭha.
The Vāsiṣṭha uses arithmetical methods for the Moon and planets akin to those of late Babylonian astronomy (linear “zigzag” functions and step functions), as transmitted in Greek and Greco-Egyptian astronomy before Ptolemy.
Romaka.
The Romaka has a tropical year of 365;14,48 days (in sexagesimal fractions, that is \(365 + \tfrac{14}{60} + \tfrac{48}{3600}\) days, about 365.2467 days), the value of the Greek astronomer Hipparchus, and a cycle of 2850 years, which is 150 times the nineteen-year (Metonic) luni-solar cycle. Its name, its year and its cycle all point to a Greek source.
Pauliśa.
The Pauliśa, associated with a Greek name, contained methods for the Sun, Moon, eclipses and planets, and Varāhamihira used it extensively. A “Pauliśasiddhānta” was also known to later authors and to al-Bīrūnī, but it is not certain that it was the same text.
Saura.
The Saura is a fully developed siddhānta, with a sidereal year of 365;15,31,30 days (about 365.25875 days), epicyclic models for the planets and procedures for eclipses. It is an earlier form of the Sūryasiddhānta, which survives in a substantially later recension (Chapter 6).
The Pañcasiddhāntikā thus shows Indian astronomy around 500 CE as a field in which older Indian, Babylonian-derived and Greek-derived systems existed side by side, and in which the most accurate of them had already taken the form of the classical siddhānta.
5.5 What Was Borrowed and What Was Made¶
5.5.1 Pingree’s thesis¶
David Pingree, who did more than anyone to edit and interpret the sources, argued that the mathematical models of the siddhāntas derived from Greek astronomy of the period between Hipparchus (second century BCE) and Ptolemy (second century CE), transmitted to India in the early centuries CE, before Ptolemy’s Almagest became standard in the Greek world. Since that pre-Ptolemaic Greek astronomy is almost entirely lost in Greek, he argued that the Sanskrit texts could be used to recover it (Pingree 1976, 1978a). On this view the Indian epicycle models, without Ptolemy’s equant, preserve an earlier Greek stage of planetary theory, and some Indian parameters go back to Greek ones.
5.5.2 Objections and qualifications¶
Pingree’s reconstruction has been criticised on several points.
-
Independent observation. Roger Billard’s statistical study of the mean motions in Indian texts showed that the parameters of several systems, including Āryabhaṭa’s, are most accurate close to the dates of the texts themselves, which implies that they were derived or corrected from observations made in India at that time (Billard 1971). Indian astronomers were not simply carrying Greek numbers forward.
-
Absence of direct evidence. No Greek text with the Indian planetary models has been found. The reconstruction of a lost Greek source is plausible but inferential.
-
Distinctive structure. The siddhānta models differ from anything known in Greek astronomy in important respects: the use of the sine instead of the chord; the variable size of the epicycles; the sequential application of the two planetary corrections; the integer revolution numbers in a great cycle; the day-count method; the sidereal frame. The case for a largely independent development was put by Indian historians of astronomy writing in the decades after independence, among them K. S. Shukla and other contributors to the history of Indian astronomy edited by S. N. Sen and Shukla (Sen and Shukla 1985). In broad terms their argument rests on these structural differences, together with Billard’s evidence for Indian observation: the characteristic methods of the siddhāntas have no counterpart in any surviving Greek text, so the planetary models should be regarded as built in India, and the borrowing as limited to a few elements, such as the zodiac of signs, the horoscope and some astrological vocabulary.
5.5.3 A balanced view¶
Most historians today take a middle position, well expressed by Plofker (2009). The Greek contribution was real and fundamental: the zodiac of signs and degrees, the geometrical model of the sphere, the idea of representing planetary motion by combinations of circles, the horoscope, and a body of technical vocabulary. Babylonian methods came both directly and through Greek intermediaries. But the classical siddhānta was not a translation. Indian astronomers created a new trigonometry based on the half-chord or sine (Chapter 6), a new form of computation based on integer periods and the day-count, new parameters derived from their own observations, and new models with variable epicycles and corrections applied in stages, procedures that some modern analysts interpret as approximating the effects Ptolemy obtained with the equant. Later they developed these further in directions without parallel in Greek astronomy, culminating in the Kerala school.
