PART THREE
A RECONSTRUCTION OF ANCIENT INDIAN HISTORY:
AŚOKA AND BEFORE AND AFTER
1
A Selective Resumé
Who was the Indian original of Sandrocottus, the adventurous youth whom the Greeks reported as having met Alexander the Great during his invasion of India in 326 B.C. and as having become king not long after? We have sought a decision between Chandragupta Maurya, the candidate of modern historians, and Chandragupta I, founder of the Imperial Guptas, the choice of those who go by the traditional-Purānic chronology.
After noting in some detail how this time-scheme can be seen to set Chandragupta I exactly where Sandrocottus stands, we have brought in a compelling logical consideration. The Purānic chronology, which refers to the Kaliyuga Era of 3102 B.C., is proved to have been in vogue by 634 A.D., the date of the Aihole Inscription of Pulakesin II which calculates both with the popular Śāka Era starting from 78 A.D. and with the Kaliyuga Era to which it links the Bharata War. So the Purānic pundits, who are said by modern historians to have arranged their chronology in the centuries just before 634 A.D., would be, according to the modern time-scheme, living face to face with the Imperial Guptas (320-570 A.D.) and yet putting these monarchs - their own contemporaries - in the far-off age of Alexander and in the centuries immediately after him.
Surely, nobody could commit such an extreme "howler". Therefore, the founder of the Imperial Guptas could not have commenced his line in the period currently assigned to him but must have been Sandracottus. Likewise the Purānic pundits must have finalised their chronology during the 4th-3rd centuries B.C.
Along with our elucidation of that chronology and with our reductio ad absurdum of the modern time-scheme for the Guptas, we have made an appeal to the Indica of Megasthenes, the ambassador sent to the court of Sandrocottus in c. 302 B.C. by
Page 229
Seleucus Nicator, the principal successor of Alexander the Great in the Orient. A thorough scrutiny of the information from Megasthenes on the chronological scheme of the Indians contemporary with him has provided us with an astonishing detailed parallel to the historical vision we can ascribe to the most ancient pundits of the Purānas from the extant versions of these writings, including the Yuga-Purāna section of the Gārgi-Sarhhitā. One of the most impressive features is the number of kings - 153 or 154 - Megasthenes counts between Alexander or Sandrocottus on the one hand and, on the other, Dionysus whom he mentions as the first Indian monarch. When, taking the help of F.E. Pargiter's collated lists from the Purānas as well as some Purānic genealogies found elsewhere, we count backward from Chandragupta Maurya through the Magadhan line and its predecessors, we do not come anywhere near Megasthenes's number - even though we count up to Manu Svayambhuva, the first traditional king. But going backward from Chandrgupta I, founder of the Imperial Guptas, we get exactly 153 or 154 kings between him and Prithu Vainya who is called the Ādi-rāja, the first consecrated king, in the Purānas and other Indian literature. On comparing what the Greek reports tell of Dionysus in India and what India's tradition says of Prithu Vainya we discover convincing reasons to see the former as variously shadowing forth the latter. Everywhere the vote of the Greek who lived with Sandrocottus himself gets cast in favour of the first Gupta instead of the first Maurya.
We may glance at another happy "coincidence". On the basis of Megasthenes and some subsequent Classical writers, Strabo has drawn attention to an important point about the royal line of the Prasii (Prāchya=Easterner) kingdom, the rulers of Palibothra (Pātaliputra) in the post-Alexandrine age. He informs us that the reigning king, "as for example, King Sandrocottus to whom Megasthenes was sent on an embassy", had, among other items of nomenclature, his family name always going with his personal one. Strabo adds: "Such is also the custom among the Parthians; for all are called Arsaces, although personally one king is called Orodes, another Phraotes, and another something else." This peculiarity means that in the name "Sandrocottus" the family name was included: "Sandro" could be the personal name and "cottus" an indication of the family. Similarly, according to Strabo, the successors of Sandrocottus were characterized: he speaks in the plural of
Page 230
the Prasii kings. Such a fact could be marked only for the Imperial Guptas. Each of them bore the family title "Gupta" linked to his own designation: Chandragupta I, Samudragupta, Chandragupta II, Kumāragupta, Skandagupta and so on. The Mauryas were entirely different in this respect. Their founder has "Chandragupta" as wholly his personal name and neither he nor any successor of his carried his family designation as inseparable from his own. In fact, none of them even refers to it. Nowhere in the numerous edicts of Chandragupta's grandson Aśoka does "Maurya" appear. The three inscriptions of Aśoka's grandson-Daśaratha are also lacking in it. Their practice is in so striking a contrast to that of the Guptas that the information proffered by Megasthenes-Strabo cannot but exclude them and focus on the latter.
Our study of the evidence relating to the accession of Sandrocottus has led us to conclude that the Imperial Gupta line was started in 315 B.C. and not, as modern historians hold, in 320 A.D. As for the alleged testimony to the second date by Albērūnī, the Arab scholar who visited India in 1031 A.D., we have found it glaringly misinterpreted by modern historians: it clearly makes the Gupta Era mark the end and not the beginning of the Imperial Gupta line. Albērūnī mentions other eras also starting at a point of destruction or disappearance. Then we dealt with Sylvain Lévi's so-called "synchronism" - based on a Chinese report - of Samudragupta (the son of Chandragupta I), whose reign is currently made to end in 376 or 380 A.D., and the Ceylonese king Sirimeghavanna, dated at present to 352-376 A.D. by the Buddha Era which Geiger postulates as 483 B.C. We cited the frank admission of one of our best historians, H.C, Raychaudhuri, that Geiger's epoch is not explicitly recognised in tradition. The eminent Ceylonese scholar S. Paranavitana was quoted to show that no such era had ever been in vogue and that the introduction of it creates several confusions in chronology. Lévi's synchronism is exposed as a self-contradiction on the part of its proposer who, on the strength of Chinese reports for the 5th-8th centuries A.D. of Ceylonese history, demonstrates the soundness of the traditional Buddha Era of Ceylon, namely, 543 B.C. According to that era Sirimeghavanna gets -dated to 304-332 A.D. But if, as generally assumed, Chandragupta I founded the Gupta Era of 320 A.D. on his accession and if, as is most likely, his accession coincided with his marriage to Kumāradevī, the Lichchhavi princess, how could
Page 231
Samudragupta who was Kumāradevī's son be a contemporary of Sirimeghavanna? His Allāhābād Pillar Inscription pictures his father as abdicating and choosing him for his prowess and ability as "worthy to rule this whole world". In 332 A.D., Sirimeghavanna's final year, Samudragupta would be merely 12 years old. Further, the same inscription classes Simhala (Ceylon) as a "vassal state": the Ceylon of Sirimeghavanna was never feudatory to any Indian king. A second smaller Samudragupta looms up in the early part of the 4th century A.D. at the tail-end of the Gupta dynasty as viewed à la Albērūnī. This possibility is regarded as acceptable for a number of reasons.
As the most important side-issue out of many, we have picked out the question of the identity of the king whom the Greeks named Xandrames and who was on the throne inland when Alexander halted at the river Hyphasis (Beās) and Sandrocottus was still uncrowned. In all Indian records there is only one name echoing Xandrames - the powerful Nāga king Chandrarhsa listed by the Purānas in the period just preceding the rise of the Guptas. Closely analysing the Greek evidence we have tried to stake a convincing claim for him.
Against all our apparently irrefutable arguments for Chandragupta I as Sandrocottus, the champions of the modern chronology are expected to point, first and foremost, to the secure seat which Aśoka seems to have around the middle of the 3rd century B.C. So solid in their time-indications the Rock Edicts of this Maurya emperor are taken to be that a veritable dilemma appears to present its formidable horns. On the one hand, our equation has an unanswerable look: on the other, the Rock Edicts give the impression of being incontrovertible.
Believing that the former horn cannot be blunted, we shall undertake to turn the latter. A new interpretation of the Aśokan pointers has to be attempted. Let us then set them forth in their full strength against us.
The Aśokan Support for the Modern Time-Scheme
The famous Rock Edict XIII first confronts us. There are two things here, both connected with the designation "Yona", the Prākrit form of the Sanskrit "Yavana". This designation is understood as derived from the Old Persian "Yauna", which connoted
Page 232
"Ionian" and denoted "Greek" in general, in the Achaemenid inscriptions starting from the one of c. 520-518 B.C. at Behistun.
Aśoka had a community of Yonas as his subjects, and in some versions of R.E. XIII he tells us that there is no country in his empire except that of the Yonas where are not found those two religious orders, the Brāhmanas and the Śrāmanas. This exclusive distinction is precisely what we should expect if the Yonas had the Hellenic and not the Hindu Āryān civilisation. But, if thus they were definitely the Greeks, what more natural than that Aśoka should have reigned some time after Sandrocottus, as the grandson of Chandragupta Maurya?
A still greater proof from the same edict is that Aśoka mentions, among the independent powers outside his empire, a group of five rulers, the first of whom he terms "Yona rāja". They are: Amtiyoka, Tulamāya, Amtikini, Magā and Alikasudara. These rulers can be equated to five Greek kings who reigned in the period after Alexander's immediate successors. They were Anti-ochus II Theos of Syria and Western Asia, Ptolemy II Phi-ladelphus of Egypt, Antigonus Gonatas of Macedonia, Magas of Cyrene and Alexander of Epirus or else his namesake of Corinth.
The reign-period of these kings were 261-246 B.C., 285-247 B.C., 278-239 B.C., 300-258 (or 250) B.C. and 275-255 B.C. or, if Alexander of Corinth is preferred, 252-247 B.C. The reign-periods of Chandragupta Maurya and his son Bindusāra (Aśoka's father) are taken respectively to be 24 and 25 years (according to the Indian Purānas) or 24 and 27 or 28 (according to the Ceylonese Chronicles). So Aśoka would come to the throne either 49 or 52 or 53 years after Chandragupta. Our historians are mostly disposed to put the latter's coronation in 321 B.C. Then Aśoka would be king in 272 or 269 or 268 B.C. According to the Purānas, his reign lasted for 36 years and, according to the Chronicles, 37. Therefore, it extended down to some year between 236 and 232 B.C. and overlapped, wholly or in part, with those of the five Greek kings. Thus there is very good chronologiocal agreement, and modern historians feel quite justified in their time-scheme.
Perhaps even more important than Rock Edict XIII is the recently discovered Aśokan inscription in two languages - Greek and Arāmaic - near Kandahār in Southern Afghānistān,1 the site
1. The most thorough-going studies, from the standpoint of modern historians, are Serie Orientate Roma XXI, 1958,8, pp. 1-35, Journal Asiatique CCXLVI, 1958,
Page 233
most probably of the ancient Alexandria of Arachosia where, as we learn from Strabo (XV.2.9), Alexander had established a Greek colony and which Seleucus Nicator later transferred to Sandrocottus.
The Greek version is in good Greek of the third and second quarters of the 3rd century B.C. (that is, c. 275-225 B.C.) and stands above the Arāmaic version. The matter of the inscription resembles that of R.E.s I and IV and Minor R.E.s I and II. A point to be especially noted is the presence of moral precepts like those in R.E. IV - precepts which in Aśokan edicts never go without a mention of respect towards Brāhmanas and Śramanas. But in the Kandahār inscription there are no Brāhmanas and Śramanas. We are reminded at once of the state of affairs in Aśoka's Yona country.
And the Arāmaic version encourages us. It is sprinkled with Irānian words in the manner of the official proclamations of the Achaemenid emperors from the time of Darius I (522-486 B.C.) onward in the territories of their Irānian subjects. Aśoka's R.E. XIII has the compound "Yona-Kamboja", showing the extremely close association of the Yonas with the Kambojas in his empire, the Kambojas who are known from linguistics as well as from literary evidence to have been thoroughly Irānianized, if they were not themselves Irānians. Hence the Arāmaic version must have been for the Kambojas, and the bilingual inscription as a whole locates Aśoka's Yona-Kamboja.
This inscription is held not only to prove that "Yona" meant "Greek" and that Aśoka's empire included Greek subjects. It is held also to provide an unchallengeable crowning touch to the chronology which places Aśoka in the India of the second generation after Alexander.
Confirmatory of the Kandahār bilingual's testimony are the implications seen in the three other epigraphs also found in Afghānistān in subsequent years. One is a fragment in Greek, picked up from the Kandahār market, translating the end of R.E. XII and the beginning of R.E. XIII.1 Once more the Greek is of
1, pp. 1-48, Epigraphia Indica XXXIV, I, p.lff., Indo-Asian Culture, April 1960, pp. 357-363.
1. Academie des Inscriptions el Belles-Lettres, Janvier-Juin, Paris, 1965, pp. 126-140; Journal Asiatique, Paris, 1966, Fascicule 2, pp. 132-157.
Page 234
the period 275-225 B.C. The remaining two epigraphs are in Aramaic. The one obtained in fragmentary form again from the market of Kandahār in the same year as the Greek inscription incorporates Indian Prākrit words in the Arāmaic text in the manner of the two Aśokan fragments found a long time back at Taxila and Pūl-i-Dārunteh respectively - and, as may be expected from a message addressed not to the Yonas but to the Kambojas, it includes an exhortation to revere Brāhmanas and Śramanas.1 The other Arāmaic edict, which came to light in 1969 on a rock-face in Laghman (ancient Lampāka) refers to the oasis of Palmyra in the Syrian desert, under its Semitic name "Tadmor", a famous city by which passed the roads leading to the realms of all the five Greek kings with whom the five monarchs listed in R.E. XIII have been identified.2 It is as if Aśoka revealed his gate of entry to the Western world where he promulgated his dharma.
Some Anomalies of Aśoka's Present Date
Before we directly tackle Aśoka's Yonas, the five Greek-sounding kings headed by a "Yona Rāja", the Kandahār inscription in Greek and Arāmaic as well as the three most recent epigraphic finds, we may formulate some general reasons to ask for a reinterpretation of this manifold evidence. There are a number of anomalies in putting Aśoka around 250 B.C.
The Greek kings with whose names the five rājas in R.E. XIII have been equated were not the only ones in the middle of the 3rd century B.C. It is not as if Aśoka listed just five rājas who might be Greek and there were just five Greek monarchs in the period allotted to Aśoka. The posing of Alexander of Corinth as an alternative to Alexander of Epirus is itself a proof that the number of Greek monarchs was more than five and it prompts us to ask why one of the two Alexanders, neighbours to each other, was left out. What could have put a check on Aśoka's mighty zeal to spread his dharma'? The question is especially pertinent when we look around a little more. R.K. Mookerji,3 after choosing on the ground of relative importance Alexander of Epirus and not his
1.Academie des Inscriptions et Belles Lettres, 1966, pp. 440-451; Journal Asiali-que. 1966, pp. 437-458.
2.Academie des Inscriptions el Belles Lettres, Janvier-Mars, 1970, pp. 158-173.
3.Chandragupta Maurya and His Times, p. 74.
Page 235
namesake of Corinth for Alikasudara, remarks: "There were several kings of Asia Minor of equal and higher status whom Aśoka should have mentioned such as Eumenes of Pergama (262-240 B.C.) or, nearer home, Diodotus of Bactria."
How natural would have been the inclusion of Diodotus is shown by F. W. Thomas's remarks1 in the course of determining the date of R.E. XIII: "The omission of the Bactrian ruler Diodotus, whose independence of the Seleucid empire dates from about 250 B.C., confirms the inference that the edict is not long posterior to the year 258."
Doubly natural would the inclusion have been in view of the recent revision of Bactrian history. C. 250 B.C. is no longer considered valid for Diodotus's declaration of independence. A. K. Narain2 writes: "Newell has very ably shown from the numismatic evidence that the break away of Bactria from the Seleucids is to be placed in 256-255 B.C." Diodotus reigned up to c. 245 B.C.3 A glance at the series of the other kings' dates is enough to demonstrate that he and Eumenes fall within the time-span measured by this series. The omission of Eumenes and Diodotus confuses the whole situation, particularly as Bactria where Diodotus reigned is not only next door to India but also known to have been in close touch with things Indian. Indeed, in the very age to which Aśoka has been assigned, we have archaeological evidence of Indo-Bactrian contact. The Bhir mound, the oldest part of the Taxila site, has yielded some fine jewellery, dating from 250 B.C. and associated with a gold coin of Diodotus.4 How could Aśoka, round about that time, have slurred over Diodotus and his Bactria?
As for Eumenes, it will not do to build a defence for his omission on the fact that he did not officially declare himself king as did his successor Attalus I. There was no question of his not being recognised as king. At any rate, the point is nominal, for Pergamon had been independent since 281 B.C.5 and, whether self-designated king or no, Eumenes ruled his state like one and
1."Aśoka, the Imperial Patron of Buddhism", The Cambridge History of India, I, p. 502.
2.The Indo-Greeks, p. 14.
3.V.A. Smith, The Early History of India, p. 235.
4.Journal of the Punjab Historical Society, Vol. Ill, p. 9; Archaeological Survey of India, 1912-13, p. 41, pi. xxxix.
5. The Encyclopaedia Britannica (13th Ed.), Vol. XXI, p. 143.
Page 236
proved his power by defeating at Sardis in 262 B.C. the accredited king Antiochus I of Syria, the father of the Antiochus our historians have made Aśoka's frontager. His case may be compared with that of Pushyamitra, founder of the Śunga line, who never took the title of King but retained his title of "Commander-in-chief even when he had obtained imperial power.1 This power was sufficiently great for him to have been treated as a regular monarch and most historians do not even know that he called himself merely Senānī.
Furthermore, Eumenes was already known in India from 316 B.C. Diodorus Siculus (XIX.34) reports that in that year the leader of an Indian contingent which had gone to fight under Eumenes in Irān was killed in battle and his two wives competed as to which was to be the sati.2
Our amazement at Aśoka's inconceivable omissions would cease if we refrained from dating him to the post-Alexandrine age. No query could be raised of any kind, were his rājas five border rulers in an age much earlier, about whom we know next to nothing.
Still more strange than Aśoka's blind spot for Diodotus and Eumenes is the absolute ignorance we find about Aśoka in the Classical records concerning the kings in whose domains he is supposed to have propagated his dharma. There was a great liveliness of Indo-Greek contacts in the 4th and 3rd centuries B.C. And three ambassadors were despatched to the Indian courts from the Greek kingdoms. Seleucus Nicator sent Megasthenes to Sandrocottus; his son Antiochus I sent Daimachus to the son of Sandrocottus whom the Greeks called Amitrachates or Amitro-chades;3 Ptolemy II Philadelphus sent Dionysius to an unspecified king who must be either Amitrachates or his son.4 Greek historians have recorded in some detail the dealings of Alexander the Great with Indian monarchs and of Seleucus with Sandrocottus. According to Athenaeus, Sandrocottus sent presents including certain powerful aphrodisiacs to Seleucus.5 Both Hegesander
1.Dikshit in Indian Culture, Vol. VI, p. 196.
2.E. R. Bevan, "India in Early Greek and Latin Literature", The Cambridge History of India, I, p. 415.
3.Athenaeus XIV, 67 for the former and Strabo III, i, 9 for the latter. Actually. Strabo has "Allitrochades", but that is obviously a mistaken transcription.
4.E. J. Rapson, Ancient India (Cambridge, 1916), pp. 102, 103. 104.
5.H. C. Raychaudhuri, The Political History of Ancient India, 5th Ed., p. 273.
Page 237
and Athenaeus are responsible for the piquant report that has become a favourite with our historians: Amitrachates wrote to Antiochus I, asking him to purchase and send him not only sweet wine and dried figs but a sophist, only to be reminded that it was not lawful in Greece to sell a sophist! Drawing upon Megasthenes or some of Alexander's generals, the Classical writers Diodorus, Strabo, Pliny, Curtius, Arrian and Aelian have left us histories chockful of facts and fables about India. But after Sandrocottus and Amitrachates we get no name of any king of their line. Such neglect is in itself a little hard to explain, but we can grant the possibility of some explanation if we do not identify, as our historians do, Sandrocottus with Chandragupta Maurya, Amitrachates with Bindusāra and the son of Amitrachates with Aśoka -Aśoka who is supposed to have established by his dharma-mission greater ties with the Hellenic world than either Sandrocottus or Amitrachates had done. With Aśoka put in the time of those five Greek kingdoms, the neglect becomes wholly unintelligible.
Ptolemy II Philadelphus not only sent an ambassador to India but, as Athenaeus tells us, included in his processions Indian women, Indian hunting dogs, Indian cows and Indian spices on the backs of camels.1 Epiphanius informs us that the Head of Philadelphus's Library at Alexandria was anxious to translate the books of the Hindus.2 But if Ptolemy was Aśoka's Tulamāya, in whose domain as well as in the domains of the four other rājas Aśoka the Buddhist emperor claimed to have spread his precepts of virtue and, as R.E. II records, even instituted hospitals for ailing men and animals, why is there not the slightest hint of Aśoka's activities in the historical passages relating to Ptolemy? How can we account for this complete lack of interest here when interest is abundantly shown in so many Indian things?
It will not help at all to deny the lack by asking: "What about the 'Buddha-head' found in Memphis and dated by Sir Flinders Petrie as far back as the 5th century B.C. when that city was the capital of Egypt under the Persians? Does it not show sufficient Buddhist influence in the land of Ptolemy II Philadelphus?" If any influence is shown by the "Buddha-head", it cannot be attributed to Aśoka's activities or to any similar missionary work which might bring the Dhamma into prominence and keep it in the public eye until the
1.The Age of Imperial Unity, p. 616.
2.D. A. Bhandarkar, Aśoka (University of Calcutta, 1932), p. 163.
Page 238
middle of the 3rd century B.C., when Ptolemy ruled over Egypt. If the head (with Gurkha features) is really of Buddha, the influence was not of the kind with which we are concerned but of an accidental, limited, private and temporary nature, for this head is connected by Sir Flinders with members of a hypothetical Indian component of the Memphis-garrison which, according to Herodotus, who visited the city in 453 B.C., consisted largely of Persians but also of "others" - a small component which T. Balakrishnan Nayar, discussing the find, considers to have left in 405 B.C. (long before the time of Philadelphus): "When the Persian occupation ceased, the Indian colony at Memphis also would have ceased."1
The utmost that has been possible to say on the positive side for our period is communicated by R.A. Jairazbhoy.2 After mentioning the statement in R.E. XIII of Aśoka's victories of morality in foreign countries, Jairazbhoy writes: "There is no direct confirmation of this from western sources, but a wheel with trisula upon it was found on a Ptolemaic gravestone in Egypt, and it is thought that the revolving wheels in Egyptian temples referred to by Heron of Alexandria (c. 250 B.C.) may be influences from Buddhist India." But what is thought of the revolving wheels is as good as nullified by Jairazbhoy's very next sentence:3 "Heron (Prop. 31) says that these bronze wheels were placed in the porticoes of Egyptian temples for those who enter to revolve, the belief being that bronze purifies." The whole emphasis shifts from wheels as such to their constituent material, bronze, which has nothing to do with Buddhism. And a little earlier in his book Jairazbhoy4 offers us information which makes it gratuitous to connect with Buddhist influences these wheels or that gravestone wheel with triśula upon it:
"The emblem of a sun disk carved in relief on Buddhist monuments, for example, at Amarāvatī ultimately originates in Assyria, and similarly the so-called Buddhist triśula ornament is patterned on the winged solar disk of Assyria. Moreover Mesopotamian sun pillars surmounted by disks (surinnu) which stood at entrances to temples, or were borne by priests of the sun gods, are prototypes
1.The Hindu, Madras (Sunday edition, date not traceable).
2.Foreign Influence in Ancient India (Asia Publishing House, Bombay, 1963), p. 63.
3.Ibid.
4.Ibid., p. 58.
Page 239
of sun wheels on pillars at the Buddhist stupas."
Whether or not we agree with Jairazbhoy on the origin of Buddhist symbols, it should be clear that even the signs picked out from Ptolemaic Egypt to exhibit in an indirect manner Buddhist influences need have no relation with Buddhism. If they are at all derived and not independently developed, they may well reflect Assyrian religious symbolism.
All in all, there are no positives to go by. Not that the ancient Western world was quite unaffected by Buddhism. The Jewish religious body known as the Essenes, as early as the middle of the 2nd century B.C., seems to have been affected in subtle ways by Buddhist monastic practices, and the early Christians themselves were perhaps tinged with a Buddhist attitude in their stress on fellow-feeling and charity. But there is no ostensible Buddhism even here. And in a still earlier age we can put our finger on nothing appreciable, however subtly, when the grand phrases of the Edicts would lead us to expect a great deal, at least in Syria, Egypt, Cyrene, Macedonia and Epirus.
Could it not then be that Aśoka was totally unconnected with Ptolemy or any Greek king of the 3rd century B.C. but dealt only with non-Greek rājas and their countries adjacent to his empire at a date removed from this period?
The Greeks and Buddhism
There is also the fact that to the whole Greek world Buddhism in any expressible form was an unknown quantity until we reach the 2nd century A.D. "Greece," says R.C. Majumdar,1 "knew nothing of Buddhism previous to the rise of Alexandria in the Christian era. Buddha is first mentioned by Clement of Alexandria (A.D. 150-218)." Can we imagine such ignorance if five Greek kingdoms received from Aśoka missionaries who, for all the general humanitarian bearing of their message, were representatives of a great Buddhist emperor?
Matching this lack of knowledge in the Greek world abroad, there is the absence of the slightest pointer to Buddhism from the account given of India by Megasthenes and the generals of Alexander. During Alexander's invasion the Greeks first saw Indian
1. "India and the Western World", The Age of Imperial Unity, p. 616.
Page 240
religious men at Taxila. From the description of them B. M. Barua1 recognises in their chief "the leader of a Brāhmanical sect of Gymnosophists (naked ascetics), evidently of Śaivite persuasion and closely allied to the Āījlvikas..." Megasthenes speaks of "Brachmanes and Sarmanes" (Brāhmanas and Śrāmanas). E.R. Bevan2 observes: "it has ... been thought that we have in the Sarmanes of Megasthenes the first mention of Buddhists by a Western writer. In the description however there is nothing distinctively Buddhist, and the term çrāmana is used in Indian literature of non-Buddhist ascetics. If therefore the people to whom Megasthenes heard the term applied were Buddhists, he must have known so little about them that he could only describe them by features which were equally found in various sorts of Hindu holy men. His description applies to Brahmin ascetics rather than to Buddhists." Mookerji3 adds that "from the description of these Śramanas" they seem to be "brahmins of the third and fourth Āśrāmas of life and known as Parivrājakas and Sannyāsīs."
In the practice of some Śarmanas, however, we have a hint of Jainism. Barua4 says: "Though the Nigranthas or Jainas are not expressly mentioned, their inclusion among the Indian Hylobioi is evident from the reference to the ascetic practice of remaining 'for a whole day motionless in one fixed attitude', say in the posture of one called ubbhatthika in Pali." And from Megasthenes we get in addition to the Hylobioi other differentiations like Pramnai (Prā-mānikas) and Gymnetai. He also mentions more than one specific religious cult - the worship of a Bacchanalian god by the Oxydra-kai (Kshudrakas) and of Heracles (apparently Shiva) by the Sibae (Sibis) and of Krishna under the name again of Heracles by the Souresenoi (Śurasenas) at Mathurā and elsewhere. But the Buddhists could not impress on him any characteristics. And this is most remarkable when we know, as J. Charpentier5 says, that Megasthenes spent years at Pātaliputra in the very province, Magadha, where Buddha had developed his chief activity.
1.Aśoka and His Inscriptions. Part I, p. 319.
2."India in Early Greek and Latin Literature'', The Cambridge History of India, I, p. 420.
3.Op. cit., p. 300.
4.Op. cit., p. 241.
5.Journal of the Royal Asiatic Society, 1925, p. 735.
Page 241
The milieu of Megasthenes in Magadha seems no part of the period between the death of Buddha and the reign of Aśoka - the period in which two Buddhist Councils had already been convened under royal patronage and the new religion had been steadily, though slowly, on the march. Even in Buddha's lifetime his teaching had extended to at least a couple of provinces outside Magadha - Āvanti and Śūurasena (the home of Megasthenes's Souresenoi) where, as Barua1 tells us, "important centres of Buddhism were founded..." Megasthenes appears to have come in an age in which Buddhism was suffering a severe slump, as it could if Aśoka (and consequently Buddha) lived centuries earlier, and if Sandrocottus was the founder of the Imperial Guptas who were mostly Vaishnavites or Bhagavatas, worshippers of Vāsudeva-Krishna, and if the Brāhmanical Śungas, Kānvas and Āndhra Sātavāhanas had already reigned before Sandrocottus. And in contrast to the insignificance of Buddhism in the time of Megasthenes we may stress the prevalence of the Guptas' own religion, Vaishnavism or Bhāgavatism, which he noted at Mathura in particular.
As a converse to the information by Megasthenes we may add what we gather from the earliest Buddhist works and the inscriptions of Aśoka. Sircar2 writes: "The Buddhist canonical work Anguttara Nikaya gives a long list of religious sects, but does not mention Vasudevaka or Bhagavata. The inscriptions of Aśoka; which speak of Brāhmana, Śrāmana, Ājīvika and Nigrantha, do not refer to the followers of Vasudeva." Aśoka's empire included the places where Megasthenes found Krishna-worship. His omission of this cult implies the practical abeyance of it in the period between Buddha's death and Aśoka's reign, and the complete separation of the time of Aśoka from that of Megasthenes.
It is only the later Buddhist canonical commentaries, the Mahā-niddesa and the Chullaniddesa, that mention "the worship of Vāsudeva and Baladeva (Samkarshana)."3 Evidently, these commentaries point away from the time of both Buddha and Aśoka towards that of Megasthenes. And what Buddha's and Aśoka's time held in contrast to Megasthenes's is indirectly admitted in
1.Op. cit., p. 322.
2."Religion and Philosophy: Vaishnavism", The Age of Imperial Unity, p. 437.
3.Ibid..
Page 242
F. W. Thomas's remark1 on the Hinduism in the environment of Megasthenes: "...the greatest share of popular adoration accrued to Ççāiva and Vishnu (under the form of Krishna), whom the Greeks report to us as Dionysus and Heracles respectively. With the former was associated Skanda or Vicakha, the god of war. The Buddhist books and scriptures , which give the preference to Brāhma and Indra, are in this respect archaising."
But why should those books and scriptures archaise? Thomas has no explanation. He thinks of them as doing so because he has put them in the time of Megasthenes and a little earlier. If we remove them from it, make them fairly anterior, no need arises to imagine any archaising. The straightforward thing to do is to accept that in their time Brahmā and Indra were actually given preference and that therefore this time was appreciably earlier than the Greek ambassador's.
Some More Reasons for Reinterpretation
In addition to all this information let us see what, on the authority of Megasthenes, Pliny says about the Āndhras in the age of Sandrocottus - the Āndhras whom Aśoka also mentions in R.E. XIII and to whom the post-Aśokan Sātavāhanas who preceded the Imperial Guptas belonged.
Sircar2 has written: "Pliny (first century A.D.), who is usually supposed to have utilised the information supplied by Megasthenes (c. 300 B.C.), speaks of a powerful king of the Āndhra country possessing 30 fortified towns as well as an army of 100,000 infantry, 2,000 cavalry and 1,000 elephants." Sircar comments: "This no doubt points to the large extent of the land occupied by the Āndhra people, and it is not improbable that Pliny actually received the information from a later source referring to the Sātavāhana kingdom."
We may remark that it is extremely improbable that Pliny should have interpolated material from another source in his present context without any warning. For, almost immediately after mentioning the great power of the "Andarae", he3 goes on to
1."Political and Social Organisation of the Maurya Empire", The Cambridge History of India, I, p. 485
2."The Sātavāhanas and the Chedis", op. cit., p. 194.
3.The Classical Accounts of India, pp. 142 & 143.
Page 243
say: "But the Prasii surpass in power and glory every other people... their capital being Palibothra , a very large and wealthy city ... The Indus skirts the frontiers of the Prasii .. ." When the Sātavāhana kingdom was at its height, the Prasii - the Prāchyas (Easterners) of Magadha - with their capital at Pataliputra were certainly not the most powerful and glorious people commanding an immense empire. This empire was in existence before that kingdom and after - in the days of the Mauryas and in those of the Guptas. But in the days of the Mauryas the Āndhras were not at all the prominent people pictured by Pliny: if they had been , Sircar who accepts the usual identification of Sandrocottus with Chandragupta Maurya would not have suspected a later source than Megasthenes for Pliny's information. They could be as in Pliny the strongest power next to the Prasii under Sandrocottus - if the Prachyas under the Guptas had just founded their new empire. Although surpassed by the Prachyas they could still rank as more powerful than any other Indian nation.
On approaching Pliny without bias and with the full sense of his context we get a clear indication of the true time of the Āndhra supremacy - namely, a little prior to the rise of Sandrocottus; in which case Sandrocottus must be Chandragupta I, and Megasthenes must have lived centuries later than the Mauryas, the third of whom was Aśoka.
The Āndhras pose a further problem, driving a wedge between the age of Aśoka and that of Sandrocottus. Thomas1 observes: "It has been actually remarked by Lassen that in a number of cases Megasthenes states the military power of particular provinces; and he infers that these are instances of independent rule. The inference may have been carried too far; but it has an undeniable validity as regards the kingdoms south of the Vindhya mentioned by Megasthenes, namely the Andhras and Kalingas, as well as their western neighbours..." Now, in R.E. XIII the Āndhras, with the Parirndas, are listed as semi-subject peoples, with just the limited independence such as we can argue for the Yavanas and Kambojas. How did they change their status so much if Aśoka was the grandson of Sandrocottus?
Like the Āndhras, the Kalingas too could not have been independent in the time of Sandrocottus if he and the first Maurya had
1. "Chandragupta, the Founder of the Maurya Empire", The Cambridge History of India, I, p. 473.
Page 244
been identical. For, the first Maurya, who took over the empire of the Nandas, must have ruled over Kaliiiga which was one of the provinces whose Kshatriya dynasties Mahāpadma Nanda had uprooted.1 The Purānas' attribution of the conquest of Kalihga to Mahāpadma is corroborated by a passage in the Hathigumpha inscription of Khāravela, from which it has been inferred that a Nanda king had conquered Kalihga.2
No doubt, Kalihga must have been lost by Chandragupta's son and Aśoka's father Bindusāra, for Aśoka's inscriptions speak of war upon and subdual of Kalihga. But it was, as Mookerji3 tells us, his one and only conquest. And this must mean an otherwise large inherited kingdom such as he could never have obtained had his grandfather been Sandrocottus whose India knew of several independent parts, the most prominent of which was the country of Megasthenes's "Andarae".
Another difference between the world of Megasthenes and that of the Mauryas may be thrown into relief. In R.E.V Aśoka declares: "For a very long time past previously there was no despatch of business and no reporting at all hours. This, therefore, I have done, namely, that at all hours and in all places... the Reporters may report people's business to me. People's business I do at all places."4 Mookerji,5 accepting the current equation of Sandrocottus with Chandragupta Maurya, is led to remark: "Aśoka's strictures against his precedessors do not apply even to his grandfather whose devotion to public work is thus described by Megasthenes: 'The king may not sleep during the day-time. He leaves the palace not only in time of war, but also for the purposes of judging cases. He then remains in court for the whole day without allowing the business to be interrupted, even though the hour arrives when he must needs attend to his person' (McCrindle, p. 72). Curtius (VIII.9) adds: 'The palace is open to all comers, even when the king is having his hair combed and dressed. It is then that he gives audience to ambassadors and administers justice to his subjects' (lb., Ancient India, p. 58, n.)."
Sandrocottus, according to Megasthenes and Curtius, did exact -
1.The Age of Imperial Unity, p. 32.
2.Ibid.
3."Chandragupta and the Maurya Empire', ibid., p. 61.
4.Bhandarkar, Aśoka, p. 315.
5.Aśoka (Rajkamal Publications Ltd., Delhi, 1955), p. 143, fn. 1.
Page 245
ly what Aśoka says his predecessors failed to do. Mookerji is surprised at the discrepancy between the two statements, for he takes Sandrocottus to be Aśoka's grandfather. But both the statements can be correct if Sandrocottus is not Chandragupta Maurya but the founder of the Imperial Guptas. And the discrepancy is an index to a background for Aśoka remote from the time of Sandrocottus.
One more index to a difference of background may be elicited from the name of Sandrocottus's son, which we have already mentioned: Amitrachates or Amitrochades. We may remember Barua's reading1 of the Sanskrit equivalent for the latter: "Amrita-khāda" ("Eater of ambrosia"). With the Allāhābād Pillar inscription's presentation of Samudragupta, the son of Chandragupta I, as a god in human form this reading accords very well, as does also our own suggestion for Amitrachates, "Amitrachchhettā" ("Mower of enemies"), which is reminiscent of the term "Sarvarajoch-chhetta" ("Mower of all kings") applied to Samudragupta by his sucessors in four inscriptions. Two other proposals have been made for both the names:"Amitraghata" ("Slayer of enemies") and "Amitrakhāda" ("Eater of enemies"). These too suit Samudragupta down to the ground. We may go even a step further and ask whether "Amitrachates" may really be not a title but the actual name of the son of Sandrocottus. The dropping of the letter S as an initial sound in foreign names was a frequent practice among the Greeks. The Sindhu became the Indus and even Sandrocottus became Androcottus in Plutarch (Life of Alexander, LXII) as well as Appian (Syriaca, 55). If Amirachates is a name with an initial S omitted, we have actually Samudragupta to deal with and it sounds very much like a variant of Samitracottus and may be equated with Samudragupta.
However, what fits the king whom Vincent Smith called "the Indian Napoleon" hardly goes well with Chandragupta Maurya's son Bindusāra. Doubtless, the Tibetan writer Taranatha writes that Bindusāra's minister Chānakya helped his monarch to be master of a vast territory. But Barua2 comments: "What reliance can be placed on the evidence of so late a chronicler as Taranatha?" And Mookerji3 asserts: "Aśoka's father Bindusāra is not known to history as a conqueror." How then will the names handed down by Athenaeus (XIV.67) and Strabo (II.1.9) be in
1.Op. cit., p. 309, fn. 4.
2."Chandragupta and the Maurya Empire", The Age of Imperial Unity, p. 61.
Page 246
tune with Barua's and Mookerji's own reconstruction of post-Sandrocottus history?
Surely, all that we have said is enough in general to show this history to have no place for Aśoka.
2
Now we may come to those particular issues that are the main supports of the modern chronology of Aśoka.
Aśoka's "Yonas"
The first of these issues is: "Should we not consider as Aśoka's Greek subjects the Yonas who were within his empire in close association with the Kambojas and whose country was the only one in which there were no Brāhmanas and Śramanas?"
The mere resemblance of the Prākrit "Yona", whose Sanskrit is "Yavana", to "Ionian" should not prejudice us. It is curious to find Daradas (Pliny's Dardae) and Kiratas (Cirrhadae of the Periplus of the Erythraean Sea) and rāmanas (the geographer Ptolemy's Rhamnai) listed among ancient Indian tribes and to encounter Dardanians and Cretans and Romans in Europe. But we cannot jump to the conclusion that these Indian tribes colonised Europe or that those European peoples came and settled in ancient India.
Again, the mere absence of Brāhmanas and Śrāmanas in the country of Aśoka's subject Yonas is no pointer to Hellenic civilisation alone. In the borderland between India and Irān, where the Yonas were situated, a country could easily be without Brāhmanas and Śramanas because of the Mazdean religion , the Zarathustrian culture and society . A completely Irānianized life would be all that is necessary.
Yavanas or Yonas in Ancient Indian Literature
Even apart from these arguments, there is sufficient matter in ancient Indian literature to demonstrate that "Yavana" or "Yona" did not originally denote Greeks and that Aśoka's subject Yonas were without Brāhmanas and Śramanas not by virtue of any Hellenism but in perfect accordance with what this literature has to say about their non-Greek character.
Page 247
At the very start we may note with Sircar:1 "The earliest use of the Sanskritized form Yavana can be traced in the Ashtādhayāyī oi Pānini (c. fifth century B.C.)..." At the moment we do not have to dispute Sircar's dating as being somewhat late. We may just point out the way in which the great grammarian brings the Yavanas into his Ashtādhyāyī. The actual word is "Yāvānānī" (V.1.49) which the commentators succeeding him - Kātyāyana and Patanjali -have interpreted as "the script of the Yavanas". If Pānini's Yavanas were the Greeks denoted by the Persian "Yauna", we have the implausibility of Pānini being already so familiar with the Greeks as to be able to cite their very script and of his readers taking in their stride his mention of it rather than a less distant mode of writing to exemplify a grammatical rule.
Before tackling the question of Pānini's date, we may also remind ourselves that "Yāvānānī" is not his sole reference to the Yavanas. There is another which has a perfectly Indian traditional bearing and makes a non-Greek connotation of "Yāvānānī" obligatory. A. K. Narain2 remarks: "Like other early Indian sources, Pānini associates the Yavanas with the Kambojas." For, as Narain says in continuation, Pānini's ganapātha 178 on his Ashtādhyāyī's II. 1.72 has the phrase: Yavanamundah kāmbo-jāmundah - "shaven-headed Yavanas, shaven-headed Kambojas". The greatest commentators, such as Vamanajayaditya of the Kāsikā and Bhattojldikshita of the Siddhānta-kaumudi, have accepted the phrase as authentic. But what is striking about it is not merely the traditional association of two peoples: it is also the curious manner in which the association is couched. Both Narain3 and Barua4 observe how the phrase links up with the story in the Purānas - e.g., the Harivarnsa (XIV. 16) - according to which tribes like the Yavanas and the Kambojas were Kshatriyas who got degraded and were variously punished by King Sagara: the punishment of the Yavanas and the Kambojas was to be always shaven-headed: munddam kritva vyasarjayat Yāvānānām sirah sarvam kambojānām tathai cha - "getting the heads of the Yavanas wholly shaved and also those of the Kambojas (he) expelled them."
It is incredible that Pānini should describe the Greeks as having
1."The Yavanas", The Age of Imperial Unity, p. 102.
2.The Indo-Greeks, p. 1.
4.Op. cit., Part I, p. 99.
Page 248
completely shaven heads and incredible too that - as Narain has suggested - his description, positively implying no hair at all, probably meant nothing more than the Greeks' custom of wearing their hair short! A straightforward look should convince us that Pānini's Yavanas no less than his Kambojas were non-Greeks.
And what is particularly noteworthy is that the very first indubitable occurrences of "Yavana" convey to us not only the notion of a script but also the sense of a degraded Kshatriya. This sense, in one form or another, persists down the ages through all ancient Indian literature that touches on the history of the Yavanas. Coming so early, in however elliptical a form, it seems to bear the stamp of a truth and keep the Greeks away.
To what extent it is early is indeed a momentous question, but even if Pānini is considered as late as c. fifth century B.C. and therefore liable to be influenced by the "Yauna" of Darius I and of the Achaemenid inscriptions succeeding his of Behistun in 520-518 B.C., is it not absolutely unthinkable that, so soon after 520-518 B.C., "Yavana" could bear an utterly different and entirely Indian sense, unless Pānini's term had nothing to do with the Greeks mentioned in the Behistun inscription and in its successors?
Of course, if Pānini could recede substantially beyond the fifth century B.C., the most spectacular coup de grace would be delivered to the common coupling of the Greeks with his "Yāvānānī". According to V.S. Agrawala,1 "the range of Pānini's date is in the opinion of scholars limited to a period of three centuries between the seventh and the fourth century B.C." Thus there has been no chronological unanimity. But most scholars would concur with Agrawala2 when at the end of a detailed discussion he places the grammarian around 450 B.C. One of the sharpest dissidents is Ram Gopal3 who has countered every one of Agrawala's arguments - the chief of which are drawn from Indian religious history. On the positive side he has attempted to estimate properly the interval between Patanjali and Katyayana and that between Katyayana and Pānini. On several well-pleaded grounds he is inclined to fix Pānini to about 600-550 B.C.
At a much later stage of our book we shall hold a brief for a date still earlier. At present, if something like this time-bracket at the
1.India as Known to Pānini (Lucknow, 1953), p. 456.
2.Ibid., pp. 456-75.
3.India of Vedic KalpaSūtras (Delhi, 1959), pp. 86-88.
Page 249
latest can be justified beyond cavil our immediate purpose will be served. Oddly enough, Agrawala himself puts before us a fact which sets Pānini clear of all Persian influence. It is about the Janapada (province) Gandhāra in which was situated the town of Salatura where Pānini lived. Agrawala1 writes: "Pānini mentions both the Vedic form Gāndhāri as the name of the janapada and its people in Sūtra IV. 1.169, and its later form Gandhāra in the ganas to IV. 1.133 and IV.3.93.... Pānini knows Gāndhāri as a kingdom (IV. 1.169). It seems that soon after the death of Darius [I] Gandhāra became independent, as would appear from the manner of its mention as an independent janapada."
Here is a piquant paradoxical situation. Gandhāra has to be under Darius I in order that Pānini's "Yāvānānī" should signify "Greek writing". Gandhāra has to be free soon after him in order that it should be the province Pānini knew. But everything cannot be so conveniently arranged. And we can demonstrate the inconvenience from Agrawala's own book.
Darius died in 486 B.C. What do we find soon after him in»the reign of his son Xerxes (486-465)? We hear from Agrawala2 that Gandhāra as a part of the Persian empire "continued under the reign of Xerxes, who recruited to his army a contingent of Indians from Gandhāra about the year 479 B.C." Again, after distinguishing Pānini's Āyudhajīvī Sahghas (warrior clans or groups) of "Parvata" from similar Sahghas of "Vāhīka" as those belonging to "the highlands of north-west", "outside the plains of the Vahika country", "which comprised the regions from the Indus to the Beas and the Sutlej", Agrawala3 remarks that the geographical clues we possess from various writers "all together point to the mountainous regions of central and north-east Afghānistān as being the Parvata country". And he4 notes: "Arrian mentions these mountainous Indians as fighting in the army of Darius [III] against Alexander at Arbela (Anabasis, III, 8.3-6). It was these Paravatīya Āyudhajīvins that offered stout resistance to Alexander in Bactria and Gandhāra." This implies that the fighters for Darius III from central and north-east Afghānistān included Gandhārians. So Gandhāra comes into the picture right at the end of
1.Op. cit., pp. 49, 446.
2.ibid.,
3.Ibid., pp. 434, 435.
4.Ibid., p. 435.
Page 250
the Achaemenids (330 B.C.) just as it does during the time of Darius I. From Agrawala's own pronouncements there seems to be no evidence at all that Gandhāra was independent at any period of the Achaemenid empire from the Behistun inscription's c. 520-518 B.C. onward.
In the interests of greater precision let us inquire about the state of affairs in the reign of Artaxerxes I (465-424 B.C.), which succeeded the reign of Xerxes and which would coincide with Agrawala's epoch for Pānini. Scholarship has disengaged an important fact in relation to the narrative by Herodotus who was in Persia during the reign of Artaxerxes I. Herodotus supplies us with a list of the tributes paid by the various provinces of the Achaemenid empire. "It is generally believed," writes S. Chattopadhyaya,1 "that the tribute list, as preserved in the account of Herodotus (III.89-95), is clearly from his own time, that of Artaxerxes I..., not, as he states, from the time of Darius." In that list the "Gāndhārians" figure in the empire's seventh province. So Gandhāra around 450 B.C., which Pānini is said to know as an independent janapada, is found to be in Achaemenid hands.
Beyond Darius I we have Cambyses (530-522 B.C.) and his father Cyrus (558-530 B.C.). We have no narrative about the former in connection with Gandhāra, but the latter is credited by Greek and Roman historians with eastern conquests.. The best conclusion from their reports is, in Mookerji's eyes,2 Eduard Meyer's: "Cyrus appears to have subjugated the Indian tribes of the Paropanisus (Hindu Kush) and in the Kabul valley, especially the Gandarians."3 From the manner in which Gandhāra stands in the Behistun inscription in a roster of twenty-three provinces in a wide-spread, stable and organized empire, with no sign of any province having been newly acquired, Gandhāra along with the other constituents of the empire seems to have been inherited by Darius from Cyrus without any break in the Persian possession of it. So the independent Gandhāra in which Pānini lived has to go past in time not only the Behistun inscription but also the year between 558 and 530 B.C. marking Cyrus's conquest of it. Darius's
1."The Rule of the Achemenids", The Indian Historical Quarterly, Vol. XXV, p. 156.
2."Foreign Invasions", The Age of Imperial Unity, p. 40.
3.Eduard Meyer, Geschichte des Altertums, III, 97.
Page 251
"Yauna" as the source of the Pāninian "Yavana" is left fairly behind.
But can we even say that the period before Cyrus's conquest of Gandhāra will allow "Yāvānānī" to signify "Greek script"? Unfortunately, a dilemma blocks our path. There was no opportunity for Gandhāra to know this script until Cyrus first came into contact with Greece on defeating Croesus of Lydia in 547 B.C. and conquering that country, a portion of Ionia. If Gandhāra was conquered after that date, then as long as it had enjoyed its independence it had come into no relation with Cyrus and consequently acquired no knowledge of the Greek script from Lydia through him. If the independence of Gandhāra was earlier than 547 B.C., there was also no chance for its Pānini to get at this script, all Persia dividing him from Greece by its own lack of connection with Lydia. Therefore, either way, an independent Gandhāra and the Greek script cannot go together during the reign of Cyrus. And, obviously, prior to his reign the second horn of the dilemma would be just as sharp. By no means and at no time can we bring "Yāvānānī" into rapport with the Greeks instead of with an Indian tribe.
A touch of scepticism may meet us here: "If the Greeks could be involved by Pānini's expression, we would have a distant people haunting our mind's eye. But if it refers to the script employed by an Indian tribe, how is it that Pānini who lists a substantial number of janapadas, including the usual companion of Yona - namely, Kamboja - fails to bring in the Yona province, a janapada called Yavana contiguous to Kamboja?"
The sceptic proceeds on the misconception that Pānini set out to give geographical, political, social or cultural information. Pānini did nothing of the sort. He compiled just a masterly grammar: his aim was grammatical analysis and linguistic investigation. Whatever finds place in his treatise serves as an example of rules and verbal structures. All the information provided is a byproduct. Nor is it likely to be exhaustive under any head. No matter how profuse, it is bound to be selective on the whole. In respect of any topic we should do well to remember Agrawal's observation1 apropos of a certain species of information: "Names of rivers, forests and mountains are noticed not in connection with any social or historical significance, but as examples of linguistic
1. Op. cit., p. 36.
Page 252
peculiarity, such as vowel-lengthening (IV.3.117-120) or cerebralisation (VIII.4.4-5)."
At this point.we may relevently comment that comparable to Pānini's overlooking of a possible Yavana janapada is his overlooking of the Ganges among the rivers, although the gana to VII. 157 names a river we know as its tributary: Rathaspā, identical with the Rathasthā of the Vedic period and with the modern Rāmagangā which is still in its upper course known as Rahut or Raput and falls into the Ganges at Kanauj.1 Again, the omission of a Yavana janapada or of Yavana as a tribe has no momentous overtones any more than has the omission, except for "Yāvānānī", of all the other scripts that, for instance, appear along with "Javanāniya" (=Yāvānānī) in the list of scripts in the Samavayari-ga Sūtra (Samavāya XVIII), to which Agrawala2 alludes. The sceptic's procedure would be a case of insensitivity to the fundamentals of the Pāninian approach.
Moreover, even the Yavana tribe cannot be considered quite out of Pānini's sight or definitively divorced by him from its usual companion. Back in 1881 Rajendralala Mitra3 noted that Pānini offers "Yāvānānī as an example to show the use of the affix ānuk to indicate the writing of the Yavanas". Thus Pānini's usage appears not to confine itself to a mere adjectival label: it implies a collectivity called "Yavanas" and so most probably a janapada of them. Mitra4 has also a mention of Katyayana, the commentator of Pānini, who is always taken to bring out the earlier grammarian's sense and not to import anything alien: "In Katyayana's Vārttika on Aphorism 175 of the first section of the fourth Book of Pānini, the Yavanas are linked with the Kambojas, showing their near relationship." In fact, for this relationship and its suggestion of a Yavana janapada, no less than of a non Greek-people, neighbouring a Kamboja janapada, we do not have to depend on implications of Varttikas. We have already commented on Pānini's Gana-pātha 178 on his Ashtādhyayī's II. 1.72 - the phrase which translates: "shaven-headed Yavanas, shaven-headed Kambojas." There assuredly is a background of a janapada for the Yavanas as much as for the Kambojas.
1.Ibid., pp. 45, 46.
2.Ibid., p. 312.
3.Indo-Āryāns (Calcutta, 1881), II. p. 177.
4.Ibid., p. 190.
Page 253
That the Yavanas were an ancient Indian tribe with a janapada of their own contiguous to that of another tribe, the Kambojas, and that both the tribes were non-Greek - such a conclusion from Pānini can be backed by all other testimony from India. The Buddhist canonical work Majjhima Nikāya is referred to in the Chullavagga which does not mention the Third Buddhist council held in Aśoka's reign, and which is therefore pre-Aśokan. The Majjhima Nikāya, on this showing, is pre-Mauryan.1 It makes Buddha tell Assalayana (11.149): "Suttante: yonakrambojesu an-hesu cha pachchantimesu janapadesu..." Mookerji2 translates; "Have you not heard this: that in the Yona, Kamboja and other countries...?" The Yonas thus are already a state on India's borders during Buddha's life - before even the Persians could develop wide contacts with the Greeks, leave aside their meeting Indians on a substantial scale during and after Alexander's invasion. The very term Aśoka uses - yonakambojesu - is here met with. Again, the Chullaniddesa, which is included in the Buddhist canon and is hence not too late a commentary though not one of the earliest compositions,3 numbers Yona no less than Kamboja as having been a country of the Indian subcontinent before Buddha's birth.4 Can we think of an organised Greek colony on a provincial scale in north-west India as early as such a period - actually before Cyrus's campaign against the Ionian Greeks of Lydia?
The Mahābhārata (XIII.33.21) and the Manusamhitā (X. 43-44) not only speak of the Yavanas, together with the Kambojas and some other tribes, as being degraded Kshatriyas: they also speak of them as having lost their Āryān status by not consulting Brah-manas. This explanation reminds us at once of Aśoka's own statement in R.E. XIII singling out the Yonas within the empire as the people who in his day were without the ministrations of Brāhmanas and Śramanas. The statement, which has been taken as a pointer to the Greek nature of Yona society and culture, is really in tune with India's own tradition and shows that what was true about all north-western border provinces, including Yona, held somehow for just that one province in Aśoka's day. Buddhist
1.Pusalker, "The Pali Canon", The Age of Imperial Unity, p. 408.
2.Aśoka, p. 161, fn. 8.
3.Pusalker, op. cit., p. 409.
4.B.C. Law, "North India in the Sixth Century B.C.", The Age of Imperial Unity, p. 1.
Page 254
literature itself has said of the Kambojas, the close associates of the Yonas, the same thing in essence as the Mahābhārata and the Manusamhitā and, by implication, the Ganāpaha. Law1 tells us: "In the Jātakas we read that the Kambojas were a north-western tribe who were supposed to have lost their original Āryān customs and become barbarous." So Aśoka is merely qualifying in a particular way a common ancient testimony, both Buddhist and non-Buddhist, which is far removed from Greek concerns. Since the condition of his Yona country remains the same as in this testimony and simply the condition of its fellow-provinces changes, no Greek shade requires to be imported into Aśoka's statement. To import such a shade is to raise an entirely false and unnecessary issue.
As to the reason why the Yonas, and not the Kambojas or any other related tribes which had all been described in old Indian books as lacking Brāhmana-ministrations, were now alone in that lack, we may propose two approaches. Indian tradition itself suggests the possibility of those ministrations returning to degraded tribes. For, in the midst of their degradation, a shadow of Brāhmanical institutions was yet allowed to fall on them. A long passage in the Mahābharata recounts what the degraded Kshatriyas were required to do by way of religious observance and social duty.2 A reversal to Brāhmanas and Śramanas via such observance and duty was always on the cards and several tribes might have made it. On the other hand, the life without the direct presence of Brāhmanas and Śramanas might lead to a complete deviation from the Vedic religion. In the area between India and Persia this deviation could have been towards no alternative religion except the Mazdean. So the Yonas, unlike their companions, may have been totally Mazdeanized by Aśoka's time and become the only people without the Brāhmanas and Śramanas.
Even the Kambojas are reported to have come under Mazdean influence. A Buddhist Jātaka3 attributes to a good number of Kambojas - kambojakānam... bahunnam - a certain religious practice - the killing of insects, moths, snakes and worms - which we may recognize as Mazdean from passages in Mazdean books like
1.Tribes in Ancient India (Poona, 1943), p. 7.
2.Muir's Original Sanskrit Texts, 2nd Edition, Vol. I, p. 484.
3.V. Fausboll, Jātaka (London, 1877-97), VI, 208, 27-30. Cf. Law, Tribes in Ancient India, p. 7.
Page 255
the Vidēvdāt (XIV, 5-6) as well as from the remarks of Herodotus (I, 140) about the Persian religion. The Jātaka's wording leaves it beyond doubt that not all Kambojas followed this practice and the picking out of the practice indicates that those Kambojas who followed it were Mazdeanized insofar as this practice was concerned. Not all of Mazdeanism need be attributed to them. But, as against the limited Irānianization of the Kambojas in the post-Buddha age, we have an extremely revealing piece of information about the Yonas in Aśoka's time, which takes us beyond the general presumption that they, like their close associates, the Kambojas, must have been Irānianized only to some extent.
In the Junāgarh inscription of the Śaka ruler Rudradaman I, the dam repaired by Rudradāman's engineer is recorded to have been dealt with in Aśoka's time by Aśoka's viceroy in Saurashtra, "the Yavana raja Tusāspha."1 V. Smith has rightly noted the Persian indication in the name ofthe Yavana (Yona) raja. One may try to make this rāja out to ·be a Persianized Greek, but that would be special pleading on the presupposition that the Yonas in Aśoka's empire must have been .Greeks. The bare fact is the Irānianized name. And Tusāspha the Yavana raja - viceroy of Saurāshtra was a dweller within Aśoka's empire, as were the Yonas who were neighbours to the Kambojas. He cannot but cast on them a Persian tinge and render them more Irānianized than the Kambojas in our eyes. In addition he could easily imply for them a full adoption of the Mazdean religion and prove Mazdeanism responsible for their uniqueness among Aśoka's subjects.
Looked at from any side, this uniqueness stands free of the Greek nuance. It can be accounted for without going outside Indian evidence. The Yonas may legitimately be taken to be Indians Irānianized in toto.
Consequently, the ancient terms "Yavana" and "Yona" must be derived without any appeal to "Ionian". The Sanskrit root yu can make it connote "keeping away, averting" or "mixing, mingling" or else "quick, swift".2 So the Yavanas could be "those who are disliked" or "a mixed people" or "men who move rapidly". The last-mentioned sense is favoured by a linguistic peculiarity which allows such alternative forms in Sanskrit as yūśa and jūśa for "soup" or yāmi and jāmi for "sister" and which in Kauśika-sūtra
1.Epigraphia Indica, Vol. VIII, pp. 36 ff.
2.M. Monier-Willaims, Sanskrit-English Dictionary, (Oxford 1899), p. 848.
Page 256
makes y do duty for j on several occasions - e.g.., yunaymi (3.1) -and which S.N. Pradhan1 very appositely elucidates: "Philologists declare that the Y in Yavana may be replaced by J; that is, both Yavana and Javana are permissible and therefore interchangeable. Javana in later Sanskrit literature means 'one who possesses speed'. (Compare Yaska's derivation of Pijavana, Nirukta II, 24, Roth's edition, Gottingen.)"
The explicit interchanging of Yavana and Javana is found in a Buddhist allusion to the script which Pānini named "Yāvānānī". Agrawala2 writes: "Yāvānānī... as the name of a script occurs in the Samavayanga Sūtra (Samavaya XVIII) under the form Javana-niya (cf. also the same list in Pannavanā Sūtra)." And in the light of this we make a surprising discovery on referring back to Yaska's derivation. For, Yaska discusses Pijavana while considering it the name of the father of Sudās, a Rigvedic hero who in the Rigveda (VII .18.23) is given the epithet Paijavana. So "Javana" and therefore implicitly "Yavana" is as old as the Rigveda.
Both the connection with "one who possesses speed" and the sense of remote antiquity like the Rigveda meet us in a different way for the Yavanas in the Mahābhārata passage (XIII.33.21) upon which we have already drawn . There the Yavanas are said to be the descendants of Turvasu. "Turvasu" is the Purānic form of "Turvas" mentioned in the Rigveda (VII .19.8) - "Turvas" which , as Pradhan3 tells us, means "one who runs fast. "
We may also associate youthfulness with rapid running and align "Yavana" with the Sanskrit "Yuvan" which has the same root yu and signifies "young". Perhaps this alignment would bring it directly into touch with the Greek "Iaones" which would connote the young or new race from the Aegean islands and the Asiatic coast as opposed to the "Graichoi", meaning "the old" and denoting the original inhabitants of the Greek peninsula. But this alignment would render it superfluous even to entertain the possibility that "Yavana" came from Europe or Asia Minor to India. For, it would be the variant of a term common to the Āryān peoples wherever they were and appearing, as Rajendralala Mitra4 enumerates, in the Zend jiwán, the Latin juvenis, the Saxon iong,
1.The Chronology of Ancient India (Calcutta, 1927), p. 95.
2.Op. cit.. p. 312.
3.Op. cit., p. 95.
4.Op. cit.. II, p. 177.
Page 257
the Dutch jong, the Swedish and Danish ung, the Gothic yuggs.
We may attend to a further fact which keeps in countenance our idea that "Yavana" or "Yona" owes nothing to the Persian or any other usage indicating the Greeks. Even outside India we have proof that the early use of a term analogous to "Yavana" was not in reference to them. Mitra1 quotes Rawlinson to show that in the inscriptions of Sargon, dated 708 B.C., the isle of Cyprus, where the Assyrians first came into contact with the power of the Greeks, appears as Yavnan or Yunan, but that the name of this country is said by Sargon to have never been heard of by his ancestors, the kings of Assyria and Chaldaea, from the remotest times. If Yavnan had never been heard of before 708 B.C. in Assyria and Chaldaea, it is not to be supposed that it was better known in the much earlier time of the XVIII Dynasty of the New Kingdom in Egypt, and we can fully credit another piece of information by Rajendralala Mitra:2 in the monuments of that dynasty under Tutmosis III and IV and Amenophis HI, the term "Uinim" which, according to Curtius' History of Greece (Ward's translation, 1, p. 45), is the oldest form of "Ionia" is used for all foreign subjects of the Pharaohs. The implication here is not only that Uinim, occurring so much earlier than 708 B.C., covered non-Greeks but that it covered no Greeks at all.
Lastly, there is the proof offered by Dr. I. Olsvanger3 of Hebrew University, Jerusalem, that scholars of Greek itself, as long ago as the early part of the third century B.C., did not consider the Greeks to have been denoted by an early use of the Hebrew analogue of "Yavana". Dr. Olsvanger writes:
"In support of your argument concerning the identity of the Yavanas, I wish to draw your attention to the name Yāvān which appears several times in the Hebrew Bible.
"First in Genesis X, 2,4 (with repetition in I Chronicles I, 5,7). The ancient Greek translation, the Septuagint,4 of the Old Testament simply transcribes the name and so does the Latin Vulgate where it appears as Javan. It follows from that that the authors of the Septuagint did not regard the Yāvāns of Genesis as meaning Ionians or Greeks.
1.Ibid., pp. 174-175.
2.Ibid., p. 169.
3.Letter to the author, dated 4.1.1957.
4.C. 270 B.C. (The Concise Oxford Dictionary, 1952, p. 1135).
Page 258
"Only at a later stage of the Books of the Bible did the word Yāvān accept the meaning 'Greeks'. Thus in the Book of Joel IV, 6: 'And you have sold the sons of Jerusalem to the sons of Yāvān', the Septuagint translates the last phrase by 'Tois Uniois ton Ellenon' and the Vulgate by 'Filiis Graecorum'..."1
When we read the note of the Jerome Biblical Commentary2 on Joel IV.6, it is not difficult to affirm with Mitra the earliest form of "Ionia" to be "Uinim". The note runs: "the Greeks: the Ionians (Yewānim), the Greeks of the Aegean Islands and western Asia Minor." The inflected Hebrew word bracketed here is surely a modulation of the Egyptians' "Uinim" which dates back to 1500-1400 B.C. "Ionian" as a name would seem to predate the Greeks with whom it was later associated. Olsvanger's information combines with Mitra's to leave not even a straw for the modern theory to catch at and save itself from drowning. When Asia Minor and Greece herself never began with identifying the Yavanas with the Greeks, we need have not the slightest hesitation in refusing to assume that India did or to brand as incorrect the old Indian view of them.
Not that "Yavana" or "Yona" never stood in India for "Greek" at any time. But this connotation is first observed in the BesNāgar inscription of the last quarter of the 2nd century B.C. And here, as also in the late Buddhist books, Milindapanha (1st century A.D.), the Dīpavamsa (4th century A.D.) and the Mahāvamsa (6th century A.D.), its connection is with the Greek rulers (c. 190 B.C.-c. 50 A.D.) over portions of India, whose dynasties hailed from Bactria, once a part of the Achaemenid empire, in which the Greeks were known as "Yauna". Naturally in India they got pinned with the labels "Yavana" and "Yona". But, as we learn from Otto Stein's detailed inquiry3 into epigraphs, even in the time of the Bactrian-Greek and Indo-Greek rulers, "Yavana" did not always or exclusively stand for "Greek". He demonstrates that, except for a few cases, personal names of Greeks do not possess the attribute "Yavana", while the term occurs again and again
1.Most scholars date Joel to 400-350 B.C. see The Jerome Biblical Commentary (Original American Edition published by Prentice-Hall, Inc., New Jersey, 1968-Indian Edition published by Theological Publications, Bangalore, 1982). The Old Testament, p. 440, col. 1.
2.The Old Testament, p. 443, col.l.
3.Indian Culture, I, pp. 351-357.
Page 259
without any Greek association. Also, "there is no proof that the Yavanas, where they appear in connection with a genitive plural, are 'Yavanas' at all, they may be personal names of members of Indian families or of some corporations."
Hence, where is the least compulsion to read "Greek" in Aśoka's chronologically earlier "Yona" subjects or "Yona rāja" -especially when we have no evidence at all that the immediate successors of Alexander the Great - originating, as they did, from outside any part of the old Persian Empire - were known in India as "Yonas" or "Yavanas"?
After the age of the Bactrian Greeks and Indo-Greeks, "Yavana" is well known to have become a very general designation. Sircar1 records: "The word Yavana was used in mediaeval Indian literature as a synonym of mlechchha and indicated any foreigner... The Muslim rulers of India are often called Yavanas..."
"Yavana" and "Yona" appear to have had varied uses and it should be nothing of a surprise that in ancient times it pertained to a non-Greek tribe, originally Indian and later fallen from its Āryān Kshatriyahood and gone foreign without the services of Brāhmanas and Śramanas.
Pointers from Greek Accounts
The non-Greek character of the Yavanas in ancient India is suggested also from the accounts of Alexander's invasion of India.
If there was a Greek settlement in India, it must have antedated Alexander's arrival in 326 B.C., for "Yavana" is a term much older. As the Majjhima Nikāya speaks of a Yona border-state in Buddha's time, Alexander should have met the settlement in the course of his advance. But there is no evidence of his meeting any colony of Greeks, whether called "Yavanas" by the natives or named anything else. No doubt, between the rivers Kophen (Kābul) and Indus the people whom the Greeks termed the Nysaioi lay in his path, and his soldiers played with the idea that they were a colony left behind by Dionysus when he returned home from the wandering which, according to a story by Euripides in his Bacchae, had taken him all over the East. However, we may heed what Strabo (XV.7) has to say on the "conquests" of both Dionysus and Heracles: "As for the stories of Heracles and
1. "The Yavanas", op. cit., p. 101 with fn. 2.
Page 260
Dionysus, Megasthenes and a few others consider them trustworthy; but most other writers, among whom is Eratosthenes, consider them untrustworthy and mythical, like the stories current among the Greeks." So we see that the mass of serious Greek minds themselves dismissed the very basis for the claim of the Nysaioi's Hellenic ancestry.
And there was even a counter-legend: Diodorus (111.63) reports that more than one Dionysus existed and that the most ancient of them was not a Greek but an Indian and moved from India to every part of the world. Thus, even on the legendary plane, there is a check.
"The Greeks," observes Bevan,1 "always experienced a keen joy of recognition, when they could connect foreign things with the figures of their legends..." Do we not know how they saw Heracles in Krishna of Mathura and found Dionysus in Shiva of the hills? The Dionysian Nysaioi are also an instance of their Hellenifying tendency. Another instance is provided by Strabo (XV.8): "They further called the Oxydrakai descendents of Dionysus, because the vine grows in their country, and their processions were conducted with great pomp, and their kings on going forth to war and on other occasions marched in Bacchic fashion, which is also a custom among other Indians." In addition to that bubble-pricking last clause, Strabo (XV.58), touching on the presence of the vine east of the Euphrates as a special proof, dryly remarks that most parts of Mesopotamia, Media, Persia and Carmania have good vines. And we may gauge the flimsiness of the Greek fancy from the fact that the Oxydrakai, who are kinned to the Nysaioi as descendants of Dionysus, are universally recognised as the Indian Kshudrakas.
In regard to the town named Nysa, M. de St.-Martin2 rightly wrote long ago: "This place ought to be of Median or Persian foundation, since the nomenclature is Irānian, the name Nysa or Nysaya which figures in the cosmogonic geography of the Zend-avesta being one which is far-spread in the countries of ancient Irān." As an alternative, we may think of a settlement of the Indian tribe named Nichya, which Pusalker,3 in detailing the Āryān communities in India during the period of the later Sarhhi-
1."Alexander the Great", The Cambridge History of India, I, p. 354.
2.McCrindle, Ancient India as Described by Megasthenes, (Calcutta, 1926), p. 183, fn.
3."Āryān Settlements in India", The Vedic Age, p. 257.
Page 261
tās, Brāhmanas, Upanishads and Sūtras, speaks of in connection with the Punjab and Sind. We may then very well echo the phrase of Vincent Smith:1 "a fancied connection with Dionysus and the sacred mountain Nysa of Greek legend."
We may safely assert: the accounts of Alexander's invasion reveal no Greek colony in India. Judged from them, the pre-Alexandrine Yavanas or Yonas whom Pānini and the oldest Buddhist books mention were not Greeks. And, since Alexander never came across any tribe he could have dubbed, after its Indian name, something like "Iaonoi", even non-Greek Yavanas or Yonas were absent as a distinguishable tribe in Alexander's time. Between Alexander's time and that of Pānini or Buddha a gap existed, during which the Yavanas somehow got submerged as a tribe. Obviously, Alexander had a different background from Pānini or Buddha. And then Aśoka, who cannot be much separated from Buddha, must have also had a background different from Alexander's.
What, in relation to the Yonas, confirms this difference is the fact that even the Kambojas, the recurrent associates of the Yonas, do not appear in the long lists of peoples provided by Alexander's historians or by Megasthenes. On the analogy of "Kambyses", the Greek version of the Persian king-name "Kam-bujiya", we might expect something like "Kambysoi" if the Kambojas were known to them. The nearest we come to it is Arrian's reference2 to a people called Kambistholi from whose dominion the river Hydraotes (Ravi) is said to flow. The geographical clue would place the Kambojas well to the east of the Indus in the Punjab - a fantastic location.
The name itself resolves best into "Kapisthali" which in Pānini (IV.1.42) can be an alternative form to "Kapistalā" (VIII.3.91), a place-name - or into his "Kapishthala" (VIII.3.91), a family-name from perhaps the name of a sage, giving rise to the forms (II.4.69) "Kapishthalāh" and "Kapishthalayah", the descendants of Kapishthala. A curious fact noted by Agrawala3 is that Pānini (IV.3.107) mentions the Kathas as northern disciples of Vaisam-payana, probably in Punjab, and that an offshoot of their school is called Kapishtala-Kathas by the Charanavyuhd, and that Arrian's
1.The Early History of India (London, 1924), p. 55.
2.McCrindle, op. cit., p. 198.
3.Op. cit., pp. 323-324.
Page 262
Kambistholi are in the same territory as the Kathioi (Kathas) who resisted Alexander's march. Arrian's tribe has nothing to do with the Kambojas. And where the Kambojas geographically should have been we find the Aspasioi (Āśvāyanas) and the Assakenoi (Āśvākāyanas) - tribes whose names were derived from the reputation they enjoyed for the excellence of their horses. The Kambojas - famed as they are in Indian tradition as expert horse-breeders - are surely among them, but their specific existence is submerged, unlike the case with Pānini's references where we find not only the Āśvāyana (IV.1.110) and the Asvakayana (IV.1.99) but also the "shaven-headed Kambojas" and the janapada (state or cultural unit) called "Kamboja" (IV.1.175).
How long the Kambojas remained submerged we cannot quite tell. But in Western reports it is only when we come to the geographer Ptolemy (140 A.D.) that we hear of a place named Tambyzoi to the south of the Oxus - Tambyzoi which Sylvain Lévi1 identifies with Kamboja on the ground that the word is a Greek transliteration of the Austro-Asiatic spelling of Kamboja. But even Tambyzoi, south of the Oxus, is a far cry from the country contiguous to that of the Yonas on India's north-west in Aśoka's day. The whole combination of Yonakambojesu seems alien as a positive entity to the historians of Alexander and to Megasthenes: it appears to be no part of the milieu of the Macedonian conquest of north-western India. Hence this milieu must belong to an epoch different and distant from Aśoka's.
The Results of Modern Research
Our conclusions from ancient Indian literature and from Greek accounts are not contradicted by any result of modern research. Some scholars have argued for a Greek colony in India before Alexander on the strength of certain coins obtained from Rawalpindi: oriental imitations of the earliest Athenian type, the "Owls", and a batch bearing a helmeted head on the obverse and a cock on the reverse together with the name "Sophytes" in good Greek characters. As Alexander in Punjab came across a king whom the Greeks named "Sopeithes" or "Sophites", the latter coins are attributed to him and he is considered the ruler of a
1. The Indian Antiquary, 1923, p. 54.
Page 263
pre-Alexandrine Greek colony in ancient India.
But none of Alexander's historians ever suggested that their "Sopeithes" or "Sophites" of Punjab was a Greek. Alexander mixed with him on very friendly terms, yet not the shadow of a hint is left to us that he was not an Indian king. About all the coins concerned, Mookerji1 supplies the decisive information: "In 1877 a large hoard of coins was found at Kabadan on the Oxus and was imported to Rawalpindi by dealers. This hoard included the oriental imitations of Owls and Eagles akin to the coins of Sophytes."
R. S. Whitehead's study2 of the last-named coins may be summed up: "Sophytes and his coins belong to the Oxus region: they are probably earlier than 320 B.C., the date given to them by Sir George Macdonald. Sophytes must have been a local satrap who asserted his independence on the fall of the Persian empire." In a more recent article than Whitehead's A.K. Narain3 has discussed with expert skill two new coins of Sophytes, one of them representing him as a very young man. Narain offers strong grounds for taking the author of the coins to be a Graeco-Irānian satrap who ruled in north-western Irān under the Achaemenids, became Governor in about 335 B.C. and got reinstated in his office by Alexander just as the Persian Governors Oxyartes and Phrat-phernes are known to have been.
Thus, from every point of view, the existence of Greek Yavanas in pre-Alexandrine India is disproved and consequently that of non-Greek ones established. And if Pānini, Buddha and, with Buddha, Aśoka recede far beyond the age of Alexander it becomes logically impossible for Aśoka's "Yona rāja Amtiyoka" no less than for his subject "Yonas" to be Greek.
But, although this "Yona rāja" outside the empire is shown to be logically impossible, we cannot be said to have historically disposed of the problem raised by him. For, the Greek-sounding Amtiyoka is not alone: he is in the company of four other king-names alleged to be equally Greek-sounding. The multiple manner in which their problem confronts us is a challenge that demands independent investigation.
1.Ancient India, p. 139.
2.Numismatic Chronicle, Sixth Series, III, pp. 60 ff.
3.The Journal of the Numismatic Society of India, 1949, Vol.11, pp. 93-99.
Page 264
3
"Yona Rāja Amtiyoka" and the Four Other Rājas
We may begin with remarking that Aśoka terms only Amtiyoka "Yona rāja" and leaves the nationality of the four other rājas unspecified. If he had wanted to term them "Yona" he could easily have done it; or else he could have used a word like "similar". Our historians would understand Aśoka's abstention from Yona-fying them to imply a tacit dittoing of Amtiyoka's nationality for them. But, with as many as four kings concerned, this abstention looks rather significant: it is as though Aśoka cared little for their great number and would have omitted to specify their nationalities even if it had been something else than Yona. Further, in R.E. II, while again terming Amtiyoka "Yona rāja", he mentions neither their names nor their number. And in both R.E.s II and XIII he names Amtiyoka twice, once in his own right and then again in relation to others, as if to give him a special, a central importance and the rest the status of mere appendages. So we may legitimately ask whether anything in the two edicts could provide a clue to a difference in status.
All the recensions of R.E. II, except the one at Girnār, employ the word sāmamtā about the four rājas. The exception at Girnār is sāmīpam, which means "neighbours". In the light of this word the other is taken to connote the same. But independently it could denote, as Bhandarkar1 reports from Buhler, a shade of inferiority. Buhler renders it: "vassal-kings." And, since it occurs in more places, should we not take sāmīpam to be a non-committal way of implying this shade? Further, should we not search the more elaborate R.E. XIII for a decisive touch?
In that edict, immediately after the mention of Tulamāya, Arhtikini, Maga and Alikasudara, comes a peculiar word.2 It is, in the version at Shāhabāzgarhī, nīche: in those at Mānsehrā and Kālsī, nīchaihth. Literally meaning "below, down, downwards", it has been translated as "southwards" because the Chodas and the Pāndyas and the Tāmraparnīyas who are next mentioned were in South India. But in R.E. II Aśoka shows no geographical concern
1.Op. cit., p. 301, n. 2.
2.The Inscriptions of Atoka, edited by DR. Bhandarkar and Surendra Nath Majumdar (1920), p. 52.
Page 265
and blandly brings in Amtiyoka and his associates, without any preparation, on the heels of "Tamraparnī" as if Amtiyoka stood still more to the south.1 Moreover "southwards" as a particular sign of direction would imply that all who were listed before were to the north, but our historians situate all the five rājas to the west: hence "southwards" loses its point. Finally, although "southwards" is not impossible for nīche or nīcham on the analogy of uttara which means "higher" as well as "north", such a use is unknown anywhere else in the whole range of Sanskrit or Prākrit literature. Mookerji2 rightly remarks: "The meaning of the word does not seem to be satisfactorily settled."
The same word occurs in R.E. VII, where the form is nīche at Girnār and nīche or nichē at five other places.3 Here Bhandarkar4 observes: "Nichā seems to have been used as an adjective... The variant nīche corresponds to Skt. nīchaihih, as suggested by Hultzch, which is often used adjectively." Bhandarkar5 and Sircar6 take it in the sense of "worthless", Luders and Hultzch7 in that of "low" or "mean". This procedure may be followed in regard to R.E. XIII also. The form nīcham is, like nīche, adverbial and equivalent to nīchaihih. If nīche can be used adjectively, why not nīcham? As an adjective, nīcham could well connote in R.E. XIII "lower" or "inferior" and make all the four rājas listed after Amtiyoka the subordinates and satellites of Amtiyoka. Considering their dependent position, Aśoka may not have felt that their nationality needed any mention: they are tagged on to Amtiyoka, so that his nationality alone is of concern to Aśoka. In a situation like this, it would be a matter of indifference to Aśoka whether their nationality were the same as Arhtiyoka's or no. Consequently, there cannot be any presumption that, just because he is called Yona, they must be Yona too if they are not distinguished as something else.
Our theory has not only the advantage of explaining Aśoka's omission: it has also the advantage of being consistent in word-interpretation and of avoiding for nīcham a totally unsupported
1.Ibid., p. 4.
2.Aśoka, p. 163, fn. 3 (contd. p. 169).
3.The Inscriptions of Aśoka,, p. 27.
4.Aśoka, pp. 319, 320.
5.Ibid., p. 319.
6.The Inscriptions of Aśoka, p. 45.
7.Cf. Mookerji, op. cit., p. 148, fn.2.
Page 266
usage. But, if our theory is to be preferred, we have at once a state of affairs unhealthy for the assumption that Aśoka was listing a series of Greek kings. Ptolemy, Antigonus, Magas and the two Alexanders were not at all subordinate to Antiochus. And if Aśoka lived in their epoch he would certainly know this through the diplomatic relations with them which our historians claim for him. Just as the father and grandfather of Antiochus II are known to have sent ambassadors (Megasthenes and Daimachus) to the Indian court, so also is Ptolemy II Philadelphus reported to have sent Dionysius there. If Aśoka reigned in the 3rd century B.C. he would not fail to be aware of Ptolemy's status and would not go out of his way to dub him a satellite of Antiochus. The sole satisfactory interpretation of nīchaihm disproves the identification of Ptolemy with Tulamāya and hence of the remaining Greek kings with Aśoka's rājas.
Yona Rāja Amtiyoka and Yavana Rāja Tusāspha
The next point we may make is: although Amtiyoka alone and none of his neighbours is called "Yona rāja" in Aśoka's reference to rulers outside his empire, there is, as we have already observed, yet another Yona rāja connected with him: "Yavana rāja Tusāspha" named, in the Junāgarh Sanskrit inscription of the Śaka Rudradāman I, as Aśoka's viceroy in Saurāshtra in-the past. The undeniable Irānianized character of this king-name has served us to give a concrete body to our suggestion that the Brāhmanas and Sramanas may have been absent from Aśoka's Yona province because the Yonas led an Irānianized and Mazdeanized life. Now we can use it as a clue to the non-Greek nature of Yona raja Amtiyoka, Also ; Tusāspha can be our link between the Yonas within Aśoka's empire and the Yona raja outside it. For, Tusāspha is a Yona who is not only within the empire but a rāja as well, and his Irānianized character .passes on to Arntiyoka the non-Greek colour of these Yonas.
No doubt, there have been Śakas with Indian or Indianized names and Tusāspha could be a Greek for all the Irānianization of his name; but, against the background we have sketched and within the context concerned, the comparison with those Śakas becomes purely academic. Even in itself it is merely suggestive and not at all determinative. Just a possibility is conjured up: to accept
Page 267
it or not depends on the attendant circumstances.
Under the circumstances attendant here, what Tusāspha drives home to us vis-à-vis Amtiyoka is that we need not go towards Greek countries at just the mention of a Yona rāja. An Irānianized Indian or an Indo-Irānian is all that Amtiyoka may be. But, of course, we must try to show that his name is not necessarily an echo of the Greek "Antiochus".
Can "Amtiyoka" be a Non-Greek Name?
The form "Amtiyoka" is not the sole reading in the edicts. Kālsī gives us "Amtiyoga" or, with one sound slurred, "Atiyoga" which may be read "A(rh)tiyoga". Seen as made up of the components "Amti" and "yoga", this reading can be conjectured as an outlandish name suggesting "he who gets close-joined" or "he who collaborates", a name perhaps not far removed in its structure from a compound Indian appellation like "Antināra" found in the Purānas.1
What is important to decide here is whether "Amtiyoka" is merely a dialectal form of an original and basic "Amtiyoga" or a name in its own right, a real alternative form. Kālsī, where it is found, has several dialectal peculiarities and on a few occasions it turns the intervocal k to g: for instance, pasopagani in R.E. II. But pasopagani occurs not only at Kālsī: it occurs also in the Dhauli and Jauguda recensions which have a dialectal affinity with the Kālsī and, by a rare influence, it occurs even in the Girnār version. "Amtiyoga", on the contrary, is only at Kālsī: Dhauli and Jauguda give "Amtiyoka" and Girnār "Arhtiyaka".2 Again, Kālsī's "yoga"-ending is not only twice in R.E. II but also twice in R.E. XIII. It is a special persistent form which seems to have little to do with dialectal variations.
Its status appears to be akin in its way to that of "Maga" at Girnār in distinction from Shāhbāzgarhī's "Maka" and Kālsī's "Maka". Girnār has hardly any dialectal tendency of its own to change the intervocal k to g. A thing like pasopagani is not in the least characteristic of it. So its "Maga", when neither Kālsī nor
1.E.g., Matsya, XXIV.
2.The Markandeya Purāna mentions a north-western tribe with the terminal sound exactly echoing this Yona king's at Girnār: "Kālatoyaka" (Barua, Aśoka and His Inscriptions, Part I, p. 102).
Page 268
Shāhbāzgarhī corresponds to this form, is either a scribe's mistake or a genuine alternative name. Historians who equate Aśoka's Maga to the Greek king Magas take it evidently as a genuine alternative name and "Maka" or "Makā" as a derivation not by a dialectal movement but by a Prākritization. Prākrit (which includes all the Aśokan dialects and several others) converts at times to k what in Sanskrit would be g. Thus the Sanskrit "Nabhaga"1 becomes "Nabhaka", a tribe-name, in all the recensions of R.E. XIII. In view of this and in view of the consistent character of the "yoga"-ending at Kālsī, we may opine that this ending, like "Maga", can have the status of an original form, a genuine alternative to "yoka" (or "yaka"), and not of a dialectal variation of it. "Amtiyoka" could be a Prākritization of "Amtiyoga" or "A(rh)tiyoga". It might even be an Irānianized aspect of the name of one who might have belonged to a group of Irānianized Indians or Indo-Irānians. In any case, a double form seems legitimate here as with the king commonly identified with Magas.
A name akin to our Yona rāja's we may note in a Kharoshthī inscription from Taxila. F.W. Thomas2 reads the inscription: "In Śira A(m)tiyoha, sister of Looda, daughter of a hamśi mother and a hamśa father, deposits relics of the Bhagavat." According to Thomas, the inscription testifies to a thoroughly Indian consciousness. The ties with India are not only in reference to the Bhagavat but also in the rich multiple relevance of the hamśi and hamśa: Thomas observes a fusion of no less than three suggestions here. Almost as significant is the precedence given to "mother" instead of to "father". Such precedence is everywhere in Aśoka's own edicts and we find it even in the Arāmaic part of the bilingual inscription at Kandahār in contrast to the reversal of it, the father-mother order, in the Greek part. However, the names in the Taxila inscription - "A(m)tiyoh"a" and "Looda" - sound strange. Thomas has asked whether they are a distortion of the Greek "Antioche" and "Leontes". Yet our historians have refrained from committing themselves. According to Sircar,3 "no Indian approximations of the names of the Taxila inscription were proposed" but "it is difficult to say whether the names in question are really Greek as F.W. Thomas suggested".
1.Barua, Aśoka and His Inscriptions, Part I, p. 92.
2.The Journal of the Royal Asiatic Society, 1916, pp. 284-285.
3.Letters to the author, dated January, 4 and 16, 1958.
Page 269
Is there any need to be committed about Aśoka's "A(m)tiyoga" which is certainly affined to the "Antioche"-resembling "A(rh)tiyoha"? Both might be ancient non-Greek names - neither thoroughly Indian nor thoroughly non-Indian - from the border-country between India and Irān.
"Amtikini" as a Non-Greek Name
What would apply to "A(rh)tiyoga" and to its alternative could apply to "Amtikini". "Amtikini" actually finds a curious approximation in the name of an Indian tribe listed by Pliny (VI.23) after Megasthenes, among those on the east side of the Indus: Antixeni. The Girnār form "Arhtikina", which is not so much a dialectal variation as a vocal modification, has an ending as of that in the name "Airikina" by which the city of Eran was known in antiquity.1
As for the equation to a Greek name, we have a most interesting note by Bhandarkar:2 "Arhtekina or Amtikini, as Buhler has remarked, corresponds to the Greek Antigenes rather than to Antigonus. But, as no king named Antigenes is known, Amtikini has been identified with Antigonus Gonatas of Macedonia." Have we not here a bit of a hurry to suit preconceived theory? Why not admit straight away that the name of no Greek king in the 3rd century B.C. answers to "Arhtekina" or "Amtikini"?
"Tulamāya" an Improper Form: no Greek possible for "Turāmaya"
A much more serious fault of procedure - sheer wishful thinking - may be exposed in the matter of Tulamāya and Ptolemy. Out of the three versions of the edict available for the name, the two at Shāhbāzgarhī and Girnār read "Turāmaya" and the one at Kālsī "Tulamāya." Kālsī exhibits invariably the dialectal change of r to l, among other changes that are all termed Māgadhisms, while there is no opposite rule of / becoming r at Shāhbāzgarhī and Girnār. A further point to stress is that at all these three places, as well as at Mānsehrā, the letter /, if originally present in a name,
1.Mookerji, The Gupta Empire, p. 48.
2.Aśoka, pp. 49-50.
Page 270
remains intact everywhere. Thus the "Alikasudara" of Shāhbāzgarhī reads "Alikyasudala" at Kālsī: the r becomes l but not vice versa and the l in the first half of the name persists as part of the original form. The Girnār version is missing, but Mānsehrā has "Alikasu..." So we may conclude that if "Tulamāya" had been the original form there would have been "Turāmaya" nowhere. And in case some freak tendency to change l to r had been operative or a scribe's mistake had been at work, we could not have had "Turāmaya" at both such widely separated places as Shāhbāzgarhī and Girnār. The double presence establishes beyond all doubt "Turāmaya" as the basic original form and "Tulamāya" as a mere dialectal variation. "Tulamāya", which seems to correspond to "Ptolemy", is an improper form for comparison with the Greek name.
It is the basic "Turāmaya" that has to be compared. But how can we ever equate "Turāmaya" with "Ptolemy"? If "Ptolemy" had been intended, surely the basic original form would have been "Tulamāya" and the l would have stayed unchanged as it does in the first half of "Alikasudara". With "Turāmaya" as basic and original, the equation with "Ptolemy" is impossible and irrational.
Perhaps we shall be queried: "Has not J. Filliozat1 written: 'The relations between Indians and Greeks were managed through Persian interpreters'? And has not P. Giles2 told us that 'in the Old Persian of the Achaemenid inscriptions l is found only in two foreign words, and has otherwise been entirely replaced by r'? So may we not suppose that a Persian interpreter conveyed 'Ptolemy' to Aśoka with an r instead of an l?"
Our answer is: "It would be highly odd that a Persian interpreter should turn a Greek l to an r when conveying names from the Greeks to the Indians and yet correctly convey from the Indians to the Greeks the Indian l by mentioning such names as the Greeks wrote down as Peukelaotis, Sangala, Patala, Malloi, Glaukanikoi, Phegelas, Kalanos. The historian Curtius (XII), recording the meeting between Alexander and the Indian prince Omphis (Āmbhi), actually says that an interpreter was procured, but the official name reported of the Indian prince was 'Taxiles' after the name of his kingdom, Taksaśila. The original l of 'Taksaśila' came
1 Political History of India, translated from the French by Philip Spratt (Calcutta, 1957), p. 119.
2. "The Āryāns", The Cambridge History of India, I, p. 74.
Page 271
unaltered to the Greeks in spite of the interpreter, who may have been Irānian. There is even a case in which an r seems to come out as / instead of vice versa: the Indian 'Kripā' appears to have been the original of 'Cleophis', the Greeks' name for the queen of the Assakenoi (Aśvākāyana). Moreover, even Persian names themselves with an l have come to us in Greek histories. Arrian (Anabasis,. I.16;VII.3,6) has written of Arbupales, the son of Darius III, and Abulites, the Persian governor of Susa appointed by Alexander, and Artiboles, the son of Mazaeus. Finally, in Giles's own statement we find mention of two foreign words which, having l in them, preserve it in Old Persian. Would we not expect the foreign 'Ptolemy' to keep its l in the mouth of a Persian interpreter - particularly when our historians assume that the foreign word 'Alexander' where both l and r are concerned was properly interpreted so as to result in 'Alikasudara'?"
Besides, with two generations of Greek settlement in Arachosia and with the Indian interior open to the settlers in consequence of the friendly relations between Seleucus and Sandrocottus, why should we not think of either a Prākrit-knowing Greek or a Greek-knowing Indian as an interpreter? And we may remember too that Ptolemy II Philadelphus sent an ambassador to India, Dionysius. Surely the ambassador would at least have seen to the correct communication of his master's name? The presence of Dionysius must make "Ptolemy" the one name that ran no risk at all of being misconveyed. Consequently, "Turāmaya" can have nothing to do with Dionysius and his sovereign: it must belong to another historical context than the age of Ptolemy.
A less trivial argument against us would be: "In R.E. XIII itself there is an l-to-r change, a rhotacism, for 'Pulinda' in the form 'Parirhda' which Girnār gives for Shāhbāzgarhī's 'Pālida' and Kālsī's 'Pālada'. Why cannot 'Turāmaya' be a similar rhotacism?"
While Sanskrit "Pulinda, Paulinda" can equal, as Sircar1 has contended, Prākrit "Pālinda" or "Pālimda", a rhotacist conversion to "Pārimda" in the Aśokan dialect of Girnār is hardly on the cards and we may with more reason postulate a pure slip on the scribe's part like "Ketalaputo" in the Girnār R.E. II, which Mookerji2 calls a mistake for "Keralaputra". Error at times in
1.In Indian Culture, Vol. VIII, pp. 399-400.
2.Op. cit., p. 130, fn. 4.
Page 272
writing is admitted in general by Aśoka himself. Sircar,1 summarizing the Girnār R.E. XIV, recounts Aśoka's mention of the various reasons why some topics have been written incompletely in some places: Aśoka cites as one reason "a fault of the scribe."
A Māgadhism, on account of the royal source in Magadha of the edicts, is more possible at Girnār or Shāhbāzgarhī or Mānsehrā than the opposite: r may turn into l or even get left out but there is no tendency of l turning into r. Bhandarkar,2 studying "Pārimda", adverts to the influence of the Magadha court language but categorically declares as a clear fact in the context of all the edicts of Girnār, Shāhbāzgarhī and Mānsehrā before him that if the original has really an l we can never have an r in any version. "Pulinda" may be accepted as the correct tribe-name provided we accept the r of "Pārimda" as "a fault of the scribe".
Even otherwise, even granting "Pārimda" to be somehow a result of rhotacism, its occurrence at Girnār alone, with nothing analogous at Shāhbāzgarhī whose dialect3 differs more from Kālsī's than does Girnār's, can supply no ground for any talk of a rhotacist tendency, however sporadic, in the Girnār, Shāhbāzgarhī and Mānsehrā versions to balance to whatever degree the wholesale contrary trait in the Kālsī.4 Matching this Kālsī trait we have wholesale rhotacism in a fragmentary version of R.E. IX discovered not long ago at Sopārā.5 In Aśokan versions that use both r and / there is no proof of rhotacism. "Turāmaya" cannot be denied its basic position - especially since, unlike "Pārimda", it occurs not only at Girnār but also at Shāhbāzgarhī.
Moreover, we have to heed Barua's warning:6. "It is still uncertain whether Aśoka's Pārindas were the Pulindas of the Purānas. The name Pārinda occurs in the Pali Chūlavarmsa (XXXVIII.29.30) in which Pārinda and Khuddapārinda are mentioned as two sons of a Pandya king. It is possible, therefore, that the Pārindas were racially connected with the Pāndyas." Again, one may see in the Kālsī reading "Pālada a Magadhi-influenced form of Parada". The Paradas, as Barua7 reminds us, were a
1.The Inscriptions of Aśoka, p. 56.
2.Op. cit., p.
3.Ibid., p.
4.For a fuller discussion of the point see Appendix at the end of this section
5.Eptgraphia Indica, Vol. XXXII, Part I.
6.Op. cit., p.
7.Ibid., p.
Page 273
north-western tribe mentioned in the Mārkandeya Purāna. Raychaudhuri1 cites also the Vāyu (Ch. 88), the Harivamśa (I. 14) and the Brihat Samhitā (XIII.9). But, as Barua2 continues, it should not worry us to find the Paradas in the south with the Āndhras in Aśoka's R.E. XIII, for several tribes like the Chullkas and Mushikas who had their original settlement in the north have been found in the south. We may add: the Pulindas themselves figure both in the Vishnu Purāna3 and in the Mahābhārata (IV.9.40) in the compound Sindhu-Pulindaka, implying that they lived in the north, near the Indus. Nor are the Āndhras in Indian literature invariably set in the company of the Pulindas. The earliest reference to them - in the Aitareya Brāhmana (VII, 18) -puts the Pundras next to them in its tribe-list. If we wish to go by some collocation of countries in Indian literature broadly resembling Aśoka's own list of southern territories - those of the Āndhras, Parirhdas, Chodas (Cholas), Pāndyas, Satiyaputras, Keralaputras and Tāmraparnīyas - we have only the phrase in the Rāmāyana (IV.41.12) - "Āndhrams cha Pundrāmś cha Cholām Pandyams cha Keralan" - followed soon after by a description of the region of the Tamraparni river (IV.41.15). But, as Barua4 notes, "we have the Pundras instead of Aśoka's Pārindas, and the Keralas in place of Aśoka's Satiyaputras and Keralaputras." May not Aśoka's "Pārindas" represent a fusion of Paradas and Pundras?
One more facet of the problem may be presented. "Tonda-mandalam... was a province in the empire of Aśoka... and the Pulindas, included in the list of his subject peoples, were perhaps identical with the Kurumbas of Tonda-mandalam. Their name is reflected in Pulinādu and Puliyūrkottam, two ancient territorial divisions of the region. The Vāyalūr Pillar Inscription of Rajasimha mentions the name Pallava after the first seven mythical ancestors from Brahmā to Aśvatthāmā and before Aśoka's name. It may therefore, be argued that there was a Pallava ruler before Aśoka. Further, Pallava may be taken to be a variant of Palada (a form of the name Pulinda in some versions of Aśoka's edicts) and
1.Op. dr., p. 313.
2.Op. cit., p. 90.
3.Law, Tribes in Ancient India, p. 17.
4.Op. cit., p. 74.
Page 274
may be regarded as the southernmost people of the Maurya Empire."1
In the face of all the various possibilities we have sketched, it would be rash to swear that the Parirhdas who go with the Āndhras in Aśoka illustrate a genuine rhotacist trend at Girnār.
The chance of a rhotacism in "Turāmaya" in R.E. XIII may be completely dismissed. We may go still further and assert in the field of epigraphy and numismatics that when Sanskrit or else Prākrit of a type where both l and r occur transcribes a Greek or any foreign name, a clearly audible l of the original as in "Ptolemy" - an l uncombined with a consonant - never changes to r in any instance known to history.
Take the Besnagar inscription of probably the last quarter of the 2nd century B.C. Its author "Heliodorus, son of Dion" becomes "Heliodora, son of Diya".2 Or take the names of two Śaka kings of a little earlier date, who use Greek on the obverse and Kharoshthī on the reverse of their coins. We see "Spalirises" of the Greek becoming "Spalirisha" or "Spaliriśsa",3 and "Asilises" becoming "Ayilisha".4 In the Besnagar inscription, even an l before a consonant retains its identity: the Prākrit corresponding to "Anti-alkidas", the name of one of the Indo-Greek kings who ruled over a part of India and has left us some coins, is "Amtalikita".5 A coin that in the near past came to light of the last of these kings, Heliocles, proves how the clearly audible l resists all pull towards rhotacism. The coin shows the l after a consonant turning into r and yet the clearly audible l remaining the same in the Prākrit for Heliocles: "Heliakreyasa."6
Correspondence to "Turāmaya" is not in "Ptolemy" but in Indian and Irānian or else Turkish names. "Tura" is the name of a Vedic seer, the son of Rishi Kavasu.7 According to the Nepalese Chronicle, Aśokavadāna,8 it is the initial component in the name
1."Dynasties of South India" by R. Sathianathaier, The Classical Age, pp. 256-57.