| Feature | Ptolemy, Almagest | Indian siddhāntas |
|---|---|---|
| Trigonometric function | chord, radius 60 | sine (jyā); radius \(3438'\) in Āryabhaṭa and the Sūryasiddhānta |
| Frame of longitudes | tropical (from the equinox) | sidereal (from the beginning of Aśvinī) |
| Mean motions | tables of motion from an epoch | integer revolutions in a great cycle; day-count (ahargaṇa) |
| First inequality | eccentric circle with equant | manda epicycle of variable size |
| Second (synodic) inequality | epicycle | śīghra epicycle of variable size |
| Combination of inequalities | single geometrical construction | corrections applied in sequence |
| Earth | stationary sphere at the centre | stationary sphere at the centre (Āryabhaṭa: rotating) |
| Order of the spheres | Moon, Mercury, Venus, Sun, Mars, Jupiter, Saturn | the same |
5.5.4 Astrology as the vehicle¶
Why did Greek mathematical astronomy enter India? Pingree’s answer was that it came with astrology. A horoscope requires the positions of the Sun, Moon and planets at the moment of birth, and the ascendant, which depends on the rising times of the signs at the place of birth. Greek astrological manuals therefore included, or presupposed, methods for computing these things, and the Indian astrologers who adopted the horoscope needed them too. The last chapter of the Yavanajātaka, with its rules for planetary positions, illustrates the connection (Pingree 1978b, 1997). If this is right, the demand for horoscopes, not curiosity about the planets for their own sake, was the main force behind the creation of the siddhānta, and the close association of astronomy and astrology in the later tradition (Chapter 15) is a legacy of its origins.
5.5.5 From chord to Sine¶
The most important mathematical change was the replacement of the Greek chord by the Indian half-chord. The Indian choice of radius, \(R = 3438'\) (the number of minutes in a radian), is itself a point of contact. The historian G. J. Toomer argued that Hipparchus’ lost table of chords, from the second century BCE, used the same radius, \(3438'\), and that the Indian Sine tables were derived from it by halving the chords of double arcs (Toomer 1973). Ptolemy, three centuries later, used a radius of 60 parts. If Toomer is right, the Indian Sine table is a transformed descendant of the oldest Greek trigonometry; the transformation, from chord to half-chord, was nevertheless an Indian step, and it made Indian trigonometry simpler and more powerful than its source. The Indian interval of \(3^\circ 45'\) corresponds to chords of arcs at intervals of \(7^\circ 30'\), the interval of Hipparchus’ table as Toomer reconstructed it.
5.5.6 Revisions after Varāhamihira¶
The Romaka and Pauliśa systems did not survive as working systems, but they left traces. Varāhamihira mentions an astronomer Lāṭadeva in connection with them, and later tradition remembered Lāṭadeva as a pupil of Āryabhaṭa. Brahmagupta rejected the Romaka as standing outside the tradition of the smṛtis, since it did not use their yugas and Manu periods; his objection shows that it was still felt to be alien to the Indian scheme of time. The Saura, revised, survived as the Sūryasiddhānta, and its claim to divine revelation, to the asura Maya at the end of the Kṛta age, gave it an authority that no human author’s text could have (Pingree 1981).
5.5.7 Heliocentric origins?¶
The Dutch mathematician B. L. van der Waerden proposed that the Indian planetary models, and especially Āryabhaṭa’s, derived ultimately from a heliocentric Greek system (Waerden 1987). His argument rested on the heliocentric ratios that can be read in the Indian parameters, which, as Section 7.3.3 shows, any accurate geocentric model must contain; the hypothesis has not been accepted by most historians (Plofker 2009).
5.6 Other Directions of Exchange¶
Contact did not run in one direction only. Indian astronomy and astrology were known in Sasanian Iran. Under the Sasanian kings, especially in the sixth century, Greek and Indian works on astrology and astronomy were translated into Middle Persian, and the Sasanian royal astronomical tables (the Zīk-i Shahriyārān, known in Arabic as the Zīj al-Shāh) appear to have used Indian parameters and methods alongside Greek ones. Through these Persian intermediaries, as well as directly, Indian elements entered the astronomy and astrology that the Abbasid caliphs inherited (Pingree 1963). Indian astrological material is also found in the Arabic astrological literature of the eighth and ninth centuries, which in turn shaped medieval European astrology. In the eighth century Indian siddhānta astronomy reached Baghdad directly, and Indian astronomy and calendars also travelled to Tang China and across Southeast Asia; these later transmissions are followed in Chapter 16.
Further Reading¶
Pingree (1978b) and Neugebauer and Pingree (1970–1971) are the essential editions; Pingree (1976) and Pingree (1978a) state Pingree’s thesis; Mak (2013) re-examines the Yavanajātaka; Plofker (2009) gives a balanced assessment; Billard (1971) is the main statistical study of the parameters.