2.Raychaudhuri, op. cit., p. 271.
3.The Age of Imperial Unity, p. 124; Raychaudhuri, op. cit., p. 428.
4.The Age of Imperial Unity, 127.
5.Ibid., p. 115.
6.The Journal of the Numismatic Society of India, Vol. XI, 1949, p. 101.
7.Aitareya Brāhmana, VIII.21; VII.34; IV.27. Also Bhagavata Purāna, IX.22.37.
8.R. Mitra, Indo-Āryāns, Vol. II, pp. 36-37.
Page 275
of one of Aśoka's own near-ancestors: Turakuri, a partly dialectal variant of which is Tulakuchi, the name of the same ancestor in the other Nepalese Chronicle, Divyavadāna.1 "Maya" as a terminal is there in "Asurāmaya", a name well-known in Indian legendary history. In the Irānian Avesta, Tura is one of the three sons of King Thraetona, and a certain Tura is also cited in Yasht XIII.2 As for the terminal "maya", E. Benveniste3 writes apropos of the Elamite name "u-ma-ya": " 'u-ma-ya'... ought to be 'humāya', known as a proper noun from Avestan Humayaka (cf. Aramaic Humayeak) fern. Humāyā (Pers. Humaya)."
"Ptolemy" for "Turāmāya" is not only impossible and irrational: it is also superfluous.
In passing, we may quote Zajti Ferenc4 to the effect that according to the interpretation of the Avesta, "Tura" and "Hūna" both mean "resistant" and indicate a strong and stubborn opponent. Ferenc traces to the former word the place name "Turan". We may, in addition, point to the name "Turfan" of a place in Chinese Turkestan: its Indian equivalent is "Turapamni"5 .We may remark further that the name "Toramāna" of the foreign invader of India, whose coins and inscriptions have been found and who is generally taken to be a Huna chief, father of Mihiraku-la, is not unconnected with "Tura". It has, no doubt, a Turkish ring, but even in the Turkish language it has the variant "Turāman"6 linking it to "Turāmaya". Wesendok7 has tried to show that "Toramāna" as well as "Mihirakula" is an Irānian name. The Jain Kuvalayamāla (c. 778 A.D.) gives for "Toramāna" the form "Toraraya"8 which seems to stand midway between "Turāmaya" and "Toramāna".
The Non-Greek Possibilities of "Magā"
We come now to "Magā, with its alternative "Maka" or
1.Ibid.
2.E. Blochet, J.R.A.S., 1915. pp. 36-37.
3.Journal Asiatique, 1958, I, p. 52.
4.Journal of the K. R. Cama Oriental Institute, No. 36, 1943, p. 13.
5.Mookerji, Ancient India, map facing p. 484.
6.Sircar, Select Inscriptions, p. 390, fn. 4.
7."Kusan Chroniten und Hephthalite", Klin (1933), p. 345.
8.Cf. Buddhaprakash's "Kalidasa and the Hunas", The Journal of Indian History, Vol. I XXXV, Part I.
Page 276
"Maka". The Irānian ring of the last two is clear. "Maka" occurs even as a place-name: a south-eastern province of the Achaemenid empire was called "Maka" (modern Makrān).1 Again, everybody has heard of the term "Magus" for a Persian priest: we read, in the Purānas, of the Śakadvīpa - Śaka-territory - where dwell the Maga dvijas who worship the sun-god;2 we have in India of later times the Maga-Brahmanas, the Persian priests (Magi) who migrated to India and appear to have contributed to the growth of the Ujjain school of astronomy.3 The Mahābhārata (VI.l2,33) enumerates the peoples of Śakadvīpa as Maga, MaŚaka, Manasa and Mandangao And among the variants of "Maga" here we find "Maka".4 The form "Maga" is a component of the full designation of the son of a queen Khribanis in Tibet in the period 635-43 A.D.: Maga Thogon Khagan. F.W. Thomas;5 remarks that not only Khagan but also Maga is a Turkish title , apparently the same as the Moho which the Chinese apply to certain rulers of Chinese Turkestan round about 700 A.D. Going to more ancient times, we read in the Cambridge History of India6 that the inhabitants of the region making up Magadha in Buddha' s time used to call it Maga, a name doubtless derived from Magadha. And we hear even of a "Maga raja" from the Satitkicca Jātaka.7 The exact form "Maka" is a constituent of the Indian name "Sivarnaka" which occurs in an Arnrāvatī inscription .8
"Alikasudara" as a Non-Greek Name
What about "Alikasudara"? About the names "Alexander" and "Alexandria", S.N. Majumdar9 observes: "The hypothetical Sanskrit forms of Alexandros and Alexandreia are, according to Sanskrit phonetics, Alakshandrah and Alakshandrā. In the vernacular, Alakkhandā or Alashanda are the forms expected." In Kautilya's
1.Richard N. Frye. The Heritage of Persia, p. 72.
2.Brāhmanda Purāna, XX. 71 f; Agni Purāna, 119.21.
3.The Age of Imperial Unity, p. 144, fn. 1.
4.Mahābharata (Critical Edition by S. K. Belvelkar), Vol. VII. p. 66.
5.Journal of the Royal Asiatic Society, 1927, pp. 63-64.
6.Vol. I, pp. 182-83.
7.Jātaka, Ed. V. Fausboll (London, 1877-97). Vol. V, p. 267-G, 103.
8.The Age of Imperial Unity, p. 205.
9.Notes to Cunningham's Ancient Geography of India (1924), pp. 692-93.
Page 277
Arthaśāstra1 we read of a place famous for corals: Alakanda at the mouth of the river SrotasI in the land of the Barbaras. Majumdar finds little difficulty in equating it to Alakshandra and identifying it as the port known as Alexander's Haven. In the Pali Mahāvamsa (XXIX, 30 ff.) and Milindapanhā we have "Alasanda" in which we can easily recognize Alexandria. Such variations from the expected form are natural. But "Alikasudara" seems too great a departure: paradoxically, it is too close to the Greek "Alexander" to be an Indian equivalent of it. And there is no need to bring in "Alexander", either. "Alikasudara" could be a perfectly Sanskrit name Prākritized. "Alika" seems a Prākritization of "Alaka" while "sudara" is easily the Prākrit for "sundara". In Sanskrit "Alaka-sundara" would describe "one whose face is beautiful with forelocks".2
There is also the possibility of a queer nomenclature. "Alika" as itself a Sanskrit term is attested as part of a compound word by what R.C. Majumdar3 writes by way of annotating the geographer Ptolemy's expression (VII.1,8) "Pseudostomos" for an Indian river: " 'Pseudostomos' means false mouth. The Dravidian literature has alemukham (Sanskrit Alikamukham)." Combined with "sudara" (="sundara"), it may point in a semi-Sanskrit semi-Prākrit way to some such concept as "Beauty that deludes and is a snare". If we doubt whether a term pejorative in immediate suggestion could ever go into a man's name, we have only to look at "Alikayu" which is part of the name of a man mentioned twice as an authority in the Kausitaki Brāhmana.4
Are the Names Too Odd as Non-Greek Ones?
No doubt, if we take the Aśokan names to be non-Greek the four that out of the five make compounds have a certain oddness. But we encounter many odd names in old historical records, defying explanation. Atiyoha and Looda of a Taxila inscription have already been pointed out: neither foreign nor Indian equiva-
1.P. 86, fns. 7, 8.
2.I owe this information to a letter dated 11.11.1972 from my friend E. Vedavyasa, I.A.S., Special Officer at the time to the Government of Āndhra Pradesh, Rājamundry.
3.The Classical Accounts of India, p. 379, note 8.
4.XXV, 51; XXVIII, 4. See The Vedic Index, Vol. I, p. 39.
Page 278
lents have been convincingly suggested. Richard N. Frye1 refers to the name of the first Kushāna ruler in India:"Kujula Kadphises, the etymology of which is most uncertain." There are also names accepted as Indian and as belonging to historical persons but strangely built and some of them never repeated. Sircar2 says: "None of the suggestions regarding the etymology of Sātavāhana and Satakarni is satisfactory." Again, he3 reports: "S.K. Chatterji and Pzryluski have written on the etymology of the name Khāravela. Their views are not satisfactory." In Aśoka's Rock Edict II itself where the Yona rāja Amtiyoka figures we have the designation "Satiyaputra" to denote another frontager. Barua4 informed us in 1946: "The name Satiyaputra or Satiyaputras is nowhere met with in Indian literature." Only recently a Tamil-Brahmi inscription was discovered in South Arcot District commemorating a gift by "Atiyan Naduman Anji, the Satiyaputra".5 The chieftain mentioned had been a subject of old Sangham poetry but distinguished in no poem by the title found in the epigraph.
If we deem all the unusual compound names which are still unexplained to be part of genuine history, why should we jib at Amtiyoka, Turamāya, Amtikini and Alikasudara when presented as being non-Greek?
Possibly we shall be asked: "Can the historicity of these non-Greeks of yours and of the remaining member of the group, Maga, be confirmed by independent testimony from outside Aśoka? If not, why go beyond Antiochus, Ptolemy, Magas, Antigonus and Alexander?" Our answer should be: "Do we ask for outside confirmation of the historicity of that other Aśokan Yavana rāja, Tusāspha, who requires no Greek correspondence but has left no evidence of his own and is mentioned only once in epigraphy and that, too, not by Aśoka himself? Again, did we not accept up to now without outside confirmation Aśoka's own Satiyaputra?"
1.The Heritage of Persia, p. 229.
2.Select Inscriptions, p. 185, fn. 4.
3.Ibid., p. 206, fn. 1.
4.Op. cit., p. 111.
5."Aśoka and the Tamil country: A New Link" by Dr. R. Nāgaswamy, The Sunday Express. Magazine Section, December 6. 1981, p. 6.
Page 279
Coincidences in History
The final objection we are likely to be posed with is: "Whatever fault you may find in relation to Turamāya and Amtikini, you must grant that Amtiyoka, Magā and Alikasudara are rather impressive in their echoes of Antiochus, Magas and Alexander. And in their company Turamāya and Amtikini also may not be rejected as entirely far-fetched for Ptolemy and Antigonus. Would you explain away all this as pure coincidence - and an almost fivefold coincidence at that?"
General echoes are not confined to Greek names. And several Indian or Persian or Perso-Indian names can accidentally echo Greek ones. We have dwelt sufficiently on such counterparts to Magas: we may list a few more. The name "Puloma" which the Matsya Purāna twice gives in its series of the Āndhra kings1 corresponds closely to "Polemo", the name of a Greek writer subsequent to Megasthenes.2 Puliśa, an astronomer mentioned in Indian books, has been equated with Paulus of Alexandria, but "Pauliśa" can be, as Altekar3 avouches, an authentically Indian name and the equation may be quite erroneous. Then there is Devamitra, a king of Ayodhyā,4 whose name looks like the Sanskrit for "Demetrius".
As for that triad of pretty close echoes - Amtiyoka, Magā and Alikasudara - in one and the same context, let no one make a unique case out of the situation. In Indian history itself we have the case of the Mahānāman inscription at Bodh-Gayā and the report of the Chinese writer Wang-Hiuen-t'se. In the former we have two monks, Mahānaman and Upasēna, from Ceylon, dedicating "a mansion of Buddha" in the year 269 of an unspecified era. In the latter we have also a Mahānaman and his colleague Upa-, two Buddhist monks from Ceylon, building at Bodh-Gayā a monastery and some stupas in the reign of an Indian king named San-meou-to-lo-kiu-to who has been equated to Samudragupta, one of the Imperial Guptas who had an era of their own to which they do not always give its proper title "Gupta". Sylvain Lévi
1.The Age of Imperial Unity, p. 705.
2.McCrindle. Ancient India as Described by Megasthenes and Arrian, p. 8, fn.
3."Education, Literature and Sciences", A New History of the Indian People, VI, p. 415. fn.l.
4.The Age of Imperial Unity, p. 174.
Page 280
observes that, if the same monks are not spoken of, it would be indeed a very odd coincidence. But Vincent Smith,1 who never doubts that Arhtoyoka and Maga and Alikasudara are Antiochus and Magas and Alexander (of Epirus or Corinth), marshals a number of arguments against the identity alleged by Lévi. And many other historians, whether agreeing or not with him about the era to which the year 269 is to be referred, are at one with him in disagreeing with Lévi's identification. The double resemblance in the monk's names, the common country Ceylon from which they hailed, the common location Bodh-Gayā for a common religious purpose, not to mention the unspecified era, are not allowed to have any weight.
Perhaps the most famous case of multiple resemblance is from Persian history. The sacred book of the Parsis, the Avesta, tells us that the prophet Zarathustra preached his religion of Ahura Mazda at the court of Vistaspa. In Persia the name of Zarathus-tra's God outside the Avesta is found for the first time beyond controversy in the inscriptions of Darius I, and it is found there frequently as if with the enthusiasm of a convert to a new cult. Darius I writes the name with a slight difference in the spelling and as a single word "Aurāmazda"; but these peculiarities may be no more than dialectal. And in one instance2 in the inscriptions of his successor, Xerxes, the name has two components, separately declined, as in the Avesta: so the knowledge of the Avestan form is directly proved. A connection for the first time between Zarathustra's religion and the Persian monarchy in the reign of Darius I may reasonably be conjectured, and the conjecture draws extraordinary strength from the fact that the father of this king is known to have been Hystaspes and this king is familiar to the Greeks as Darius Hystaspes. The Greek form "Hystaspes" is the precise equivalent of the Avestan "Vistaspa".
Impressed by these facts as well as several other features of the whole case, including a tradition in Persia about Zarathustra's date, scholars like Jackson, Hertel and Herzfeld have placed the prophet in the 6th century B.C. But Haug, Geiger, Andreas, Meyer, Moulton, Keith, Soderblom, Charpentier, Geldner, Bartholomae, Mills and Benveniste carry him to a fairly earlier date - scholars who yet would not think twice about the identifica-
1.The Indian Antiquary, 1902, pp. 192-97.
2.Ibid., p. 376.
Page 281
tion of Aśoka's rājas with post-Alexandrine Greek kings.
Indian scholars too who accept that indentification are seen disagreeing with Jackson, Hcrtel and Herzfeld. Thus Ghosh, in his article on Indo-Irānian relations in the compilation, The Vedic Age,1 writes about the Irānian tradition which put Zarathustra in the 6th century B.C.: "... according to Eduard Meyer (Geschichte des Altertums, second edition, third volume, p. 110, fn. 3) it is an inexplicable thing that anybody should think so. That Eduard Meyer was right can hardly be doubted, although weighty opinions have been raised against his view. The mention by Assurbanipal about 700 B.C. of Assara Mazas along with seven good angels and seven bad spirits is a clear indication of acquaintance with the reformed Zarathustrian pantheon (See Cambridge History of India, I, p. 76). It is impossible therefore to suggest that Kavi Vistaspa mentioned in the Avesta as the patron of the prophet was no other than the father of Darius I (522-486 B.C.), for in that case the Zarathustrian pantheon could not have been known in Assyria in the days of Assurbanipal." With Meyer, Ghosh carries the Gathas of Zarathustra to c. 1000 B.C. Altekar is prepared to go even a little further back, though still on the hither side of 1500 B.C. Parsi scholars are mostly disposed to agree in general with Herodotus and Aristotle and other Greek writers before Christ, who place the Persian prophet several thousands of years before their own time. AH leave Darius Hystaspes centuries behind.
We also need to make no bones about disregarding the apparent multiple correspondence on our hands and the "weighty opinions" in its favour. We are not forced to accept Aśoka's five rājas as Greeks.
"Even as far as 600 yojanas"
Possibly a last shot may be fired against us: "Aśoka in R.E. XIII mentions the distance within which his five rājas had their kingdoms. He declares: 'Here and in the bordering dominions, even as far as 600 yojanas, where (dwell) the Yona king called Amtiyoka and, beyond this Amtiyoka, the four kings...'2 Now, a yojana in ancient India, as J.B. Fleet calculates from Kautilya's Arthaśāstra3
1.P. 224, Note 18.
2.Bhandarkar, op. cit., pp. 330-331.
3.Note on p. 541 of R. Shama Sastry's translation.
Page 282
meant approximately 41/2 miles. So we get about 2,700 miles - the very distance roughly between India's western borders and Macedonia, t he farthest of the Greek dominions concerned. Who else could the five rājas be save the five Greek kings with whose names theirs have been identified?"
This poser is pretty easy to meet. First, it is not necessary to go 2,700 miles from Aśoka's western borders towards Asia Minor and Europe. His enumeration of bordering dominions does not stop with the lands of those five rājas. After mentioning them he continues with "the Chodas, the Pāndyas, as far as the Tāmraparniyas... "2. The 600 yojanas could be in reference to all the countries outside the empire, where the dharma was spread. That is how Barua3 understands the passage. Sircar4 also asserts that the construction of the passage allows the interpretation either that together the five kingdoms in the West and the three in the South covered the distance or that the distance was covered only by the western kingdoms.
Even if we go westward or in anyone direction, we are not obliged to go 2,700 miles. The yojana was a varying measure. Monier- Williams4 equates it not only to about 4 or 5 English miles but also to about 9 on the one hand and 21/2 on the other. Fleet's equation seems to have been favoured at one time especially because the Arthaśāstra was taken to be a work contemporaneous with Chandragupta Maurya but now, according to Mookerji, "many scholars .. . regard the present text as of much later date" .5 In parts of it we have a clear difference from Aśokan practice. Barua6 points out: "Its mode of dating a record in terms of the regnal years, month, half-month and day ... differs appreciably from that in the inscriptions of Aśoka... It counts the seasons as six and defines each of them (11.20), while Aśoka's phrases tim chātumrnasīsu (P .E. V) clearly suggests the adherence of his inscriptions to the tradition of three seasons." So there is nothing to bind us to 41/ 2 miles for Aśoka's yojana. And historically we should even be justified in choosing 21/2 miles. Basham7 says apropos of the Arthaśāstra: "It would seem that for practical
1.Bhandarkar, op. cit., p. 331.
2.Aśoka, Part I, p. 109. See also p. 131.
3.Letter to the author, dated 14.2.1957.
4. Op. cit., p. 858.
5."Chandragupta and the Maurya Empire", The Age of Imperial Unity, p 66
6.Atoka, Part II, p. 43.
7. The Wonder that was India, p. 66.
Page 283
purposes the shorter yojana was more used than the longer, especially in ancient times." If the Arthaśāstra was posterior to Aśoka's time the still shorter yojana may be more appropriateto his period. 600 yojanas might have been merely (600 * 2 1/2 = ) 1,500 miles.
Even if the Arthaśāstra be proved contemporaneous with Chandragupta Maurya, whatever differences may be thought of between the circumstances, physical or psychological, of Aśoka's lifetime and those of his grandfather's can permit us to ascribe to Aśoka the shortest yojana. Rapson1 has a significant remark: "The yojana has very different values according to the period and the locality in which it is used; but there is good evidence of the use in Buddhist books of a short yojana, equal to about two and a half English miles..." The equation of 600 yojanas to 1,500 miles would be quite appropriate in the case of a king like Aśoka who was a Buddhist par excellence.
Again, we cannot be sure that Aśoka's "Here" connotes the empire as a whole and that the 600 yojanas started from its frontiers. The word which Bhandarkar translates as "bordering dominions" is amta which strictly speaking, connotes "ends, extremities, borders" and J. Bloch2 actually translates the opening Aśokan phrase - "hida cha savesu cha amtesu" - by: "Here and on all my borders..." So the contrast may be between some central point within the empire and the empire's peripheries where the foreign dominions commence. This central point could most appositely be Aśoka's own capital city, Pātaliputra, the seat of government which was the one point common to the whole empire and to which everything was related and referred and from where Aśoka issued his edicts as well as sent out his dharma-messengers. In terms of the spread of dharma which is the theme of Aśoka, hida ("here") and amta could represent the starting-point and the ending-line of this propagation as if to say: "Beginning with my own capital and terminating in all lands on my frontiers." Such a contrast would be quite as pertinent as a contrast between the empire as a whole and the bordering dominions.
Moreover, we may note that hida appears twice in this passage.
1."The Successors of Alexander the Great", The Cambridge History of India, I, p. 550.
2.Les Inscriptions d'Aśoka, p. 125, quoted in translation in Basham, op. cit, p. 54.
Page 284
After referring to the amtas where Amtiyoka and others live, Aśoka goes on to use two expressions together: "hida rājavisavajri" (or "rājavisaye")1 - that is to say, "here in the king's dominions".2 Can we hold the two expressions - the opening word and the rest of the phrase - to be synonyms just because there is no cha ("and") connecting them? If they are synonyms no purpose is served by employing them both now when one of them has already appeared without the other at the very beginning of the passage. Again, if they are synonyms and the second is added to explain the first, why does the explanation come at this place instead of at the very beginning of the list of territories where the dharma-victory was achieved? In place of "Here and in all the bordering dominions" we should have had "Here in the king's dominions and in all the bordering kingdoms" and afterwards the simple "here" should have done duty. Everything seems to indicate a difference between "here" and "the king's dominions" - a technical difference.
The same difference appears to be meant between "the king's dominions" and the succeeding phrases which also begin without a cha: "among the Yonas and Kambojas, the Nābhakas and Nābhapamtis, the hereditary Bhoja rulers, the Āndhras and Pārimdas."3 Why are these tribes particularly named? Mookerji4 suggests that the tribal areas indicated by those phrases were separately mentioned in spite of their being included in the king's dominions, because they were not directly under Aśoka's rule but enjoyed a certain degree of autonomy. Barua5 subscribes to the same view: "Even within his own empire, we are to discriminate the portion which was at first entirely under his direct rule and subsequently under the rule of himself and his Viceroys, from that which was occupied by the semi-independent tribal territories or states."
If this view is correct, the identification of "here" with "the king's dominions" at the beginning of the passage would hardly cover the whole empire and the identification would be useless. Altogether we may posit three categories. (1) the semi-independent tribal territories under Aśoka's suzerainty, (2) the royal realm in the narrow sense of direct rule, and (3) the capital
1.The Inscriptions of Aśoka, p. 53.
2.Bhandarkar, op. cit., p. 331.
3.Ibid., p. 331.
4."Aśoka, the Great", The Age of Imperial Unity, pp. 78-79.
5.Op. cit.. Part I, pp. 53-54.
Page 285
city where he himself resided and from where he exercised rule over both the portions of his empire. To read any other sense in the three divisions into which his statement falls would be to rob the statement of its pertinence in details.
The third category, the capital city of Pātaliputra, would alone be connoted by "here". And in his two uses of "here" he would be referring from the same capital place, his own seat, first to the various regions outside his empire and, next, to the various regions inside it.
In favour of understanding "here" in this way there is actually an inscription where Aśoka leaves us in no doubt that his capital city is meant. In R.E. V, when the Mānsehrā, Shāhbāzgarhī, Kālsī and Dhauli versions read "here and in all the outlying towns", the king's own town must be their "here". What is more, the version at Girnār reads "at Pātaliputra" instead of "here". Nowhere else does Aśoka give an explicit alternative substitute for "here". So, in its absence, we should stick to the clear equation - made in at least one place - of "here" with his capital, unless the context anywhere compels us to abandon it. Is there any such context?
Apart from R.Es XIII and V, there are three places where "here" occurs: R.E. I, R.E. VI and Minor R.E. of Rūpnāth. In R.E. VI Aśoka speaks of his motive in spreading dhamma -namely, "that I may render some happy here and that they may gain heaven in the next world."1 Nobody can misunderstand this "here": it is the present world as contrasted to the next. But R.E. I and the Rūpnāth Minor R.E. bring "here" to mean something less than the whole present world, the earthly habitat of all life. In the first,2 Aśoka talks of many hundreds of thousands of animals killed formerly in the royal kitchen. Surely they were not killed for himself or his own family, however large it may be. Immediately we think of king Rantideva of the Mahābhārata's Vanaparvan, who was in the habit of distributing meat among his subjects.3 So the people of Pātaliputra get involved in Aśoka's curry-making. They are involved also in the matter of samājas to which the same inscription refers. The samājas are public shows such as Kharavela in the Hāthigumphā inscrip'ion and Gautamiputra Śātakarni in a Nasik cave insciption mention as having been organised in their
1.Bhandarkar. Aśoka. p. 316.
2.Ibid., p. 297.
3.Ibid., p. 21.
Page 286
own capitals.1 Hence when we read, "No life should here be immolated as a sacrifice; nor should any samāja be held",2 the natural notion is that "here" which goes with both animal-sacrificing and samāja-holding refers to Aśoka's own capital Pātaliputra.
In the Rūpnāth Minor Rock Edict, we have the phrase: "Have this matter engraved on rocks. Here and far off where there is any stone column, have it engraved on the stone column."3 The turn of expression is analogous to that in R.E. V - "here and in all the outlying towns..."Also, the engraving on stone-columns can be done only within Aśoka's empire and not in foreign countries. If we take "here" to mean the empire, "far off" must point to foreign countries. Consequently "here" can mean nothing except Pātaliputra and "far-off" indicate "outlying towns" or at least other places in the empire.
A general point supporting a uniform interpretation of "here" is that the moment we admit the possibility of an interpretation varying from context to context we land ourselves in uncertainty and controversy. Different scholars have held different opinions about one and the same "here": thus, to Mookerji,4 the "here" of R.E. I is Pātaliputra, to Bhandarkar5 it is the palace or royal household, to others it is the whole empire or, broadly, the world.6 But if even acute scholars cannot agree, what would be the state of mind of average Indians, subjects of Aśoka, reading their great king's announcements? It is difficult to believe that Aśoka left them in doubt, drifting in argument. Either his "here"s clearly signified different things or they signified an identical place no matter how unclearly. If he intended a diversity of meaning he was under the obligation to be exact in his indication: if he was known to have intended the same place he could be as careless as he liked. The very fact that scholars can differ shows his carelessness and this carelessness is, paradoxically, the proof that there was an understanding between him and his common subjects.
If we start from Pātaliputra and adopt the yojana of Vh
1.Ibid., p. 20.
2.Ibid, p. 247.
3.Ibid., p. 370.
4.Aśoka, p. 107, note, 4.
5.Op. cit., p. 298, note 4.
6.Ibid.
Page 287
and understand the 600 yojanas to be concerned only with Amtiyoka and his associates, we have for these yojanas a distance of 1,500 miles from the capital. Of course, they are not to be counted as the crow flies: they must be thought of as lying along a road which could not have been quite straight. Pānini (V.1,77) evidently refers to such a road when he mentions Uttarapatha and of goods gathered by that route (Uttarapathena āhritam). "Within India," says Mookerji,1 "this overland trade-route... must have passed through and linked up her chief cities mentioned by Pānini and Patanjali..." Both Patanjali (commenting on Pānini's rule III.3,136) and the author of the Kāsīkā speak of inland journeys along it, the former naming Saketa and the latter Kausambi as the starting-point of a journey which ended with Pātaliputra.2 Some idea of what the road must have been like can be had from the description left us by the Greek historians of India in the immediately post-Alexandrine period. Mookerji3 writes: "We may note the interesting reference made by the Greek writers to the royal road leading from the north-west frontier to Pātaliputra, the precursor of the modern Grand Trunk Road, with a length of 10,000 stades - about 1,150 miles. Megasthenes must have travelled down this road in joining his duties at Pātaliputra as ambassador. 'Every mile of this road was marked by a stone indicating the by-roads and distances.'" Nor is this all that we learn from the Greeks. The road is described4 as one existing from earlier times and as having been constructed in eight stages. The first of them started from Peukelaotis (Sanskrit Puskalāvati or Puskaravati, the capital of Gandhāra, modern Charsadda), lying a little to the north-west of Aśoka's Taksaśila, almost at the location of Shāhbāzgarhī. Thus we know that out of the 1,500 miles 1,150 would lie within Aśoka's own empire. Just 350 more remain and they would be precisely the land-route distance we should need in order to put his Yavana rāja Amtiyoka and his associates where they ought to be between his empire and Persia.
However, if we suppose "here" in R.E. XIII to mean Aśoka's empire and if 600 yojanas have to be counted away from it on the west or north-west instead of making them cover foreign lands
1.Chandragupta Maurya and His Times, p. 334.
3."Chandragupta and the Maurya Empire", The Age of Imperial Unity, p. 67.
4.Mookerji, Chandragupta Maurya and His Times, p. 330.
Page 288
both on that side and in the south, we still do not need to move towards Asia Minor and Europe. Going from the north-west frontier through Afghānistān to the territories beyond the Hindu-kush mountains, we can cover by 1,500 miles of winding road from Aśoka's north-west a terrain which could easily have been the field of his foreign missions. Majumdar1 tells us: "We have evidence to show that Buddhism, and along with it Indian culture, was spread among the Parthians, the Yueh-chi, the Sogdians and various other peoples of Central Asia before the beginning of the Christian era... The Greek writers always cite Bactriana with India and state that thousands of Brāhmanas and Samanas (Buddhist monks) reside there. The recent excavations in Chinese Turkestan have revealed the existence of a large number of flourishing cities with rich sanctuaries... Buddhism was the prevailing religion of all these localities." All the Central-Asian documents found are supposed to be of post-Christian date, but it is of great interest to note that Khotan, which "seems to be a particularly important centre of Indian colonisation", is said in Tibetan tradition to have had its royal dynasty founded by Kustana, a son of Aśoka, and in other Buddhist traditions to have been colonised and ruled by Kunala, another of Aśoka's sons.2 Aśoka's five rājas may have included, besides Indo-Irānian ones, some Central-Asian.
Whatever way we look at the problem of these rājas, there is no call to bring in post-Alexandrine Greek kings. And their rejection imports a total unity into the Aśokan passages: the Yonas within the empire who cannot have been Greeks but were degraded Kshatriyas totally Irānianized are uniform now with the Yona rāja Amtiyoka. Arhtiyoka's territory may be considered as an extension of the Yona country in India's north-west and as holding a people akin to or continuous with the Yonas of the empire.
If we may borrow for our purposes from the Buddhist Mahāniddesa3 a term whose exact indication is yet unknown, Amtiyoka was the king of the "Parāmayonas", the farther Yonas. The term is rather apt in view of the one in the Mahābhārata (11.27.23-26) about the extension of the Kambojas: the "Parāmakambojas." According to the Epic (VII.4.5), the Kambojas had their home in
1."Colonial and Cultural Expansion", The Age of Imperial Unity, pp. 635-639
2.Ibid., p. 640.
3.Pp. 155, 415 (The Pali Text Society).
Page 289
Rājāpura. Raychaudhuri1 comments: "The association of the Kambojas with the Gandhāras enables us to identify this Rājāpura with the territory of that name mentioned by Yuan Chwung [=Hiuen-Tsang] which lay to the south or south-west of Punch." The Parāmakambojas would seem to have been where the geographer Ptolemy (1st century A.D.) put the "Tambyzoi" (identified by Lévi as "Kamboja") - a place to the south of the Oxus. Just as the Yonas of Aśoka's empire neighboured the Kambojas in it, may we not surmise that Amtiyoka's subjects whom we have dubbed the Parāmayonas neighboured the Parāmakambojas?
If we may, we can move on to another plausible conjecture. The immediate neighbourhood of Amtiyoka's territory to that of the Prāmakambojas could prompt us to take as king of the latter the rāja who is the first to be named after the phrase "Param cha tena Amtiyokend''' ("and beyond this Amtiyoka"): Turāmaya.
Thus two out of the five Aśokan kings under discussion would acquire, besides "a name", "a local habitation". But we should not be dogmatic. It is not necessary for our ends to locate any of these rājas in a particular spot: it is only necessary to show Aśokan evidence to be all of a piece and such that its reference to "Yona" in any context would not connect Aśoka with the period after Alexander.
Whatever way we look at the problem of his rājas, strength flows from every quarter to uphold our contra-Greek attitude to their names. Nothing invalidates our finding "Antiochus" indecisive, "Ptolemy" impossible, "Antigonus" inaccurate, "Magas" arbitrary and "Alexander" both unlikely and superfluous.
Appendix
We have dismissed the talk of a rhotacist tendency, in even a sporadic form, in the Girnār, Shāhbāzgarhī and Mānsehrā versions to balance, in however slight a degree, the Kālsī's wholesale contrary trait. To bear out this dismissal in detail, we need a short discussion.
The idea is sometimes entertained that the word arabh four times in the Shāhbāzgarhī, Mānsehrā and Girnār versions of R.E. I and the word arambha twice in these versions of R.E. IV and
1. Op. cit., pp. 148-49.
Page 290
once there in R.E. XI are rhotacisms from ālabh and ālarhbha found in the Kālsī versions. There is no justification for this idea. Of course, ālabh or ālambha in the sense of "slaughter for sacrifice" is a legitimate Sanskrit term. It has also the sense of "killing, slaughter" and Monier-Williams (A Sanskrit-English Dictionary, p. 150) dubs ārambha erroneous for ālambha in such a sense. But, when ārabh and ārambha occur seven times between themselves at three distinct places and not at all sporadically and exceptionally but systematically and inevitably, the simplest and most rational hypothesis is surely the bonafide use of them as genuine forms by the scribes concerned: they are not a rhotacist conversion from ālabh and ālambha in a manner which can never be expected in versions where both r and / have their own rights and where, as a rule, there is no change of r to l or vice versa.
Besides, ālabh (Monier-Williams, p. 153) carries always "sacrificing" as an alternative sense to "killing" and every Sanskritist knows that together with ālambha it generally associates "sacrifice" with "killing". But R.E. I (Bhandarkar, op. cit., p. 297) speaks three times of killing for food: no sacrificial association is there, and it is ārabh, as Bhandarkar (p. 299, note 7) points out, that has this exclusive meaning and Monier-Williams says the same, as we saw, about ārambha. In fact, when this edict at its very start speaks of slaughtering and sacrificing, it employs two distinct words in all the versions, words formed from ārabh on the one hand and on the other from hotavya (="to be offered or sacrificed") (Monier-Williams, p. 1301). Thus a word unassociated in itself with sacrifice seems deliberately employed because a special one for the other idea is brought in. This means an intentional avoidance of ālabh. So, instead of ārabh and ārambha being rhotacisms, ālabh and ālambha in Aśoka's usage seem to be not legitimate alternatives but dialectical changes in which l does duty for r. Barua, no less than Bhandarkar, treats ārabh and ārambha as natural forms in Aśoka, needing no comment (Aśoka and His Inscriptions, Part II, pp. 51, 52; Part I, pp. 139, 269).
We may add that Monier-Williams himself speaks only of these forms being erroneous and does not at all suggest a dialectal change - a rhotacism - as causing them. In explanation of the label "erroneous" he puts the letter L which, as p. xxxv clarifies, means "Lexicographers (i.e., a word or meaning which, although given in native lexicons, has not yet been met with in any published text)."
Page 291
But actually we have in Aśoka's edicts both the forms. And Barua (Part II, pp. 70, 92) quotes, as parallels to Aśoka's pranarambho (R.E. IV), the phrases "bijagāmalbhūtagāmasamārambha, pasu-nam cha samdrāmbho" from the Buddhist canonical work Digha I, p. 5, the phrase "pasunam samārambha" from the Buddhist Sutta-nipdta, 311, and the phrase "mriga-pasu-pakshi-byāla-matsyārambhān" from Kautilya's Arthaśāstra, IV. 3 (prose treatise). Evidently, ārambha has been used in perfect good faith as an authentic locution and the very question whether it is a rhotacist product appears irrelevant. What applies to it must apply by analogy to ārabh.
4
When was R.E. XIII Inscribed? - Deb's Logical Revision
R.E. XIII is said to have been inscribed when all the 5 foreign kings were alive to Aśoka's knowledge. We have quoted Thomas's inference from the chronology of Antiochus, Ptolemy, Magas, Antigonus and Alexander (of Epirus) that the edict was not long posterior to the year 258 B.C. Aśoka's accession is usually dated to c. 269 B.C. Can we show that the edict in question was inscribed just when, in view of those 5 kings, it should have been? If we can, at least some presumption may be created against us that they might be Greeks. Emile Senart suggested from what he considered the internal evidence of the edict that 13 years were the interval between it and Aśoka's coronation. Such an interval would suit very well the proponents of the modern chronology. Of course it would not be any kind of decisive hammer-blow to the arguments we have adduced but it certainly would not leave these arguments absolutely at ease. A vague touch of the unresolved may linger. On the other hand, if we can substantially undermine Senart's suggestion, we shall set the proposed Greek kings hanging totally in mid-air, for the conjunction seen between them and the date of the edict will vanish.
Bhandarkar,1 no champion of the traditional-Purānic time-scheme, has yet taken cognizance of a cogent objection against Senart by Hrit Krishna Deb, author of Aśoka's Dhammalipis. He sums up the objection as follows:
1. Aśoka, p. 50. fn.l.
Page 292
"Pillar Edict VII is dated [by Aśoka himself] in the twenty-seventh year [of his reign] and is admitted by all to be a resumé of the multifarious measures which Aśoka adopted up till that year for the dissemination of his Dharhma. It is puerile to suppose that P.E. VII is a mere resumé of the domestic measures he adopted for his subjects, and not also of what he did for foreign peoples. The field of Aśoka's activities is all mankind, and he never draws any factitious distinction between his subjects and foreign peoples so far as Dharhma is concerned, his ideal being that of Chakravarti Dharmika Dharmarāja. The carrying out of philanthropic works (R.E. II) and the propagation of Dharhma (R.E. XIII) in the realms of the Greek rulers are such important things that Aśoka would most certainly have made mention of them in P.E. VII, if he had heard, when it was engraved, that they had met with any appreciable measure of success in those foreign countries. The omission is significant, and shows that Rock Edicts II and XIII could not have been promulgated prior to P.E. VII, that is, the 27th regnal year."
Vincent Smith,1 though a follower of Senart, was still aware of the difficulty and confessed: "I cannot explain the failure to commemorate the foreign missions which occupy a prominent place in the Rock Edicts."
Bhandarkar's Three Arguments for Deb
Bhandarkar adds three arguments in Deb's support. The first points out that Senart's "thirteen years after coronation" marks only the lower limit: what is proved is merely that Rock Edict XIII could not have been promulgated before this time and hardly that it could not have been promulgated after it. "It is true," says Bhandarkar2 about the Fourteen Rock Edicts, "that no less than four different dates are found mentioned in this series (R.E. IV, V, VIII and XIII), but it is nowhere stated that this whole set of Dharhma-lipis or any component thereof was inscribed in any particular year. They are dates of the different events alluded to in the different parts of this series, and not of the actual engraving. The latest of these is the thirteenth year of Aśoka's reign, and this has been proposed by Senart as the date when the fourteen Rock
1.The Early History of India, p. 169.
2.Op. cit., pp. 286-87.
Page 293
Edicts were incised. The French savant, it is true, has been followed by other scholars, both Indian and European. But this date cannot reasonably be taken as the actual date of the inscribing. All that we can logically conclude is, not that the whole set was engraved in the thirteenth regnal year, but only that it could not have been engraved before that year. We have, therefore, to fix the date for this series on independent grounds."
The second argument1 runs more directly in favour of Deb's conclusion: "After giving the date of its promulgation P.E. VII ends thus: '...This the Beloved of the gods saith: this Dharhmalipi should be inscribed where there are stone pillars or stone tablets so that it may long endure.' This clearly shows that whatever records [Aśoka] had already engraved were engraved on stone pillars and stone tablets only. There is indeed no reference here to the inscribing of Dharhma-lipis on parvatas or rocks. The idea does not seem to have occurred to him till after the twenty-seventh year of his reign, the date of Pillar Edict VII. This shows that all his Rock Edicts, whether they are the Fourteen Rock Edicts or the Minor Rock Edicts, must have been engraved when the work of inscribing the Seven Pillar Edicts came to an end."
The essential reasonableness of such a sweeping conclusion comes home to us after Bhandarkar2 has reflected on the word likh: "...the word likh which means 'to engrave' also means 'to write', and sometimes it is very difficult to understand which sense is intended. Thus R.E. IV has idam lekhāpitarh twice, e.g., in Girnār (J) and (K). Here it no doubt seems tempting to take likh in the sense of 'to engrave'. But if we do so, we shall be compelled to suppose that while this edict was inscribed in the 12th, the next was in the 13th year, because, as a matter of fact, R.E. IV is dated in the 12th and R.E. V in the 13th year of Aśoka's reign. And Senart has adduced some solid grounds to prove that this series of Rock Edicts forms one ensemble and was not engraved in successive additions. Similarly, there is nothing to show that the term likh is necessarily used in the sense of 'to engrave' in the case of iyarh dhammalipi likhitā which occurs about the end of R.E. XIII. It seems safer and better to take likh in the sense of 'to write' in all cases except where there are definite indications pointing to the other sense, such e.g., as where the words chira-thitikā-hotu ['that
1.Ibid., p. 288.
2.Ibid., pp. 283-84.
Page 294
it may long endure' or 'in order that it may endure for a long time'] occur as they do in R.E.s V and VI and P.E.s II and VII."
Bhandarkar adds: "In one instance even the engraver, we find, confounded these two senses of the word. Thus in R.E. I, the Dh[auli] and Jaug[uda] versions have parvatasi before lekhāpita in the very first line, and one is therefore inclined to translate the latter by 'caused to be engraved' instead of merely 'caused to be written'. It seems, however, that this reference to parvata was not in the original draft of Aśoka and was inserted locally when it was incised at Dh. and Jaug. For in the same edict further on are repeated the words iyarh dhammalipi.... lekhitā, just where Aśoka says that two peacocks and one deer alone shall be killed for the royal table. He could not have said this, if lekhita had meant engraved', for this series, viz., the Fourteen Rock Edicts, must have been inscribed at different times as their find-spots are situated at different distances from Pātaliputra. They could not have been engraved at one and the same time, that is, at any particular time when Aśoka could make that reference to his royal table. This is possible only if by lekhita is meant 'the writing' and issuing of that edict, because it is only when the edict was drawn up in the palace that he could say that only two pea-fowls and one deer were being slaughtered that day in the royal kitchen. This shows that the reference to parvata at the beginning of R.E. I in Dh. and Jaug. was inserted by local officers at the capital town of Kalihga without understanding what the following word likhāpitā really meant there."
Seeing that the question when the different edicts were inscribed is far from easy to answer in many cases, we should do well to give weight to the formula Bhandarkar quotes: chira-thitikā-hotu. This formula points definitely to engraving on stone, be it rock, pillar or slab; but no regnal year is associated with it in R.E.s V and VI, while R.E. IV's regnal year which is associated with the word "likh" is not associated with an expressed plan for the long endurance of the message, a plan which would directly imply rock-engraving. Hence no sign exists in any R.E. to give the lie to what Bhandarkar reads in P.E. VII's omission of the word parvata, and to prove that any R.E. was engraved before Aśoka's 27th year.
Barua1 has raised a voice of dissent over likh: "Professor
1. Aśoka and His Inscriptions, Part II, p. 11.
Page 295
Bhandarkar opines that wherever a record is said to have been written with a view to its permanence, there the expression 'caused to be written' must he taken to mean 'caused to be engraved.' I have sought to show that wherever a dhammalipi is said to have been written there are phrases to indicate Aśoka's motive to see it long endure (Inscr., ii, p. 22). Behind the very idea of causing something to be engraved in stone was the certainty of its durability (cf. Aiiguttara I., p. 283: pāsāna lekhā chiratthitikā). When Aśoka by way of an apology said (R.E. XIV) that it was not possible to cause Dhammalipis to be 'written' all over his empire owing to its vastness, he must have meant by 'written' engraved, promulgated."
Apart from the plausible suggestion of "engraved" by "written" in the phrase from R.E. XIV, one is not certain what exactly Barua is driving at. Does he mean that whenever Aśoka writes something he implies the wish for its long duration and consequently for its being engraved on long-enduring stone? In that case likh would invariably connote the desire for engraving. This may be so, but the difference can still remain between an explicit and an implicit reference to getting a thing engraved, and the latter will imply an unengraved written order. Barua1 does not appear to deny such a difference when at one place he touches on the chronology of Minor Rock Edict (Rūpnāth, Sahasram) in relation to P.E. VII: "...Whatever the actual date of the dhamma-savana [religious message], the date of its engraving is posterior to the erection of pillars by Aśoka..." Here sāvana connotes the still unengraved writing. A little later Barua has the phrase: "R.E. VI speaks indeed of giving verbal orders for announcement or proclamation (srāvāpakam)..." Here word of mouth followed by document to be read is the sense, and whether the document be intended to persist permanently or not, Aśoka in the same edict evidently feels the need and the advisability of engraving to ensure permanence, for he ends: "And what little effort I make, - what is it for? (in order) that I may be free from debt to creatures, that I may render some happy here and that they may gain heaven in the next world. For this purpose have I caused this document of Dhamma to be engraved: what for? - in order that it may endure for a long time and that my sons, grandsons and great grandsons
1. Ibid., p. 15.
Page 296
may similarly follow me for the welfare of the whole world...."1
Ultimately, Bhandarkar's distinction is to be adjuged by (1) concentrating on P.E. VII's omission of parvata, (2) realising that in general any vocable in the R.E.s which is understood to connote "engraved" may linguistically be just as well construed "written", and (3) attending to the peculiar contradiction between R.E.s IV and V, in the context of Senart's demonstration of the ensemble engraving of the 14 R.E.s, if likh is not equivalent to "written". This triple test leaves Bhandarkar fundamentally unrefuted by Barua.
Bhandarkar's third argument2 is related again to Aśoka's engraving plans. "Aśoka," he writes, "does not seem to have been able to carry out his whole programme of inscribing his edicts. He orders P.E. VII to be engraved on stone pillars and stone slabs. But, so far as we can see, it was incised only on one pillar, viz., the Delhi-Topra Pillar. As to stone slabs, none has been found with this edict inscribed on it. Probably none was engraved. Similarly, he intended inscribing Minor Rock Edicts on stone columns, but this too does not seem to have been done. It does appear after all that Aśoka commenced this work rather late in his life and was not thus able to carry through his whole programme of inscribing his Dharhma-lipis."
Bhandarkar's reference to Aśoka's intention to inscribe Minor Rock Edicts on stone columns can lead us to formulate a fourth argument on his behalf. Earlier we have quoted a phrase from his translation3 of the Rūpnāth Minor Rock Edict: "have this matter engraved on rocks. Here and far off where there is any stone column, have it engraved on the stone column." P.E. VII, aware of stone pillars, is unaware of rocks. Now both rocks and stone pillars are in sight. Is this not a sign of the Rūpnāth Minor R.E. being later than P.E. VII? With it a firm ground can be thought of as cleared for other R.E.s to be later as well.
Mookerji versus Bhandarkar
Mookerji has made an attempt at controverting Bhandarkar by way of refuting his whole position about the Rock Edicts. We must
1.Bhandarkar, op. cit., p. 316.
2.Ibid., pp. 289-90.
Page 297
in fairness quote it. "Arguments from silence," says Mookerji,1 "are always unreliable and in the present case seem to be specially so. What is stated about the scope or intention of P.E. VII is not stated in the Edict itself, but is a mere assumption. It is inferred from the contents of the Edict. The contents would rather lead one to infer that the Edict was meant to be a resumé of the various domestic measures introduced by Aśoka for the moral uplift of his own people, and not of what he had done for foreign peoples. All these, which are also mentioned in the R.E. [II, V and XIII] are mentioned in this Edict with a degree of elaboration and generalisation that is almost a sure indication that it was issued later than the R.E. The chief officers mentioned in the R.E., viz. the Rajukas and the Dharma-Mahāmātras, are also mentioned in P.E. VII; and the function of these newly created Dharma-Mahāmātras which are fully described in the R.E. are summarised in P.E. VII; the information about the public works of utility consisting of 'wells and trees planted along the roads for the needs of both man and beast', as given in R.E.II, is elaborated and supplemented in P.E. VII in a manner that undoubtedly points to the latter being later than the former; lastly, as instances of generalisation and reference in the P.E. VII to the R.E. may be mentioned the statement that for the spread of the Dharma Aśoka had religious messages (dhamma-sāvanāni) proclaimed (sāvdpitāni) and religious injunctions (dhammānusathini vividhāni) ordained (ānapitāni); along with his Pillars of Piety (dhammathambhāni) and the special officers to preach and expound (paliyovadisamti pavithali-samti) the Dharhma. There is also the other generalised statement in the same Edict that the growth in Dharma of the people may be accomplished in two ways, by Dhamma-niyama, by regulation, and by Nijhati, reflection. Thus the contents of P.E. VII, viewed as a whole, show without doubt that it is meant to sum up Aśoka's moral measures for his people and, as such, it was the last of his Edicts. Above all, arguments from the inclusion or omission of certain matters in the two classes of Edicts cannot be conclusive. Does not Aśoka himself warn us on the point - 'Nacha sarvam sarvatra ghatītam', - 'nor is all suitable in all places'?"
1. Aśoka, pp. 42-3.
Page 298
A Refutation of Mookerji
In appraising the force of Mookerji's criticism we may begin by noting that Barua1 who does not see eye to eye with Bhandarkar is yet unconvinced that Pillar Edict VII is the last edict of Aśoka. He finds reason to place the two inscriptions that are called Separate Rock Edicts later than it. And he also places what is known as Minor Rock Edict between it and the first six Pillar Edicts. Bhandarkar therefore does not appear to be completely vulnerable at Mookerji's hands. Is he vulnerable even where Barua and Mookerji agree? Are there factors in Bhandarkar's favour, which they have failed to reckon with? We are concerned mainly with Rock Edicts II and XIII. In relation to them we may say, first of all, that if Aśoka's domestic measures for the sake of dhamma deserve in his twenty-seventh year a resumé and if the foreign missions date so far back as his thirteenth year we have every right to expect a resumé of them also at more or less the same time. If P.E. VII was meant to be confined to domestic measures of dhamma and hence not to summarize the foreign ones, then where is the summary of the latter? The absence of such a summary is a vacuum that demands to be filled. We have here no simple argument from silence: the silence is logically significant and suggestive - and the vacuum we have spoken of can be filled only by shifting the two edicts from their supposed place to after P.E. VII. There is no summary of the foreign missions either in P.E. Vll or separately because there could have been no success worth mentioning earlier and summarizing afterwards: the success must have come later and was embodied in what may be regarded as at once announcement and resumé.
Mookerji's argument from "elaboration and generalisation", though not quite negligible, is far from happy, for it is not consistently enforced by him. If elaboration is, as he holds on comparing P.E. VII and R.E. II, a sign of the former being subsequent, we should consider R.E. V later than P. E. VII because, on his own showing, the description of the newly-created Dharma-Mahāmātras" is full in the former and skeletonic in the latter. Whether a piece of information is full or skeletonic is hardly a vital issue: both are consistent with the nature of a resumé. What a resumé is essentially intended to do is to gather together scat-
1. Aśoka and His Inscriptions. Part II, pp. 10-18.
Page 299
tered things in a final totting up: it may discharge this function in any way it likes. Nor is it always necessary that for it to mention things the things must already have been mentioned elsewhere. If a new work has been done after the mention of several others in earlier edicts, it may figure straight away in the resumés as an additional feature. All that is needed is previous execution rather than previous mention: Aśoka is not gathering up things so much from his edicts as from his life. And Mookerji admits this when he. talks of information being "supplemented". So the fact that P.E. VII informs us about "wells and trees planted along the roads for the needs of both men and beast" should not make us expect a similar piece of information before Aśoka's twenty-seventh year. An edict like R.E. II which does contain such information is not necessarily earlier than P.E. VII. It may very well be later and repeat the information if a special context calls for repetition - as evidently the context of R.E. II does with its combination of the foreign works and the domestic. As for generalizations about the proclamation of religious messages and the ordainment of religious injunctions and special officers to preach and expound the dhamma, they can only involve that the matter of some Rock Edicts precede in a written form P.E. VII and not that P.E. VII was subsequent to the engraving of them. Bhandarkar's distinction between the two meanings of likh should be remembered.
The plea that, since Aśoka warns us that all is not suitable in all places, we should not make much of inclusion or omission of certain matters in the two classes of edicts - the plea is as weak in the present case as the playing down of the argument from silence. All depends on the particular situation in which inclusion or omission is done. And against Mookerji's closing quotation from R.E. XIV (Girnār) we may pit another from the same source: "Puna puna vitum tasa tasa athasa madhurataya kimti tatha pad pajetha" - "Something is said again and again for the sweetness of the topics concerned..." Surely, to a king who said in the Kalinga R.E. I (Dhauli), "All men are my children", if the work of dhamma had been accomplished abroad before P.E. VII, it would have been sweet enough for a resumé repetition just as much as dhamma-work achieved at home.
Mookerji's answer to Bhandarkar cannot be pronounced sufficiently effective. And it loses whatever small value it may theoretically have here and there when we weigh against it some consi-
Page 300
derations relating to the query: "What number of years could the foreign missions have taken to reach the results mentioned in R.E.s II and XIII?" Mookerji1 himself says that the missionary activity in the foreign countries must date from an earlier time than the year assigned to these edicts and more so because it is described in the edicts as having been successful and borne fruit. Face to face with the statement in R. E. XIII that the Kaliriga war, whose cruelties turned Aśoka's mind actively and enthusiastically towards Buddhism, was waged in his 8th regnal year, Mookerji who puts this edict in Aśoka's 13th year knows that he cannot have had more than 5 years for the development of the successful results. Five years seem too few to us, but Mookerji finds them adequate and we may grant that they cannot be immediately dismissed. What can be said by us legitimately is that Mookerji, on his own premises, is illogical in allowing even as many as five years on the evidence of the edicts. Apropos of a statement in R.E. V, he2 has the following to tell us.
When Aśoka "had been consecrated thirteen years" he instituted "a special body of officers... called by the new and appropriate title of Dharma-Mahāmātras.... The activities of this department extended over a wide field, even beyond the limits of his direct jurisdiction or administration.... The department had also to send out some officers as Dutas, envoys or ambassadors, carrying Aśoka's religious message to foreign countries both in the north and south". The point now is: If Aśoka's special dhamma-work commenced after the appointment of the Dharma-Mahāmātras and if they alone exercised the office of sending out religious ambassadors to foreign countries and if they were appointed only in the thirteenth year of his reign, how could Aśoka's missionary activity in these countries be described, in an edict promulgated in the very year, as having been successful and borne fruit in a spectacular manner? Assuredly, the time of less than a year is preposterously short. But nothing longer can be allowed if the edict came out in the thirteenth year which is the earliest possible for the origin of the Dūtas. Here Senart's theory meets with a reductio ad absurdum.
In another context Mookerji3 himself has scored a similar point.
1 Aśoka. p. 32.
2.Ibid., p. 29.
3.Indian Culture, Vol. I, pp. 66-68.
Page 301
About the translations and interpretations of a certain passage in Minor Rock Edict, where Aśoka mentions an achievement of his own and offers himself as an encouraging example to even the least, Mookerji has declared: "All these translations and interpretations, however literal or true to text, do not yield much sense or significance. They only agree in ascribing to Aśoka the credit of achieving the impossible. It is claimed for him that within the short space of 'a little more than a year' he was able to convert the wild tribes of India to Brāhmanism and to impart to them a knowledge of its gods. It is also as an alternative claimed that within that short interval he was able to make the millions of Indians distinctly religious and godly, or to abolish the battle of creeds, the strife of sects, for which India is notorious to this day. In my opinion, these interpretations, besides condemning themselves by suggestions of the impossible, do a great injustice to the character of Aśoka by letting him appear as a man given to vainglorious boasting." So Mookerji chooses to give a "subjective" interpretation: he makes Aśoka say, "By a little more than a year's exertion, lo! I have made such progress: it is, indeed, the men in Jambudvīpa [the best country, according to the sacred texts, for spirtitual life] who could thus have commerce with gods in such a short time. But let it not be understood that such progress is only for the great, like me..." Now, if a year's religious propaganda is taken to be absurdly inadequate for turning godlike the entire people of just one country, India, or even for making the wilder tribes acquainted with the gods or for harmonizing the warring creeds, how infinitely unreasonable it is to believe Aśoka to be recording a resounding triumph in dharhma-work not only at home but abroad in five countries in less than a year! Here we cannot even choose a "subjective" interpretation. We have either to accept him literally and objectively as claiming the impossible and hence acting as a champion among fatuous braggarts or else dismiss unreservedly the thirteenth year for R.E.s II and XIII.
Mookerji does not touch at all on Bhandarkar's gloss on Aśoka's engraving plans, to which we on our own have sought to bring extra strength. P.E. VII knows only stone pillars: it does not have its eye on rocks, whereas the Rūpnāth Minor R.E. mentions both rocks and stone columns. A before and an after in time seem pointedly suggested as between P.E. VII and R.E.s. In the present context he has not even tried to retort: "If P.E. VII ignores rocks,
Page 302
does not your minor R.E. ignore stone slabs? Does not this tend to indicate its earlier existence and tend to blunt your point and render it irrelevant?" But Mookerji elsewhere has tackled the issue in an indirect way, and Barua has met his position. After controverting Hultsch's contention that this Minor R.E. connects with the events of Aśoka's 10th year after accession and that R.E. XIV which is ascribed to the 13th year must have meant the text of this inscription, Barua1 writes:
"Professor Mookerji employs a much subtler argument to prove an early date of M.R.E., namely, that when it was promulgated, the engraving of a record on a rock or a stone-pillar was just an idea and not a fait accompli. But this is falsified by the fact that the direction as to engraving on Rock or Pillar, occurring at the end of M.R.E. (Rū[pnāth], Sa[hasarām]) is in language and effect the same as that at the end of P.E. VII... Comparing them, none can fail to notice that the direction appended to M.R.E. presupposes the existence of rocks and stone-pillars, while that appended to P. E. VII presupposes the existence of stone-pillars and stone-slabs as engraving material. The direction in M.R.E. does not suggest the use of stone-slabs as engraving material, but that in P.E. VII does, and the reason undoubtedly is that already prior to the engraving of P.E. VII these were used as material for the engraving of the Bhābru Edict."
One may adjoin two further considerations. Stone-slabs, also called stone-blocks, are an item of much less importance than rocks or pillars and their mention or non-mention is not particularly significant. Besides, Minor Rock Edict, which is in 10 recensions, has itself one of its recensions - that at Bairāt - on a slab or block of stone near the place where the Bhābru Edict on a small stone block stood before its removal to the Indian Museum, Calcutta.2 So there should really be no question of M.R.E. not knowing of stone-slabs.
Perhaps a final effort will be made to date all the R.E.s earlier than P.E. VII by citing Barua's statement:3 "The Rock Edicts speak nowhere of the stone-pillars (silāthambā)." Here the decisive question is "Do the R.E.s speak at all of any engraving material?" If they do not, their non-mention of stone-pillars has
1.Op. dr..
2.Ibid., p. 5.
3.Ibid., p. 13.
Page 303
no relevance to our discussion. Bhandarkar1 informs us that they do not refer to any specific material but only tell us at times of Aśoka's engraving them in stone for the purpose of making them last. So - "stone" being a blanket-term - even the final effort fails. And it is illuminating to note the reason Bhandarkar gives for the non-mention of any specific material. It comes up apropos of the question whether the Fourteen Rock Edicts or the Minor Rock Edicts were first inscribed. Bhandarkar2 writes: "... it deserves to be noticed, as we have seen, that in the Sahasram and Rūpnāth epigraphs Aśoka orders that edict to be inscribed on a parvata, and on a stone pillar wherever it exists. The phrase 'on a stone pillar wherever it exists' shows that these objects were engraved after he had set up his pillars. This also shows that the idea of inscribing parvata or rocks was new to him at that time. It seems therefore that soon after the pillars were engraved Aśoka took up the work of incising Minor Rock Edicts, which must have been followed by that of the Fourteen Rock Edicts. When the latter series was being inscribed, the idea of engraving rocks as well as pillars had become so familiar that Aśoka makes absolutely no reference at all to such material, he makes the general remark that those (Fourteen) Rock Edicts were engraved on stone in order that they might endure permanently."
A psychological sequence is conjured up vis-à-vis Aśoka's attention to the various kinds of material for edict-engraving. Thus one more point is scored in favour of putting R.E.s II and XIII after P.E. VII. In summing up the discussion we may quote Bhandar-kar's words:3 "We are... compelled to infer that Rock Edics II and XIII, in fact the whole set of the Fourteen Rock Edicts, came to be engraved after the Seven Pillar Edicts were promulgated." In connection with his reference to "the whole" we may repeat his statement:4 "Senart has adduced some solid grounds to prove that this series of Rock Edicts forms one ensemble and was not engraved in successive additions."
All in all, the case for Deb's contention is quite substantial. Bhandarkar5 reverts to it and reaffirms it before trying to fix the
1.Op. cit., p. 289.
2.Ibid., pp. 288-89.
3.Ibid., pp. 287-88.
4.Ibid., p. 283.
5.Ibid., p. 287.
Page 304
date for the Fourteen Rock Edicts on the basis of the impossibility of R.E.s II and XIII having been engraved before P.E. VII, which is obviously a summing-up of Aśoka's Dhamma-work before his 27th regnal year. "But we find absolutely no reference made in it to the works of charity he executed in and outside India and which have been described in Rock Edict II or to the successes which crowned his missionary efforts, as we learn from Rock Edict XIII, not only in his empire but also in the dominions of his neighbouring sovereigns, Greek and Indian. Both these matters are of such paramount importance in Aśoka's estimation that he would never have failed to make mention of them in Pillar Edict VII, if he had known about them before the twenty-seventh year, the date of that edict."
Revolutionary Consequences of Deb's Contention -
Bhandarkar's Futile Attempt at Readjustment
If we put Aśoka around the middle of the 3rd century B.C. and date his accession to c. 269 B.C. R.E.s II and XIII must be ascribed to c. (269-27 = ) 242 B.C. Then it becomes irrational to think in terms of Antiocbaus II, Ptolemy II, Magas, Antigonus and Alexander of Epirus as being all alive, as they have to be, along with Aśoka. In brief, even on the current chronology these Greek monarchs must be ruled out - and there are none others to be set in their place. With their exit, no excuse is left for this chronology at all.
Bhandarkar,1 however, makes a gallant effort to save the chronology: "Supposing that both these Rock Edicts were issued in the twenty-eighth year, the date must correspond to a year when the five Greek rulers were alive. If Alikasumdara of Rock Edict XIII is Alexander of Epirus, this year would fall between 272 and 255, but if Alexander of Corinth is intended, then between 252 [his accession year] and 250 [the year in which Magas most probably died]. The latter supposition is more probable.2 So we may take it that the twenty-eighth regnal year of Aśoka corresponds to 251 B.C. If this calculation is correct, Aśoka probably ascended the throne circa 279 B.C. Whatever the actual result of such a calculation may be, it is based upon two things, namely, the date of Rock
1 Ibid., pp. 50-51.
2. Journal of the Royal Asiatic Society. 1914. p. 945.
Page 305
Edicts II and XIII, and the identification of the Alikasumdara of the latter edict."
Bhandarkar has done his best in respect of those "two things", but "the actual result" of his calculation must play havoc with the modern historical perspective. If Aśoka ascended the throne c. 279 B.C., his father Bindusāra by the estimate which is lower than that of the Buddhist records would be king in (279+25=) 304 B.C. and Aśoka's grandfather Chandragupta would wear the crown, by the estimate accepted by modern historians, in (304+24=) 328 B.C. This is 2 years prior to Alexander the Great's invasion of India and anticipates by 7 years the most eligible date among our scholars for the coronation of Sandrocottus: 321 B.C. The Buddhist figures for the reigns, towards which our historians lean, would make Bindusāra king in (279+27 or 28=) 306 or 307 B.C. and Chandragupta in (306 or 307 +24=) 330 or 331 B.C., even if the Mahāvamsa's 34 years are corrected to 24 as usually done. How then are we to identify Sandrocottus with Chandragupta? And, when Seleucus Nicator attempted to invade India in c. 305 B.C., he must have confronted not the founder of the Mauryas but Bindusāra. Again, Megasthenes who from c. 302 B.C. stayed at the court of Palibothra (Pātaliputra) would have not Sandrocottus but his son as his royal host. All our historical data will have to be jettisoned.
Yes, neither the currently favoured time-scheme nor Bhandar-kar's emendation of it will help. With Aśoka's R.E. XIII, as well as his R.E. II, shifted, as it must be, beyond his 27th regnal year, he ceases to fit into the niche modern historical opinion gives him and the five foreign kings get de-Grecized. Deb's logic and the development from it by Bhandarkar reveal to an extreme degree what we have already laid bare acutely enough - the misfit of Aśoka and his 5 rājas in the post-Alexandrine century.
However, we shall not rudely brush aside all conceivable further objections to our new chronology but accord them full hearing and criticize each within its own particular universe of discourse. In that manner our reconstruction of India's ancient history will acquire a multiple concreteness.
Page 306
The Kandahār Inscription
We are now up against the last and, in some eyes, the greatest obstacle to any rejection of post-Alexandrine Greek kings as the rulers of the bordering dominions mentioned in Aśoka's R.E. XIII: the bilingual Aśokan inscription discovered in April 1958 at Kandahār. The obstacle may be summed up: "The two languages are Greek and Aramaic. The Greek is of the type current in the period c. 275-225 B.C. and the text in it is engraved above that in Aramaic. Linguistically the latter resembles in general the epigraphs of the Achaemenid empire in which the Aramaic is sprinkled with Irānian words. It may be taken to have continued the practice after this empire met its doom at the hands of Alexander the Great in 330 B.C., except that the Irānian words here have an Avestan - i.e., Mazdean religious - character. So in general it tallies very well with the chronological indication of the other text, and both of them put us pretty precisely in the post-Alexandrine era to which Aśoka has been assigned by modern historians who identify Sandrocottus of the Greek accounts with Chandragupta Maurya and the Aśokan R.E. XIII's five named rājas, of whom the first is the Yona rāja Arhtiyoka, with five of the Greek monarchs who ruled various parts of Alexander's empire in the wake of its founder's death."
J. Filliozat1 compares the Kandahār inscription with Aśoka's R.E.s I and IV. Like R.E.I, it speaks of abstention from animal-slaughter and, like R.E. IV, it announces a new and more prosperous era by the establishment of a new behaviour consisting in cessation of animal-slaughter and in obedience to parents and elders. The sole difference between the two announcements is that R.E. IV enjoins good conduct towards Brāhmanas and Śramanas, while the bilingual inscription omits reference to them. As Brāhmanas and Śramanas are prominent again and again in Aśokan texts and would be most expected - nay, inevitable - in prescriptions as here of meritorious behaviour, their omission is too glaring to be accidental. Filliozat informs us that they are omitted because, as Aśoka himself declares in one version of R.E. XIII, they did not inhabit the Yona country and because the
1. Epigraphia Indica, Vol. XXXIV, Part I, p. 1 ff.
Page 307
present inscription was engraved for the Yonas - the Yonas who are Greeks and who, by being Greeks, are without Brāhmanas and Śramanas. Filliozat further asserts that the find-spot of the inscription indicates the region of the Alexandria of Arachosia, the colony-city which Strabo (XV.2,9) has reported.
But those who are of Filliozat's mind have to account for the fact that these two religious orders are omitted from the Aramaic version no less than from the Greek. If the Greeks were the Yonas, who were the people addressed in Aramaic and why are they also treated as though they lived without Brāhmanas and Śramanas?
E. Benveniste1 argues out three propositions, after reflecting on the Aśokan compound yonakambojesu in R.E. XIII, which seems to take the Yonas and the Kambojas together as one whole. The three propositions may be briefly stated thus:
1.The Yonas (Yavanas) within Aśoka's empire were the Greeks settled in the country of the Kambojas, with whom they -though culturally different - formed a single population.
2.The Kambojas were the Irānian autochthones, following the Mazdean religion, of the country round present-day Kandahār.
3.The Greek text of the inscription was intended for the Yonas, and the Aramaic for the Kambojas.
Against these three propositions, which are linked with Filliozat's stand, we may put up four of our own which are at utter variance with them as well as with that stand:
1.The Yonas within Aśoka's empire were not the Greeks and they did not form with the Kambojas a single population.
2.The Kambojas were not Irānians but Irānianized Indians who neither occupied the country round Kandahār nor followed completely the Mazdean religion.
3.The Greek text was intended for the Greeks and the Aramaic for the Yonas who, like the Kambojas, were Irānianized Indians but, unlike the Kambojas, inhabited the Kandahār region and were fully the followers of Mazdeanism.
4.Only the Aramaic text is Aśokan and it was set up in a much earlier age than the Greek, which was engraved in the period c. 275-255 B.C.
1. "Une bilingue gréco-araméene d'Aśoka", Section IV, Journal Asiatique, CCXVI, 1958, 1, Paris, pp. 47-48.
Page 308
Could the Aramaic Text have been for the Kambojas?
Benveniste assumes that Indian tradition knew the Kambojas as a foreign people. We have already seen the Manusamhitā (X.43-44) grouping them with some other tribes as degraded Kshatriyas, tribes who lost Vedic rites and customs and thereby became "Dasyus". The Mahābhārata (XIII. 33, 21) calls them an Indian people that became Vriśalas (degraded castes) from seeing no Brāhmanas. Buddhist literature too tells the same story, a loss of original Aryan customs leading to a barbarous state.1 To clinch all this evidence against original foreignness, we may add Pusalker's pronouncement:2 "The earliest mention of Kamboja occurs in the Vamśa Brāhmana of the Samaveda where a teacher Kamboja Aupamanyava is referred to. The sage Upamanya mentioned in the Rigveda (X. 102.9) is in all probability the father of the Kamboja teacher. From the fact that Kamboja Aupamanyava is stated to be a pupil of Madragara, Zimmer infers that the Kambojas and the Madras were close neighbours in north-western India. The speech of the Kāmbojas is referred to by Yāska as different from that of other Aryans and Grierson sees in this reference the Irānian affinities of the Kāmbojas, but the fact that the Kamboja teachers were reputed for their Vedic learning shows them to have been Vedic Aryans, so that Kāmboja was an Aryan settlement. Later on Kāmbojas settled in the north-west of the Indus, and were the Kambujiyas of the Old Persian inscriptions."
The last point is an error, since Benveniste3 categorically tells us in all honesty that although he sees the Kambojas as Irānians the Persians do not know of the name Kamboja at all and he rejects the rapprochement proposed by J. Charpentier between the Kambojas and the name of the Persian king Cambyses, Old Persian Kambujiya. But all the rest of Pusalker's information is incontrovertible. And what Benveniste categorically tells us is extra evidence against the Kambojas being not Indian but Irānian.
On two points Benveniste relies for Irānianizing the Kambojas. First, the passage of the Jātaka which we have dealt with and which ascribes a certain Mazdean custom to "many Kambojas", Kambojakānam... bahunnam. Surely, not all Kambojas and not all
1.Tribes in Ancient India, p. 7.
2."Aryan Settlements in India", The Vedic Age, pp. 259-60.
3.Op. cit., p. 48.
Page 309
of Mazdeanism are involved here. The evidence is markedly insufficient. Benveniste's second point is connected with Grier-son's observation on the Kambojas' speech. Benveniste cites Yāska's Nirukta (II, 2.8) to the effect that the Kambojas spoke a language with Irānian words in it (e.g., śavati, "to go"). But does Yāska imply that this language was Aramaic with Irānian words strewing it? Not at all. His distinction of the Kambojas' speech from other speeches known to him can only mean that this speech, as distinguished from those others, carried Irānian words within its Indianness. What Grierson1 himself deduced from Yāska's context and other data was that the Kambojas must have used either Sanskrit with an infusion of Irānianized words to which they gave Indian inflexions or else a language partly Indo-Aryan and partly Irānian. Grierson is pretty far from Benveniste's linguistic generalization. His Kambojas would never qualify as the people for whom Aśoka got the Aramaic version engraved.
And we should remember Agrawala's remark:2 "The Nirukta... is a commentary on a traditional list of Vedic words known as the Nighantu." The Nighantu itself (11.14) lists śavati among the verbs meaning "to go". So the implication of Yāska's statement is: śavati in the sense "to go" is an ancient Vedic verb once in general Indian use but now restricted to the Kambojas who, unlike other sections of the Indian people, have somehow preserved it. Whether originally Irānian or no, this verb cannot make the Kambojas Irānian any more than it can make Irānian the Vedic literature of which the Nighantu took stock. And for Yāska it was not in the least a pointer to any non-Indianness.
How little he considered the Kambojas non-Indian is further evinced by his etymology (doubtless "folk-etymology") of their name - "Kambhoja" to him - in the same context as his comment on śavati. He writes: "Kambhojāh kambalabhojāh kamanīyabho-jāh vā. Kambalah kamanīya bhavati" - "The Kambhojas are fond (bhojāh) of blankets (kambala) or they are fond of pleasant (kamanīya) things. The blanket is a pleasant (kamanīya) thing." Yāska takes "Kambhoja" to be composed of bhoja preceded by an abbreviation, kam, of kambala which is at the same time
1.The Journal of the Royal Asiatic Society, 1911, pp. 801-802.
2."Yāska and Pānini", The Cultural Heritage of India (Calcutta, 1958), Vol. I, p. 294.
Page 310
kamaniya.1 This etymology, drawing confirmation from the fact that the Kamboja country was noted for many of the best kinds of blankets,2 and basing itself entirely on Indian words, could never have been pertinent to a people thought to be Irānian.
So Benveniste's outlook on the Kambojas and on their relation to the Aramaic text may be wholly ruled out. With its exclusion no intrinsic reason exists to see the Kambojas as devoid of the services of Brāhmanas and Śramanas and therefore to see any inscription omitting these services as having been meant for them. We have also Aśoka's explicit assertion, emphasised by Filliozat, that only the Yona country lacked Brāhmanas and Śramanas. The presence of these religious orders amongst the Kambojas is thus implicitly indicated by Aśoka. We have thus a strong prima facie case against Aśoka addressing the Aramaic version at Kandahār to the Kambojas as well as against their forming a single population with the Yonas.3 Benveniste's proposition that the Yonas were settled in the country of the Kambojas must go by default - unless an incontrovertible argument is found for it.
What does Aśoka's "Yonakambojesu" signify?
Benveniste urges on us the compound yonakambojesu of R. E. XIII. But there is nothing specially Aśokan in this expression. Have we not seen the very same combination in the pre-Mauryan Buddhist canonical work Majjhima Nikāya (11.149), where Buddha, telling Assalayana of several border countries, brings in yonakambojesu? No hint is here of a single Yona-Kamboja country. Nor are the two terms always put together in a compound. For example, the Manusamhitā (X.44) has Kāmbojā Yāvānā.
The Aśokan edicts themselves have a number of compounds putting together various entities without implying the slightest fusion. Can anybody say that the Gandhāras were a single population with the Yonas and the Kambojas because in R.E. V (Girnār) we have the compound yonaKambojagamdharānām? In the same edict the Mānsehrā version reads rathikapitinikana while the Sha-bāzgarhī reads rathikanam pitikanam, clearly implying that the
1.Barua, op. cit., Part I, p. 99.
2.Ibid., referring to B.C. Law, op. cit., p. 2.
3.When we touch on the epigraph known as Kandahār II we shall see what an Aramaic text from Aśoka for the Kambojas would be like.
Page 311
two peoples combined in the Mānsehrā are really distinct. In R.E. XIII (Girnār) we find Āndhrapārimdesu: should we say that the Āndhras and Pārimdas had no separate provinces but were inhabitants of the self-same province? In R.E. V (Mānsehrā) we have the word bramanibhyesu where, according to Mookerji,1 two entirely different and contrasting categories - Brāhmanas and householders - are combined. The Brāhmanas and Śramanas too, for all their common religious purpose, are not a single body and their distinctness is preserved in the passage in R.E. XIII to which Filliozat has referred: bamhane ch[a] samane (Kālsī); but in R.E. IV to which he also alludes we get the combination Brāhmanasa-manānam (Girnār). The combinations in the edicts serve various ends, of which perhaps the convenience of the moment is the most telling. None keeps in countenance the interpretation Benveniste puts on yonakambojesu. Rather, everything makes this word signify that the country of the Yonas was merely contiguous to that of the Kambojas.
This word cannot be cited to explain the absence of the Brāhmanas and Śramanas from the Aramaic text.
The Greeks, the Kambojas, the Brāhmanas and Śramanas
Even supposing the Yonas and the Kambojas occupied the identical province, we cannot account for the absence on the assumption that the Yonas were Greeks. We should have to accuse the Greeks of banishing religious ministrations natural to the natives of the territory the Greeks had settled in. But they are known to have been extremely tolerant of alien creeds and cults. We have already enumerated a host of testimonies to their keen interest in things Indian in both the Alexandrine and the post-Alexandrine epochs. Megasthenes has written in detail of the Brāhmanas and Śramanas themselves. And the very proclivity of the Greeks to look for Hellenism in diverse garbs among other nations would lead them to encourage and keep alive the religious legends and practices of the natives. If there was a Greek colony in Arachosia and there were Brāhmanas and Śramanas among the local Indian population, they would be free to pursue their vocations and would even succeed in Indianizing some of the interested foreigners.
1. Aśoka, p. 140. fn. 1.
Page 312
Of course, these facts would not entitle us to say that in an inscription for the Greeks, who had a non-Indian civilization, Aśoka would go out of his way to mention Brāhmanas and Śramanas. But they do entitle us to say that the omission of these religious ministers in the Aramaic text would never be due to the Greeks of the province.
All such omission would be due only to the Yona-character of the very people addressed in that text. So this people could not have been the Kambojas.
Now the problem faces us of two texts in two distinct languages omitting the Brāhmanas and Śramanas. Surely, both the peoples concerned could not be Yonas. And, if that is so, we have to ask: "Could one of the texts be Aśokan and the other not?"
Here our first guide has to be linguistics. The Greek employed is indubitably of c. 275-225 B. C, which includes the time usually assigned to Aśoka's reign. Is the Aramaic of the same period? If the answer is "No", Aśoka may not have belonged to those fifty years.
The Aramaic Text and Its Chronological Pointers
The expert study of the Aramaic text by A. Dupont-Sommer and Benveniste, who have themselves no idea of separating the two texts in time, has brought out three characteristics of the language which cannot but prompt in us such an idea:
(1)Benveniste1 notes: "Even leaving aside the rather unclear ptytw, we gain five Irānian words of good standard quality: patiz-bata, frabasta, frabasti, hupatyasti, mazišta, a high proportion for a relatively short text. The gain is as appreciable as it was unexpected." This comparatively large number of Irānian words embedded in the Aramaic would seem to distinguish the present text from the usual Aramaic inscriptions found in provinces once ruled by the Achaemenids of Persia from the 6th century down to 330 B. C.
(2)What distinguishes it yet more is the type of the Irānian words. Benveniste continues: "At the date of the inscription, in the middle of the 3rd century B. C, these words indicate, by their form, an Irānian language of ancient type, like Avesta or Old Persian, and not of the Middle Irānian stage: the notation of pati
1. Ibid., pp. 44-45. The translation from the original French is mine.
Page 313
with the mute and the final vowel, the notation of the final vowel -i in prbsty, are the principal proofs of this. To define the dialectal position of the language, we have a significant pointer in the form mazišta- which was restricted till now to Avestan." Elsewhere Benveniste1 speaks of "the rather unclear ptytw" as a word which is definitely Irānian but cannot be elucidated in the light of post-Alexandrine or even Achaemenid Irānian. And the difficulty is even greater with the possible alternative reading: ptytz. Thus there are forms which seem to set the Irānian part of the Aramaic text outside and beyond the whole range of the language known in the period not only after the Achaemenids but also during their rule.
(3) When we look at the non-Irānian element in the text - the strictly Aramaic element - we come across several signs-of archaism. We find Dupont-Sommer2 writing things like: "to note the archaic form", "as an archaic ittafal of this root", "it may be explained as corresponding to a form archaic or dialectal", "this form is altogether unprecedented... Is it the survival of an archaic form... ?" Here, again, are surprises in an inscription said to be later than any of the Achaemenid ones - surprises that look very much like indications of an antiquity answering to the suggestions of the Irānian and Avestan component.
The Correct Conclusion
Dupont-Sommer and Benveniste, arguing within the framework of the modern chronology, draw no revolutionary inferences from the above three characteristics of the language of the Aramaic text. But the signs are wide enough for a cleavage between this text and the Greek. Benveniste3 feels uneasy and asks whether we should conclude that in the middle of the 3rd century B.C. the population settled at Kandahār spoke the same language as the Avesta. He answers guardedly: "We need not yet go so far. All that seems allowable to affirm, on the indirect testimony of the inscription, is that the religious language coincided with that of the Avesta, and that Mazdeanism prevailed in this region."
But there is no reason why "the religious language" here should
1.Ibid., p. 36.
2.Section III, ibid., pp. 25, 27, 28, 30.
3.Ibid., p. 45.
Page 314
be so much like that of the Avesta rather than some other less ancient form which we might expect in the middle of the 3rd century B.C. The common people would be reached far more readily through the latter language. Benveniste himself cannot avoid the impression of the unexpected and says1 about the word hupatyasti which he2 recognises as composed with hu added to the Avestan Mazdean term patyasti: "That it should appear now at this date..." And when we consider the implications of mazišta and ptytw (or ptytz) we can hardly escape the hypothesis that the Aramaic version belongs to an earlier date than the Greek of the period c. 275-225 B.C. Also, we have to bear in mind that ptytw is not a religious term. It follows the phrase which in Aramaic signifies "ten years"3 and the word in the same place in the Greek text is "elapsed".4 So it is not only the religious language that carries the atmosphere of an Avesta-like tongue more ancient than even the Old Irānian of the Achaemenids.
Aśoka and the Language of the Avesta
The prevalence of a language of the Avestan type for all uses in the areas adjoining north-westem India in Aśoka's day would best explain some general peculiarities in the dialect of the Shāhbāzgarhī and Mānsehrā redactions of his edicts and even in that of the Girnār redaction. Here we are much nearer to Sanskrit than in other versions, which are more Prākritized, but Mookerji5 well remarks:
"As Michelson has pointed out, this dialect cannot be regarded as a mere lineal descendant of Sanskrit. It presents certain forms which establish its affinity to Avestan rather than Sanskrit: e.g. atikratām (G); susrusā, susrusarām (G) corresponding to Avestan susrusemno; G. srunāru, Shb. śruneyu and M. śruneyu, which agree with Avestan surunaoiti in structure as opposed to Sanskrit śrinoti."
What, of course, concerns us is not Avestan rather than Sanskrit forms but Avestan rather than later Irānian. And their presence
1.Section IV, ibid., p. 42.
2.Ibid., p. 43.
3.Section III, ibid., pp. 22-23
4.Section II, ibid., pp. 36-37.
5.Asoka, p. 244.
Page 315
indicates not only the former's general prevalence we have spoken of: it indicates also the antiquity of the Aśokan dialects involved.
Can the Greek Text be Much Later than the Aramaic?
If uninhibitedly we go by all the technical evidence before us we should be obliged to date the Aramaic text far earlier than the Greek and, associating it with the Aśokan language, separate the Greek text from Aśoka. Then the Aramaic would be the sole inscription set up by him. The Greek would be nothing more than a free translation and partial adaptation made of it under different circumstances in the second and third quarters of the 3rd century B. C. by Greek colonists round Kandahār and engraved by them in the empty space above the Aramaic on the rock-face.
But here two searching questions have to be met:
(1)Do not Dupont-Sommer and Benveniste, as well as all the other scholars who have written on the inscription, affirm that the Greek scribe no less than the Aramaic is dealing at first hand with an original Indian model? If the scribes do not depend upon each other, Aśoka cannot be divorced from the period 275-225 B.C. and we must be contented with a cautious and moderate conclusion from the technical evidence we have laid out.
(2)Is it not fantastic to think of a later-date addition to an epigraph - and particularly of such an addition by the Greeks to an epigraph of Aśoka?
Old Epigraphs, Additions to Them, the Greeks and Aśoka
We shall take up the second question first. To suppose a later-date addition to an epigraph of the remote past is not to imagine a state of affairs unparalleled in history. The Allāhābād Pillar Inscription of Samudragupta, son of Chandragupta I, is right under an old one - and the latter is actually an Aśokan epigraph. The inscriptions of the Śaka Rudradaman I and of Skandagupta, great-grandson of Samudragupta, are both on the same Junagarh rock - and they share this rock again with an edict of Aśoka. When two inscriptions of Aśoka himself have received additional matter from other ages, we should not be astonished if a third is supposed to have done the same.
The sole novelty in the present instance would be that the
Page 316
matter is a translation and that the later epigraph is put above rather than below the earlier even though it merely translates it.
But when the rock-face shows space available both above and below the Aramaic text, the Greeks' proverbial pride in their own language and their equally proverbial eye for artistic effect can easily be conceived as choosing the upper space even though the lower might be equally convenient. The fact of mere translation need not be given too much weight.
A close look at the lines of the Greek text1 seems to show that the first four lines are less interspaced than the rest, as though the engraver put them one after another somewhat near lest sufficient space should not remain for the ten more to follow and, then, finding space enough, put those lines a little apart. If the Aramaic were not already engraved, there would be no need to crowd the lines even slightly, for there is space on the rock for an Aramaic text twice the present length.2 The tentative crowding would be required only if a fixed height of space were available because the Aramaic stood already there in the middle of the rock.3
Perhaps we shall be asked: "Why should the Greeks of a later age have been interested to put up a version in their own language?" We know that during the seventy-five years or so after Alexander's invasion they were extremely interested in India and things Indian. We may particularly hark back to the eagerness by the Head of Philadelphus's Library at Alexandria to translate the books of the Hindus. To get a translation done of an old Indian semi-religious inscription discovered in Arachosia and to have it engraved on the same rock as the original - this is precisely what we may expect of the Greek Arachosian colony.
L. Robert4 tells us that a Greek emigration, in the Hellenistic epoch, always included intellectuals, schoolmasters, rhetoricians, poets, actors, philosophers, doctors, professors of every sort and every qualification, athletes, artists, marble-cutters, stone-engravers. He5 adds: "One may be sure that there was in Arachosia, as at Babylon, a gymnasium, with its young men, meeting-
1.Journal Asiatique, CCXLVI, 1958, I., Plate IV, facing p. 8.
2.Ibid., Plate III, facing p. 6.
3.In connection with an Aramaic Aśokan text called the Laghman inscription we shall see the non-sequitur in thinking the Aramaic text to be a sequel to the Greek just because the latter is above it.
4.Journal Asiatique cit.. Section II, p. 13.
5.ibid.
Page 317
place for festivals and conferences and conversations." The cultural character natural to a Greek colony is enough to account -more than ever in a period when Hellenic eyes were keenly turned on India - for the engraving of a Greek recension of an Aramaic inscription in the Kandahār-region.
A further query may be posed to us: "If we keep Aśoka where he is generally dated - the middle of the 3rd century B.C. - an inscription of his for the Greeks is natural to that period in the province of Arachosia. But if we shift him considerably back, is it not unnatural to think of the Greeks discovering one of his inscriptions in just that period? Are you not over-exploiting the play of coincidence?"
Not at all if, following the traditional-Purānic chronology, we have put Chandragupta I instead of Chandragupta Maurya in Alexander's time and dated his acession to 315 B.C. For then we get, some time before c. 250 B.C. his son Samudragupta and a little later Skandagupta of the same line. We have noted that both these kings set up their inscriptions together with Aśoka's. So an awareness of Aśoka's inscriptions was common in their age, and if their age was the 3rd century B.C. and the first half of the 2nd the post-Alexandrine Greek colony at Kandahār would not require to live under Aśoka to know of an Aśokan epigraph. What is more, they would be living in an age in which the practice of accompanying Aśokan epigraphs with new ones was nothing unnatural. No, we are not over-exploiting the play of coincidence. Everything falls into a design which is most fitting.
Are Both the Texts Aśokan?
We come now to our first question: the equal, direct, independent relation to Aśoka, which the scholars have affirmed for both the Greek and the Aramaic texts. Yes, they have affirmed such a relation essentially but they have also admitted certain differences and a scrutiny of these can lead to a contradiction of the former.
The Greek has a Hellenic touch more marked than the Irānian touch of the Aramaic which, says Dupont-Sommer,1 the Aramaic scribe has adapted from the original Indian model "to a lesser degree than the Greek scribe". What is of still greater import is
1. Section III. Ibid., p. 34.
Page 318
presented by Dupont-Sommer1 in the observation: "The Aramaic recension, a little longer, is manifestly closer to the Indian text..." Indeed he is so impressed by this manifest closeness everywhere that even when in one single place the Aramaic is a little shorter he cannot bring himself to infer in the Greek a greater fidelity to the Indian original but postulates a shortness in the latter to match it in the Aramaic and says:2 "the Greek scribe... may have wished... to make explicit the passage he was translating." But the closeness appreciated by Dupont-Sommer goes beyond the issue of length or brevity, likeness or unlikeness to Indian ideas and expressions. It extends to being detailed in conscious conformity with Aśokan edicts found elsewhere. There is the capital instance, which Dupont-Sommer3 notes: while the Greek version states in general the king's abstention from slaughter of animals, the Aramaic specifies the killing of only "a few" animals for the king's sake and is thereby almost exactly in accord with the datum of R.E. I that only three animals instead of many were killed daily for the king's curries at the time of the inscription. Such a direct and intimate relation to the contents of a known Aśokan edict is nowhere disclosed by the Greek text.
Thus it is impossible to cut apart the Aramaic text from Aśoka. But, if the Greek is briefer, more generalised and less Aśokanly suggestive in content, why should we not derive it from the Aramaic instead of from an Indian model sent by Aśoka? Dupont-Sommer and Benveniste hold that the Greek no less than the Aramaic makes adaptations straight of Indian concepts and phrases from a common model. But is it not possible that the adaptations by the Greek scribe according to the spirit of his culture are from the Aramaic adaptations? If we can think of him as Hellenizing Indian concepts and phrases, we can equally think of him as Hellenizing those proper to the Avestan-Aramaic mind and tongue.
Dupont-Sommer4 writes: "'Piety', ... under the pen of the Greek scribe is a transposition of the idea of dhamma, just as 'Truth' under the pen of the Aramaic scribe." But to adapt dhamma to "Piety" is on a par with adapting "Truth" to "Piety"
2.Ibid. p. 28.
3.Ibid., p. 26.
4.Ibid., p. 24.
Page 319
especially as Dupont-Sommer1 informs us about the word for "Truth" in the Aramaic text: "We translate therefore 'Truth', but this word means practically 'religion, faith, piety'."
Benveniste2 draws our attention to the sequences "mother-father" and "father-mother" in the Aramaic and Greek versions respectively. About the former he says that it proves for the Aramaic a direct translation from an Indian model, for Aśoka always writes "mother-father", continuing the old Vedic dvandva mātára-pitárā. Then Benveniste argues that since the Greek text is in disaccord with the Aramaic and gives the traditional Greek order of the words, we have a proof that the two versions are independent of each other and go to a common Indian model.
But is there not a slip in Benveniste's logic? If the Aramaic order of the words establishes a direct derivation from an Indian model, how can the reverse order do the same? Conversely, if the reverse order implies a direct derivation, its difference from the Aramaic text which in this case is exactly like the Indian model should also imply a direct derivation from the Aramaic text. Logically, the true inference from Benveniste's observation about the Aramaic phrase is that since the Greek sequence differs equally from the mutually agreeing Aramaic and Indian it could derive from either.
Indeed, every concept and phrase in the Greek can be unobjectionably related to the Aramaic in one way or another. But the opposite of this, though theoretically valid in several instances, is ruled out on the whole because there is no typically Aśokan particularity in the Greek comparable to the Aramaic's detail about only "a few" animals slaughtered. The absence of such a particularity not only makes us unsure that the Greek goes back straight to the Indian original: it also leads us to ask how and from where the Aramaic could get the detail if the Greek's generality was all that was before the Aramaic scribe.
Even the one place at which the Greek is slightly more particular does not carry any Aśokan stamp for Dupont-Sommer to recognize, and he puts the Greek's length over against the compressed and the implicit which he feels obliged to ascribe to the Indian text from what he finds in the Aramaic. So here too we can take the Greek scribe as rendering explicit nothing else than what
1.Ibid., p. 23.
2.Section IV, ibid., p. 42.
Page 320
the Aramaic holds in brief. On Dupont-Sommer's evidence we can relate the Greek to the Aramaic just as legitimately as to the Indian.
Thus no sign exists anywhere, in the matter of concept and phrase, that the Greek scribe worked from a non-Aramaic model framed by Aśoka and that his post-Alexandrine age must be Aśoka's. Aśoka and his Aramaic text could be much earlier.
The Sole Objection Possible: Piodasses and Prydrš
Only one proper noun - the name of the king - can provoke a doubt. And as it occurs at almost the commencement of the inscription it has made a strong impression, for the whole text, of a first-hand contact by both the Greek and the Aramaic with Aśoka's own message. Benveniste1 comments: "The difference between Piodasses of the Greek and the prydrš of the Aramaic shows already that the two texts do not depend on each other. The difference attests two distinct dialectal traditions. The Greek form Piodasses is imitated from the Middle Indian Piyadassi, with a phonetic peculiarity, the internal -o- ... On the contrary the Aramaic prydrš agrees with those edicts (notably of Shāhbāzgarhī) which have the quasi-Sanskrit form Priyadraśi (Skt.-darśi)."
Certainly two distinct dialects are observable in the two forms, but what is implied in the situation cannot be gauged unless we try to reconstruct how the assumed original model could have led to them.
Dupont-Sommer2 says: "... every Greek colony, from archaic times down to the Byzantine period, had around it the group of those 'hellenised' by the culture, and it is from them most often that the translators and interpreters were recruited." Such being the case, there would be for the translation and interpretation here a "Hellenised" native of Arachosia, the Kandahār-area in antiquity, knowing Greek as well as the Indian language in which the original model came, just as for the Aramaic version there might be an Irānianised Arachosian knowing Aramaic and the Indian language. But Arachosia belongs broadly to the same region as Shāhbāzgarhī and Mānsehrā where quasi-Sanskrit was in vogue, and actually in the neighbourhoods of Shāhbāzgarhī and Mānsehrā
1.Section IV, ibid., pp. 37. 38.
2.Section III,
Page 321
Pūl-i-Dārunteh and Taxila - Aramaic fragments of Aśokan edicts have been found, the one from Taxila using prydrš itself. In this region Priyadraśi would be in the air as the name of Aśoka. How could the name become Piodasses in agreement with another dialect - the language in vogue in Middle India?
Suppose the Greek version was framed not in Arachosia but in Pātaliputra by a Greek-knowing Yukta (Officer) of Aśoka and despatched to its destination. In that case the king's name in it would still be meant for Greeks in Arachosia and assume some form resembling what constantly fell on their ears. The same would be true, were the framer of the version an educated Greek living at Aśoka's court and knowing the court-language. Most probably he would be a member of a family settled in Arachosia since Alexander's time - in other words, for two generations - and there would be no possibility of his employing for the king's name any form except the one he and his fellow-Greeks in Arachosia had heard. Even if by any chance a freakish Piodasses from his hand or from a Greek-knowing Yukta's did issue from Pātaliputra, it would immediately be set right by the Arachosian Greek supervisors to something less unnatural.
It is true that the Greek transcription of Indian proper nouns bears some touches of a Middle Indian medium but never un-mixedly in the age with which we are concerned. These touches are exhibited by at least two out of the three examples quoted by Benveniste1 to illustrate one of the ways in which the internal o could have entered: Taprobanè for Tāmraparni, Sandarophagos for Chandrabhāgā, Erranoboas for HIrānyavāha. Yet it is notable that the r in the first part of each Sanskrit name is preserved. On the analogy of Taprobanè and Sandarophagos and; we may add, Sandrocottos (Chandragupta), we should expect Priodasses or Piriodasses. Something Sanskrit mixes with something Middle Indian.
The initial Pr itself of a Sanskrit name is seen preserved in the well-known Greek Prasioi for Prāchya, the name of the "Easterners", the people whose capital was Pātaliputra (Greek "Pali-bothros"). Its preservation is also noticed where Strabo (VI.22), based on Megasthenes, mentions a class of Indian philosophers, the Pramnai, who are obviously the Prāmanika. And, in both
Prasioi and Pramnai, even the original a associated with the r 1. Section IV, ibid., p. 38.
Page 322
comes through instead of turning into o. Thus Priadasses too could be expected, rendering Piodasses all the more an anomaly.
There is a small yet not insignificant further matter to be reckoned with. Piyadassi is indeed the Indian equivalent of Piodasses, but in no inscription of Aśoka's do we find Piyadassi. At Kālsī and several dialectally related localities, as well as occasionally at Girnār, we have Piyadasi: we never have the form with a double s. Of course, the single s is an orthographic convention of the time , adopted in order to save the trouble of engraving a conjunct which would abound in Prākrit: the actually spoken Middle Indian for the Sanskrit Priyadarśi must have been Piyadassi or Piyadāsi. However, in all copies in Middle Indian of an edict distributed for engraving, the written form would be Piyadasi. Such a copy handed for the Greek text to a translator - be he a Hellenised Arachosian in Arachosia itself or a Greek-knowing Yukta at Pātaliputra or a Prākrit-knowing Greek there - could never have read Piyadassi. And there is no reason why the translator should ignore the universal practice of ,Piyadasi for epigraphs and, in spite of the same form confronting him in his copy, Grecise the name as Piodasses rather than Piodases. Benveniste1 has been very scrupulous about the phonetic value of prydrš: "The Aramaic script naturally does not let us decide between Priyadraśi and Priyadarśi, but as priyadraśi is alone attested (to the exclusion of -darsiv in the Indian epigraphy of Aśoka, it is for this form that we opt." It is surprising how Benveniste, though never doubting the Greek version to be Aśokan, forgot that Piyadasi is the sole attested form in the epigraphs of Aśoka in India and that it is the form relevant to the discussion of Piodasses as an epigraphic fact. Epigraphically , Piodasses should be considered non-Aśokan.
Hence, even on the assumption that a Middle Indian form might have got into a text intended for Arachosia, Piodasses would be anomalous. And the form most natural for a version got ready for Arachosian Greeks from an Indian original, composed either in quasi-Sanskrit or in Middle Indian for the purposes of engraving, would be - according to all the circumstances we have surveyed -Priadases, if not Priadrases by Priyadraśi's direct influence.
Evidently a man altogether from Middle India and totally ignorant or negligent of the linguistic practice in a place like
1. Section IV, ibid., p. 38.
Page 323
Arachosia and of the linguistic trend of the Greeks in regard to Indian proper nouns was responsible for Piodasses. Such a man can hardly fit into a picture of Arachosian Greeks living under Aśoka in the 3rd century E.C., side by side with those for whom the Aramaic text was intended. To explain the Greek text's Piodasses as nothing more than the result of a Middle Indian dialectal tradition is too simple, too abstract. This tradition can be thought of as influencing the Arachosian Greeks only if we separate them from Aśoka and from Aśoka's Aramaic-speaking subjects in Arachosia, and imagine an unusual state of affairs in which a natural equivalent of Priyadraśi would be pushed out.
We can suggest what that state of affairs could be Suppose the Aramaic text is already there from the past and the post-Alexandrine Greeks in Arachosia discover it. An interpreter, knowing both Aramaic and Greek, translates it for them. The Greeks vis-à-vis the alien and unheard-of name Priyadraśi as given by the interpreter, will inquire into the identity of the king bearing it. An inhabitant of Arachosia at a period sufficiently removed from Aśoka's will scarcely be able to throw light on the subject. Several Indians will be called in. The interrogation will go on until the arrival of one who hails from Magadha, the home of Middle Indian. He will be most likely to know about the king who called himself Lājā Māgadhe in the Bhābru Edict. Speaking Middle Indian, this man will, in the course of his informative talk, keep repeating to the Greeks a form which they may represent by Piodasses. Naturally they will defer to his greater acquaintance with Indian things and select his version rather than the other.
On such lines of thought alone can Piodasses be explained, and they leave no ground for tracing the Greek text to an original Indian model despatched by Aśoka. Thus the sole possible linguistic objection fails.
The Aramaic Text's Mazdeanism and the Brāhmanas-Śramanas
We may now positively affirm: "The Aramaic text omits the Brāhmanas and Śramanas because it is meant for the Yonas. The Greek does so because it simply makes an adaptive translation of the Aramaic."
We may further elucidate the raison d'etre of the original omission. The psychology of the Aramaic text with its Avestan
Page 324
words is definitely Mazdean and directed towards followers of Mazdeanism. As such, it fully accounts for the omission from it of the Brāhmanas-Śramanas - in contrast to the psychology of Greek culture which never interfered with the religious customs of the non-Greeks and would set no bar to the Brāhmanas-Śramanas being amongst them. Historically, the Mazdean religion was to a great extent a forcible break-away from the Vedic cult. Although retaining the worship of fire, it stressed more than ever the pristine Irānian reversal of the meaning of the Vedic word "deva" and, with a larger insistence, made it signify "demon" instead of "god". We may attend to the remarks of Raychaudhuri1 on one of the Achaemenid emperors, Xerxes (485-465 B.C.):
"Xerxes refers to the expression of rebellion in lands 'where, before, the Daivas were worshipped; then, by Ahurmazda's will, of such temples of the Daivas, I (the King) sapped the foundations'. The Daiva-worshipping lands may have included the Indian satrapies."
All that the Brāhmanas and Śramanas stood for would be repelled by an over-orthodox section of the ancient Mazdeans.
The Aramaic Text and the "Yavana Script"
When the Yonas are demonstrated to be the recipients of the Aramaic text, a flood of light is immediately shed on Pānini's word: "Yāvānānī." As we have seen, this word of the oldest Indian grammarian has been translated as "the Yavana script" and interpreted as "the Greek script". Now we may well credit the suggestion, first made on the discovery of the Aramaic stone-inscription at Taxila but unsupported by any proof so far, that "Yāvānānī" pointed to the Aramaic script. And, if we attend to the verbal contents of Pānini's Ashtādhyāyi, we shall see that not a single word appears to have been borrowed from the Greek language whereas some contact with the Semitic tongues is undeniable. Let us quote Agrawala:2
"Jābāla denoted a goatherd, and Mahājābāla (VI.2.38) one who was the owner of a big sheep-run. Jābāla does not seem to be a word of Sanskrit origin. It may be traced to a Hebrew word yobel or jobil, signifying ram's horn, whence 'Jubilee'... Hailihila and
1.An Advanced History of India, p. 64.
2.India as Known to Pānini. pp. 220-221, 124.
Page 325
mahā-hailihila are words of unknown meaning and origin, mentioned by Pānini as special names of some article (VI.3.38). The word is not explained in any Sanskrit dictionary, nor is there any instance of its being used in literature. It appears that hailihila was a Semitic word appearing in a Sanscritised form, as the name of a poison which was imported from the West. In Arabic halāhila means deadly poison (cf. Hebrew halūl, deadly poison). Steingass derives without reason the Arabic word from Skt. halahala (F. Steingass, Persian-English Dictionary, p. 1506). The Skt. word itself is exotic as shown by its variant spelling, e.g. hālāhala, halāhala, hālahala, hālahāla, hāhala, hāhāla (Monier-Williams, A Sanskrit-English Dictionary, 1899, revised edition, p. 1293). Pānini's hailihila seems to come nearest to the original Semitic form of the word, which may have been Aramaic, the international language of trade and commerce in the Achaemenian world from Syria to Gandhāra. Pānini refers to poisons in general called visha and to the third degree methods of liquidating particular persons marked out as vishya by the administering of poisons."
Agrawala points to the Achaemenid world; but it is unnecessary in order to account for Pānini's Semitic contacts across the frontier. They were a natural continuation of Indo-Semitic exchanges from very ancient times. We may discount as legend the statement of Arrian (Indica, I) that to the west of the river Indus, up to the river Cophes - the river Kābul by which Pānini's own native Salatura stood - the two Indian tribes, the Astakenoi (the people of Pānini's Hāstināyana mentioned in sutra VI.4.174) and the Assakenoi (the Asvakayana of Nadādi gana IV. 1.99) "were in old times subject to the Assyrians" before the period of Median and Persian rule over them. But it is certain that in the times preceding those of the Persian (Achaemenid) Cyrus and the still earlier Medes we get two bits of information proving Assyria and India to have been in touch. Jairazbhoy1 writes: "The cotton tree was introduced from India into Assyria (c. 700 B.C.) by Sennacherib (704-681 B.C.), who is reported to have said that trees that bear fleeces were sheared and shredded for garments... The earliest date [for the Indian peacock's arrival in Assyria] may be 738 B.C., when there is a possible reference to a peacock among the wonderful birds received as tribute by Tiglath Pileser III." Majumdar,2
1.Foreign Influence in Ancient India, p. 30.
2."India and the Western World", The Age of Imperial Unity, p. 612.
Page 326
gathering "archaeological evidence" of India's contacts with Western Asia, refers "to the figures of apes, Indian elephants and Bactrian camels on the obelisk of Shalmanesar III (860 B.C.)."
Actually, right back to the age of the Indus Civilisation (broadly dated 2500-1500 B.C.) go India's contacts with the Semitic world. A sign in Pānini himself of ancient Indo-Semitic exchanges is the word gonī mentioned as denoting a container or sack (āvapana) made from gonī IV. 1.42), obviously a cloth. "It is unknown in Vedic literature," comments Agrawala,1 "but occurs in the Brahmajāla Sūtta XV as gonaka, explained as a woollen cloth made from the hair of long-haired goats. It was probably the same as kaunakes, a one-piece loin-cloth worn by the early Sumerians and Accadians, and made of suspended loops of wool hanging from a woollen skirt (Marshall, Indus Valley Civilization, 1.3.3, 342; pi. 95, fig..10). The word seems to have travelled to India through commerce in pre-Pāninian times."
Perhaps it will be said: "To date the Aramaic script so far back in time - and, with it, the Kharosthī and Brāhmī used by Aśoka elsewhere in his empire - is rather unrealistic. Further, can we take the Aramaic language and not just a few Semitic words as current in India's borderland so long ago?"
We find Aramaic fairly on the move when in the year 731 B.C., in the reign of the Assyrian Tiglath Pileser III, there comes, as Philip K. Hitti2 tells us, the representation of a scribe recording in Aramaic the plunder from a captured city. But previous to that time the Aramean merchants had already spread their language far and wide. "The Arameans traded in purple from Phoenicia, embroideries, linen, jasper, copper, ebony and ivory from Africa, and in the 'products of the seas', perhaps pearls for which throughout the ages the Persian Gulf has been famous."3 And if we believe, as Hitti4 does, that the first mention of the Arameans by Tiglath Pileser I (c. 1100 B.C.), in close association with the word "Akhlamu", merely throws into prominence a people already subsumed under the latter designation in the centuries before, then we may recognise their presence in remote antiquity whenever the Akhlamu - otherwise known as the Ahlamû - are epigraphically
1.India as Known to Pānini, p. 146.
2.A History of Syria (London, 1957), p. 168.
4.Ibid., p. 162.
Page 327
attested. Here Georges Roux has a lot of information to provide.
"Tukulti-Ninurta I (1244-1208 B.C.) claims that he conquered... 'the mountains of the Ahlamû'."1 "Shalmanesar I (1274-1245 B.C.)... attacked Shatuara, 'King of Hanigalbat' and his Hittite and Ahlamû mercenaries and defeated them."2 "The Ahlamû are first mentioned in a mutilated letter from el-Amarna [c. 1400-1350 B.C.3] alluding to the King of Babylon; during the same period their presence is attested in Assyria, at Nippur and even at Dilmun (Bahrain)..."4 Bahrain means down the Persian Gulf. Overseas as well as overland the Arameans, in the interests of trade, are likely to have travelled with their language and their script in the direction of Arachosia at an early date.
What is their ultimate antiquity? We may attend to the words "Aramû" and "Arimi", which occur without association with "Ahlamû" in documents subsequent to the time of Tiglath Pileser I.5 These words have their counterparts far earlier: "In texts of the Akkadian, Ur III and Old Babylonian periods occasional mention is made of a city Arami and of individuals by the name of Aramû..."6 This may be no more than a phonetic resemblance, but the resemblance is so close that we are not debarred from tracing the Arameans even to that period: c. 2371-1595 B.C.
We may add that the association with "Ahlamû" is not quite broken in the times when "Aramû" and "Arimi" stand alone. Ashurnasirpal II (884-859 B.C.) has still the expression "Ahlamû-Aramean."7 There seems to be an organic relation between the terms, a doubling as if they were fully designative when both stood together, though either would suffice on its own to designate the same entity.
We may justifiably conclude that the Arameans and their script were known over a wide stretch of the Orient in at least the middle of the 2nd millennium B.C. Here some observations by John Gray8 will be in place:
1.Georges Roux, Ancient Iraq, (Pelican, Harmondsworth, 1966), p. 248.
2.Ibid., 236.
3.Ibid., Chronological Table 5, col. 6.
4.Ibid., p. 248.
5.Ibid.
6.Ibid., p. 247.
7.Ibid., pp. 259. 261.
8.Archaeology and the Old Testament World (Harper Torchbooks, New York
Page 328
"It was the contacts of Palestine and Syria with Mesopotamia and Egypt that awakened the inhabitants to the advantages of a convenient way of writing. The respective systems of cuneiform and hieroglyphics, though great achievements, were yet not the most expeditious way of communicating men's thoughts. Both were syllabic, not alphabetic, and the cuneiform might even be ideographic, a single sign representing a whole word. Thus both were highly complicated and the medium of specialists. In the Ras Shamra Tablets we see a great advance in the simplification of the cuneiform to an alphabetic script of thirty ciphers which, as may be seen from the private correspondence of one of the princes of Ugarit with the queen-mother, could be used for short, informal, personal correspondence. By the middle of the second millenium a similar experiment was made on the basis of Egyptian hieroglyphics, and the resulting alphabet was widely used, evidence being found at Byblos, Gezer, Shechem and Lachish, and from the Sinai Peninsula, where the Egyptians mined turquoise and copper. From the Sinai inscriptions, fragmentary and uncertain as they are, a direct line of development may be discerned to legible alphabetic inscriptions in Syria and Palestine in the second half of the second millennium B.C. This group begirfs with the inscription of the king Shaphat-ba'al of Byblos, which is dated by Dunand in the 15th century B.C.1 though he does not exclude the possibility of a date two centuries earlier."
So the Aramean and allied scripts can go back even to the 17th century B. C. This dating seems a fair middle course between the usual belief that the Aramaic language "appeared at the end of the second millennium B. C. as the tongue of the Aramaeans who had settled in Damascus" and the opinion of "some philologists [who] affirm that it was already in use in the third millennium B.C.".2 In regard to the possible touch of Aramean and allied scripts with India, we have to take into account a fact long ago spotlighted by G. Buhler apropos of the Brāhmī script which Aśoka used along with Kharosthl. Today the former is sought to be linked with the as yet undeciphered script of the Indus Valley Civilization. But Biihler's fact has still a genuine force. He pointed out that a certain
and Evanston, 1965). p. 24.
1.M. Dunand. Byblia Grammata (1936), pp. 146-51.
2. Encyclopaedia Americana (New York, 1966), Vol. II,
Page 329
proportion of the oldest Brāhmī letters are practically identical with letters on some Assyrian weights and on the Moabite Stone of King Mesa1 which bears the longest and one of the earliest Hebrew inscriptions.2 The striking resemblance must have come about by the contact of India with the Semites (Hebrews, Arameans, Phoenicians) before and not after the time of the Moabite Stone; for the later Semitic forms show insufficient agreement.3 The Moabite Stone was erected, as Hitti4 notes, shortly before 842 B.C. We must, therefore, pass beyond the 9th century B.C. for the contact of the Arameans and their likes with Arachosia and its neighbourhoods. And the contact must have been long in order to mould so markedly the Brāhmī script.
The Indus Valley script is itself comparable in part to Semitic signs. To trace links between it and Brāhmī is not to give the lie to Buhler altogether. We merely show the Semitic influence to have been more complex than Buhler suspected.
Kharosthī, of course, is clearly derived from Aramaic. What Buhler drew our attention to in connection with Brāhmī serves to set Kharosthī's derivation against the proper chronological background: the pre-9th-century B.C. period.
There seems no deubt, in view of a variety of considerations, that the Pāninian "Yavana script" is the one we meet with in the Aramaic text at Kandahār, which our analysis of the bilingual inscription has determined to have been for the Yonas.
The True Historical Perspective for the Aramaic Text
Now what remains to be done in relation to this inscription is to formulate the true historical perspective for that text which alone we can attribute to Aśoka and then to arrive at the most likely date for him and for that text.
Benveniste5 has expressed his view on the first question thus: "To put this Aramaic text in the historical perspective which accords with it, we should compare it to the documents of Achaemenid date found in Egypt. Two centuries after the Great
1.Rhys Davids, Buddhist India (Calcutta, 1950), pp. 70-1.
2.Hitti, op. cit.. p. 196.
3.Rhys Davids, op. cit., p. .71.
4.Op. cit., p. 195.
5.Section IV, op. cit., pp. 43-44.
Page 330
Kings, at the other extremity of the Irānian world, we rediscover the same type of wording, the same titling, the same language equally stuffed with Irānian words. The surprise is just that an inscription set up in a region bordering on India and subject to the power of an Indian sovereign should be in the Aramaic language and find its analogues only in the writings of the ancient Persian satrapies. The use of Aramaic shows that we are in reality in an Irānian province, where were maintained the traditions of the Achaemenid chancelleries. One could conjecture this after the discovery of the Taxila inscription. It is now a certitude."
If our thesis in correct, the Aramaic version still keeps a relationship with the writings of the ancient Persian satrapies, but the relationship gets altogether reversed. It is not the documents of the Achaemenid times discovered in Egypt that provide the precedents of this version: this version itself manifests a style and texture that are repeated in those documents. The use of Aramaic may show that we are in a province, if not Irānian, at least Irānianized, but not that the traditions of the Achaemenid chancelleries are maintained here. These traditions may have formed themselves on an antecedent practice such as we find in our inscription. We know that Xerxes and Darius Hystaspes and perhaps even Cyrus were masters of Arachosia and other regions west of the Indus: even some parts about the Indus were under Persian control and this control continued up to 330 B. C. when Alexander defeated Darius III. We know too that the traditions of the Achaemenid chancelleries in the form that Benveniste has in mind do not go beyond Xerxes (486-465).1 They could easily have started after the Persian conquerors had seen Aśokan inscriptions in the territories ruled by them. Or else, if they did not see these inscriptions, they must have got acquainted with the official practice in those territories, inherited from earlier Aśokan times.
That the historical perspective of Benveniste needs to be reversed is deducible from the very observations by him on the linguistic pecularities of the Aramaic recension.2 In the Middle-Irānian of the post-Alexandrine epoch where this perspective would place that recension, terms like patyasti have not survived and even the notation of pati with the mute and the final vowel and
1. The Encyclopaedia Britannica (13th Edition), Vol. XIX, p. 619; also A. T. Olmstead, History of the Persian Empire (Chicago, 1948), pp. 244-45.
2. Section IV, op. cit., p. 36.
Page 331
the notation of the final vowel -i in prbsty do not exist, nor does the Irānian of the Achaemenid period itself have a form like mazišta. Benveniste has further underlined the obscure status of the recension's language when he speaks of "the rather unclear ptytw" which is definitely Irānian but cannot be elucidated in the light of post-Alexandrine or even Achaemenid Irānian. And the difficulty seems still greater with the possible alternative reading: ptytz. Benveniste1 confesses; "ptytz (or ptytw) sets us at the very start face to face with a difficulty which will be in fact the most serious of those we"shall encounter in our study." There is also his statement:2 "The word 'dwsy' (Vivify'), supplied with an Aramaic plural, is not Aramaic, but one hardly ever sees an Irānian word, like 'dwš or 'rwš that answers to the sense which the context would suggest: 'disagreement, misfortune' or 'bad action, impiety' or a similar action... One does not find in Irānian the matter for a useful hypothesis." All these puzzles, and the purely Avestan nature of some of them, plus the archaic nature emphasised by Dupont-Sommer of many Aramaic expressions, carry us clean beyond the Achaemenid chancelleries and their Irāno-Aramaic inscriptions in Egypt.
Nor is it true that only in "the documents of Achaemenid date found in Egypt" do we have "the same type of wording, the same titling" as in the Kandahār inscription. Benveniste's "Egypt" is correct, but his "Achaemenid date" is not at all binding for the official verbal turns by which he sets store. The most typical expressions of this type in our text are: "our lord Priyadraśi the King" and "our lord the King". The basic formula here is actually found in Egypt 900 years before the Achaemenid Xerxes. Leonard Cottrell3 mentions the urgent appeals which in c. 1400 B.C. poured forth for troops into Akhnaten's Foreign Office from the Governors of some threatened provinces. The Governor of the coastal city Tunip wrote: "...And when Aziru enters Simyra [another coastal city] he will do as he pleases in the territory of our Lord the King... For twenty years we have been sending to our Lord the King, the King of Egypt, but there has come not a word, no, not one." This type of wording and titling continued in later yet still
3.The Anvil of Civilization (Mentor, New York. 1956), p. 144.
Page 332
pre-Achaemenid epochs, and even outside Egypt. Thus Roux1 quotes from one Zakir's correspondence with Ashurbanipal (668-631 B.C.): "The disturbance is the fault of the King of Amurru and his land for allowing the money of the king, my lord, to be in the land of Amurru. Let the king, my lord, do as he wishes. The hand of the king, my lord, shall capture him..."
If India's vicinity before the epoch of Xerxes was in touch with the Near and Middle East, as it certainly was, Aśoka's Aramaic text - provided it does not go back too much into antiquity - could certainly reflect the formula employed in Akhnaten's Egypt and elsewhere. The formula seems to have been the standard one all over the Orient and the ancient Persian satrapies merely copied it. These satrapies themselves are not required in the least to account for it.
Like other countries of the pre-Achaemenid Orient, an earlier Persia than of these satrapies comes into the picture conjured up by the Avestan and archaic Aramaic language in our epigraph - a Persia with Semitic connections - and, across this Persia, we see Semitic connections with India.
We have already brought together Persia and Assyria by referring to the tablet of Ashurbanipal which mentions the Zarathus-trian pantheon. There are other facts as well. Sargon (722-705 B.C.) records in a clay prism the names of two conquered princes: Mazdaku and Mastaku - names derived from the Zarathustrian God Mazda. R. Ghirshman2 cites the annals of Shalmaneser III to show that the Assyrians knew of the Medes (Madai) in 836 B.C. and of the Persians (Parsua) in 844 B.C. The Parsua are said to have conquered Elam and Anshan. These territories came to be called "Parsa" and from them the Achaemenids hailed.
As for India and Persia, there are passages in the Avesta seeming to indicate a political hold by old Irān on Northern India before the Achaemenids. As Mookerji3 reports, although corroboration or details of this hold have hitherto been lacking, some scholars have taken these passages as genuine evidence. Particularly apt is the Vendidad's reference (1.73) to Hapta-Hindu (Sapta-Sindhu) - the land of the seven-rivered Indus-system - as one of the countries Ahura Mazda created for his own people.
1-.Op. cit., p. 312.
2.Irān (Pelican, Harmondsworth, 1954), p. 90.
3."Foreign Invasions", The Age of Imperial Unity, p. 39.
Page 333
Juxtaposed, the Avestan passages and our inscription may be seen to throw light on each other.
What we have said about the introduction of Indian cotton and the Indian peacock into Assyria, and shown from the obelisk of Shalmeneser III - all this belongs to the same pre-Achaemenid period. Its testimony to Indo-Assyrian contacts completes our picture.
We have now the period c. 559-1400 B.C. - the time from before Cyrus's accession up to the al-Amarna Letters. And it is meaningful that our epigraph's style of titling and the reference to the Ahlamû, the Arameans, should be found in the same group of documents that are as early as c. 1400 B.C. Obviously, it is not for the first time that such titling was used or that the Ahlamû were on the march. If the latter were already as far south-east as Bahrain in the same age, they must have been spreading out fairly before. We should have nearly eight hundred and fifty years prior to 559 B.C., within which our Aramaic text may have taken shape. In about the middle of these thousand years we have the two earliest dated inscriptions in Aramaic. A votive stele, unearthed four and a half miles north of Aleppo and bearing the name of Ben-Hadad I, is of c. 850 B.C.1 Earlier yet is a short epigraph from Tell-al-Halaf (Gozen) in North Syria: it dates back to the beginning of the 9th century B.C.2 Still more early epigraphic evidence may be indirectly cited. Hitti3 brings the information: "The annals of Ramses III (1198-1167 B.C.) give the Aramaic spelling of the name [for Damascus] 'Tiramaski' corresponding to Aramaic Dar(fortress)-Mesheq." So the inscription at Kandahār which has characteristics tending to place it a little earlier than the votive stele and the Tell-al-Halaf message need not be denied its proper antiquity. The Egyptianized echo of the Aramaic for "Damascus" allows, by its promise of epigraphic possibility, the Aśokan edict to find its true historical perspective.
The Palaeographical Question
Perhaps we shall be challenged: "Can our inscription be regarded palaeographically as more primitive than the Egyptian
1.Hitti, op. cit.. p. 170.
3.Ibid., p. 165.
Page 334
documents mentioned by Benveniste?"
One should not be discomfited if the palaeographical pronouncement were unfavourable to one's thesis. We may pertinently repeat here what we have remarked in a much earlier context -the Gokāk plates of Dejja Mahārāja. Palaeography - the comparative study of old writings - is not an exact science. We may recall Sylvain Levi's opinion1 that palaeographical tests have little independent value (autorite absolue) although they may be used as a check (controle) upon or a guide (indice) to the interpretation of positive history. In a narrower context Jouveau Dubreuil2 has also put palaeography in its true place: he has remarked that it is a bad auxiliary to the chronology of Indian history as very often the documents dated in the same reign differ much from each other and thus its evidence would be mainly suggestive or corroborative but at any rate by itself not conclusive.
Altekar provides us with a significant "slant" on questions palaeographical. Discussing the dynasty known as the Maghas, he3 mentions that "there is a great controversy about the dates of the rulers of this dynasty". Some scholars think the Maghas date themselves in the Chedi Era starting in 248 A.D.; others go in for the Gupta Era of 320 A.D.; still others opt for the Śaka Era of 78 A.D.4 Thus the datings would vary by 72, 178 or 242 years. Altekar5 favours the Śaka era for reasons which to his mind outweigh the most impressive-looking palaeographical considerations. What is especially to be marked is that such considerations can be countered within the realm of palaeography itself, for there is no final standard in it. Altekar writes: "The most cogent argument in favour of the Chedi or the Gupta era is palaeographical; there is no doubt that the characters of the Magha inscriptions are almost the Gupta characters. This argument however is not a convincing one; for many of the Gupta forms of characters are to be seen in an inscription of Kanishka, dated in the 14th [Śaka] year (92 A.D.); see Epigraphia Indica XXI, 2."
In the case of the Kandahār bilingual, all that we have to go by is
1 Quoted in English by Vincent Smith in The Indian Antiquary, Vol. 31, 1902, p. 196.
2.Quoted in English by Govind Pal in The Journal of Indian History, August 1935. p. 197.
3."New Indian States in Rajputana and Madhyadesa", A.New History..., p. 41.
4.Ibid., fn. 2.
Page 335
the analogy between our Aramaic version and the documents to which Benveniste refers. The analogy per se does not determine what came earlier and what later. Among the factors that are determinative, palaeography has as little force as in the case of the Magha inscriptions.
The Real Date of Aśoka and of the Kandahār Inscription
Our new historical perspective - pointing possibly beyond even c. 900 of the Tel-al-Halaf inscription - appears to launch us on the way to the traditional-Purānic chronology for Aśoka. Yet we cannot go ahead light-heartedly. For, with Chandragupta Maurya's accession fixed Purānically in 1538 B.C., his grandson Aśoka's has to be after the lapse of his grandfather's reign of 24 years and his father Bindusāra's of 25 by the Purānas. We reach (1538-49=) 1489 B.C. Purānically, Aśoka reigns for 36 years. So his regnal period will be from 1489 to 1453 B.C. The Ceylonese Chronicle Mahāvamsa allots 34 years to the first Maury a, 27 or 28 to the second and 37 to the third. Then Aśoka will reign from 1477 (or 1476) to 1440 (or 1439). But to locate in the 15th century B.C. the Mauryan and his huge empire stretching beyond the Indus to places like Kandahār would fly in the face of all ascertained fact of history and archaeology, including most prominently the end of the Harappā Culture in c. 1500 B.C., the far-flung Indus Valley Civilization which had its reach eastward up to Alamgirpur near Delhi. Neither does Mahāpadma of the Nanda dynasty about 150 years earlier, with his conquests - by the Purānic testimony - of the Aikshvākus, Panchālas, Kāśīs, Haihayas, Kalingas, Aśmakas, Kurus, Maithilas, Sūraśenas and Vitihotras seem the right contemporary of the Indus Valley Civilization in its later days of the 17th century B.C. We have to situate Aśoka as well as his immediate predecessors where they would fittingly stand.
To set him at the right distance from the Imperial Guptas whom we have started in 315 B.C., we have first to consider the proper place of the Āndhra Sātavāhanas between him and them. The Purānas reckon these dynasts as Magadhan monarchs and we have Purānically accorded them 412 years in all. After them we have traced, on the basis of both Megasthenes and the Yuga-Purāna section of the Gārgi-Samhitā, an interval of 75 years in Magadha, during which period - called a "republic" by the Greek annalist -
Page 336
the republican Lichchhavi clan into which Chandragupta I married had held sway over Pātaliputra, as shown by an inscription pubished by Pandit Bhagwanlal Indrajit in The Indian Antiquary (XI, p. 7). Down to 490 B.C. from 802 the Āndhras were, in the eyes of the Purānas, masters of Magadha, the province whose capital was this city. But can we accept such a picture, historically and archaeologically?
Modern historians, by and large, refuse to count the Sātavāhanas as Magadhan monarchs. Indeed, in the present state of research there is no evidence in the form of inscriptions or coins for their kingship in Magadha. However, we have hardly anything else to put here in place of their supposed rule. Even in the current time-scheme we are obliged to confess:1 "The history of Magadha, from the end of the Kānva rule to the rise of the Guptas, three hundred years later, is very obscure." It may be more satisfactory to combine something of the general modern view of the Sātavāhanas with the Purānic and say that, while their seat of empire was the western and central Deccan - Āndhradeśa - they exercised to a fluctuating degree an overlordship of Magadha. At least the Yuga-Purāna, by which modern historians set some store,2 states that a "Sātarājā" not only defended the territory of Kalinga against foreigners but also pursued them to Pātaliputra, their recently established headquarters, drove them out and ruled there for 10 years.3 "Sāta," M. Rama Rao4 assures us, is a well-known contraction of "Sātakarni", the name of several members of the Āndhra Sātavāhana dynasty.5 Sircar6 informs us of "Sāta" contracting "Sātavāhana" no less than "Sātakarni", and refers to "certain coins of the so-called 'Malwa fabric' with the legend indicating 'of the illustrious king Sāta'." So some connection of the Āndhras with Magadha may be conjectured. Rama Rao7 seems right in inferring: "The Empire of the Sātavāhanas... included at
1 . The Age of Imperial Unity, p. 100.
2.Ibid., p. 106.
3.Studies in the Early History of Āndhradesa by M. Rama Rao (University of Madras, 1971), p. 24. Cf. also Journal of the Bengal and Orissa Research Society, XVI, pp. 18-60 referred to by Rama Rao.
4.Ibid.
5.Ibid., pp. 26-27. Also "Genealogy", The Age of Imperial Unity, pp. 706-07.
6."The Sātavāhanas and the Chedis"; The Age of Imperial Unity, p. 198, with fn. 4.
7.Op. cit., p. 5.
Page 337
times Magadha..." - and V. Smith1 justified in granting the possibility that "the Āndhras may at some time or other have controlled Magadha".
Besides, if we believe, as we must, that Megasthenes acquired his chronological knowledge from Purānic pundits in the age of the founder of the Imperial Guptas, it is difficult to think of those pundits fabricating baseless tales about a dynasty which preceded the Guptas by a mere 75 years. They may have represented overlordship as direct sovereignty because of adventures like that of Sātarāja, but we cannot conceive them as inventing a connection with Magadha where there was absolutely none. Surely they could not take such liberties with events still fresh in men's memories? We must also remember that the Āndhras were still a powerful people in the period of Sandrocottus. From what Pliny (VI, 22),2 founding himself on Megasthenes, reports, the "Andarae" were actually next in power to that king of Palibothra (Pātaliputra). A sheer fantasy about them in relation to Magadha is scarcely credible. And the power they still retained would certainly not entitle the chronologists to speak of the end of their dynasty -except in the very special sense that for the last 75 years before the accession of Sandrocottus to the throne of Pātaliputra they had ceased to have any connection, however indirect, with that throne.
Rapson3 has well conjured up the circumstances in general which led the Purānas to count the Āndhra dynasty as rulers of Magadha. He does not evince any awareness of "Sātarāja" and hence states that "the Āndhras had probably no connexion with Magadha", but explicates what he calls "their only possible claim to a place in its records". He attends to "a conquest which transferred to them the suzerainty previously held by Magadha". His insight runs:
"In order to understand the situation we must consider what the consequences of a triumph of this kind must have been. Under the Nandas and the Mauryas Magadha had established a suzerainty which passed by conquest to the first Çunga king, Pushyamitra, and was solemnly proclaimed by his performance of the 'horse-sacrifice' (açvamedha). This suzerainty, and with it the proud title of chakravartin, 'universal monarch,' was contested successfully by
1.The Early History of India, 2nd Edition (Oxford, 1924), p. 119.
2.The Classical Accounts..., p. 342.
3."The Purānas", The Cambridge History of India, I, p. 318.
Page 338
the Āndhra king who, as is known from the Nānāghāt inscription of his queen, Nāganikā, celebrated the Açvamedha on two occasions; and... there is good reason for believing that the genealogies preserved in the Purānas have their origin in the proclamation of the king's lineage which accompanied the performance of the sacrifice.
"The rank of a chakravartin must, at this period, have conferred on his family an hereditary distinction which entitled all his successors to be commemorated in the records of Magadha. Imperial and royal dignities of this kind, when once established, are not easily abandoned, however shadowy and unreal they may have become. It must be remembered that the sovereigns of our own country continued to use the title and the arms of France until the beginning of the nineteenth century, nearly two centuries and a half after the loss of Calais, the last of their French possessions. Regarded as historical documents, the British coin-legends of the eighteenth century, with their purely hereditary titles, are as misleading as the Purānas, which, arranging all in one long series, ascribe to Magadha both its own kings and the families of the suzerains of Northern India."
Especially with knowledge of the inroads into Magadha made by personages like Sātarāja and of the obscure fluctuating indefinable conditions in post-Kānva Magadha except for the 75 years of "republican" Lichchhavi rule just before the rise of the Guptas, the background of broad suzerainty Rapson sketches would sustain the Purānas' claim to count the long duration of the Āndhras as a part of Magadhan kingship in the lengthy time-scheme the pundits drew up from the birth of Parīkshit at the end of the Bhārata War to the emergence of the Guptas.
However, modern historians are at variance with our Purānic stand on the length of Āndhra monarchy no less than on the Āndhras' connection with Magadha. They grant them no more than 257 years: the dynasty starts in c. 30 B.C., with Simuka and ends in c. 227 A.D. with Puloma.1 We have seen no reason to cut down their period so drastically. The Purānas give the impression of being the most historically minded where the Āndhras and the kings of more or less the same time are concerned, which is not unnatural since the pundits were nearest to the Āndhras and to these kings when the dynastic lists were being finally prepared.
1. The Age of Imperial Unity, "Chronology", pp. 701. 702.
Page 339
Raychaudhuri,1 who has freely criticized the Purānas on occasion, has yet this to aver: "The epigraphic or numismatic records of the Sātavāhanas, Ābhiras, Vākatakas, Nāgas, Guptas and many other dynasties fully bear out the observation of Dr. Smith that 'modern European writers have been inclined to disparage unduly the authority of the Purānic lists, but closer study finds in them much genuine and valuable historical information.' " Dr. Smith2 was himself inclined to accept "four and a half centuries in round numbers" for the duration of the Sātavāhanas. Nor has he lacked followers in refusing to reduce their reign to a mere 257 years. "There is a wide difference of opinion regarding the duration of the dynasty," reports Rama Rao3 in 1971. In 1974 Mahājan4 refers to "a lot of confusion regarding the chronology of the Sātavāhanas." Rama Rao5 opts for 456 years, while Mahājan6 writes: "we accept the view of Dr. Smith and Dr. Gopalachari that the Sātavāhana dynasty lasted for 460 years."
When we pass beyond the Āndhras the uncertainties increase. Several historians7 take the Śungas, the Kānvas and Āndhras to be successive post-Maurya lines as presented in the Purānas in their main statements. But these very historians8 who make much of these statements do not hesitate in connection with the earlier Pradyotas and Śiśunaga dynasties to declare that "the Purānas give a distorted account of the political vicissitudes that took place in Magadha after the fall of the Bārhadratha dynasty... and we propose to treat the history of Magadha on the basis of the Buddhist texts, notably the Sinhalese Chronicle Mahāvamsa, rather than the Purānas." In fact, all historians are in a mist here and are free to find whatever way strikes them as the most likely. The authorities we have just quoted have themselves admitted: "we know very little of the history of the Śunga dynasty"9 - and confessed about the Kānvas: "We really know nothing of their
1.The Political History of Ancient India, pp. 6-7.
2.The Early History of India, 2nd Edition, p. 119.
3.Op. cit., p. 5.
4.Ancient India, p. 307.
5.Op. cit., pp. 26-7.
6.Op. cit., p. 308.
7.The Age of Imperial Unity, pp. 99-100.
8.ibid., p. 18.
9.Ibid., p. 98.
Page 340
history."1 Apart from the need to counterbalance with a long preceding period the short duration preferred for the Sātavāhanas, there is no particular cogency in counting the Śungas, Kānvas and Āndhras in succession. Smith, Gopalachari and Mahājan see the Sātavāhanas as running side by side not only with the Śungas and Kānvas but also with the last few Mauryas. Rapson,2 while holding that the Śungas did come after the Mauryas, chooses yet to look on them as contemporaries of both the Kānvas and Āndhra Sātavāhanas.
"The Purānas," he3 writes, "have been edited, and, in the process, much of their value as records has been destroyed. Certain incidental statements, however, have escaped the editor; and these seem to show that the Kānvas and the Çungas were contemporary. The Kānvas, who are expressly called 'ministers of the Çungas,' are in some versions, said to have become kings 'among the Guhgas;'4 and... the Āndhras are credited with sweeping away not only the Kānvas, but 'also what was left of the Çungas' power'5"
Evidently, if the editing was done, the process was already complete by the time of Megasthenes; for, only by taking the three dynasties as successive could the total number of kings which he transmits to us accumulate. The succession must have been necessary in order to support the idea of great antiquity for the initiation of the line leading to the Magadhan kings and to fill the large space between that antiquity and the age of Sandrocottus. Once a very remote period had been posited as the starting-point of Indian history, diverse tactics had to be adopted to get down from the Purānic point of departure through the enormous stretch of years to the Purānic point of arrival, namely, the rise of the Imperial Guptas in 315 B. C.
Beyond the later Āndhras, who must have had something to do with Magadha, the Purānas are in a semi-legendary domain of chronology, and there is no reason why we should always stick to their pronouncements instead of profiting by one or another school of modern scholarship. We may accept the 412 years the
1.Ibid. , p. 99.
2."The Purānas", The Cambridge History of India, I, p. 317.
3."The Early History of Southern India", ibid., p. 522.
4.Pargiter. The Purāna Text of the Dynasties of the Kali Age, pp. 38. 71.
Page 341
Purānic calculations yield to us for the length of the Āndhra dynasty, and we may accept the post-Āndhra "republic" of 75 years in Magadha or at least in Pātaliputra which the Purānas of Megasthenes's day must have led him to report and which may be equated to the pre-Gupta rule there of the republican Lichchhavi clan. But we may well credit Rapson as far as the Śungas and Kānvas are involved. Sir R. G. Bhandarkar, facing the Purānic statement that the Āndhra king destroyed not only the Kānvas but also whatever was left of the power of the Śungas, arrived at the same picture as Rapson: he1 concluded that "when the princes of the Śunga family became weak, the Kānvas usurped the whole power and ruled like the Peshwas in modern times, not uprooting the dynasty of their masters but reducing them to the character of nominal sovereigns". He urged this view in spite of believing that the Purānas' 30 Āndhra rulers did not reign in a line to make up more than 4 centuries but held sway collaterally in different parts of the country during this period and that only about 19 members belonging to the main branch ruled for about 3 centuries (c. 75 B. C.-A.D. 225).2 Naturally both he and Rapson considered that the period of 112 years which the Purānas assign to the Śunga dynasty included also the 45 years of rule attributed to the Kānvas: the latter were the de facto rulers during the final 45 years of the Śunga sovereignty.
Such a computation would do away with the earlier "republic" à la Megasthenes, which we, on the suggestion of the Yuga-Purāna, assumed to divide the Kānvas from the Śungas - unless we could call the last 45 years of the latter dynasty's rois fainéants a "republic" in relation to this dynasty. What about the "republic" still earlier of Megasthenes, which, again on the strength of the Yuga-Purāna, we put between the Śungas and the Mauryas? Here too the term may be understood in an indirect sense to cover what Mookerji3 describes as "the disintegration of Aśoka's empire, which was too large to be kept together by his unworthy successors". His several sons set themselves up independently in different parts of the empire.4 Even his grandson Dasaratha who has left three short dedicatory inscriptions on the walls of rock-cut
1.Early History of the Dekkan, Section VI.
3."Aśoka, the Great". The Age of Imperial Unity, p. 89.
4.Ibid., pp. 89-90.
Page 342
caves on the heights of the Nāgārjunī hills can be deemed one of his immediate successors in the eastern provinces, just as another grandson, Samprati, son of Kunāla, stands out as such in the western.1 On this showing it is very probable that all the names given as those of his successors denote either his sons or grandsons. Then the story of the supposed last Maurya Brihadratha having been assassinated by that king's commander Pushyamitra, who became the first Śunga monarch, could pertain to the period immediately post-Aśokan and need not prevent us from starting the 112 Śunga-Kānva years in a portion of the old empire at the end of Aśoka's life rather than somewhat later. If there is nothing to stop this vision of ours, we can count back by 112 years from the commencement of the Āndhras in 802 B. C. to reach the death of Aśoka in 914 B.C. With 36 years to his reign according to the Purānas (and modern historians) we get his accession in 950 B. C.
If we accept this date we can approximate the year of the Kandahār epigraph. Going by the Deb-cum-Bhandarkar thesis, which we have found most logical, that the 14 Rock Edicts were engraved after Aśoka's 27th regnal year when Pillar Edict VII had been engraved, we can place the bilingual inscription between 923 and 914 B.C.
Thus the very record which was declared to be the crowning touch for the modern identification of Sandrocottus has raised problems whose resolution has enabled it to be the most potent and precise weapon against that identification and in favour of the one to which we are conducted with the help of modern research joining hands with the ancient Indian tradition which would identify Sandrocottus otherwise.
The Date of Pānini's Word "Yāvānāni"
Our new chronology for Aśoka's reign and for the Aramaic component of the Kandahār inscription can serve as a general signpost towards the possible date of the Pāninian "Yāvānānī" which we have referred to the Aramaic script. It shows that the word can go back in time far beyond Cyrus's conquest of Gan-dhara between 558 and 530 B.C. How far it can go has to be determined from several indications. Aśoka has the expression
1. The Oxford History of India by the late Vincent E. Smith. 3rd Ed., edited by Percival Spear (1970), p. 137.
Page 343
yonaKambojagarhdharānām in R.E. V (Girnār), which means that Gandhāra no less than Yona and Kamboja was part of his empire, though not a section directly ruled by him but a section in which he had a Viceroy: the inscriptions mention four Viceroys, one of whom was posted at Taxila, the eastern capital of Gandhāra.1 So Gandhāra was not an independent kingdom such as Pānini has noted; and as Aśoka records only the conquest of Kalihga, Gandhāra must have come to him from the first Maurya who had got possession of the far-flung empire of Mahāpadma Nanda and extended it. Pānini therefore must have lived before the first Maurya or after Aśoka. The known disintegration of the empire on Aśoka's death seems the right milieu, particularly in view of the tradition transmitted by Tāranātha that a "successor of Aśoka, Virasena by name, set up at Gandhāra"2 and even more in view of what scholars call "the Mauryan passage" in Patanjali. Bali Nath Puri3 writes apropos of it: "Commenting on the Sutra Jivikārthe cāpanye (V.3.99) [of Pānini] 'in the case of a life-sustenance, serving an object which is an image (pratikrti) the affix ka is not used except when the object is saleable', Patanjali cites as examples the images of Śiva, Skanda and Viśākha, where the rule of affixing ka does not apply. The gold-coveting Mauryas had caused images of the gods to be prepared, but the rule applies only in such cases where these provide living for the person who exhibits them to the householders. Weber was of opinion that 'Pānini in referring to images (pratikriti), that were saleable - that is by their afforded sustenance of life (jivikārthe) - had in his eye such as these that had come down from the Mauryas.' This he cites as the opinion of Patanjali."
It appears very likely that Pānini in a post-Mauryan age was looking at images made notorious in late Mauryan times. In Gandhāra late Mauryan times would be those started by whoever declared this province independent. The grammarian may be put at the end of a period counted from Aśoka's death in 914 B.C. To arrive at some precision we need to consider some other words than "Yāvānānī" in the Ashtādhyāyī and ascertain their chronological bearings.
One of them is concerned with a subject we have touched upon:
1.Mookerji, "Aśoka the Great", op. cit., p. 79.
2.Ibid., p. 90.
3.India in the Time of Patanjali (Bharatiya Vidya Bhavan, Bombay, 1958), p. 10.
Page 344
Bhāgavatism, the cult of Vāsudeva-Krishna. And here we have to keep in mind three things. First, none of the pre-Mauryan Buddhist books nor any inscription of Aśoka has the slightest cognizance of the Vāsudeva-cult in the midst of the several religious practices they report of their time. Secondly, Pānini touches on what can be called a tinge of devotion to Vāsudeva without giving him supremacy as an object of devotion or even definitely involving religiousness in the devoted attitude. Here only an elementary tendency towards Bhāgavatism can be affirmed. Hence Pānini must be post-Aśokan by all means yet need not be very much so. Thirdly, what is the indication in the other documents which in the post-Aśokan period can be seen as referring to Bhāgavatism? As we shall note in a later chapter, where we shall cite Pānini's reference too in its proper context, they come subsequent not only to the Mauryan epoch but also to the Sutiga-Kānva epoch, pertaining as they do to the rule of the Āndhras through a number of centuries. They suggest the initial stage of the Vāsudeva-cult but in a more definite manner than Pānini. So we may aver, in anticipation of our later chapter, that the Ashtādhyāyī must stand around 802 B.C., a little before or after the rise of the Āndhras to power-a date too early to allow us even to dream of any connection between "Yāvānānī" and the Greeks.
Pānini, Patanjali and the Mauryas
We may be questioned: "Would not making Pānini practically post-Śunga-Kānva create a veritable paradox with regard to the other famous grammarian, Patanjali? Does not Patanjali mention King Pushyamitra, the first Suiiga, as his contemporary? If the Mauryas' empire became fragmented in 914 B.C., as you have said, and if the Śungas commenced with Pushyamitra at almost the same time, Patanjali turns out to be a predecessor of Pānini. But Patanjali actually comments on Pānini's grammar as on an old document, as you have yourself already shown. Are you not absolutely in the soup?"
Is it really certain that Patanjali makes Pushyamitra Śunga his contemporary? Even after quoting the relevant phrase from Patanjali, Mookerji1 is content to say no more than that Patanjali was "most probably" a contemporary of Pushyamitra. Vincent
1."The Fall of the Magadhan Empire", op. cit., p. 98.
Page 345
Smith,1 knowing full well the same phrase, is even less sure: Pushyamitra, to him, is "perhaps" Patanjali's contemporary. Puri flatly affirms the contemporaneity of the two in several places -though in one place, after considering all the references to Pushyamitra, he merely says that the grammarian "may have been" the contemporary of the Śunga monarch2 - but when Puri brings up the phrase concerned we have a statement which is hardly reassuring: Patanjali, "suggesting the use of the present tense (lat) to denote an action or undertaking which has begun not finished,... cites as an instance: Here we dwell, here we perform (as priests) the sacrifice instituted by Pushyamitra (iha vasāmah -iha Pusyamitram yājayāmah)"3 If this is Patanjali's meaning, we can credit Weber's view, as quoted by Puri,4 "that Patanjali did not live at that time, but the memory of the king was still cherished by the Brāhmanas".
All scholars of Sanskrit are cognizant of the Pushyamitra-references and yet several eminent names have shifted Patanjali away from the date assigned to Pushyamitra by most modern historians - c. 187-151 B.C. by Mookerji's reckoning5 or 185-149 B.C. by Puri's6. Peterson was for the 4th century A.D. for the grammarian, Weber for 25 A.D., Goldstiicker for 140-120 B.C., Max Miiller for 200 B.C.7
We are free to relate Patanjali to Pānini by whatever chronology strikes us as appropriate. We shall resolve the problem with greater exactitude when we discuss the growth of Bhāgavatism. But when on the strength of a comment by Patanjali we make Pānini look back in a post-Mauryan age at those notorious images of late Mauryan times, we must forget the modern chronological scheme. In that scheme Patanjali's comment cannot be interpreted à la Weber, for it would plump Pānini straight into the two and a half centuries before Christ - an un-Pāninian context of Bhāgavatism, in which this cult was not budding but in full flower. Patanjali, as we shall see, can and must have his place there, though without necessarily implying Pushyamitra Śunga in that period. The
1.The Oxford History of India, 3rd ed., p. 139.
2.Op. cit., p. 10.
3.Ibid., p. 9.
4.Ibid., pp. 6-7.
5."The Fall of the Magadhan Empire", op. cit., p. 97.
6.Op. cit. p. 11.
7.Ibid., pp. 5-7 with fn. 4 of p. 7.
Page 346
Sunga-tradition happened to be popular in those centuries. Witness Kālidāsa, who flourished either under the third Gupta emperor, Chandragupta II Vikramaditya - that is, within a hundred years of Sandrocottus, according to us - or under that hero of legend, Vikramāditya of Ujjayinī, connected with the founding of the era of 57 B.C. One of Kālidāsa's most celebrated plays, Mālavikāgnimitra, deals both with Pushyamitra's son Agnimitra and with Pushyamitra himself.
6
As a pendant to our discussion of the Kandahār inscription we must build a proper perspective of the two Aramaic and one Greek epigraphs discovered soon after that bilingual.
"Kandahār II"
In 1966 a slab broken off from a rock was picked up in the bazaar of Kandahār, mentioning "Priyadraśi" and containing a part of what may be called an Indo-Aramaic edict of Aśoka, whose provenance is supposed to be the old city of Kandahār itself. It has been christened "Kandahār II". It is Indo-Aramaic in the sense that though the script as well as the basic language is Aramaic it incorporates several Indian words. In this it resembles the two fragmentary Aramaic inscriptions that came to light long ago at Taxila and Pūl-i-Dārunteh respectively. It has Irānian words, too, like those inscriptions and like the Aramaic version of the bilingual "Kandahār I"; but the use of Prakrit words shows that it could not have been meant for people completely Irānianized. A semi-Irānianized people must have been addressed. So we have to think what should have been evident from the texts of Taxila, Pul-i-Dārunteh and Kandahār I - namely, that Aramaic was employed for at least two kinds of Aśokan subjects.
We have considered the Aramaic version of the 1958 bilingual to have been for the Yonas. The new epigraph may be regarded as addressed to those close associates of the Yonas: the Kambojas. The proof is this: the new inscription, reproducing in its own way the eighth section of Pillar Edict VII, openly advocates reverence for Brāhmanas and Śramanas in its 6th line.1 We know from one
1. Journal Asiatique (Paris, 1966), text facing p. 462.
Page 347
version of an Aśokan edict that in the country of the Yonas no Brāhmanas and Śramanas moved. Kandahār I omitted these religious orders even where on account of the moral precepts we should most expect a reference to them. Kandahār II does not hesitate to bring them in: bābhanasamanesu. In view of their appearance we should imagine the text's provenance to be not Kandahār itself where the Yonas must have lived but its neighbourhood where the Kambojas were settled. Its coming to hand in the Kandahār bazaar does not necessarily imply the old city as its source. Curios and antiques are breught wherever they would most draw attention, e.g., the coins of Sophytes which were collected in the Peshawar market but which were ultimately shown to have had their origin in Irān. For convenience's sake, however, we may continue to speak of Kandahār II.
The Most Important Issue
The most important issue to resolve is: "Does K II suggest Aśoka's date to be 269-232 B. C, as at present believed or does it support our idea of a substantially greater antiquity?"
Like K I's Aramaic, the language here also exhibits traits which cannot be compared with the known Aramaic inscriptions of the Achaemenid period. Dupont-Sommer1 dwells on the expression KN' found joined to the Irānian word patiasti which occurs as hūpatiasti in K I. He says that the interpretation of these letters is very difficult. They cannot be explained by resort to Irānian nor can they be taken as a terminal or else as an independent word: in Aramaic they correspond absolutely to nothing known. Dupont-Sommer asks: "One thinks of the word kēno 'right, just, true' attested in Syriac (cf. Hebrew Kēn: Akkadian Kēnu 'firm, right', Kettu 'the truth, the right'); is it a qualifier of patiasti: 'the just obedience'; expression equivalent to hupatiasti 'the good obedience', with this difference that the Irānian prefix hu would be transformed into an Aramaic pseudo-suffix? It would make of patiastikena a hybrid word, trans-Aramaic, rather monstrous, or the element kēna would play, we repeat, the role of a suffix, a role which it never has in Aramaic... " Dupont-Sommer revolves the possibility of reading for the second letter not N but L, making the word mean kulla "all", but he cannot have this reading consistent-
1. Ibid., pp. 458-59.
Page 348
ly with another part of the text. He concludes by saying: "We must then consider KN' as a last syllable or an inflexion intimately tied to the word patiasti but what kind of last syllable or inflexion?"
Our scholar's puzzlement is enough to suggest that we are in a world quite apart from the post-Achaemenid where we should have an Aramaic speech affined to that which had prevalence in the Achaemenid empire. We have on the contrary Syriac affinities and the resultant Aramaic is sui generis, carrying us beyond the period in which Aśoka is situated by our historians.
A factor clinching our contention is brought to our notice by Dupont-Sommer himself though without realizing its true import. He1 tells us: "...in many places a vertical bar appears, which we have to interpret as a bar of separation between the words. The employment of a bar to separate words from one another is an ancient procedure attested in the very old Phoenician inscriptions (for example, the inscription of Ahiram of the 11th century B.C.) and Aramaic ones (for example, the inscription of Zakir, round about 800 B.C.); this procedure is found resumed in our II Kandahār, though in a very irregular fashion - one does not know too well under what influence." Dupont-Sommer adds that the engraver indicates word-separation also by leaving a little space blank or using terminal forms, but in spite of knowing how to combine all these three ways of separation he sometimes entirely neglects them and reverts by a sort of atavism to the ancient breakless script (scripta continua) which creates most grave and annoying difficulties of interpretation for the reader of an almost vowel-less language like Aramaic.
Clearly, the vertical bar and the scripta continua put on our inscription the stamp of a time far beyond the Achaemenid and post-Achaemenid and of a relationship to places far other than the Persian empire and the vicinity of India of the 3rd century B.C.
The Laghman Inscription
1966 disclosed another Aramaic epigraph of Aśoka - now in the valley of Laghman in North Afghanistan. Except for two words -KNPTY and SHYTY - it is all in Aramaic. Its main interest lies in two features: (1) certain numerical indications which seem connected with topography, and (2) the name inscribed at the end.
1. Ibid., p. 445.
Page 349
The numerical indications of supposed topography run: "At 200 'bows' it is down there, (the place) called TDMR. It is the route KNPTY, that is to say, (the route) of the Garden (?): more than 120 ('bows'). At TRT', here: 100. Below: 80."1
Dupont-Sommer2 reads TDMR as "Tadmor" and finds this reading, which points to the place more generally known as Palmyra near Antioch, rather appropriate in the inscription. Tadmor, he says "is the distant oasis, the great caravan city which was situated some 3800 kilometres from the place where the inscription has been engraved, and by which passed the old road of India, the old silk road." He continues: "The Aramaic inscription indicates without ambiguity, according to us, one of the most famous relays - 'down there' (tammāh), towards the West - of this famous route. One knows that the name of Tadmor appears in the cuneiform texts, about the beginning of the second millennium B.C., that it figures in the Bible, that Tadmor (or Palmyra) knew in the Hellenistic epoch and finally in the Roman epoch an extremely prosperous period. It is surely a little surprising, at first glance, to encounter this name in an inscription which has for its author the Indian emperor Aśoka, but, to tell the truth, it is not more strange than to read in other inscriptions of the same emperor the name of Antiochus, King of Antioch (Antiochus II Theos), that of Ptolemy, King of Egypt (Ptolemy II Philadelphus), that of Magas, King of Cyrene, brother of Ptolemy, that of Antigonus (Gonatas) of Macedonia, lastly, that of Alexander (either Alexander of Epirus or Alexander of Corinth). To reach Antioch, Alexandria, Cyrene, Macedonia, Epirus or Corinth from India, it would be necessary to pass by Palmyra, and the mention of this celebrated oasis of the Syrian desert in the new Aramaic inscription confers on it a special lustre."
Dupont-Sommer blithely goes on thus because he has accepted the five king-names as undoubtedly those of post-Alexandrine Greek monarchs. But actually the whole numerical and seemingly topographical passage is shrouded in mystery. The curious term SHYTY, which occurs in several places in the Aramaic epigraphs of Aśoka but is an unknown word, is properly interpreted by Dupont-Sommer as "that is to say". For, it always comes between
1.Académie des Inscriptions el Belles-Lettres: Comptes rendus des Séances de l'année 1970. Janvier-Mars, Paris, p. 163.
2.Ibid., pp. 166-67.
Page 350
an Indian term and an Aramaic one. Here the unknown quantity KNPTY answers to what in Aramaic seems to be GNT which follows SHYTY. Dupont-Sommer1 writes: "if this Aramaic word is to be read thus, it means 'the Garden' or 'the Gardens'. Does not this term describe the very valley of Laghman, well irrigated and particularly pleasant, resembling one vast grove?" But Dupont-Sommer has a twinge of doubt: "The exact sense of the Aramaic word cannot be fixed until the term KNPTY, which it translates, is understood. We may note that the first letter, in place of G, could, if required, be read as Y, and the second, in place of N, be read as D or R. The translator here remains embarrassed." So, in fact, we are in uncertainty as to what Aśoka is alluding to. And can this uncertainty leave TDMR in such absolute light as Dupont-Sommer tries to make out?
The shadow upon it takes solid shape as soon as we turn to examine the numerical indications. The word after the number 200 is QŠTN, literally "bows" or else "archers". Dupont-Sommer2 observes: "..,the context forces us to give the word the sense of a unit of length serving to measure distances. What exact value? It is indeed difficult to determine." Then the author refers to the well-known Aśokan unit of length, yojana, and, quoting Bloch, comments: "Uncertain measure, varying with the Buddhists from a league to a dozen kilometres." He goes on: "This measure is then rather imprecise. In the Anabasis, Xenophon speaks of 'parasangs', a Persian road-measure of thirty stadia, round about 6 kilometres, whose name is transcribed in Persian farsang. In the Aramaic inscription, QŠTN ought to designate a greater distance, round about 15 to 20 kilometres, perhaps the distance an archer on horseback would cover on the great roads of the Achaemenid empire from one relay to another."
Dupont-Sommer's passage is shot with self-contradictions. He begins by affirming the genuine difficulty of determining the value of the unit of length. He then mentions the Aśokan unit and states its variability. It could be very short or it could be longer, but the utmost length is 12 kilometres. From all this he jumps to the conclusion that the length-unit in the Aramaic inscription must be 3 to 8 kilometres more than the utmost extent we can attribute to the yojana and that the excess would be included in the relay-intervals on the Achaemenid roads. Dupont-Sommer has no
1. Ibid., p. 168.
2. Ibid., pp. 165-66.
Page 351
reason or right at all for this leap. It is a sheer non-sequitur.
The ground for it is nothing else than the assumption that TDMR is the far-off Syrian oasis which is some 3800 kilometres from the place where the inscription stands. What we know of the Aśokan length-unit or the Persian "parasang" renders Palmyra impossible. It calls for some other reading of TDMR.
If no such reading is available, we need not be disconcerted. We are in the midst of unknowns: neither KNPTY nor GNT has been clarified. To connect them with the valley of Laghman, as Dupont-Sommer is disposed to do, lands us in incongruities. For, apropos of the expression "YTRY 120", he1 has to admit in the teeth of his conceiving a connection with Laghman: "120 (bows) more. It seems that the length of the route called KNPTY is indicated from its starting-point up to its end-point. We do not know, alas, either the one or the other. The distance, in any case, must be considerable, if one compares it to that of 200 (bows), which is the distance from the valley of Laghman to Palmyra."
Mark that crucially significant "if". Once we drop the hypothesis of Palmyra we do not have to pass beyond Laghman. When we refuse to do so, we have to end up with a confession like Dupont-Sommer's2 in his role as an honest researcher: "All the numerical indications of lines 2 and 3 (bis) remain for us rather obscure."
"Tadmor" has no real raison d'etre. Even acceptance of it would not necessarily force us into the 3rd century B.C. Dupont-Sommer has himself traced the name to a greater antiquity, right back to nearly 2000 B.C., that is about 1050 years earlier than the time we have chosen - 950 B.C. - for Aśoka's accession. But the likelihood is more that the Aramean influence came towards Aśoka than that he extended himself to the Palmyra oasis.
Finally, we may ask: "If Aśoka lived in the 3rd century B.C. and if his outstanding foreign connections in the West were with Greek kings whose names he has Indianized, why does he not use the term known to those kings - namely, 'Palmyra' - rather than the non-Greek 'Tadmor'?" It was precisely the Greek rulers known as the Seleucids, contemporary with the early members of the dynasty founded by Sandrocottus - the Greek rulers of Syria from Seleucus Nicator onwards for three generations - who made old Tadmor prominent under the name "Palmyra" as a site on the
1.Ibid., p. 168.
2.Ibid., p 169.
Page 352
east-west trade-route.1 If Aśoka was the grandson of Sandrocottus and if Tadmor was concerned in the inscription, an Indianized form of "Palmyra" is just what the inscription would yield. To find "Tadmor" instead would itself point us away from connections with the Seleucid kings and from their period in which Aśoka is presently situated.
Now for the name of the "judge" with whose help the inscription was made: "Wašu." Dupont-Sommer2 is perplexed by it. It is not an Indian or an Irānian appellation. Considering the language of the epigraph, we should expect it to have an Irānian air about it. But Dupont-Sommer reports: "The name Wašu is not found noticed in the work of Emile Benveniste, Titles and Proper Names in Ancient Irān3 (Paris, Klincksheck, 1966). One encounters in Nabatean, in the inscriptions called Sinaitic, the proper name W'ŠW (CIS, II, 2136, 2315) and also 'WŠW (ibid., 325); one supposes generally that these forms are incorrect for 'WŠW 'Ušu', a proper name extremely frequent in Nabatean (cf. J. Cantineau, Nabatean,4 Vol. II, p. 88 sq.)."
We may recollect Gray's information5 in regard to the Sinaitic inscriptions. The legible alphabetic inscriptions in Syria and Palestine, beginning with "the epigraph of the king Shaphatha'al of Byblos, which is dated by Dunand in the 15th century B.C., though he does not exclude the possibility of a date two centuries earlier" - these Syrian and Palestinian inscriptions develop along a direct line from "the Sinai inscriptions, fragmentary and uncertain as they are". Thus "the inscriptions called Sinaitic" go back beyond the middle of the second millennium B.C. And such antiquity and association for the name of the "judge" who served to make the Aramaic epigraph are in keeping with the kind of language we have here. Dupont-Sommer6 informs us: "As the other Aramaic inscriptions of Aśoka show, this Aramaic language is generally of a rather awkward style, and, at times, of an incorrect grammar." In the post-Achaemenid post-Alexandrine age, and in the proximity of Irān, we should have a very different
1.The Encyclopaedia Britannica, Micropaedia, 1976, VIII, p. 703.
2.Op. cit., p. 169, fn. 1
3.Original French title: Titres el noms propres en Irānien ancien.
4.Original French title: Le Nabatéen
5.Archaeology and the Old Testament World, p. 24.
6.Op. cit., p. 163.
Page 353
quality of expression and syntax. All our Aśokan epigraphs in Aramaic betray an archaic gaucherie, a linguistic primitivism, taking us back towards the beginnings of a tongue and that too in an alien locale far from its home. They do not at all suggest an Aśoka of 269-232 B.C.
A few words by way of a general criticism of Dupont-Sommer's interpretation may be added apropos of his assumption that the numerical pointers refer to a unit of length in the word Q§TN. He renders this word as "bows" (or "archers") and starts inquiring what length it connoted. But have we at all a length-unit here? Dupont-Sommer is unaware of the origin and history of the word and of its significance.
Richard N. Frye, the authority on Persia, has an illuminating passage:
"The Achaemenids, by establishing colonies of soldiers in conquered lands and by giving land to civil and military servants, favoured and promoted feudalism. From Akkadian records we hear of two terms in the feudal relationship, ilku and qashtu. The former was the old Babylonian fief or feudal service. But under the Achaemenids, and not attested before Darius I, we find the word qashtu, originally meaning a 'bow' or 'division of land for the support of an archer', given to persons in return for military obligations to the ruler. It would appear that the institution of qashtu was introduced by the Persians while the Babylonian ilku then becomes a monetary or silver obligation, and comes to mean 'ground tax' or simply 'tax'. The qashtu, given as fiefs to families, however were subsumed under a hatru, which was primarily a tax collecting institution with an important official called a shaknu over it."1
What we gather from Frye is that qashtu stands for a certain piece of territory and is not anything comparable to a yojana or a parasang. As Dupont-Sommer's Aśoka is nearly 300 years later than Darius I, the Persian sense of the term must hold for the Aśoka of modern historians.
With this sense we do not move into greater ultimate clarity in respect of the inscription but we take away all plausibility from Dupont-Sommer's approach which in itself too we have found unacceptably tending, despite his hesitations here and there, to be "slanted" under the compulsion of a particular chronology. And
1. The Heritage of Persia, p. 138.
Page 354
all talk of Tadmor - rather arbitrary already - becomes irrelevant. TDMR remains completely cryptic.
How does qashtu bear upon our chronology for Aśoka? Can it at all figure in an inscription dated centuries before Darius I? Although Frye says that the word is first attested in this king's time, he disavows absolute certainty about the earlier time by saying: "It would appear that the institution of qashtu was introduced by the Persians." The opening phrase evinces some doubt. The very fact that qashtu, like ilku, is heard of in Akkadian records suggests a non-Persian background and the possibility of its employment earlier than the Achaemenids. Whether the possibility has been realised or not in the Laghman epigraph depends on whether or not this epigraph has signs of a pre-Achaemenid antiquity which would tend to make qashtu also older. As such signs exist, we may aver that the term is found first occurring in this Aramaic inscription.
However, its meaning might not be the same as in Achaemenid and post-Achaemenid ages. Separated from ilku and neither coloured by nor interfering with this word's sense of "the old Babylonian fief or feudal service", it may even do what it cannot in a post-Achaemenid context. On the analogy of the Sanskrit dha-nuh (bow) which was a measure of length equal to 4 hastas (i.e., 4 cubits or 6 feet), it may even be a unit of measure.1 Of course, on the same analogy it would hardly provide such length as could carry us to a far place like Tadmor.
Anyway, the term can be present in a document of the late 10th century B.C. We do not need to fear for our Aśokan chronology.
Two Meaningful Facts
In passing we may mention two facts that held a lot of meaning in an earlier discussion of ours. When we treated the bilingual inscription we noted sufficient space above and below the Aramaic text for other texts to be accommodated. We argued that just because the Greek text was above the Aramaic it need not be regarded as contemporaneous with it. It may have been put there later. The present Aramaic text has below it an inscription in quite different characters, a specimen of the famous "petroglyphs" of
1. I owe this idea to my friend David Hopkins.
Page 355
Laghman.1 Above, there is a blank space where a Greek translation or paraphrase can easily be introduced. An Aśokan inscription does not always start right at the top of a rock-face and its being in the middle does not imply that it is there because it follows a Greek or any other version of it. Its position allows several possibilities.
Secondly, if the Greeks were in Aśoka's time as his subjects, why is there in Laghman, which is part of the locality where the Greeks are known to have settled in the 3rd century B.C., no Greek version set up by Aśoka of the Aramaic text although room is available both above and below? The petroglyphs must have come much later than the original Aramaic text. Aśoka could have added a Greek version in the same place or in the place at the head of the rock-face. The absence of it argues against the Greeks having been present in Laghman in Aśoka's time and against this time having been post-Alexandrine in which the Greeks were actually in Laghman.
The case of this inscription is different from that of the epigraphs of Taxila and Pul-i-Dārunteh. The latter two are on stones leaving hardly any room for a Greek version. The opportunity provided by the epigraph and the wasting of it are significant.
The New Greek Inscription
The new Greek inscription was discovered in the Kandahār bazaar in the same year as Kandahār II. It consists of a fragmentary stone piece holding 22 lines which translate or adapt the end of R.E. XII and the beginning of R.E. XIII. There is no Aramaic text accompanying it. Again, Aśoka is called "Piodasses". Most tantalizingly the fragment of R.E. XIII breaks off before the phrases which would settle once for all whether Aśoka's Yonas were Greeks or not. As Emile Benveniste2 writes: "At the end we have only the first words: 'and, as with all the peoples, there are found...' One easily completes: 'these groups, Brāhmanas and Śramanas, etc' But here the Indian text presents an interesting discordance between the version of the North-west and the others." We may recall that one of the versions announces that
1.Académie des Inscriptions..., p. 159, PI. II.
2.Le Journal Asiatique, CCL II, Année 1966, Fascicule No. 2, Paris, "Edits d'Aśoka en Traduction Grecque", p. 156.
Page 356
these religious communities are found among all the peoples "except among the Yonas". Benveniste1 comments that this restriction does not appear at Shāhbāzgarhī and he would like to know how the phrase was conceived in this version which he takes to have been meant for the Greeks. What is of moment is to wonder whether the text would have the term "Yonas" or there would be a substitution openly indicating "Greeks". If "Yonas" occurred, we would be certain that the Aśokan Yonas never were Greeks. And we should be certain that the Greek version was an adaptation made of an ancient original by the Indological Greeks themselves of the 3rd century B.C. On the other hand, if we came across an open indication of "Greeks" instead of "Yonas" we could wager that Aśoka had Greeks as his subjects and then our theory would fail. Perhaps a further chance discovery will settle the issue. Meanwhile we must remember how heavily the dice are loaded against them and we have to go by a number of pointers in the new Greek epigraph.
The mere mention of Brāhmanas and Śramanas would not mean that the Greeks might not have been Aśoka's subjects. What we have here is an adapted reproduction of matter pertaining to conditions in far-away Kalinga. We read: "All those who dwell there (in Kalinga), Brāhmanas and Śramanas and others consecrating themselves to Piety..." The Greeks are merely told of happenings elsewhere: they are not themselves asked to revere Brāhmanas and Śramanas. The really important factor is something else and relates to the material on which the Greek version appears and the context of the material.
From Benveniste's articles we have to go to Daniel Schlumber-ger's communication: "Une Nouvelle Inscription Grécque d'Açoka."2 Schlumberger conveys the information:
"The text of 1958 was a rock inscription; that of 1964 is engraved on a block which surely belonged to a building.... It is a novelty, a wall inscribed in the Mauryan epoch. Among the texts of Aśoka recovered up to now, the majority are engraved on rocks; and the rest on 'pillars', that is, on monolithic columns of rose sandstone, all drawn from the same quarry, near Benares. One does not know of any inscription of Aśoka engraved on the walls of buildings, for
2.Académie des Inscriptions el Belles-Lettres. Comptes rendus des Seances de I'annee 1964, Janvier-Juin, Paris, 1965. pp. 129, 133, 134.
Page 357
the simple reason that one does not know of any buildings of Aśoka beyond the stūpas said to be Aśokan, which are in brick and whose Aśokan date would besides, demand to be verified in each case.
"Aśoka got built at Kandahār a construction in freestone, or at least certain parts of it were in freestone: that is what we know today. What type of contruction? Temple, stūpa, palace, rampart or simple support of some terrace or of some utilitarian building, we cannot say. It would be of great interest to know."
The long and short of Schlumberger's reflection is: we are faced with a completely non-Aśokan feature. None of the numerous edicts of Aśoka gives us the slightest hint that he ever got his messages engraved as here on freestone - that is, fine-grained sandstone or limestone easily sawn to specification for construction-purposes. Here obviously is a portion of a building. But even the buildings attributed to Aśoka - a few stūpas - are never in any kind of stone. If anything could render a so-called Aśokan edict non-Aśokan, it would be this unique feature which we have come up against. This feature would be quite natural if we thought of the Indological Greeks of the 3rd century B.C. doing what they might on their own if they had no connection with Aśoka and his practice but were contemporaries of the Guptas.
Stone-structures begin to be in use only in the Gupta period. K. A. Nilakanta Sastri1 says: "Stone temples of the time are rare. They are unpretentious flat-roofed structures about ten feet square with a porch of still smaller dimensions; their masonry, however, is excellent, the stones being finely dressed and held together with no mortar. The Daśāvataratemple of Deogarh forms a transition to the later style with high śikharas. This temple had a śikhara of about 40 feet. It stood on a raised plinth in the centre of the open terrace. It had a plain interior, but its entrance was exquisitely carved and decorated, the figures of Ganga and Yamuna being carved on the jambs, a typical Gupta feature. When complete, this was doubtless a monument of rare merit, and its sculptural panels were the most superb of their kind."
Sastri's picture shows us the two extreme bands of the Guptan architectural spectrum: stone-building art practised on a semi-primitive scale on the one hand and splendidly deployed on the
1. History of India, Part I - Ancient India (S. Viswanathan, Madras, 2nd Ed., 1953), p. 162.
Page 358
other though still not in wide dimensions. What we may infer is that for the first time in India stone-structures came into use, particularly for religious ends, in the Gupta age and was soon developed with artistic skill. The Deogarh temple must not be regarded as a solitary example of a transition to a more sumptuous style. V. Smith1 perhaps strikes the right note when he tells us: 'The next important and interesting extant stone temple of the Gupta age is one of moderate dimensions at Deogarh.... The panels of the walls contain some of the finest specimens of Indian sculpture.... Fragments, including some beautiful sculptures, indicate that magnificent stone temples of the Gupta age stood at Sārnāth near Benares and elsewhere."
Of course, the Greek building of the 3rd century B.C. could not have been a temple or stūpa, but the non-Aśokan fact spotlighted by Schlumberger is decidedly pro-Guptan.
Conclusion
Looking back at the two inscriptions in Greek mentioning "Piodasses", we may sum up that one of them embodies a special message adapting an Aramaic text which could be an ancient original with the Greek merely a much later paraphrase, while the other is a paraphrase of two already existing Aśokan edicts, made by the Greeks likewise in a period far posterior to the period we have assigned to Aśoka: 950-914 B.C. Neither epigraph can claim to be a particular communication for the Greeks alone. So far no inscription in Greek only and uniquely directed to them has come to hand. If they were the Yonas, the people to whom the Brāhmanas and Śramanas did not minister, we should expect from the great Buddhist emperor some message suited to them and none else. In this respect we stand before a total vacuum.
The two Aramaic epigraphs have also introduced no element encouraging us to bring Aśoka down to the post-Achaemenid post-Alexandrine epoch. Whatever notable traits they have displayed set us in the reverse direction.
Page 359
7
All Aśokan issues that have to do with the Greeks have now been tackled. They are indeed the main issues and the ones with which we are concerned. But it may be both interesting and useful to tackle in brief whatever problem directly related to Aśoka is left over.
Aśoka and the Date of Buddha's Death:
the Ceylonese Traditions
The first problem is the date of Buddha's death (Nirvāna or Parinirvāna, as it is often called). Since Aśoka definitely comes after this event and since this event is taken to have occurred in either 543, 486 or 483 B.C., we may be asked: "Should we be surprised if Aśoka's grandfather Chandragupta Maurya emerged as Sandrocottus and Aśoka stood in the post-Alexandrine epoch along with Antiochus II and company?" What we have to do is to throw serious doubt on all those three dates of the Nirvāna.
543 B.C. was for long the sole accepted date on the strength of the Ceylonese Chronicles and a Ceylonese tradition. But with its help Aśoka can never be satisfactorily made post-Alexandrine. For, in these Chronicles Aśoka, here named Piyadassi, has his coronation 218 years later - that is, in (543-218=) 325 B.C., a date which fails to render him a contemporary of Antiochus II and company.
To avoid this inconvenience, scholars have urged that when the Chronicles spoke of Piyadassi's coronation they had Chandragupta Maurya and not Aśoka in mind. The argument is: "Is not Chandragupta once called Priyadarśi in Act VI of the late Indian drama Mudrārākshasa?"1 With Chandragupta coronated in 325 B.C., as is surely possible, and with his own reign of 24 years as well as his son Bindusāra's of 27 intervening between it and Aśoka's, Aśoka will naturally fall into just the period wanted. But the fact stares us in the face that nobody except Aśoka is known to be designated Piyadassi in the rest of the Chronicles. Hence the Mudrārākshasa is of no avail.
Besides, now on the top of a sandstone hillock or range of Vindhyan hills near Bhopal we have found edicts that, as H. D.
1. Cf. R. C. Raychaudhuri in Indian Culture, II,
Page 360
Sankalia1 tells us, "make specific mention of the emperor's name as 'Priyadarśi Nama Rājā'. 'Aśoka' was his assumed name." So the rest of the Chronicles is right, and most unlikely to employ "Piyadassi" anywhere as a general honorific for any other king.
Even if the Mudrārākshasa were of any use, the very starting-point - 543 B.C. - of the calculation would still be an artificial construct. Let us not forget what Max Muller2 said long ago in comparing the chronologies of the Southern and the Northern Buddhists:
"The Northern Buddhists found their chronology on a reported prophecy of Buddha that 'a thousand years after his death his doctrines would reach the northern countries'. Buddhism was definitely introduced into China in the year 61 A.D.: hence the Chinese fix the date of Buddha's death about one thousand years anterior to the Christian era... If... the starting-point of the Northern Buddhist chronology turns out to be merely hypothetical, based as it is on a prophecy of Buddha, it will be difficult to avoid the same conclusion with regard to the date assigned to Buddha's death by the Buddhists of Ceylon. The Ceylonese possess a trustworthy and intelligible chronology beginning with the year 161 B.C.3 Before that time, their chronology is traditional and full of absurdities. According to Professor Lassen, we ought to suppose that the Ceylonese, by some means or other, were in possession of the right date of Buddha's death; and as there was a prophecy of Buddha that Vijaya should land in Ceylon on the same day on which Buddha entered Nirvāna, we are further asked to believe that the Ceylonese historians placed the founder of the Vijaya dynasty of Ceylon in the year 543 B.C., in accordance with their sacred chronology. We ate not told, however, through what channel the Ceylonese could have received their information as to the exact date of Buddha's death, and although Professor Lassen's hypothesis would be extremely convenient, and has been acquiesced in by most Sanskrit scholars, it would not be honest were we to conceal from ourselves or from others that the first and most
1.Indian Archaeology Today (Ajanta Publications, Delhi, 1979), p. 130.
2.A History of Sanskrit Literature (Edition 1912), pp. 136-137.
3.Cf. Sylvain Levi, as quoted by Vincent Smith in The Indian Antiquary, 1902, p. 199: "There is not, I believe, any reason to doubt the substantial accuracy of the Ceylonese dates... from... the... time of Dutthagāmanī, about 161 B.C., although the dates prior to his reign are not to be trusted." (K.D.S.'s note)
Page 361
important link in the Ceylonese as well as the Chinese chronology is extremely weak."
Rhoads Murphy1 has subsequently expressed the same cautious attitude: "The written history of Ceylon begins with an account of the landing of Vijaya, an Aryan-speaking prince, and his seven hundred followers somewhere on the north-west coast at about 543 B.C.... although both the date (which is artificially fixed) and the story are of later origin and border on legend."
Fleet and Geiger, dissatisfied with 543 B.C. for Buddha's death, have tried to prove that this era is a comparatively modern fabrication and that the true date of the event to which it is tied is 483 B.C. "Geiger's date, however," remarks Raychaudhuri,2 "is not explicitly recognised in tradition." A. Paranavitana3 rejects the theory that there was a Ceylon era reckoned from 483 B.C. So it would seem that this era is still more a modern fabrication than the other. But even if it were not and had its source in tradition, the tradition would not per se confer historical value upon it any more than on 543 B.C.
The truth in this matter is threefold. First, although the tradition is linked with the fairly detailed treatment of Aśoka by the Ceylonese Chronicles, the latter can hardly be regarded as very trustworthy in matters Aśokan. Sircar4 has declared: "There is some palpable and irreconcilable difference between the Ceylonese accounts and the information supplied by Aśoka's inscriptions." One important instance is: the inscriptions reveal Aśoka as directly responsible for the Buddhist missions, whereas these accounts completely ignore his part and attribute the missions wholly to the Buddhist Church. An equally important example may be culled from Barua:5 "The main drawback of the traditional narratives is that they have nothing to say about the Kalihga war which was waged in the 8th year of abhisheka [= coronation] and marked the real turning-point in Aśoka's life and career (R.E. XIII)."
Secondly, there is a wide diversity in the dates which have come down to us of Buddha's death from ancient sources: the years
1.The Journal of Asian Studies, Vol. XVI, No. 2, February 1957, p. 183.
2.An Advanced History of India, p. 58.
3."History of Ceylon: Appendix", A New History..., Vol. VI, pp. 262-63.
4.Letter to the author, dated 7.11.1956.
5.Aśoka and His Inscriptions, Part I, p. 31.
Page 362
derived from the Ceylonese tradition are just two out of several in our hands. We shall shortly deal with the diversity.
Thirdly, scholars choose either 543 or 483 B.C. by establishing to their satisfaction some sort of relation between it and the date they accept as certain for Aśoka's accession, a date within the period 275-268 B.C. Once this date is challenged, there is no reason to set any store by the Ceylonese tradition.
How about the "Dotted Record" of Canton?
The same holds for what is called the "dotted record" of Canton, about which Sircar1 writes: "The most reliable tradition about the date of Buddha's death is 486 B.C., supplied by the Cantonese dotted record of 489 A.D., taken from India to China by Sungabhadra... although scholars now usually accept the date 483 B.C. proposed by Fleet and Geiger."
It is the Chinese priest Tao-Suen who speaks in 664 A.D. of this record2 in which up to the year 489 A.D. a dot is said to have been put for 975 years for each year after Buddha's death. Thus the first and only reference to it comes eleven hundred and fifty years after the alleged time of the Nirvāna and a hundred and seventy-five years after the reported dotting ceased. Further, Geiger,3 while appreciating the closeness of 486 B.C. to the date favoured by himself (483 B.C.), makes the observation: "I- would not for my part attach too much importance to the 'dotted Record'. It is singularly improbable that in the course of time - it is a question of nearly a thousand years! - not a single error or oversight should have occurred."
The very fact that most scholars agree with Geiger's and Fleet's date implies their assumption of an "error or oversight" of three years in the "dotted record". But the moment such a slip is admissible, there is no intrinsic ground to think only of three years: any number of slips may have taken place. A restriction to three years is made simply because, in the opinion of most scholars, 483 B.C. gives the best relation of Buddha's death to Aśoka's accession.
1.Journal of the Royal Asiatic Society, B Letter, Vol. XV, No. 2, 1949, pp. 127- 128.
2.Indian Culture, Vol. V, p. 316.
3.The Mahāvamsa (Colombo, 1950), Introduction, p. xxvi.
Page 363
K. A. Nilakanta Sastri, reviewing a new book on Aśoka1 in the Sunday Hindu of Madras some years back, wrote quite frankly: "Chronology is always a puzzle in ancient Indian history and the author does not seem to be aware that 483 B.C. for the parinirvāna of the Buddha is itself based on an assumed date for Aśoka's consecration; and the only merit of the 'dotted record' of China is that it yields a date close to that assumed for Aśoka."
The Large Number of Dates in the Field
Moreover, we may suspend belief in the "dotted record", no less than in the two other dates currently in the field, because there are a large number of different dates actually available for Buddha's death. The concentration of scholars on 486 or 483 as alternatives to 543 B.C. has prevented the common student of history from scrutinizing the credentials of the other dates and thereby the possibilities of various epochs for Aśoka.
Particularly the "dotted record" as an article carried from India by a Chinese visitor falls under suspicion. Even apart from the extreme improbability that immediately after Buddha's death the document started being written - unknown to any canonical Buddhist book - and even apart from the extreme likelihood of the document's having been based merely on some tradition of the Parinirvāna, according to which the dots were entered by hindsight from a certain year and carried forward up to 489 A.D. -there is an amazing fact to be noted from what we learn from another Chinese visitor to India: the famous Hiuen-Tsang. Hiuen-Tsang was in India during 630-643 A.D. and not only was a contemporary of Tao-Suen (664 A.D.) in China but also collected all traditions current in his time for the epoch of the PariNirvāna. Sircar2 cites four different traditions reported by Hiuen-Tsang: (1) about the end of the 3rd century B.C., (2) about the middle of the 6th century B.C., (3) about the middle of the 7th century B.C. and (4) about the middle of the 9th century B.C. The 5th century B.C. in which Tao-Suen's "dotted record" puts Buddha's death does not figure at all in Hiuen-Tsang's four-faceted report made in the very period in which Tao-Suen lived.
As Hiuen-Tsang gleaned his information during his extensive
1.Romila Thapar's Aśoka and the Decline of the Mauryas (Oxford, 1961).
2."The Yavanas", The Age of Imperial Unity, p. 114, fn. 1.
Page 364
travels, we may well conclude that the "dotted record" 's date for Buddha's death had very little vogue in the India of Hiuen Tsang's and Tao-Suen's age. This means also that the closely related 483 B.C., accepted by most scholars, had then no Indian standing to speak of.
Fa-Hien was a Chinese traveller over India in the 4th century A.D., a period when the supposed dots were still being put and which was only a century earlier than Sungabhadra, the alleged Dearer of the Record from India to China. Fa-Hien has no notion of either 486 or 483. Nor does his information tally with any of Hiuen-Tsang's four dates. He refers to an image of Maitreya Boddhisatva having been "set up three hundred years after the Nirvāna of Buddha, in the time of Ping-Wang of the Chou family".1 H. A. Giles,2 his most recent translator, interprets him to mean that Buddha died during the reign-period of Ping - that is, in 770-719 B.C. James Legge,3 one of the early translators, would seem to understand that the death was placed three hundred years before Ping - that is, in the 11th century B.C.
This is a far cry from the Ceylonese tradition as generally known. But that tradition itself is not free from a discrepancy which would render the 11th century B.C. a still more differing chronology. E. J. Thomas4 has pointed out that the Sarvastivadins among the Buddhists take Aśoka to have flourished one century after the Nirvāna and that such a tradition may be traced in the Ceylonese Chronicles themselves. Within the current framework of Aśokan history the Nirvāna, according to this date, would fall in the 4th century B.C. and a Japanese scholar quoted by Thomas places it in 386 B.C.
Thomas5 has also set down "the tradition of Khotan as reported in Tibetan books".6 It places the 50th year (out of 55) in the reign of Aśoka at an interval of 234 years from the Parinirvāna, which means 180 years earlier than Aśoka's accession in contrast to the
1.Beal, Si-yu-ki: Buddhist Records of the Western World (London, 1906), p. xxvi.
2.Records of the Buddhist Kingdoms, Chapter VII.
3.A Record of Buddhistic Kingdoms by Fa-Hien (Oxford, 1886).
4.B. C. Law Volume, II, pp. 18-22.
5."Aśoka, the Imperial Patron of Buddhism", The Cambridge History of India, I. p. 503.
6. W. W. Rockhill, The Life of the Buddha and the Early History of His Order (Triibner & Co., London, 1884), p. 233 and the Tibetan texts there named.
Page 365
218 mentioned in the Ceylonese Dīpavamsa (VI, 1). The present-day date of Aśoka's accession is c. 269 B.C. So the Parinirvāna would happen in c. 449. But surely the "Tibetan books" do not see Aśoka's time eye to eye with modern historians?
Sircar1 speaks of "Buddhist traditions of north-western India as recorded in the Milinda-panha", according to which the Indo-Greek king "Menander flourished 500 years after the PariNirvāna, i.e. in the sixth century after Buddha's death". Then Sircar2 writes: "It is interesting to note in this connection that Kielhorn suggested an epoch of the PariNirvāna falling in 638 B.C. with which the astronomical details of the date of an inscription (List of Northern Ins. No. 575) work out satisfactorily." Such an epoch would agree with the third out of the four different traditions Hiuen-Tsang reports about the Parinirvāna, and that third tradition, says Sircar,3 would place Menander "between the middle of the second and the middle of the first century B.C." - in consonance with the present-day unanimous dating of "Menander's reign... after [the Bactrian Greek king] Demetrius's death which took place about 165 B.C."4
The Ceylonese Dīpavamsa and Mahāvamsa correspond with Hiuen-Tsang's second tradition, but the correspondence is of no use to modern historians who pin their faith on 486 or 483 rather than 543 B.C. On the contrary, if it has any value, it disqualifies their faith by a double disagreement. But actually there is no particular intrinsic value in it any more than in 486 and 483 being so close to each other.
Hiuen Tsang's first and fourth traditions lack parallels - except that his middle of the 9th century seems near to the chronological implication of what the Arab traveller, Albērūnī(1031 A.D.), who made a point of collecting eras and dates, has testified. He5 wrote: "In former times Khurāsān, Persis, Irāk, Mosul, the country up to the frontiers of Syria, was Buddhistic, but then Zarathustra went forth from Ādarbaijan and preached Magism in Balkh (Bactria). His doctrines came into favour with King Gushtasp, and his son Isfandiyad spread the new faith both in East and West, both by
1."The Yavanas", The Age of Imperial Unity, pp. 113-14.
2.Ibid., p. 114, fn. 1.
4.Ibid., p. 113.
5.Sachau, Albērūnī's India, p. 21.
Page 366
force and by treaties. He founded fire-temples through his whole empire, from the frontiers of China to those of the Greek empire. The succeeding kings made their religion (i.e., Zoroastrianism) the obligatory state-religion for Persis and Irāk. In consequence, the Buddhists were banished from these countries and had to emigrate to the countries east of Balkh... Then came Islam."
Evidently, with his reference to the frontiers of China and the Greek empire, Albērūnī is following the popular Persian tradition which places King Gushtasp (or Vistašpa, as in the Avesta) in the 7th century B.C. - the tradition mainly supported in our own day by A. V. Williams Jackson who dates Zarathustra to c. 660-583 B.C. So Buddha, for Albērūnī, must antedate the 7th century B.C. and, if by the time of Zarathustra, Buddhism had spread from India to "Khurāsān, Persis, Irāk, Mosul, the country up to the frontier of Syria", there must have been a sufficient period between Zarathustra and Buddha to accommodate so much missionary activity. So Buddha must go a few centuries upward from the period of Zarathustra - at least to the middle of the 9th century B.C. if not even to the Chinese date mentioned by Max Muller as being equally credible as the Ceylonese, the date reached by counting 1000 years before 61 A.D.: 939 B.C.
In fact, by the Chinese signpost itself we are not confined to this year. In the eyes of Majumdar,1 just as historical as the event indicated by 61 A.D., is the relation between China and Buddhism in the account telling us that the Yueh-chi rulers in the Oxus Valley presented some Buddhist texts to the Chinese court in the year 2 B.C. This would fix the Nirvāna in 1002 B.C. But is there any weighty reason to ignore those two other Chinese traditions noted with doubt by Majumdar,2 according to which (1) Buddhist missionaries from India proceeded to China as early as 217 B.C. and (2) a Chinese general who led a military expedition in Central Asia in 121 B.C. brought home a golden statue of Buddha and thus made the Chinese first know of Buddhism? These traditions would give us 1217 and 1121 B.C.
Kalhana (Rājatarangini I, 172), states that the PariNirvāna preceded Kanishka by 150 years. Modern historians mostly put Kanishka in 78 A.D., which is dubbed the Śaka Era. Hence, by Kalhana's information combined with their chronology, Buddha
1."Colonial and Cultural Expansion", The Age of Imperial Unity, p. 646.
Page 367
died in 72 B.C. However, for Kalhana Kanishka comes before Gonanda III of Kāshmir, whom he dates 2330 years before his own time which is the 1070th year of the Śaka Era (Rājataranginiī, 52-53) - i.e., (1070+78=) 1148 A.D. So Kanishka passes beyond (2330-1148=) 1182 B.C. Then the Parinirvāna, by its 150 years' precedence, would occur before 1332 B.C.
The Purānic tradition and the Mahābhārata pointers would carry us still more back in time. As we know, the latter sets the Bhārata War 36 years before the hypothesized Kaliyuga of 3102 B.C. And the Purānas give dynasty-lengths as well as reign-lengths from the birth of Parīkshit in the year of the Bhārata War. The first Magadhan dynasty is of the 22 Bārhadrathas who reign for 1000 years. It is followed by the 5 Pradyotas with their sum of 138 years. Then come the 10 Śiśunagas whose total duration is 362 years. All these numbers make up exactly the 1500 years which are said in some versions - as against 1015, 1050 and 1115 in others -to elapse from Parlkshit's birth to the rise of the Nandas, the dynasty next in time to the Śiśunagas. The Nandas rule for 100 years. As they begin in (3138-1500=) 1638 B.C., the Mauryas' founder Chandragupta mounts the throne in (1638-100=) 1538. With Chandragupta's reign being 24 years and his son Bindusāra's being 25, Aśoka's accession is in 1489 B.C. We are not told, as by the Buddhist books, the interval between his accession and the death of Buddha nor the name of the pre-Aśokan king during whose reign Buddha died. Hence we cannot calculate on the Buddhist basis1 of the 8th year of King Ajātaśatru's reign. But if we did we would not be able to go by the Buddhist 218 years as the interval between that year and the accession of Aśoka. In the Purānas Ajātaśatru is the 6th Śiśunaga king. As this dynasty succeeds the Pradyotas, they start after the Bārhadrathas' 1000 years and the Pradyotas' 138: namely, in (3138-1138=) 2000 B.C. The reign-periods of the 5 pre-Ajātaśatru Śiśunagas total (40+ 36+26+40+38=) 180 years. Adding Ajātaśatru's 8, we get 188 for deduction from 2000 B.C. We reach 1812 B.C. for Buddha's death - creating an interval of (1812-1489=) 323 years between this event and Aśoka's Purānic enthronement. But, truly speaking, all we can affirm from the non-Buddhist Indian document is that Buddha died a few centuries prior to 1489 B.C.
Even the Purānic tradition is beaten hollow by the farthest term
1. Ibid., p. 36.
Page 368
of the Tibetan chronology. We have already marked Thomas's information from "Tibetan books". Further details emerge from A. Weber's History of Indian Literature:1 "Among the Northern Buddhists fourteen different accounts are found, ranging from 2422 B.C. to 546 B.C." Max Miiller's History of Sanskrit Literature2 lists all of them as having been "current in Tibetan chronology: 2422, 2148, 2139, 2135, 1310, 1060, 834, 882, 880, 837, 752, 623, 576, 546 B.C."
On their own merits most of the computations we have set down so far appear to be as good as any in vogue today. For, we have no certainty at all about the epoch of the Parinirvāna. As K. A. Nilakanta Sastri has stressed, historians have to date Buddha's death by Aśoka's consecration and not vice versa. Buddha will simply move backward or forward with Aśoka. If we identify Sandrocottus with Chandragupta I and put the accession of Aśoka, as we have done, in 950 B.C., the Parinirvāna can be carried backward by either 218 years to 1168 B.C. or fixed in the more plausible of the two Chinese dates distant enough from Aśoka: 1121 B.C., 1217 B.C.
Since 1168 is almost exactly midway between the latter pair, we may for the sake of convenience as well of historical likelihood take it as the most probable and, making it our point de départ, place Buddha's life of 80 years in the bracket: 1248-1168 B.C.
(By the way, on this showing, Ajātaśatru's reign, taken on the Buddhist information, began in (1168—8=) 1160 B.C. and the death of Mahāvlra, founder of Jainism, occurring, as it is said, 3 years before Buddha's, was in 1165 B.C., and his life of 72 years started in 1237 B.C.)
Aśoka and King Devānampiyatissa of Ceylon
The point which, next to the date of Buddha's death, we have to cope with is that the Ceylonese Chronicles make Aśoka a contemporary of King Devānampiyatissa who, by the traditional chronology of Ceylon, reigned in "the period 308-7 B.C. to 268-7 B.C." and, according to Geiger's modification, "from 247 to 207 B.C."3 This king is said to have sent "an embassy with rich presents to the
1.Translated by J. Mann and Th. Zacharias, 2nd Edition, London, 1882, p. 287.
2.P. 139.
3.Ibid., p. 237.
Page 369
Mauryan emperor", got investitured as king with the help of "various essentials" despatched by Aśoka for his "formal consecration" and finally accepted Buddhism from "Aśoka's son (or brother) Mahendra, who had become a monk" and "was sent to Ceylon as a Buddhist missionary".1 We shall be asked: "Do you doubt that Devānampiyatissa lived in the post-Alexandrine age or that he was associated with Aśoka? If you do not, Aśoka must be a post-Alexandrine."
If, as Paranavitana argues, it is legitimate to accept only the traditional Ceylonese chronology, Devānampiyatissa's reign-period 308-7 to 268-7 B.C. immediately puts out of court any association with Aśoka who mounted the throne in c. 269 B.C. by the modern time-scheme and who became a Buddhist after the Kalinga war which took place in the 8th year of his reign. Even by Geiger's chronology the situation hardly improves. For, Aśoka's inscriptions supply no reference either to King Devānampiyatissa or to Mahendra. And the whole story of their relationship may be included among the several points of "palpable and irreconcilable difference between the Ceylonese Chronicles and the information supplied by Aśoka's inscriptions". In that case, may not these Chronicles have foisted on Aśoka's reign Indo-Ceylonese events belonging to the reign of some other Indian king?
A mistaken association of a Ceylonese with an Indian king can be shown as quite possible to the Chronicles. There is the instance of the transfer of a Buddha-relic to Ceylon from Dantapura in the Āndhradeśa by permission of the king of a Nāga country called Manjerikā near the Diamond Sands, i.e., Kanchi. The Buddhist traditions of both Ceylon and Siam attest to it. But, although Siam has adopted the Ceylonese Buddha Era, its dating of the transfer differs widely. "The epoch to which the two traditions refer," says Sircar,2 "are irreconcilable. The Ceylonese tradition gives the date as 157 B.C., while the Siamese tradition gives A.D. 310-313." There is a gap of 467-70 years between the two chronologies. In Geiger's scheme, of course, the dates are somewhat later, but the gap is reduced only by 60 years. And, in either view of Ceylonese history, it is Dutthagāmanī who is king of Ceylon at the time whereas by the Siamese dating it is Sirimeghavanna. Historians
1.Ibid., pp. 236-37.
2."The Early Pallavas", The Journal of Indian History. August 1957, p. 183.
Page 370
accept the Siamese dating. Thus both Paranavitana1 and Sircar2 tell us that in the reign of Sirimeghavanna the Buddha-relic came to the island from Dantapura. The Chronicles of Ceylon have matched the Indian Nāga king with the wrong Ceylonese monarch. The converse of this would be their matching a Ceylonese monarch with the wrong Indian king - Devānampiyatissa with Aśoka. And we should not be surprised if the substitute for Aśoka were divided from the Mauryan emperor by several hundreds of years just as Dutthagāmanī was from Sirimeghavanna.
Only one question needs to be answered in order to place all these rational possibilities beyond contradiction. Do Aśoka's terms "Tambapamnī" (R.E. II) and "Tambapamnīya" (R.E. XIII) indicate Ceylon, which the Chronicles call Tambapamnī as well as Lankā and Sihala and which Megasthenes calls Taprobanè?
Aśoka's "Tambapamnī" and "Tambapamnīya"
Two scholars - Vincent Smith3 and B. A. Saletore4 - have opined that Aśoka was referring not to Ceylon but to the river Tāmraparnī which watered the southernmost region of India, the present Tinnevelly district. Barua5, taking stock of Smith's position apropos of the Aśokan word "Tambapamnī", pronounces: "The Tambapamnī of R.E. II may indeed be taken to stand both for the river Tāmraparnī and for the Tāmraparnīs as a people and their territory. But the Tambapamnīya of R.E. XIII stands certainly for the Tamraparnyas as a people and their territory. And all that R.E. XIII contains is nothing but a restatement of what Aśoka has said in R.E. II. The question still is: where to place the Tamraparnyas and their territory?"
Thus Smith and Saletore are technically in error in speaking exclusively of a river; yet the question they have raised by looking to South India instead of to Ceylon is a live one and concerns a riverine region of the mainland and that region's inhabitants as opposed to the island across the sea and the inhabitants of the island.
1."History of Ceylon", A New History..., pp. 257-58.
2."Ceylon", The Classical Age, p. 284.
3.The Indian Antiquary, XLVIII, pp. 48-49; Aśoka (3rd edition), p. 162.
4.Indian Culture, Vol. I, 1934-1935, pp. 669-672.
5.Op. cit.. Part I, p. 112.
Page 371
Smith's argument is based ultimately on chronology.1 In terms of the latest date-calculations by historians we may formulate it as follows. Aśoka's relations with Ceylon did not begin until after the accession of Devānampiyatissa who despatched an embassy to him and to whom Aśoka sent coronation-presents which were followed soon after by the Buddhist mission under Mahendra. As we saw, the date accepted for Devānampiyatissa's accession is, according to Geiger and his school: 247 B.C. (Smith says: "about 251 B.C.") But if five Greek princes - Antiochus, Ptolemy, Antigonus, Magas and Alexander - were also recipients of Aśoka's dhamma-dutas, the Rock Edict in which they are all mentioned must have been engraved when they were alive together. "Of these," writes Mookerji,2 "the date of the fourth king and the identity of the fifth are somewhat uncertain. According to some scholars Magas of Cyrene died about 250 B.C., while others push back the date by at least 8 years.... As regards Alexander, there were two contemporary rulers of that name, one in Epirus (272-c. 255 B.C.) and the other in Corinth (252-247 B.C.). But if we accept the earlier date of Magas, Alexander can only refer to the king of Epirus. Thus all these kings were jointly alive up to 258 B.C. in or before which one of them died. If the news of his death reached Aśoka two years after the event in, say, 256 B.C., R.E. XIII, which contains this reference and is stated to have been issued in the 13th year of his coronation, could not have been issued later than 256 B.C...." As R.E. XIII is but a restatement of R.E. II, the same must hold a fortiori for the latter. Even if we make Magas die in 250 B.C., the two R.E.s cannot be put later than 248 B.C. Devānampiyatissa is still left out in the cold by the latest calculations. And, as the majority of scholars favour Alexander of Epirus who died in c. 255 B.C., we can take 253 B.C. as the latest date even if Magas died in 250 B.C. This means that six years are yet to go before Devānampiyatissa can appear on the Aśokan scene. (By Smith's date, Devānampiyatissa is too late by a couple of years.) Chronologically, Aśoka's Tambapamnī and Tambapamnīya can never be Ceylon.
Within the conventional framework of Aśokan history, Smith can lead to an extremely credible case. If one opts for the traditional date of Devanarhpiyatissa (308-268 B.C.), Ceylon dis-
1.The Oxford History of India, 3rd ed., p. 121, fn. 2.
2."Aśoka, the Great", The Age of Imperial Unity, p. 88.
Page 372
appears all the more along Smithian lines. Aśoka turned to Buddhism after the Kalinga war which took place in the eighth year after his coronation and it was subsequent to this conversion that he sent out his dhamma-missions and could have been asked by Devānampiyatissa to consecrate him and so Aśoka would have been able to carry out the consecration not before the period 267-261 B.C. since his accession is dated between 275 and 269 B.C. The period thus obtained would find Devānampiyatissa already dead for one to seven years.
Nobody seems to have given a proper answer to the Smithian argument. And the conclusion from it can be supported in a number of independent ways which have a value of their own and are in themselves convincing and would not be affected if fault could be found here. Barua1 lays before us a host of facts by which we can definitely rule out Ceylon.
In both the Rāmāyana (IV.41.18) and the Arthaśāstra (11.11) Tāmraparnī is associated with Pāndyakapāta (-kavāta) and is obviously a South-Indian river. The Rāmāyana (IV.41.15) also singles out the South-Indian Malaya mountain as the rocky landmark of the Tāmraparnī-region. The Mahābhārata (VI.6.56) does the same and the Mārkandeya Purāna2 too traces the river to this mountain. The Rāmāyana (IV.41.16-17) in addition speaks of dvīpa (island or islands) in the river Tāmraparnī, covered with beautiful sandal woods. Thus there was at least one Tāmraparnī-dvīpa in South India opposite Ceylon. And, in Ceylon, which the Chronicles of that country call Tambapannī, the Chronicle known as the Mahāvamsa (VII.39,43) refers at the same time to Tambapannī as a district, with a town of that very name as its capital. "From these facts," observes Barua, "one cannot but be led to think that Tāmraparnī (better Tamravarni from having copper-coloured sand-beaches), which was originally a riverine region in the southernmost part of South India,... came to denote afterwards, probably in about Maurya time, also the north-western sea-coast of Ceylon between the Nāgadīpa and the river Kalyānī, and ultimately the island of Ceylon."
Thus, unless a country outside India is specifically indicated, the logical inference from a mention of Tambapamnī and Tambapamnīya in the Rock Edicts is to a populated portion of South India,
1.Op. cit., pp. 113-115.
2.B. C. Law, Geographical Essays, p. 101 f.
Page 373
the primary Tāmraparnī region, including an island or islands in the river of this name.
The logical inference is all the more binding because, as Saletore points out, the Tāmraparnī river of South India is no obscure geographical entity like the Tāmbaparnī of Gōkarna, which is "a small stream that falls into the sea near Gōkarna in the North Kanara District". "The Tāmraparnī of the South was celebrated in Indian history. The Epics and the Purānas describe it;... Kālidāsa noted it; and it figured in the epigraphs of well-known Karnātaka monarchs... The Brihatsamhitā... mentions one distinctive feature of the river, viz. that pearls were obtained in it." The Rāmāyana (IV.41.17) also associates it with "the coast adorned with pearls"; the Arthaśāstra1 speaks of the pearls from the river at the place where it falls into the sea; and the Karnataka epigraphs2 mention "people purified by the Tāmraparnī".
Why then should Ceylon be at all thought of in Aśoka? Is there a specific indication of it? If Ceylon were meant, we would find Aśoka bringing in the invariable pointer that goes with it in all Indian writings: "wherever by Tambapamnī or Tāmraparnī is meant Ceylon," we learn from Barua, "the word dīpa (dvīpa) or 'island is associated with it." This holds true even for foreigners who drew upon Indian information: "Megasthenes and other Classical writers speak of the sea-girt island of Taprobanè". In one Indian instance,3 "the Nāgārjunikonda inscription, marked F by Dr. Vogel, Tambapamna (Tamraparna) is clearly distinguished from Tambapamnī-dīpa". This clinches the contention that no Indian reference to Tāmraparnī as the country of the Tāmraparnyas or as the people of that country can mean anything except the old equivalent of the present Tinnevelly district or else its ancient inhabitants. And, since Aśoka's edicts are Indian writings, their omission of the word dīpa makes them fall logically in line with all other mentions of Tāmraparnī by writers in India.
As there was an island in the Tāmraparnī-region of South India itself, even the phrase "Tāmraparnī-dvi/?a" need not always de-
1.Mookerji, Chandragupta Maurya and His Times, p. 337.
2.Epigraphia Camatica, IX, Dg., 5, p. 26; Dg., 39, p. 49.
3.Epigraphia Indica, XX, p. 22: Tambapamnīdīpa-Pasadakanam theriydnam Tambapamnakanam suparigahi. Barua's translation:"...dedicated to the Theriyas who were converters of the Island of Tāmraparnī (and other countries named) (and those who were Tāmraparnīyas)" {op. cit., p. 115, fn. 2).
Page 374
note Ceylon, though generally it must. But, in the absence of the island-qualification, it is impossible that anybody in India writing of Tāmraparnī could have Ceylon in mind.
At most we can argue that the district called Tambapannī in Ceylon may have been intended. But, when the famous Tāmraparnī-region of South India is already there before this district and when this district has for an Indian a much smaller historical importance, we cannot have the slightest doubt that Aśoka referred to the country or the people of India's southernmost part.
Saletore, concentrating on the river Tāmraparnī and its pearls, is led even to question whether Megasthenes could have meant Ceylon. Saletore says: "... it was this same Tāmraparnī thus associated with pearls which Megasthenes had in view when he wrote the following: 'that Taprobanè is separated from the mainland by a river, that the inhabitants are called Palaiogonoi, and that their country is more productive of gold and large pearls than India.' " The value of Saletore's scepticism is vitiated by the fact that'Megasthenes,1 contrary to McCrindle's translation and interpretation, does not write of Taprobanè being separated from the mainland by a river but refers merely to its division by a river, and the full context of Pliny (VI.e.22[24]), where Megasthenes's report occurs, mentions that "the age and achievements of Alexander the Great made it clear that it (Taprobanè) is an island" and Pliny also mentions from "old writers" "the sea between the island and India". No, Megasthenes's "sea-girt island" is certainly Ceylon. Although the mention of the river and of the large pearls makes it highly probable that Megasthenes is mixing up two accounts, one about the pearl-producing Tāmraparnī with its island or islands and the other about the Tāmraparnī-island across the sea, we cannot blot Ceylon from his view.
However, there is no reason to put Ceylon from Megasthenes into Aśoka's edicts: it is only our present chronology, dating Aśoka after Megasthenes, that suggests for the former's edicts the latter's Taprobanè. And surely this chronology on top of the idea of integrating history with the Ceylonese Chronicles induces Barua to assume Aśoka's connection with Ceylon in the teeth of all the literary and epigraphic data gathered by himself and pointing in the opposite direction. The same two factors compel Smith, for all
1. Majumdar, The Classical Accounts of India, pp. 345-346.
Page 375
his anti-Ceylon reasoning from the time of the edicts, to talk of Aśoka's relations with Devānampiyatissa.
The Only Aśokan Argument for Ceylon
Actually, just one single argument for Ceylon has been built directly on Aśoka's edicts. Controverting Smith's position that the Girnār text "ā Tambapamnī" indicates the river Tāmraparnī in Tinnevelly and not the island of Ceylon, Raychaudhuri1 reminds us that this phrase in R.E. II comes after the mention of Ketalaputra (Keralaputra) and not that of Pāndya. He urges that the expression, "Ketalaputo as far as the Tāmraparnī" is hardly appropriate, because the Tāmraparnī is a Pāndyan river. He concludes: "We, therefore, prefer to take Tāmraparnī to mean Ceylon."
This argument is entirely off the mark on several counts. Smith's argument, as stated by Raychaudhuri2 himself, is: "Dr. Smith lately took the word [Tambapamnī] to mean not Ceylon but the river Tāmraparnī in Tinnevelly." According to Raychaudhuri, Smith has not spoken of "a Pāndyan river". How then can Raychaudhuri introduce a concept which is not involved in the discussion and, on the strength of it, refute his opponent? If we begin by calling the Tāmraparnī a river of the Pāndyas, naturally what occurs after the Keralaputras in R.E. II cannot be this river. Smith too must have known a thing so obvious. And there is no indication in Aśoka that Aśoka considered the river Tāmraparnī Pāndyan. What he gives us are just two expressions: the one in R.E. XIII (ā Tambapamnīya) comes after a mention of the Pāndyas, the other in R.E. II (ā Tampbamni) occurs after a mention of the Keralaputras. In Aśoka there is nothing Pāndyan about Tambapamnīya any more than there is anything Kerala-putran about Tambapamnī. Aśoka in fact permits us even less to speak of Tāmraparnī as a river of the Pāndyas than to speak of it as a river of the Keralaputras; for, the word "Tambapamnī" which he employs after mentioning "Keralaputra" is really the one allowing the sense of a river over and above the sense of a people and their territory whereas the word "Tambapamnīya" after "Pāndya" can only mean a people and their territory. To talk, in the Aśokan context, of "a Pāndyan river" is to sidetrack the controversy,
1.The Political History of Ancient India, p. 331.
Page 376
prejudge the issue and demolish a dummy.
Secondly, Raychaudhuri has proceeded on the same wrong line as Smith and Saletore, taking the issue to lie between a South-Indian river and the island of Ceylon. As Barua has stressed, the "Tambapamnī" of R.E. II and the "Tambapamnīya" of R.E. XIII must mean the same thing. The latter, even if derived from a river's name, cannot denote a river: "it stands certainly for the Tamraparnyas as a people and their territory." Hence "Tambapamnī" too must be accepted in the same denotation, no matter if it allows the "river"-meaning as well as this. Aśoka's term in R.E. XIII renders Raychaudhuri's discussion of a river - be it Pāndyan or non-Pāndyan - absolutely irrelevant and inaccurate. No legitimate conclusion to the island of Ceylon can proceed from it. Raychaudhuri has pursued an altogether false issue.
The true issue is: whatever be the people and the territory meant by Tambapamnī and Tambapamnīya, how does Aśoka relate them geographically to the Keralaputras and the Pāndyas? Since in R.E. II they come after the former and in R.E. XIII after the latter, Aśoka must be putting them - for all practical purposes - equally to the south of the Keralaputras and of the Pāndyas: he intends them to be an entity distinct from both Keralaputra and Pāndya and to be geographically related to both in the same way. This entity may be extreme South India or it may be Ceylon. All depends on whether the one or the other permits the sort of geographical relation Aśoka implies.
Here the deciding factor will be the view we take of the South-Indian river Tāmraparnī. And here we reach the third count on which Raychaudhuri's argument is entirely off the mark. If this river is assumed to be Pāndyan, the whole southernmost portion of India opposite Ceylon gets practically covered by the kingdom of the Pāndyas. For, this river rises in the Western Ghats and flows due south-east. The land for the Keralaputras is just the strip of coast to the west of the Ghats, a strip completely cut off from the neighbourhood of Ceylon. Perhaps, with the Tinnevelly district in the hands of the Pāndyas, we have little reason to keep South Travancore and Cochin out. From coast to coast the Pāndyas would be dominant. In any case, it would be impossible to speak of Ceylon after mentioning Keralaputra: the kingdom of the Pāndyas would stand right between and cut the two apart. And then the geographical order can never be as in R.E. II: "Choda-Pāndya,
Page 377
Satiyaputra-Keralaputra, as far as Tambapamnī." It will have to be: "Satiyaputra-Keralaputra, Choda-Pāndya, as far as Tambapamnī." Thus the very assumption made by Raychaudhuri defeats his own conclusion.
Only if the river Tāmraparnī is not Pāndyan we can have the geographical order of R.E. II just as validly as the one in R.E. XIII: "... Choda-Pāndya, as far as Tambapamnīya." But, if the river is non-Pāndyan, it must be in a region of its own, with an independent people about it. In that case, we do not get at all to Ceylon. The Tamraparnyas and their territory, watered by the South-Indian river Tāmraparnī, stand in southernmost India between Ceylon and both the Keralaputras and the Pāndyas, with the same geographical relation to these two South-Indian peoples - a relation exactly answering to what Aśoka implies. Thus Raychaudhuri's conclusion is again falsified, and southernmost India is proved to be visioned in Aśoka's Tambapamnī and Tambapamnīya.
The River Tāmraparnī and the Pāndyas
Here the sole rearguard action open to the school of Raychaudhuri would be his challenge:1 "Even those who prefer to see in the passage a reference to a kingdom in the Valley of the Tāmraparnī river, have to prove that such a kingdom did exist in the Maurya age apart from 'pāda' [=Pāndya] and Taprobanè, and to explain the particular way in which it is mentioned in Edict II." We may add a further possible protest: "Aśoka would be quite isolated if we understood him to speak of a South-Indian Tāmraparnī-region distinct from the Pāndya country. He gets no support from anywhere - and consequently his Tāmraparnī-region cannot be South-Indian."
But, even if Aśoka were to be quite isolated, we should have to weigh two things: his alleged isolation on the one hand and on the other the impossibility of speaking of Ceylon after Keralaputra when the Pāndya country holding a South-Indian Tāmraparnī river would stretch"right up to almost the west coast and divide Ceylon completely from Keralaputra. Against this impossibility can we say it is equally impossible for that river to have been in a country
1. Ibid., p. 331, fn. 2.
Page 378
which was non-Pāndyan? After all, geographical positions change with time: certain parts belong once to one kingdom and then to another and even become independent. S.N. Majumdar1 remarks on the Kerala and Pāndya territories: "It is needless to add that the boundary varied from time to time; and thus Travancore and Cochin were sometimes in Kerala, and sometimes in Pāndya." It is perfectly on the cards that in Aśoka's day the Tāmraparnī-region may have belonged neither to the Keralaputras nor to the Pāndyas.
In fact, Barua believes that Aśoka does receive support from a couple of sources. He criticises Raychaudhuri2 when the latter, before discussing Aśoka's Tāmraparnī, chalks out the extent of the Pāndya country, saying: "The Pāndya country corresponded to the Madurā, Ramnad and Tinnevelly districts, and perhaps the southern portion of the Travancore state... The rivers Tāmraparnī and Kritamāla or Vaigai flowed through it." Barua3 comments: "But it would seem that there was a separate Tamraparnya territory in the southernmost part of the Deccan below those of the Pāndyas on the east and the Keralaputras on the west - a fact which is well attested by Hwen Thsang." Barua4 directs us to Beal's translation of the Chinese pilgrim in Buddhist Kingdoms (II, p. 230) "where the country of Malayakūta, i.e. the South Indian Tambapamnī, is placed opposite Simhala and below Drāvida or Southern Chola. In explanation of "Drāvida", we may quote Barua5 saying: "The pilgrim's Chola and Drāvida constitute together the territory of the Cholas, better the Cholas and Pāndyas..." If Drāvida can include the Pāndyas, as it should if the Cholas are already listed separately and only their southern section falls under the other term, Malayakūta may well be a distinct Tamraparnya territory. We know from the Mārkandeya Purāna, the Rāmāyana and the Mahābhārata that the Malaya mountain is the source of the Tāmraparnī river and the main landmark of its region. Hiuen Tsang also refers to "the Sandal-producing Malaya mountain"6 as being in Malayakūta. The very name "Malayakūta" is approached
1.Cunningham, Ancient Geography of India. Notes, p. 743.
2.Barua is quoting from the 1932 edition. The edition of 1950 has a slightly different wording which yet changes nothing of the essential substance.
3.Op. cit.. Part I, p. 111.
5.Ibid., p. 106.
6.S. N. Majumdar, op. cit., p. 741.
Page 379
the nearest by "Malayakoti" which, as S. N. Majumdar1 informs us, the commentator on the Arthaśāstra gives as a synonym of Kautilya's "Pāndyakavata" (associated with his "Tāmraparnī"). Of course, Pāndyakavata indicates "the door of the Pāndyas", the entrance of the Pāndya country, implying a close historical connection between this country and the river Tāmraparnī. Nobody can deny the connection, but it may be significant that Hiuen Tsang should employ a name which does not remind us of it. Had the Pāndyas lost their prominence in his age so far as the Tāmraparnī-region was concerned?
The possibility of an affirmative answer inheres in the fact that this region, although mostly under the Pāndyas, was one of several chieftaincies and that the Pāndyas ruled it by an overlordship and not as an area indissolubly integrated with their kingdom. We learn from S. Krishnaswamy Aiyangar2 that their empire included the chieftaincies of (1) Aay round Podyl hill in the Western Ghats south of the Palghat gap and west of Tinnevelly, (2) Evvi round about the port of Korkai in Tinnevelly and (3) Pehan round the Palnis. In Ptolemy's Geography3 (c. 140 A.D.) we get not only the "Land of Pandion" with its "Cape Kory" (the promontory in the island of Rāmeshwaram) and its "royal city Modoura" (Madura) but also two neighbouring "countries" in which we may recognise two of Aiyangar's "chieftaincies": "country of the Aioi" with "Kottiara, the metropolis" (in South Travancore) and "country of the Kareoi" with "Kolkhoi, an emporium" (Korkai, at the mouth of the river Tāmraparnī) and "the Kolkhic Gulf, where there is the Pearl Fishery". The Periplus of the Erythraean Sea4 which is three quarters of a century earlier than the Geography, speaks also of "Colchoi, where the fishing for pearls is carried on", but it ends: "King Pandion is the owner of the fishery." The Geography has no explicit mention of Pandion's ownership of the emporium of Kolkhoi. In any case, a chieftaincy, whether called Evvi or Kareoi, could at any time become independent, and there should be nothing incongruous if in Aśoka's day a chieftaincy called Tāmraparnī or Tamraparnya were independent to the south of the Choda-Pāndyas and the Satiyaputra-Keralaputras.
1.Ibid., p. 740
2.Beginning of South Indian History, pp, 126-128.
3.VII. 1.9, 10, 11.89.
4.59.
Page 380
Barua1 claims that such independence at some time is suggested by the Mahābhārata:2 "The Great Epic distinctly speaks of Tāmraparnī as a country south of Pāndya, which in some contexts is included in Drāvida, and locates in it the Gōkarnatiratha and the hermitage of Agastya and his disciples."
A Clue from the Rāmāyana
The Rāmāyana, associating, like the Arthaśāśtra, the Tāmraparnī river with Pāndyakapata (-kavāta) would seem to place it in Pāndyan territory. But its geographical vision of South India is precisely the one which would appear to settle for good the problem whether Aśoka's Tāmraparnī is Ceylon or no. In our discussion of his Parirhdas we quoted from the Rāmāyana. This passage has a collocation of countries remarkably resembling the list of southern territories provided by the edicts - those of the Āndhras, Parirhdas, Chodas (Cholas), Pāndyas, Satiyaputras, Keralaputras and Tambapamnīs (or Tambapamnīyas). The Rāmāyana passage (IV.41.12) runs
Andhrāmścha Pundramś cha Cholām Pāndyamś cha Keralān and is followed soon after by a description (IV.41.15) of the region of the Tāmraparnī river. The sole difference, as Barua3 notes, is: "we have the Pundras instead of Aśoka's Parirhdas, and the Keralas in place of Aśoka's Satiyaputras and Keralaputras." The general succession is too obviously the same to allow the thought of Ceylon even though the Keralas come between the river-region and the Pāndyas. The very word that Aśoka uses after the Keralaputras in R.E. II is the same as the Rāmāyana's after its Keraldn: Aśoka has Tambapamnī, which can stand not only for a people and their territory but also, like the Tāmraparnī of the Rāmāyana, for a river. It is as if Aśoka had the Rāmāyana at the back of his mind or the Rāmāyana echoed Aśoka.
The Ceylon-Mission in the Chronicles
From various directions we converge on the same conclusion.
1.Op. cit.. Part I, p. 115
2.III.88.13-18: Kumāryah kathitāh punyāh Pāndyeshveva nararshabha, Tāmra-parnin tu Kaunteya kirtayishydmi lach chhrinu. Cf. Ibid., III. 118. 3, 4, 8.
3.Op. cit.. Part I. p. 79.
Page 381
And with Ceylon removed from the history and geography disclosed by the Aśokan edicts we begin to be struck by some peculiarities of the history and geography of the Buddhist missions disclosed by the Ceylonese Chronicles. All the Chronicles1 put all the missions together except the one to Ceylon. The latter is said to have been despatched separately, a year after the rest. It seems to belong to a different context and appears as though tagged on here out of piety towards the Buddhist emperor and a sense of consistency with the other missionary activities in his reign.
The impression of a different context increases when we scrutinize the single other country than Lankā, which is taken to be foreign: Suvarnabhūmi, generally identified with Burma. Suvarnabhūmi belongs to a group of countries where the missions went a year before Mahendra went to Lankā. There is some incongruity here. How does a foreign country like Burma get dissociated from Ceylon and assembled with those that are clearly called "adjacent"? Barua2 balances against the name "Suvarnabhūmi" the name "Suvarnagiri" which is mentioned in the Mysore copies of Aśoka's Minor Rock Edict as the seat of an Aryaputra Viceroy. Could we not assume that the original name in the Chronicles was "Suvarnagiri" and that it was replaced afterwards by "Suvarnabhūmi"? Barua remarks: "According to the Samanta-pāsādikā and Mahāvamsa, Suvannabhumi was a country which bordered on a sea and which was under the sway of a ferocious rakkhasi, who was evidently its presiding female diety. The text of the Dīpavamsa, as it appears in Oldenberg's edition or in the Samanta-pasadika, preserves the earlier tradition which does not connect Suvannabhumi with any sea or ocean and represents it as a place inhabited by the Pisachas. None need be surprised, therefore, if Suvannagiri were the intended name instead of Suvannabhūmi."
What makes Barua's substitution extremely plausible is a fact which he3 mentions as having been noted by Vincent Smith apropos of the Buddhist missions in Aśoka's time: "... the propaganda in Lower Burma seems to have had little effect. The earliest form of Buddhism in that country, so far as definite evidence goes, was of the Mahāyana kind."
1.Dīpavamsa, VIII, 1 f, XII, 16 f; Mahāvamsa, XII, 1 f; Samanta-pāsādikā, I, pp. 63 f, 69 f.
2.Op. cit.. Part I, pp. 57, 325; II, p. 8.
3.Ibid., Part I, p. 325.
Page 382
Barua's substitution would make the first group of countries all of a piece, uniformly "adjacent", and throw Lankā into a totally contrasting category, rendering it still more suspect by isolating it geographically no less than chronologically.
Furthermore, the substitution creates the possibility - so far as Aśoka's own empire is concerned - of passing beyond the discrepancies between the Aśokan account of the dhamma-dutas and the Chronicles' story of missions, and bringing the places mentioned by the story into complete rapport with the find-spots of the edicts. Already a good deal of rapport has been obvious. Kandahār points to the Chronicles' "Yona country"; Shāhbāzgarhī and Mānsehrā to "Kāshmir and Gandhāra" of the Chronicles; Kālsī to their "Himālaya country"; Girnār, Sopāra and Kophal to their "Apar-antaka" and "Mahārāshtra"; Hyderabad and Northern Mysore to their "Vanavāsi and Mahishamandala". "Suvarnabhūmi and Tāmraparnī," says Barua1, "are the two places that are not explicable by the find-spots of the hitherto discovered inscriptions. Once we assume that Suvarnagiri2 was the earlier name on the list replaced later by Suvarnabhūmi, the case for both the places can be met by the inscriptions in which they find mention, one as the official headquarters of the most southern province, and the other as the most southern of the independent countries in South India."
We can now realise that the historical and geographical information in Aśoka's edicts is such as to leave no need or room for the extrapolation of a Ceylon-mission into his life-work from the Chronicles. So Devānampiyatissa of the 3rd century B.C. fades out from the Aśokan scene. And with his disappearance there remains not the shadow of an indication anywhere that Aśoka was contemporaneous with the Greek princes of the 3rd century B.C.
8
A final objection against our chronological changes may be raised apropos of Aśoka's own monuments. It may be alleged that they show distinct derivation from those of the Achaemenids of Persia who immediately preceded the age of Sandrocottus and that
2.This may be the same as Suvarnavati, which finds mention in the Hitopadesa, II. 1 as a city in South India: asti Dakshinapathe Suvarnavati noma nagari.
Page 383
therefore Aśoka must remain where he is at present situated in time.
The Achaemenids and the Aśokan Monuments
Some resemblances between Achaemenid art and Aśokan are understandable, but Havell and K. de B. Codrington1 have stressed so many essentially Indian characteristics in both subject and inspiration that the charge of direct derivation cannot be upheld. The resemblances should be explained by the admittedly common traditions inherited by Indian and Irānian cultures. The hypothesis of direct influence occurs only against a background of the current supposition about Sandrocottus.
Even the palace, excavated at Pātaliputra and recognised as the one described by Megasthenes and said by him to surpass those of Susa and Ecbatana, is not taken as directly influenced by those earlier royal dwellings abroad. Spooner suggested that a close parallel to the excavated palace may be perceived in the Achaemenid Hall of Hundred Columns at Persepolis. But his views have not been generally accepted.2
Raychaudhuri3 goes pointedly further in connection with the Pātaliputra palace. After quoting the phrase that with this palace "neither Susa, nor Ecbatana, can vie", he continues with the side-remark found in the Greek reports: "for methinks, only the well-known vanity of the Persians could prompt such a comparison", and adds his own footnotes: "The statement should be remembered by those modern writers who find traces of Persian influence in Maurya architecture." By "Maurya", of course, he means "pertaining to the dynasty of Sandrocottus".
As for the formula of Aśoka's edicts, "Thus saith King Priya-drasi," it may seem an echo of "Thus saith King Darius", and again a direct influence of the Achaemenids may be asserted. But there is really no ground for the assertion. One of the set phrases prescribed by Kautilya in his Arthaśāstra for what is called a royal prajna-pana-śāna ("writ of information") is evam aha ("thus saith").4 So Aśoka's formula is definitely of indigenous origin.
1."Aryan Rule", Ancient India (London, 1926), I. pp. 18, 19.
2.The Age of Imperial Unity, p. 486, fn. 2.
3.The Political History of Ancient India, Fifth Edition, p. 275 with fn. 3.
4.Mookerji, Aśoka, p. 108, fn. 3.
Page 384
And in any case it is such as the literary sense of any country, coupled with the high status of the speaker, would independently invent. Nobody has Suggested that if Darius had not set up his inscriptions Nietzsche could never have entitled his masterpiece Also Sprach Zarathustra (Thus Spake Zarathustra) or far earlier the Buddhist anthology's name ltivutaka (Thus Spake the Buddha) would never have seen the light of day.
Indigenous origin can be thought of, too, for the Aśokan words lipi and dibi, meaning "writing", for his edicts. They have been compared to the Persian word for "inscription", but the Aśokan terms have a clear Indian parentage. Pānini (IV, 1.49) has used the term Yāvānānī which has been taken to imply lipi and mean "writing of the Yavanas"; he (III, 2.21) has used as well the forms lipikara and libikara for the scribe, and Sanskrit has the word dibira, for .lipikara, from the root diba or dīpa, to "write".1
An Advanced History of India, edited by Majumdar, Raychaudhuri and Datta, though unaware of the objections to the current chronology and disposed to see several influences emanating from Persian and Greek officials in the Indus-provinces during the Achaemenid rule and the post-Alexandrine period, has yet declared2 about these officials: "Whether some important features of the architecture of the Maurya period and certain phrases used in the Aśokan edicts are... to be attributed to their enterprise, is a highly debatable question."
Wood-work and Stone-work at Pātaliputra
A subtler argument to make Aśoka, in relation to the architecture of Pātaliputra, post-Alexandrine may be formulated: "According to Megasthenes, cities on the banks of rivers and in other low-lying spots were built of wood to ensure safety against floods. So Pātaliputra, at the confluence of the rivers Ganges and Son, must have been all of wooden construction in Megasthenes's time. But the royal hall of pillars excavated by Spooner is of stone. And S. K. Saraswati3 has well said: 'Fragments of stone pillars, including one nearly complete, with their round tapering shafts and smooth polish indicate that the great Aśoka was responsible
1.Ibid., p. 136, fn. 5.
2.P. 19.
3."Art: Architecture", The Age of Imperial Unity, p. 486.
Page 385
for the construction of the hall, or at least for the stone columns which replaced the earlier wooden ones.' Now, if Megasthenes saw only wooden pillars, Aśoka who must have built the stone ones similar in general to his own stone columns came after him and cannot be shifted away from the third century B.C."
We may reply as follows.
The reference by Megasthenes to the wooden construction of river-side cities can pertain only to the basic erections exposed to the water rising at flood-time. It cannot apply to everything within a city. A massive wooden palisade protecting Pātaliputra has actually been dug up, fully bearing out Megasthenes's mention of a huge wall of wood around Palibothra, with towers at regular intervals. With regard to the interior of the city and especially to the royal palace, we have no indication from Megasthenes of the exclusive use of wood. Aelian1 reports from Megasthenes: "The palace is adorned with gilded pillars clasped all round with a vine embossed in gold, while silver images of those birds which most charm the eye diversify the workmanship." There is no reason to consider these pillars to have been wooden. We must keep in mind that, according to Aelian,2 "neither Memnonian Susa with all its costly splendour, nor Ecbatana with all its magnificence" could vie with the palace of Sandrocottus in imperial Palibothra. Surely, those splendid and magnificent cities of Persia were not built all of wood, least of all were their palaces wooden and nothing else. To surpass them, Palibothra and its royal abode must have been a mighty glory of stone-work. And the proper meaning of Megasthenes in the matter of wood is brought out by what Saraswati3 himself records of- the excavated palace: "The palace appears to have been an aggregate of buildings the most important of which was an immense pillared hall supported on a high substratum of wood." The discovery of stone-work in the Pātaliputra-palace, on top of wood-work, need not fix Aśoka to a period subsequent to Megasthenes.
What is more: even if Aśoka belonged to such a period, the stone-work of Pātaliputra would not automatically be Aśokan. For, the very nature of the Aśokan monuments prompts us to go farther back for their stone-work than the period of their inscriptions.
1.Mookerji, Chandragupta Maurya and His Times, p. 104.
3.Op. cit., p. 486.
Page 386
K. A. Nilakanta Sastri1 has written: "It has been suggested very plausibly that some of these pillars, specially those bearing animals sacred to the Hindu gods - e.g. the bull of Shiva, the elephant of Indra, the garuda of Vishnu (Lauriya Araraj) - may be pre-Aśokan in origin... Aśoka might have used these columns for the propagation of Dharma which was marked by his readiness to bring together all that was best in every faith." Nor is the pre-Aśokan origin of some of the pillars a fair conjecture alone. Aśoka himself (Minor Rock Edict in Rūpnāth Text and Pillar Edict VII) declares that stone columns were already there from the past in his time. And Mookerji2, with fine logic, brings us full circle with his remark: "Some of the polished pillars were rightly ascribed by Aśoka to his predecessors who could build a great city, and the most gorgeous palace of the East."
So, in any case, the excavated hall could very well have been there in the day of the Greek ambassador as a wonderful non-Aśokan monument and has no determinative bearing on Aśoka's date. Since polished stone pillars existed before Aśoka, nothing points to the building of this hall by him. And, even if he did install the pillars here, the existence of stone-structures in the period of Megasthenes would allow us to think in the light of our Purānic chronology that what had been erected by Aśoka was according to a practice coming down to this period from more ancient times. Their presence as such lays no compulsion on us to make Aśoka live after Megasthenes or to identify Sandrocottus with Chandragupta Maurya rather than with the founder of the Imperial Guptas.
The Aśokan Monuments, Achaemenid Art and the Middle East
On the strength of Minor Rock Edict in Rūpnāth Text and Pillar Edict VII, we may turn the tables on Achaemenid art. For, we have no means of deciding how far back in time the stone-columns go. We can quite conceive pre-Achaemenid Indian polished pillars being the inspiration to Achaemenid art through Darius's possession of "Ga(n)dara" and "Hi(n)du". And, if their presence in Megasthenes's day takes away all possible cause to render Aśoka post-Megasthenes, we may trace Achaemenid art to Mauryan
1.History of India, Part I (1950), p. 92.
2.Aśoka, p. 94.
Page 387
influence itself from an earlier age instead of vice versa.
A clear index to an age like our 950-914 B.C. for Aśoka can be found also in what Jairazbhoy has to say. He is anxious to derive Aśokan art from Achaemenid as well as Hellenic, but the matter he presents actually points us away from both of the latter. After making his comparisons with Persian features, he1 writes: "We may interpose a note here that the Persian ornaments of the high base and double wreath of sepals at the capital are themselves now known to be of western, ultimately of Egyptian, origin, while the double spiral volutes in the Achaemenid columns have been traced back from the double and triple volutes on the glazed facade of Nebuchadnezzar's palace at-Babylon, to those in the capitals of pilasters in Cyprus and Megiddo."
This admission carries us at once to a remoter time than that of the Achaemenids for parallels to Aśokan art. And, with Aśoka dated to 950 B.C., we get new terms of comparison for the ornaments of his columns - Egypt, Babylon, Cyprus, Megiddo come into view. If those ornaments are at all due to foreign influence, we have sufficient pre-Achaemenid factors and all we need to make them operative is to find relations between India and the Middle East before the Achaemenids conquered "Ga(n)dara" and "Hi(n)du". These relations we have already found when discussing Assyria and India's immediate neighbourhoods.
Apropos of a pair of stone figures of winged lions which were excavated at Patna (old Pātaliputra) and whose purpose, according to Jairazbhoy, "may have originally been to support the throne in the Mauryan palace" as in Persia under the Achaemenids - and apropos of a throne with a lion leg found at Amarāvatī in the first century A.D. but considered by Jairazbhoy as having been "introduced in the Mauryan period from Persia", Jairazbhoy adds: "the theme was endemic in the Middle East much earlier." In nothing here concerned is Achaemenid Persia necessary - provided we can date the Mauryas to a pre-Achaemenid epoch and show a connection in that epoch between the Middle East and India.
It is again with the Middle East, though under Persian rule, that we have analogues to the arrangement of animals and sun-wheels seen on Aśoka's Sārnāth Pillar. Jairazbhoy writes: "On the abacus of the Sārnāth capital the animals consist of a horse, a humped bull, a lion, and an elephant, and they are following the disks
1. Foreign Influence in Ancient India, pp. 44-6, 58.
Page 388
round in a circle. Now we have precisely these animals - horse, winged bull, and lion - revolving round a hub in the Nippur tablets from the reigns of Artaxerxes and Darius II (464-405 B.C.). The resemblance would seem to be fortuitous, but for the fact that here too a solar idea is probable, since in other tablets the revolving beasts are simply horses, or in other words the team that drew the chariot of the sun. These tablets must have travelled widely since they were the documents of bankers, Mureshu and sons. We know in fact that among their business tablets are records of their dealings with certain Indian merchants, and there is a reference to a settlement Hi-in-da-ai."
Jairazbhoy suggests that the Aśokan pillar copied the Nippur tablets; but, if Aśoka cannot be fixed down to the middle of the 3rd century B.C., Mureshu and sons are quite likely to have received the design of animals and wheels from the merchants of Hi-in-da-ai who were aware of it from monuments of Mauryan rule in the later half of the 10th century B.C.
We may hark back to a passage we have already culled from Jairazbhoy while discussing alleged Buddhist influence in Ptolo-maic Egypt. It ran: "the emblem of a sun disk carved in relief on Buddhist monuments, for example, at Amarāvatī ultimately originates in Assyria, and similarly the so-called Buddhist trisula ornament is patterned on the winged solar disk of Assyria. Moreover, Mesopotamian sun pillars surmounted by disks (surin-nu) which stood at entrances to temples, or were borne by priests of the sun god, are prototypes of sun wheels on pillars at the Buddhist stūpas." The point we wish to stress is that Aśokan affinities are essentially with Mesopotamian symbolism. Whether derived from it or no, they direct our attention to a time before the Achaemenids and are in no way such as to make Aśoka post-Achaemenid.
As a final touch of irony to the usual comparisons of Aśokan art to Achaemenid we may listen to S. Chattopadhyaya:1 "It was believed that the lion-figures of the Mauryan age drew their inspiration from the lion of Hāmādān which was taken to be a work of the Achaemenid period. The theory must now be discarded, for it has been conclusively proved that the lion of Hāmādān belongs to the Arsacidan and Sassanid periods." The Arsacids
Page 389
start in 208 B.C. and the Sassanids in 224 A.D.1 They are post-Aśokan by any chronology.
The Aśokan Pillars, Hellenistic Influence, the Indus Civilization
Of course, the art of the Achaemenids is not the sole source conjectured for the Aśokan pillars. Jairazbhoy has a footnote on the other source. First he makes the statement: "In so far as the actual carving is concerned, the Aśokan lions by no means emulate the artistic conventionalization of those of the Achaemenids." The footnote2 goes: "If anything, Hellenistic influence is manifest in the Mauryan lions, in the treatment of cheekbones, moustaches, deeply embedded eyes and freely rendered manes, as others have already observed. Hellenistic rather than Achaemenid origin for the Mauryan lions has been suggested for stylistic reasons. (N. B. Ray, Maurya and Sunga Art, 1945, p. 42f.)."
How far the comparison with Greek art is valid we may decide best by a few references to A. L. Basham. We shall start with his general observations:3 "The capitals of Aśoka's columns, some of which were perhaps made before his reign, are the earliest important sculptures after those of the Indus cities."
Mark first the phrase: "some of which were perhaps made before his reign." It is on a par with Mookerji's assertion of the existence of polished pillars before Aśoka's time. With such pillars possible before Aśoka, Aśoka's own structures are set free not only from the post-Achaemenid world but also from the world of Hellenic sculptures.
Secondly, there is Basham's pointer to "the Indus cities". This pointer to the Harappā Culture is here in a chronological context, but it sets us thinking on many lines. If the Indus cities came to an end in c. 1750 B.C., as is now believed in some quarters, and if Aśoka lived round about 250 B.C., we get a huge gap between the works of Harappān art and the next "earliest important sculptures". We are led to ruminate: "How much more natural it could be if there were not such a wide hiatus in time!" Our 950 B.C. for Aśoka gets an indirect encouragement - all the more if the old
1.The Heritage of Persia by Richard N. Frye, pp. 208 and 236.
2.Op. cit., p. 46.
3.The Wonder that was India, p. 364.
Page 390
date for the end of the Indus cities - c. 1500 B.C. - still holds, as Sankalia1 has cogently argued.
But that is not all. Basham supplies us with a further cue. He2 starts with noting: "the famous lions of the Sārnāth column and the less famous but more beautiful bull of the column of Rampur-va are the work of realistic sculptors, owing something to Irānian and Hellenist tradition." However, immediately on the heels of subscribing to the current opinion, Basham makes the revealing observation: "Yet, if we did not know that the possibility of Western influence existed, we might suggest that the animal sculptures of the columns were those of a school directly descended from the engravers of the Indus seals, which show a realistic treatment very unusual for so early a civilization."
The full force of this observation will be realised if we juxtapose the two expressions: "... owing something to Irānian and Hellenist tradition" - "directly descended from the engravers of the Indus seals". Evidently, the more concrete and comprehensive relation is with Harappān art. Now we have not just a chronological context: we have the context of artistic quality and manner. Aśokan art and Harappān would go together - but for the time-gap we put between them, permitting "the possibility of Western influence". What Basham has said implies that only because we place Aśoka in the 3rd century B.C. we are prompted to trace Greek as well as Irānian influences. Transpose him to the second half of the 10th century B.C. and at once we shall connect him with the Indus seals and fully explain thereby whatever in his sculptures is not characteristic of post-Mauryan Indian art.
And Basham's reference to the Indus Valley Civilization is apt in more senses than one. For, this civilization has several features which, like the realistic treatment of the animals on its seals, are unusual for an art so early. Basham3 himself has the following comment on the striking bronze "dancing girl": "this young woman has an air of lively pertness, quite unlike anything in the work of other ancient civilizations." He4 also writes: "The red sandstone torso of a man is particularly impressive for its realism.
1.Prehistory and Protohistory of India and Pakistan (Deccan College, Poona, 1974), p. 283, col. 2.
2.Op. cit., pp. 364-65.
3.Ibid., p. 121.
4.Ibid., p. 20.
Page 391
The modelling of the rather heavy abdomen seems to look forward to the style of later Indian sculpture, and it has even been suggested that this figurine is a product of a much later time, which by some strange accident found its way into the lower stratum; but this is very unlikely, for the figure has certain features, which cannot be explained on this hypothesis." Lately Sankalia1 has written of "a fresh study": "it was definitely proved on technological grounds that the small sandstone torso from Harappā was Harappān. We can thus confidently say with Sir John Marshall that the Indus artist had anticipated the Greek artist by 2,000 years." In addition we may note the comment of Stuart Piggott:2 "the treatment of the heavy lines of the abdomen... is astonishingly similar to some work of Kushan date [c. 40-220 A.D. by the present chronology]."
All this shows that ancient Indian pieces of art could not only be unique in their own time in world-history but also seem of an age more than a couple of millennia later in the same country as well as carry the suggestion of a strong foreign influence from a western country's art that was not even born for the next twenty centuries. The mention of Greece itself as the western country is all the more eye-opening for our discussion of Aśokan art. And when the Aśokan sculptures can be compared in their realistic treatment to the Indus Valley art-pieces which, in Piggott's words,3 have "caused doubts to be thrown on their authenticity as third-millenium work", it is not at all necessary to situate Aśoka for his stone monuments in the post-Alexandrine era rather than in the second half of the 10th century B.C.
How very natural and necessary, on the other hand, it is to set him in that earlier age we may drive home to our minds by keeping before us another of Basham's clear-cut artistic judgments. Referring to Mauryan sculpture he4 pronounces: "the earliest sculpture of historical times... shows a generic likeness to that of Harappā."
Here we may appropriately digress a little to a subject we have already touched upon: the pillared palace excavated at Pātaliputra, which Spooner tried to parallel with the Achaemenid Hall of Hundred Columns at Persepolis. We have cited the general criticism
1.Indian Archaeology Today (1979), p. 102.
2.Prehistoric India (Pelican. Harmondsworth, 1960), p. 185.
4.Op. cit., p. 364.
Page 392
of the suggested parallelism. Now we may point out a possible Indian origin of such a structure.
Walter Fairservis, Jr.,1 describing the Harappān site of Mohenjo-daro, has dwelt on a structure "known to the excavators as the Assembly Hall". He2 writes: "Badly preserved, it is nonetheless one of the most striking monuments at Mohenjodaro. It consisted of a broad pillared hall opening principally to the north, i.e., towards the highest part of the site. Twenty rectangular pillars approximately five feet by three feet in size supported the roof. The pillars were arranged in rows of four with five pillars to each row." After detailing the rest of the important features of the building complex containing the pillared hall, Fairservis3 comments on this complex: "One cannot help but speculate... that it was constructed in response to a formality urged by religion or government. Was it indeed a place of assembly or perhaps a place of audience? Wheeler rightfully refers to the Achaemenid pillared hall of audience, the apadana, in this context, and such a comparison is certainly called to mind.4" Thus the Pātaliputra hall, seeming to resemble the Persepolitan apadana, needs really to look back to an original in the Indus Valley Civilization.
Perhaps we shall be admonished: "What about the recent digging at Pātaliputra which, with the NBP [Northern Black Polished] sherds, yielded a good number of polished sandstone pieces, including one large slab showing the typical Mauryan or Persian polish and bearing a palmette and head-and-reel pattern of appropriate type? Doesn't this fix Mauryan art to the NBP period which, by the C-14 test, covers two and a half centuries or so before the Christian era?"
First of all, we may answer that the palmette and head-and-reel pattern is no proof of Achaemenid Persian influence. N. R. Ray5 tells us after mentioning the debt which, in his view, Mauryan art owes in some details to Persia and Greece: "The twisted rope design, the acanthus leaf and palmette designs, etc. may, however,
1.The Roots of Ancient India: The Archaeology of Early Indian Civilization (George Allen & Unwin, Ltd., London, 1971), p. 245.
2.Ibid., pp. 245-46.
3.Ibid., p. 246.
4.Sir M. Wheeler, Early India and Pakistan (Frederic A. Praeger, New York, 1959).
5."Art: Sculpture", The Age of Imperial Unity, p. 509.
Page 393
have been derived from an older common West-Asiatic heritage."
Secondly, we may answer that, just as Aśokan art has strong antecedents in the art of the Harappā Culture which was many centuries earlier, so also can it be thought to have had a continuation in some respects. Whatever has high polish and fine finish typical of the Mauryan school need not belong to Mauryan times. To quote Basham1 again: "One beautiful figure, the 'Didarganj Yaksī, bears the distinctive brilliant polish of the school, but the treatment of the figure suggests that it is post-Mauryan." Its date is almost 200 years after Aśoka as at present dated.2 So the finding of the polished sandstone pieces together with the NBP sherds at Pātaliputra need not mean that they were Mauryan and that the Mauryas were ruling in the 3rd century B.C., to the middle of which these sherds at their earliest can belong. Room is left for the idea that Sandrocottus was Chandragupta I and that the Imperial Guptas who were aware of Aśoka's fine stone monuments (e.g., the Allāhābād Pillar where the inscriptions of Samudragupta and Skandagupta stand with Aśoka's own) encouraged Mauryan polish. The idea would be debarred only if something definitely sets the Guptas in the 4th-to-6th centuries A.D.
Now a last word in the sphere of art - in relation to another phrase from Basham:3 "the Mauryan school, with its high polish and fine finish." Will Durant4 quotes John Marshall as saying about the Harappā Culture's gold and silver bangles, ear-ornaments, necklaces and other jewellery: "so well finished and so highly polished that they might have come out of a Bond Street jeweller's of today rather than from a prehistoric house of 5,000 years ago." Not only the energetic realism in animal-depiction but also the drive towards high polish and fine finish of Aśokan art can be linked with the creative or productive movement of the ancient Indus Valley.
The Aśokan monuments in any aspect do not stand in the way of our reinterpretation of Aśokan data to fix the Guptas where the Mauryas are placed at present.
1.Op. cit., p. 365.
2.Ibid., pi. facing p. 105; also The Age of Imperial Unity, pp. 516-17.
3.Op. cit., p. 365.
4.The Story of Civilization, I. Our Oriental Heritage (Simon and Schuster, New York, 1954), p. 395.
Page 394
In terms of Aśokan evidence our case against the identification of Sandrocottus with the founder of the Mauryas is completed. By way of a supplement with a positive pointer towards making the founder of the Imperial Guptas the contemporary of Megasthenes we may hark back to our treatment of the Aśokan background as being incompatible with the background of the post-Alexandrine epoch in India. We indicated, among other things, the insignificance of Buddhism in the time of Megasthenes and stressed the prevalence of Krishna-worship in it. We also noted how the canonical Buddhist work, the Anguttara Nikāya which is post-Buddha but anterior to the Mauryas and which lists various religious cults is utterly silent on the Vaishnava and Bhagavata cult of Vāsudeva-Krishna. Aśoka's own inscriptions were found equally silent although they mention several other religious denominations. This is most surprising and must show that Aśoka no less than the Anguttara Nikāya must have come before the day of Megasthenes and Sandrocottus. Thus his grandfather Chandragupta Maurya moves in time still further back beyond the 4th century B.C., where he is usually placed. Now we may briefly glance at the history of the Krishna-worship which Megasthenes speaks of, and consider it in relation to the Guptas. This procedure will determine in a positive way the precise bearing of the Sandrocottus-era.
Sircar provides us with a general look at the development of Vaishnavism and Bhāgavatism.
The Historical Picture of Krishna-worship
The earliest reference to the Vāsudeva-cult "may be traced to the Ashtādhyāyī of Pānini... which offers the rule for the formation of the word 'Vāsudevaka' in the sense of 'a person whose object of Bhakti is Vāsudeva'.1 Although "Vāsudeva was probably deified at least partially... as early as the age of Pānini,... he may or may not have been regarded as the supreme god".2 Along with him was "his friend the Pāndava Arjuna", as seen from "Pānini's rule for the formation of the word Arjunaka in the sense
1."Vaishnavism", The Age of Imperial Unity, p. 432.
2. Ibid., p. 437.
Page 395
of a person whose object of Bhakti was Arjuna".1 Besides, it is only likely, not certain, that Bhakti "is to be taken in the sense of religious adoration", for "the word is also used in the Ashtādhyāyī in connection with cakes, and the possibility of the meaning 'fondness' is not altogether precluded".2 No more than a devotional tinge can be positively asserted in the attitude to Vāsudeva.
But by the 4th century B.C. at least, Vāsudeva is a full-fledged object of whole-hearted devotion as "is proved by the statement of Megasthenes that the Souresenoi, i.e. the people of the Mathura region, held Herakles in special honour; for there is no doubt that Herakles was the Greek analogue of Vāsudeva-Krishna".3
Originally "a hero of the Yādava clan", Vāsudeva-Krishna, when deified, was "styled Bhagavat".4 He was identified with "the Vedic deity Vishnu".5 "This was accomplished by the time the Bhagavadgītā was composed, and henceforth the Vāsudeva cult or Bhagavata religion was known also as Vaishnava dharma."6
"A Besnagar (old Gwalior State) inscription of the last quarter of the second century B.C. refers to a garuda-dhvaja (column surmounted by the figure of Garuda [Eagle] conceived as the emblem or vahana of Vishnu) raised at Vidiśa in honour of Vāsudeva, the deva-deva (the greatest god), by his Yavana or Greek devotee Heliodorus, an inhabitant of Takshaśilā in Gandhāra, who calls himself a Bhāgavata... Another inscription from Besnagar speaks of the erection of the Garuda column of an excellent temple (prāsādottama) of the Bhagavat (Vāsudeva).
"The Ghosundi (Chitorgarh District, Rājputāna) inscription of the first century B.C. refers to the construction of a pūjā- śilā-prākāra (stone enclosure for the place of worship, or better, an enclosure for the sacred stone called Śālagrāma believed to be typical of Vishnu as the Linga is of Śiva), probably styled Nārāya-navataka, by a Bhagavata performer of the Aśvamedha sacrifice, in honour of Sarikarshana and Vāsudeva who are called Bhāgavat, anihata (unconquered or respected), and sarvesvara (supreme lord)...
1.Ibid., p. 451, with fn. 1.
2.Ibid., 482, with fn. 2.
3.Ibid., p. 432.
5.Ibid., p. 435.
Page 396
"These epigraphs support what is already known from literary evidence as regards Vāsudeva's association with Garuda [Eagle], and therefore with Vishnu, with the Vrishni hero Śankarshana, as well as with Nārāyana, long before the birth of Christ. He is not called Krishna in the earlier epigraphic literature of India, but the use of Krishna as another name of Vāsudeva in such Works as the Mahābhārata, the Ghata Jātaka and the Mahābhāshya should probably be ascribed to a pre-Christian date."'1
We may observe that when Megasthenes associates not only Methora (Mathurā) but also Cleisobora with the Souresenoi's worship of Heracles we have in the second city-name the word "Krishnapura" and that the name "Heracles" can reduce to "Hari-Krish(na)". Thus, from the 4th century B.C. onwards to the birth of Christ, direct Krishna-worship is certainly in full maturity. It is also widely disseminated. "The Bhagavata religion, which originated with the Yādava-Sātvata-Vrishni people of the Mathura area, appears to have spread to western India and the northern Deccan with the migration of the numerous Yādava tribes".2 "The introduction of Bhāgavatism in the far south at a much earlier date is indicated by the relation of the Pāndyas with the Pandavas and Śūrasenas, alluded to in the confused stories narrated by Megasthenes about Heracles and Pandaia, and in the grammatical work of Kātyāyana, and by the name of the Pāndya capital Madurā, adapted from that of Mathurā, the original home of the Vāsudeva cult.'"3
Contemporary with the efflorescence of Bhāgavatism evident from the 4th century B.C. to the beginning of the Christian era we need to have kings quite the reverse in religious environment to Aśoka. No doubt, there is one Aśokan edict on a pillar bearing on its top the figure of a Garuda, but this figure, though suggestive of Vishnu, could not have had such Vaishnava significance in Aśoka's time, as it had later, for the pillar was freely used by him for his own message and there was no preceding Bhagavata association with it as there was with the later Garuda-dhvajas of Besnagar. No counter-indication can be read here to what we gather from Aśoka's inscriptions. These inscriptions prove the abeyance if not absence of Bhāgavatism in his day. Unlike him, the Imperial
1.Ibid., pp. 438-9.
2.Ibid., p. 437.
3.Ibid., p. 439.
Page 397
Guptas not only show ample knowledge of Bhāgavatism but are themselves mostly Vaishnavas1 eager to call themselves Parama-bhāgavatas.2 And, what is most suggestive, they reveal the Garuda as peculiarly their own symbol. There is the famous phrase in the Allāhābād Pillar Epigraph in celebration of Samudragupta, the second emperor of the dynasty: Garut-mad-anka, meaning3 either the Gupta coin bearing the Garuda symbol or simply the Garuda seal (for imperial charters). The Meherauli Pillar inscription of King Chandra, which is generally attributed to Chandragupta II, the third emperor in the line, though Chandragupta I may be as suitable a candidate, "refers to the erection of a Vishnu-dhvaja" .4 And except for some new types of gold coins issued by Kumara-gupta I and the silver coinage which he introduced for the first time in the central provinces, bearing the symbol of the peacock, the Garuda meets us on the reverse of many Gupta coins.5
Looking at the prevalence of Bhāgavatism in both Vidiśa and Takshaśila, over and above Mathurā and the southern Pāndya country, in the four centuries before Christ, and looking at the diametrical opposition of testimony as between Aśoka and the Imperial Guptas about Bhāgavatism, and looking finally at the Guptas' Garuda-emblem, we can hardly escape the impression that they and not the Mauryas were the ruling power in the immediately post-Alexandrine epoch in India.
Particularly strong is this impression when we remember that the Śungas and Kānvas, just like the Mauryas whom they succeeded, had nothing to do with Bhāgavatism and cannot be connected with the growth of this religion in the post-Alexandrine epoch.
Still stronger becomes the impression with the knowledge that even the inscriptions of the Āndhra Sātavāhanas, who succeeded the Śungas-Kānvas and preceded the Guptas, disclose Bhāgavatism in a developing stage which is short of the full efflorescence soon after. Sircar6 writes: "...the Sātavāhana record at Nana-ghat... shows that some people regarded Vāsudeva... not as the
1.A New History..., p. 372.
2.Ibid., p. 371.
3.Ibid., p. 148.
4.Ibid., p. 372, fn. 1.
5.Ibid., pp. 304-5.
6."Vaishnavism", The Age of Imperial Unity, p. 439.
Page 398
greatest of all gods, but only as the equal of Indra and other gods. That another Sātavāhana king... claimed to be equal to Rāma (Baladeva or Sarikarshana) and Keśava (Vāsudeva-Krishna) is also noteworthy." The first inscription is quite early in the dynasty (late 1st century B.C. in the most-favoured modern chronology which starts the Sātavāhanas there), the second not very far from the tail-end (early 3rd century A.D. by the same chronology).1 Surely, all the Sātavāhanas must come before the period of the Heliodorus-inscription where Vāsudeva is "the greatest god" and even the period of Megasthenes and Sandrocottus when the "Souresenoi" of Mathurā worshipped Heracles and the Vāsudeva-cult had reached as far as southernmost India. The proper place for the Sātavāhanas is more or less with Pānini and the period just after him.
The impression, that the Guptas should be where now the Mauryas are put, grows not only very strong but also inevitable when we observe how the art ascribed to the time immediately after Aśoka - that is, the 3rd and 2nd centuries B.C. - and reckoned as Mauryan in general with regard to the themes concerned has features joining it at once to the military descriptions of Megasthenes and to certain characteristics of Gupta kingship.
Mookerji2 says: "A nearly life-size figure of an infantry soldier armed as described by Megasthenes appears among the sculptures of Bhārhut which are generally taken to date from the age of Aśoka." But the true connection of the sculpture emerges as soon as we listen further to Mookerji. He3 first refers to Strabo's account (XV. 1.55) from Megasthenes: "When (the King) goes to hunt, it is in a kind of Bacchic procession, surrounded by women who form a circle... some of the women are in chariots, some on horseback, some on elephants, fully armed as in war." Mookerji then remarks: "It is interesting to note that there is a representation of a procession in a Bharhut Sculpture (c. 2nd century B.C.) of the figure of a woman riding a horse fully caparisoned and carrying a standard, the garudadhvaja (A Guide to Sculpture in the Indian Museum, I, 24)." So we have Megasthenes confirmed by a sculpture dated to a time shortly after his, but bearing a
1.Ibid., and "Chronology", pp. 701. 702: c. 30 B.C. Simuka and 219-227 A.D. Puloma (Sātavāhana).
2.Chandragupta Maurya and His Times, p. 276.
3.Ibid., p. 98.
Page 399
prominent characteristic, both political and religious, of Gupta sovereignty: the Garuda-emblem. The royal line of the time appears to have been of the Gupta Vaishnavites or Bhāgavatas.
Bhāgavatism and the Time of Patanjali
While we are about the subject we may touch on Patanjali's relation to Bhāgavatism just as we have done with Pānini's. Pānini with his vision of a side-by-side worship of Krishna and Arjuna bespeaks an initial phase of the new cult. When discussing the age of his word Yāvānānī ("Yavana script") we also indicated how a certain reference to the Mauryas in Patanjali suggested for Pānini a post-Mauryan date within a time-scheme such as we have set up. Since along Rapson's line of interpreting the Purānas we have made the Śungas and the Kānvas contemporaneous with each other and even with the last Mauryas to some extent, Pānini's being post-Mauryan could amount in the abstract to his flourishing in the time of the Śungas and Kānvas but on concrete considerations he could at the earliest flourish only in the opening period of the Āndhra dynasty - the period of the Nanaghat Inscription. His date is most likely to be around 802 B.C. In our estimate of Patanjali's date we made the general observation that he belonged to the period when Bhāgavatism was in full flower. This means the time-bracket 300 to 1 B.C. Now we must attempt a greater precision.
Sircar has alluded to Patanjali in relation to the association of the name "Krishna" with "Vāsudeva". But as the name was already there in the age of Megasthenes we have here no exact chronological pointer except that Patanjali could be prior to the Christian Era. Puri1 supplies us with a sharper focus. After mentioning the two Besnagar inscriptions he writes about the Ghosun-di stone slab: "It records the erection of a stone enclosure of worship for Bhagavat Samkarsana and Vāsudeva, within the enclosure of Nārāyana, by Bhagavata Gajāyana, son of Parāśati. The Nārāyanavata or the enclosure of the Lord, denotes the compound of a temple or place of worship, while Pujāśilaprākāra stands for Bhagavat Sankarsana and Vāsudeva - evidently referring to a smaller stone enclosure, probably round the images representing Sahkarsana and Vāsudeva within the Nārāyanavata. The cosmic
1. India in the Time of Patanjali, pp. 185-6.
Page 400
philosophic god Nārāyana, whose name is not traced in the Mahābhārata, thus completes the trivenī, or the three streams of thought mingling together to form the cult of Vaisnavism."
Here is the only inscription in which we have what Puri terms the trivenī of Vaishnavism. Also, as we know from Sircar, both Sahkarshana and Vāsudeva are called "supreme lord" in it. There is no question of either of them being merely the equal of Indra and other gods as in the Nanaghat inscription which, as Sircar1 elsewhere says, "begins with an adoration to the gods Dharma, Indra, Sahkarshana and Vāsudeva, the Moon and the Sun, and the four lokapalas, viz. Yama (differentiated from Dharma), Varuna, Kubera and Vāsava (differentiated from Indra)". But now Sankarshana has the precedence of Vāsudeva, though both are equal in divinity and represent one and the same supreme power.
The order is reversed in two late Buddhist manuals. Sircar2 writes: a passage occurring in the Buddhist canonical commentaries (c. first century B.C.), called the Mahāniddesa and the Chullaniddesa, mentions the worshippers of Vāsudeva and Bālā-deva (Sahkarshana)." "The association," says Puri,3 "of Vāsudeva with Bālādeva... is also noted by Patahjali." A little earlier Puri4 tells us, "Patanjali, besides mentioning the names of Krsna and Janardana, the synonyms of Vāsudeva, also referred to the festive gatherings in the temples of Kesava (Vāsudeva) and Rama (Bālā-rama [Bālādeva]).... There is a reference to the Vyuha [form or phase of the conditioned Spirit] of Krsna and his acolytes..." But Patanjali, as Puri has noted, is silent on Nārāyana, just as are the Mahāniddesa and the Chullaniddesa. So he would seem to come before the Nārāyana-strain of Vaishnavism matured, as well as before Sankarshana could occasionally enjoy prime place, while not being superior to Vāsudeva. He should stand in c. 75 B.C. while the Ghosundi inscription might be put in the period from c. 50 to 1 B.C.
In the same period as this epigraph though a little earlier than it, we may place the Nārayanīya section of the Mahābhārata's Śanti-parvan where "Vāsudeva is identified with Paramatman (Supreme
1.Op. cit., p. 438.
3.Op. cit., p. 187.
4.Ibid., p. 184.
Page 401
Soul) and Sahkarshana with Jivatman (individual soul)".1
Patanjali, with his allusion to the Vyūha-doctrine, must be posterior to the Bhagavad Gītā, in which there is no reference to the Vyuhas.2 The Gītā assorts best with the period from the time of Mêgasthenes to that of the Besnagar inscriptions. And it is also in this period that we may see "the growing popularity of the worship of the avatāras of Vishnu, which became the dominant feature of Bhāgavatism in the Gupta age".3 Perhaps the reign of Samudragupta, whose Allāhābād Pillar Inscription as good as makes him out to be an Avatar, is the right background for the Gītā's final recension.
The Guptas and Buddhism
We may now pick up again the theme of the Guptas' emblem having been displayed in the royal procession depicted at Bhārhut. We should be ready for the challenge: ''Bhārhut, like Bodh-Gayā and Sānchī, is, from the viewpoint of subject-matter, predominantly Buddhist. How do you reconcile this with the sovereignty of the Gupta Vaishnavites and Bhāgavatas?"
Just because of their Vaishnavism and Bhāgavatism, the Guptas need not have been intolerant fanatics. The facts are exactly the reverse.
Altekar4 reports: "The Guptas were orthodox Hindus, but the best tribute to their administration has been paid by some contemporary Jain records." Altekar adds that among the officers of the Hindu Gupta emperors "there were some who were Buddhist and one of them is seen making a donation to the Buddhist establishment at Sānchī for the spiritual benefit of his Vaishnava sovereign Chandragupta II. Vainyagupta, one of the later Gupta kings, was a Śaiva and yet he gave a donation to the Mahāyana Buddhist establishment known as Vaivartika Sanghā.
It is possible, however, to exaggerate the Guptas' patronage of Buddhism. We cannot remain quiet when Altekar says: "It is well known how the Buddhist University of Nālandā owed most of its
1.The Age of Imperial Unity, p. 449.
2.Ibid., p. 447.
3.Ibid., pp. 449-50.
4."Religion and Philosophy", A New History..., p. 366.
Page 402
prosperity to the patronage it received from the Hindu Gupta emperors....1 Several monasteries were built to accommodate the growing population of monk-students."2 The connection of the Gupta dynasty with Nālandā University is urged on the strength of four pieces of information.
The Chinese scholar Hiuen Tsang, who travelled in India during 630-643 A.D., has written: "Bālāditya-rāja, king of Magadha, profoundly honoured the law of Buddha."3 He also calls Bālāditya "the builder of a great sanghārāma at Nālandā".4 Next we learn from Majumdar:5 "An inscription found at Nālandā, and belonging to about the middle of the eighth century A.D., also refers to 'Bālāditya, the great king of irresistible valour' who 'after having vanquished all the foes and enjoyed the entire earth, erected a great and extraordinary temple at Nālandā'." The third datum is again from Hiuen Tsang who narrates a victory of his Bālāditya over an impious invader named Mihirakula whose capital was Śakala.6 The fourth piece of information is about a late member of the Imperial Guptas:7 "Narasimha-gupta issued gold coins of a single type which show that he assumed the title Bālāditya." What are we to conclude?
Majumdar8 is right to say that we need hardly doubt that Bālāditya of the Nālandā inscription is the same as Hiuen Tsang's Buddhist Bālāditya. Is Majumdar9 right in saying: "His identification with the Gupta emperor cannot be regarded as certain, but seems to be very plausible"? Majumdar speaks of the plausibility because, according to his chronology, the fight between Narasirhhagupta and Mihirakula, if it happened, took place about 530 A.D.,10 and this was also the age when Nālandā was flourishing as a university centre though a monastery had been there for a long time, as witnessed by the Chinese pilgrim Fa-hien who
2."Education, Literature and Sciences", ibid., p. 398.
3.Majumdar, 'The Imperial Crisis", ibid., p. 200.
4.Ibid., p. 198.
5.Ibid., p. 192.
6.Ibid., p. 196.
7.Ibid., p. 192.
8.Ibid., pp. 192-9.
9.Ibid., p. 193. 10. Ibid., p. 197.
Page 403
"visited Nālandā in c. 410 A.D." but saw "no educational activity at the place".1
Had Hiuen Tsang known of Narasimhagupta's coins he would have agreed with Majumdar. The sole difference would have been in the dating. For, as Majumdar2 is himself alert enough to note, Hiuen Tsang, narrating the defeat of Mihirakula, places "this defeat 'some centuries ago' i.e., several hundred years before 633 A.D., when he visited Śakala", whereas by the modern computation the event is to be dated to about 103 years earlier. But perhaps if Hiuen Tsang had been told that a Nālandā flourishing as a university centre could not be shifted backward by hundreds of years or even to Fa-hien's period he would have separated Narasimhagupta Bālāditya from the Buddha-loving patron of Nālandā who appears to belong to that town's university-phase.
What influenced Hiuen Tsang in identifying the latter with the conqueror of Mihirakula was the confusion he made about the Imperial Gupta emperor Budhagupta, who preceded Narasirhha-gupta Bālāditya.3 He read "Buddha" into the first part of this emperor's name and wrote of Bālāditya being the grandson of Buddhagupta and brought in even a typically Buddhist name like "Tathagatagupta" as that of Bālāditya's father.4 Gupta history does not recognise any Tathagatagupta nor is it aware of anything Buddhist about Budhagupta. So, even with the modern computation in the field, we have no reason to relate Narasimhagupta with the university-phase of Nālandā. Besides, as we shall see a little later,5 apropos of another sweeping assertion of Altekar's, there were several kings named Bālāditya who lived in the period of Nālandā University and had nothing to do with the Gupta dynasty but to whom a deep interest in Buddhism can be attributed. One of them, instead of Narasimhagupta, could have been the Buddha-lover and the patron of Nālandā University.
Yes, we must refrain from exaggerating the patronage of Buddhism by the Guptas. But it cannot be denied that to find Buddhism developing and expanding within India in the three centuries before Christ, we do not need an Aśoka in the 3rd
1.Altekar, "Education, Literature and Sciences", ibid., p. 398.
2."The Imperial Crisis", ibid., pp. 196-7.
3.Ibid., pp. 191-2.
4.India through Chinese Eyes (1957), pp. 135-9, 162.
5.A New History..., p. 156, fn (last para).
Page 404
century B.C. It is sufficient to have the Guptas then and later. They perfectly account for the revival of Buddhism after the slump it seems to have suffered for quite a time lasting down to the age of Mêgasthenes. Aśoka is unnecessary for the revival. Besides, it is absurd to talk of Aśoka having to revive Buddhism. He came within a few centuries of Buddha, and two Buddhist Councils had already been convened before his time. The Middle Country in which he lived was the seat of active Buddhist work, the religion had extended to both Śūrasena and Avanti (the former being the same as the home of Mêgasthenes's Souresenoi) where important centres were founded in Buddha's own lifetime, and an attempt had been made also to spread the Gospel of piety and liberation to the people of Śunāparānta.1 In the age preceding Aśoka, Buddhism should have been in a state not at all requiring revivalist energy on his part. He had merely to continue extending it further. To revive it, one would have to come in another era than the era between Buddha's death and Aśoka's accession, and one would have to be another king than Aśoka.
The very type of king necessary on the heels of the non-Buddhist Mêgasthenes-era to revive Buddhism appears in those who succeeded Chandragupta I whom we have identified with Sandrocottus. Samudragupta, son of Chandragupta I and celebrated lover of learning, may have already started the revival in the period just after Mêgasthenes. But we should be wary of accepting a statement about him like Altekar's:2 "Samudragupta was no doubt a staunch Hindu, who took peculiar pride in resuscitating the Aśvamedha sacrifice, but he entrusted the education of his son to Vasubandhu who was a famous and erudite Buddhist scholar." We should be wary because of several reasons.
To realise them we have only to attend to the information supplied by Majumdar:3 "The references to Vasubandhu in Buddhist works throw light on the literary patronage of a Gupta king whose identity cannot be established with certainty. It is said by the rhetorician Vāmana that the son of Chandra-gupta, known as Chandra-prakāśa, was a great patron of letters, and appointed the famous Buddhist scholar Vasubandhu as his minister. If Vasubandhu flourished in the fourth century A.D. and died after the
1.Barua, op. cit., p. 322.
2."Religion and Philosophy", A New History.., pp. 365-6.
3."The Foundation of the Gupta Empire", ibid., p. 155.
Page 405
middle of that century, as is generally held, we have to take Chandra-gupta as Chandragupta I and regard Chandra-prakāśa as another name of Samudragupta Vāmana's reference to his patronage of letters would be quite in keeping with what Harishena says of the great Gupta emperor."
First of all, there is the question of Vasubandhu's date. In a long footnote, Majumdar1 writes:
"The whole question has been fully discussed by V. A. Smith in EHI. pp. 328 ff. Takakusu held that Vasubandhu lived from about 420 to 500 A.D. (JRAS. 1905, pp. 43 ff). Against this M. Peri maintained (BEFEO.XI, 339 ff.) that Vasubandhu lived in the fourth century A.D. and died soon after the middle of that century. This view is generally accepted. Takakusu opposed it and reaffirmed his old view (Indian Studies in honour of C. R. Lanman, pp. 79 ff.) For other views see V. A. Smith, op. cit.
"In addition to the passage of Vāmana referred to above in the text we have reference to Vasubandhu's relation with the Guptas in Paramārtha's biography of that Buddhist scholar, which may be summed up as follows:-
"King Vikramāditya of Ayodhya became a patron of Buddhism on account of Vasubandhu's success in religious activity. He sent his crown-prince Bālāditya to Vasubandhu to learn Buddhism, and the queen, too, became one of his disciples. When he came to the throne king Bālāditya in conjunction with his queen-mother invited Vasubandhu (who had gone to his native place, Peshawar) to Ayodhya and favoured him with special patronage (JRAS, 1905, pp. 33 ff.).
"It is generally accepted that Vikramāditya and Bālāditya refer to two Gupta emperors, but it is not possible to identify them so long as the date of Vasubandhu is not definitely fixed.
"An interesting side-issue arises out of the statement in Paramārtha's biography of Vasubandhu that the city of Ayodhya was the residence of both the kings Vikramāditya and Bālāditya. It has been inferred from this that the Imperial Guptas had a secondary capital at Ayodhya, for which, however, there is no other evidence.
"An inscription found at Sārnāth mentions a royal dynasty in which there was more than one king named Bālāditya (Cit. III, 284). It is not altogether impossible that Vasubandhu's patron
1. Ibid., pp. 155-56.
Page 406
belonged to this or a similar local dynasty of Ayodhya."
Obviously there are a number of serious uncertainties, which apply to the Imperial Guptas even if they are put in the period following 320 A.D. Vāmana speaks of Chandragupta and Chan-draprakāśa, Paramārtha of Vikramāditya and Bālāditya. To fix on Samudragupta as Vasubandhu's patron is absolutely arbitrary. The uncertainty about Vasubandhu's own date can be gauged by the fact that the only Gupta assuming the title Bālāditya was Narasirhhagupta who, by the modern chronology,1 lived between 500 and 550 A.D., far removed from Samudragupta's time. Again, it is not even certain that the Gupta line was in the least involved: some local line of Ayodhya is perfectly conceivable. Vasubandhu cannot at all be connected with the Imperial Guptas in general or with Samudragupta in particular merely because of some name-echoes. Even the most striking of these echoes - Chandragupta -has no convincing force in the absence of a clear support by other similar sounds; for in the somewhat elastic period within which Vasubandhu is sought to be placed we have at least three non-Gupta Chandraguptas. There is King Chandragupta of the Pandu-varhsīs, brother of King Tivara of Kosala who flourished c. 565-80 A.D.2 There is a prince of the same name mentioned in the Sanjan grant as defeated by Rāshtrakuta Govinda III (c. 794-814 A.D.).3 There is also Chandragupta, a petty ruler at Jālartdhar,4 in the very time at present allotted to the Guptas: 320-c. 570 A.D.
Everything considered, nothing obliges us in the smallest degree to associate Vasubandhu with any Imperial Gupta. The Gupta aid to Buddhism must be kept severely clear of Vasubandhu.
The same warning holds for the involvement of the Buddhist Chinese pilgrim, Fa-hien. We are often told how prosperous this pilgrim found Chandragupta IPs empire and how flourishing the followers of Buddhism. But in fact there is no evidence that Fa-hien travelled through the country when the third Gupta emperor was on the throne. Just by assuming Fleet's epoch our historians link Fa-hien's report with the state of India under that king. We understand this feat of fancy the moment we read Majumdar's
1."The Imperial Crisis", ibid., p. 192.
2."Eastern Deccan", ibid., pp. 90, 91.
3.Ibid., p. 91.
4."Religion and Philosophy", ibid., p. 373.
Page 407
words:1 "An idea of the peace and prosperity in the vast empire may be had from the account of the Chinese pilgrim Fa-hien who travelled through Chandragupta's wide dominions for more than six years. Unfortunately Fa-hien does not give any account of the political condition of India - he does not evert mention the name of the great Gupta emperor."
We have to stick only to the Gupta inscriptions for our estimate of what condition Buddhism was in during the Gupta regime. And they are quite sufficient to explain both Bhāgavatism and Buddhism in the few centuries before Christ, whereas Aśoka and his successors cannot explain the former at all. Even the latter is somewhat inexplicable if we realise that the dynasty succeeding the Mauryan was of Pushyamitra Śunga and this was contemporaneous with the Kānvas according to Rapson and his school or, according to some other historians, followed by the Kānvas, two dynasties strictly Brāhmanical and unlikely to foster any considerable Buddhist outflowering.
The Buddhist outflowering under the Guptas was a wholly Indian affair and can claim no missionary endeavour in the Western world. We should therefore not be surprised by lack of knowledge of it in the countries of the West, which we know to have been in connection with India in the post-Alexandrine epoch. The ignorance, as we have already said, is amazing if we think of Aśoka sending out dharma-mhsions to Greek States. Moreover, Buddhism under the Guptas would not stand out in opposition to Hinduism: there was a wide-spread spirit of tolerance and harmony. Doubtless, Aśoka too fostered such a spirit, but the very fact that the king was an enthusiastic Buddhist convert would give some front-position to Buddhism, just as the Vaishnavite persuasion of the Guptas did throw Bhāgavatism into relief in their day, making it the chief note in the manifold religious concord.
Unquestionably, all historical issues have not been laid at rest when we have fitted the founder of the Guptas into the Mega-sthenes-era in which, as we have seen F. W. Thomas2 remarking, "the greatest share of popular adoration accrued to Śiva and Vishnu (under the form of Krishna)"; but the varied information we have set forth goes to make a huge and many-sided argument
1."The Expansion of the Gupta Empire", ibid., p. 172.
2."Political and Social Organisation of the Maurya Empire", The Cambridge History of India, I, p. 485.
Page 408
suggesting a solid case for reinterpreting Aśokan evidence in favour of identifying Chandragupta I with Sandrocottus.
10
Now we may hark back to a statement we made in a context indirectly related to Aśoka. Not to deal with the issue raised by it would be to leave a loose end. The issue concerns the Ceylonese Devanāmpiyatissa whom we considered to have been wrongly connected with Aśoka by the Ceylonese: our submission was that what was relevant to another Indian king had been foisted on to Aśoka who can be shown to have had nothing to do with Ceylon. We must now demonstrate how the information the Chronicles offer suits the king who would occupy in our revised vision more or less the same chronological position that Aśoka holds in the current historical scheme.
The Reign-length of Chandragupta I
To ascertain who exactly would be ruling Magadha in c. 269 B.C. and some of the subsequent years now allotted to Aśoka, we must first decide the reign-length proper to Chandragupta I who for us replaces Chandragupta Maurya as king of Pātaliputra in post-Alexandrine India.
Fifteen years or twelve or even less is a common count for him. But if, as is mostly admitted, he and not his son Samudragupta founded the Gupta Era and if his marriage with the Lichchhavi princess Kumāradevī is closely linked in time with this occasion, helping, as that marriage did, to confirm his rise to sovereignty over Pātaliputra, Samudragupta can never be brought to the throne until he had grown sufficiently to deserve the description of him in the Allāhābād Pillar Epigraph. For, there Chandragupta chooses his son as his successor in the midst of a public assembly, joyfully exclaiming, "Thou art worthy, rule this whole world." Majumdar1 well observes: "We can hardly believe that Samudragupta gave evidence of his prowess and ability and was selected by his father as the fittest prince to succeed him directly before he attained the age of twenty-five to thirty years." So a reign of twenty-five to thirty years for Chandragupta is the minimum we
1. "The Foundation of the Gupta Empire", A New History,... p. 158.
Page 409
have to grant. It may be even longer.
All that we know is the interval between his coronation and the coronation of his grandson Chandragupta II. The earliest inscription of the latter is dated the 60th year of the Gupta Era (though without the designation "Gupta") and his own regnal year in it is read by some as pratham (first) and by others as pancham (fifth). With Majumdar we favour the second reading. This means we have 55 years for the reigns of Chandragupta I and Samudragupta. Out of them we are not restricted by any circumstances from allotting to Chandragupta I at least 3 decades.
Samudragupta and Devānampiyatissa
Even if he reigned for no more than 30 years we should have Samudragupta on the throne in (315-30=) 285 B.C. and continuing for 25 years - that is, down to 260 B.C. He would thus cover part of the reign of Devānampiyatissa which, by Ceylon's traditional chronology, would be 308-7 to 268-7 B.C. And, in general, he is fit to be there because he is the one king in Indian history who is explicitly connected with Ceylon by his inscriptional record. There is no question here of anything like Aśoka's ambiguous Tāmraparni and Tāmraparnīya. We know definitely from the Allāhābād Pillar that "the peoples of Simhala and other islands" figured among those about whom, judging by the language employed, Majumdar1 is led to say: "it may be taken for granted that they sought to win the favour of the great emperor by personal attendance in his court, offering daughters in marriage, and asking permission for the use of imperial coins or soliciting imperial charters confirming them in the enjoyment of their territories."
Not only in general but indeed in particular is Samudragupta fit to be related to Ceylon's Devānampiyatissa. It is the last category of homage in Majumdar's list that points to this fitness - the category of peoples "soliciting imperial charters confirming them in the enjoyment of their territories". For, the situation of Devānampiyatissa, as presented in the Ceylonese Chronicles, is strongly suggestive of a status feudatory to the North-Indian king of his day. S. Krishnaswami Aiyangar,2 telling us of this monarch's
1."The Foundation of the Gupta Empire", The Classical Age, pp. 10-11.
2."Southern India and Ceylon", The Age of Imperial Unity, p. 236.
Page 410
sending an embassy with rich presents to the Mauryan emperor and receiving from the latter the various essentials for his own formal consecration, which was celebrated with great splendour, gives the footnote: "It looks as if the Ceylonese king sought for investiture at the hands of Aśoka, though the reason is not apPārent..." What Devānampiyatissa did is a spectacular equivalent of Majumdar's final category of homage. The reason which to Aiyangar is "not apPārent" becomes clear in the light of the Allāhābād Pillar. There we have the only North-Indian king known in history to have claimed Ceylon as a vassal state.
And we may add: "Samudragupta's claim not only agrees with our knowledge of that unique event in the annals of Ceylon - the formal consecration of a Ceylonese king by an emperor of North India - but also lends a new colour to the famous 'synchronism' alleged by Sylvain Lévi of Samudragupta and Sirimeghavanna of Ceylon. This 'synchronism' is said to stand side by side with Fleet's determination of the Gupta Era as starting in 320 A.D. It is said to confirm Fleet in framing the first Imperial Guptas in the 4th century A.D. We have shown how Albērūnī's information has been completely twisted: he distinctly declared that the Gupta Era commenced with the end of the Guptas - a celebration of the disappearance of a king-line that had become unpopular.1 The discussion of Devānampiyatissa naturally brings in Lévi's claim and gives us a chance to deny it."
Already in Part I of our book we have dealt with it in some detail: now we may briefly touch on it by focusing on the issue of Ceylon's status which is relevant to our treatment of Devānampiyatissa.
Samudragupta and Sirimeghavanna
The Chinese writer Wang Hiuen-t'se who was in India in 647-48 A.D. has stated that the king of Chen-tsen (Ceylon) named Chi-mi-kia-po-mo (Sirimeghavanna) offered presents to the Indian king San-meou-to-lo-kiu-to (Samudragupta). Sirimeghavanna, by the traditional Ceylonese Era assuming 543 B.C. as the date of Buddha's Parinirvāna (or death), ruled from 304 to 322 A.D. and, by the Fleet-Geiger hypothesis of an earlier tradition which put the Parinirvāna in 483 B.C., from 352 to 376 A.D. In the current
1. Sachau, Albērūnī's India, II, p. 7.
Page 411
chronology, Samudragupta's reign is taken to have ended in 375-76 A.D. So the two monarchs are regarded as contemporaries. The pertinent question for us is: "Have we any sign that Siri-meghavanna's Ceylon was feudatory to a North-Indian king?"
The story1 about Chi-mi-kia-po-mo and San-meou-to-lo-kiu-to has it that the Ceylonese king, in order to remove all inconveniences to Ceylonese pilgrims visiting Bodh-Gayā, sent a mission to the Indian king with rich presents and asked for permission to build a monastery and a rest-house at that holy place. The Indian king granted the permission, and the Ceylonese king built a splendid monastery to the north of the Bodhi tree. When Hiuen Tsang saw it, it had developed into a magnificent establishment. Referring to the old history of its foundation he says that the Ceylonese king "gave in tribute to the king of India all the jewels of his country". Majumdar2 comments: "It is likely that Samudragupta's courtiers regarded the rich presents as tributes, and construed the Ceylonese king's prayer for permission to build a monastery as an 'application for a charter confirming him in the enjoyment of his territories', one of the forms of homage paid by the category of vassal states in which Simhala is included."
The courtiers of Samudragupta must surely know the exact situation. Besides, the direct and king-pinpointing story of Wang has no word like "tribute". Majumdar has taken up Hiuen Tsang's word and turned its particular application into a general feudatory usage. His surmise is quite fanciful in view of the actual facts. If Chi-mi-kia-po-mo was allowed to build a monastery and a rest-house on the territories of San-meou-to-lo-kiu-to by the latter monarch, how do the territories of Chi-mi-kia-po-mo - that is, Ceylon - come into the picture? However rich the presents in appreciation of San-meou-to-lo-kiu-to's generosity with his own territories, they cannot convert far-off Simhala into a vassal state. Sirimeghavanna was a king in his own right, owing allegiance to no other sovereign, and during his rule it is Ceylon that got honoured by India instead of the other way round by the arrival there of the Tooth Relic of Buddha from Dantapura in Kalihga which was a part of Samudragupta's empire. Ceylon under Sirimeghavanna does not and cannot ever equate with the Simhala of the Allāhābād Pillar Inscription.
1."The Foundation of the Gupta Empire". The Classical Age. pp. 11-12.
2.Ibid., p. 12.
Page 412
Consequently, Wang-Hiuen's San-meou-to-lo-kiu-to must be distinguished from the Allāhābād Pillar's Samudragupta. We should not hesitate to place the latter where we have done - in the age usually allotted to Aśoka. And some other Samudragupta than the great one looms up - a minor namesake of his who would synchronize with Sirimeghavanna before 320 A.D. à la Albērūnī, one of the members of the Imperial Gupta dynasty at the tail-end of its career. Sirimeghavanna as dated by the traditional Ceylonese chronology would be just right with the first 16 years (304-320 A.D.) out of his 28 to fit such a tail-end San-meou-to-lo-kiu-to.
Evidence for Another and Smaller Samudragupta
Do we have any epigraphic or numismatic record to render plausible the figure we have postulated? We have dealt with this theme too in Part One. But, for the sake of completeness, we need to take it up again with some variations to be played upon it.
In general there is a sign of Gupta rulership before 320 without its coinciding with the known predecessors of Chandragupta I, his father and grandfather whom modern historians date before that year. Raychaudhuri1 writes within the context of the current time-scheme:
"Traces of 'Gupta' rule in Magadha proper, or some neighbouring tract down the Ganges, are found as early as the second century A.D. I-Tsing, a Chinese pilgrim, who travelled in India in the seventh century A.D., mentions a Mahārāja Śrī Gupta who built a temple near Mrigasikhāvana 'which was about forty yojanas to the east of Nālandā, following the course of the Ganges'.2 I-Tsing's date would place him about A.D. 175.3 Allan rejects the date, and identifies Śrī Gupta with Gupta, the great-grandfather of Samudragupta, on the ground that it is unlikely that we should have two different rulers in the same territory, of the same name, within a brief period. But have we not two Chandraguptas and two Kumāra Guptas within brief periods? There is no cogent reason for identifying Śrī Gupta of cir. A.D. 175, known by
1.Op. cit., pp. 528-29.
2.Dr. Majumdar in A New History..., VI, p. 129; Dr. D. C. Ganguli, Indian Historical Quarterly, XIV (1938), 332. (Raychaudhuri's footnote)
3.Allan, Gupta Coins, Introduction, p. xv. Cf. Indian Antiquary, X (1881), 110. (Raychaudhuri's footnote).
Page 413
tradition, with Samudra Gupta's great-grandfather who must have flourished a century later."
Raychaudhuri's confident conclusion and his pointer to repeating Gupta names embolden us, in our search for a second Samudragupta, to ponder critically a piece of information by Altekar apropos of the "Punjāb during c. 230 to c. 340 A.D." Altekar1 tells us: "The numismatic evidence shows that the Western and Central Punjāb was being governed during this period by three Scythian houses." "The western part of the province was being ruled" by "the Shāka (not Śaka) dynasty", while "the Shiladas and the Gadaharas were holding sway over the Central Punjāb". "Both these dynasties were ruling... down to the days of Samudra-gupta; for a Gadahara chief is seen putting the name of that emperor on his coins." What we have to weigh with care is the phrase: "down to the days of Samudragupta."
Altekar2 continues: "The rule of the Shākas and the Shiladas came to an end in c. 340 A.D. with the rise of a tribe, which is sometimes described as the Little Kushāna and sometimes as Kidāra Kushāna. Its chief Kidāra, whose capital was at Peshawar, was originally a feudatory of the Sassanians.... In course of time, however, Kidāra conquered Kashmir and the Central Punjāb. He now felt himself strong enough to assume independence.... This step, probably taken in c. 355 A.D., evoked imperial anger and reprisal... We find Shapur II encamped at Kabul in 356-57 A.D.... Kidāra was compelled to acknowledge his suzerainty.... We find him sending assistance to his liege lord in 359 A.D.... The closer association with the imperial army probably enabled Kidāra to realise its weak points and he began to mature plans for reasserting his independence. He secured the good will of Samudra-gupta-, who had by this time extended his sphere of influence to the Punjāb, by sending him presents along with professions of allegiance, and delivered his blows against Shapur II in 367-8 A.D."
Let us follow the progress of Kidāra. His starting-point is Peshawar. Next he annexes Kāshmir and the Central Punjāb. The rule of the Shiladas who were in the latter locality ends there in c. 340 A.D. Obviously, the Gadaharas who were in the same locality lost their authority also. At this time there is no presence of
1."The Punjāb, Sindh and Afghanistan", A New History..., p. 20
2.Ibid., pp. 22-23.
Page 414
Samudragupta in the vicinity of the Central Punjāb. He appears fairly later at the time Kidāra is preparing to shake off for good the Sassanian overlordship. This means about 360-365 A.D. Kidāra's kingdom would be a "vassal state" then. It is a far cry from the period of the Gadaharas, which is over by c. 340. Samudragupta has nothing to do with it. And the coins bearing the name "Samudra" must be unrelated to him and could have come out anywhere between 230 A.D. and the end of the Gadahara-rule. The rule cannot be said to have continued "down to the days of Samudra-gupta". What is at the back of such a statement is simply the assumption that the "Samudra"-legend coinage is bound to refer to none save the great Samudragupta. A scrutiny of the progressive historical situation as unfolded by Altekar himelf contradicts his own statement. The logic of events drives us to posit a smaller Samudragupta who belongs not to the start but to the termination of the Gupta-line and has some sort of hold over a Gadahara chief. He is one of the last successors of the Śrī Gupta of c. 175 A.D. and the true contemporary of Sirimeghavanna of 304-332 A.D. partly or wholly in the latter's earlier 16 years.
We may be sure that there is a smaller Chandragupta too, either the father or the son of this Samudragupta. For, matching the coins with the latter's name, there are - as we learn from Majumdar1 - a "Kushāna type of coins" with the name "Chandra".
The final test of whether the Gadaharas could exist in the time modern chronologists fix for the eminent Samudragupta would be our ability to spot them in the Allāhābād Pillar Inscription. If we cannot spot them, they have to be completely severed from him.
The inscription has only one phrase denoting frontier kingdoms of foreign rulers in the north or the west under his suzerainty. It is the compound: Daivaputra-Shāhi-ShāhānuShāhi-Śaka-Murunda. With either side of mid-4th century A.D. for Samudragupta, the first three words of the compound are accepted as "well-known titles borne by the Kushān kings, and may refer to one of them... and the probability is that reference is made here to a Kushān ruler exercising sway over Kabul and a part of the Punjāb, and possibly other territories further to the west".2 The choice in particular falls, as we have already seen, on Kidāra of the Little Kushānas.
1."The Foundation of the Gupta Empire", Ibid., p. 150.
2.Ibid., p. 147.
Page 415
Thus the opening trio of vocables excludes the Gadaharas. What about the expression Śaka-Murunda?
It is understood as meaning either "Śakas and Murundas" or -with "Murunda'' serving as a title - "Śaka lords". Jayaswal1 proposed to include in it not only the Śakas of Western India but also the smaller Śaka chieftains who were the Gadaharas and the two other Scythian tribes in the Punjāb along with them. His view is not accepted. For, the close-knit single compound phrase within which, at the end, the expression Śaka-Murunda occurs suggests rulers contemporary with one another at the time when Samudragupta was suzerain over them. This interpretation is in keeping with that of the two earlier compounds in the inscription, listing 9 tribes in all. Those tribes are understood to be co-existent. Similarly, Śaka-Murunda implies co-existence and contemporaneity with Daivaputra-Shāhi-Shāhanushāhi. In that case, the Gadaharas and the two other Scythian tribes cannot be included in Śaka-Murunda, for they had no kingdoms when the Kushāna Kidara is stated to have acknowledged Samudragupta's overlordship: that is after 360 A.D. Śaka-Murunda is taken to cover either (1) the Śakas ruling in Western India as well as those ruling in and about Sānchi and a foreign Murunda tribe ruling in the Ganges Valley, or (2) simply the Western Śakas.2
It is impossible to accommodate the Gadaharas anywhere in the inscription. They belong to a different historical context: so the "Samudra" coins must connect with some minor Samudragupta synchronizing with Ceylon's Sirimeghavanna - and by the same token the "Chandra" coins go with either his father or his son who was a minor Chandragupta.
We are now free of all immediate hindrances to the synchronization of the great Samudragupta with Devānampiyatissa of Ceylon. And we may remark that in the epoch common to these two a monk named Mahendra might very well have brought Theravadin Buddhism to Ceylon. Such an event would be quite in accord with the Guptas' general policy towards Buddhism and with Samudragupta's patronage of all learning as well as with his declared relations with "Simhala" and "other islands". It would be natural that Buddhism should get encouraged under him and launch a mission to this island in particular.
1.Journal of the Bihar and Orissa Research Society (Patna), XVIII, p. 210.
2."The Foundation of the Gupta Empire", A New History.....pp. 147-48.
Page 416
Of course, by removing the son of Chandragupta I to the 3rd century B.C. from the 4th A.D., we shall have to throw new light on the meaning of Daivaputra-Shāhi-ShāhānuShāhi-Śaka-Murunda making the phrase suitable to the post-Alexandrine epoch. We shall do the needful in the next chapter. Here we may meet an objection likely to arise regarding the subject with which we have been mainly concerned: Ceylon.
Ceylon in the 3rd Century B. C. and Samudragupta's Simhala
We are sure to be told: "Samudragupta calls Ceylon Simhala. But this name is not as old as the 3rd century B.C. Mêgasthenes who wrote his Indica at the beginning of the same century called this island Taprobanè. And the author of the Periplus of the Erythraean Sea (c. 70-80 A.D.) states that the island was called Taprobanè by the ancients but was known in his time as Palae Simoundou. So even up to the 1st century A.D. there is no sign of Simhala. Simhala in the 3rd century B.C. is an anachronism."
The situation is much more complex than this objection indicates. We must bear in mind that Mêgasthenes speaks of the island Taprobanè in connection with a river as well as with pearls. It is indeed very probable, as we have already suggested, that he is mixing up a report of the pearl-producing river Tāmraparnī with a report of the Tāmraparnī-dvīpa. Proceeding to the island of Ceylon in relation to the South-Indian river, he would naturally use for that island the name related to the river. So his choice of "Taprobanè" does not mean that the island had no other name.
As for the Periplus's information, the section in which it is given is acknowledged to be hopelessly corrupt.1 We can build little on it. But whatever we can build must be after we appreciate the exact state in which the text exists. Hjalmar Frisk2 has published a critical edition and there the Periplus (Ch. 61) reads: "the island anciently Simoundou but among the ancient Xareris probane." At the start we must note that the Periplus does not have one single word "Palaesimoundou" but two separate words Palae Simoundou", of which the first is a Greek adverb of time: "anciently".
1.McCrindle, Ancient India as Described by Ptolemy, edited by S. N. Majumdar, 1927, p. 252.
2.Le Periplus de la mer Erythre'e suivi d'une etude sur la tradition et la langue ("Goteborg Hogskolas Arsskrift", XXIII, 1927, 1).
Page 417
Next, there is no Greek word like "Xareris". Again, "Taprobanè" is not present but only the last 7 letters of it. We may legitimately reconstruct it from them, but we cannot convert that non-existent Greek word "Xareris" into "it was called". Actually the Periplus suggests two old names: "Simoundou" and "Taprobanè."
The very same information is proffered by the geographer Ptolemy (c. 130-140 A.D.) plus something else. In Louis Renou's critical edition Ptolemy (VII,4,1) says: "Taprobanè, which anciently was called Simoundou, now Salike." Here "Taprobanè" does not figure as anything particularly ancient. It is a name persisting to Ptolemy's time, though the name in currency was "Salike". "Simoundou" is considered an ancient name - "palae", a separate word, serving once more as an adverb of time.
Both the Periplus and the Geography commit a mistake in separating "palae" from "Simoundou" and using the former to mean "anciently". Pliny who lived in 79-23 B.C., about 150 years before the Periplus and nearly 200 before the Geography has the true form. In his Natural History (VI, 85) he makes "Palaesimun-dus" the island's capital town as well as one of its two rivers, on whose bank that town stood. As Pliny drew much of his information on India from Mêgasthenes, it is very likely that the term did occur in the Indica but subsequently got ignored because of Mêgasthenes's emphasis on "Taprobanè". Henceforth the latter was for the Greeks the sole name. But the presence of "Palae-simoundou" is strongly suggested by the term "Palaeogoni" which Mêgasthenes1 uses for the inhabitants of the island of Taprobanè. No doubt, the two terms are not variants of each other and must mean different things, but they have a structural similarity which makes them tend to imply each other.
S. N. Majumdar,2 in consonance with other scholars, has tried to explain the Periplus's "palae Simoundou" by an Indian original: "In Kautilya's Arthaśāstra (II, xi) [Ceylon] has been referred to as Pāre-samudra (an alternative form of which is, according to a sutra of Pānini, Pāre-sindhu) which means 'of the other side of the ocean'. Now a confusion of Pāre-samudra and its synonym Pāre-sindhu produced the Palae Simoundou of the Periplus."
Majumdar has certainly a linguistic possibility on his side; but neither Pāre-samudra nor Pāre-sindhu is a specific term which can
1.Ancient India as Described by Mêgasthenes and Arrian, p. 62.
2.Ancient India as Described by Ptolemy, pp. 392-93.
Page 418
be unique to any country, much less to any island: it is a mere generality for all countries, including islands and non-islands, beyond the ocean. It can get attached to Ceylon, as it has in the Arthaśāstra, merely because Ceylon is for India the most important neighbouring place across the sea. It does not suggest a special individual name for Ceylon, nor does it carry any indication of Ceylon's island-character. If we can find another explanation fulfilling these conditions, we may be nearer the truth. And, what is more pertinent to our purpose, we may show the Allāhābād Pillar's "Simhala" to be sufficiently ancient.
The earliest known mention in India of "Simhala" is taken to be in a Nagarjunikonda inscription dated about a 100 years before the time currently ascribed to the Allāhābād Pillar record. But the testimony of Ptolemy should give us pause. Judging from Indian evidence we would not consider "Simhala" older than c. 250 A.D., the present dating of the Nagarjunikonda epigraph. If Ptolemy could precede Indian testimony by over a 100 years we should hesitate to rule out further antiquity.
When we examine the component "Simundus" in the word employed by Pliny, we see that it resembles one of the several antecedents of the modern "Ceylon". These antecedents, according to McCrindle,1 are: Sailan, Zeilan, Serendiba, Serendivus and Ptolemy's Salike which McCrindle2 conjectures to be a mistake for Saline, as it very well may be since Ptolemy's "Salai" for the inhabitants is clearly allied to "Simhala"'s Pali form "Sīhala".3 S. N. Majumdar,4 as well as McCrindle, regards all those antecedents as derived from "Simhala". Pliny's "Simundus" is not dissimilar to "Serendivus" which, like "Serendib" and "Sirlediba", has in its last part the Sanskrit "dvīpa" ("island") or the Prakrit dīpa.
Now, the main word "Smha" ("lion") from "Simhala" may be used to build a contracted compound Simhadvīpa instead of Sirhhaladvīpa. Such a compound would be just as valid as the familiar "Simhapura" for a kingdom mentioned by Hiuen Tsang in the Punjāb region.5 And it is most interesting to remark that this
1.Ibid., p. 252.
2.Ancient India as Described by Mêgasthenes and Arrian, p. 62.
3.Op. cit., p. 252.
4.Ibid., pp. 252, 391.
5.Walters, I, p. 240.
Page 419
kingdom was also called "Simhala". The suffix "la", as is well known, indicates "city", "kingdom", "country", and here it gets dropped when "pura" is used: we do not have "Simhalapura" but "Simhapura". Analogously we may have "Sirhhadvīpa" rather than "Simhaladvīpa" for the island of Ceylon. Then "Simundivus" or "Simundus" would not be an illegitimate foreign transcription: the insertion of the nasal (n) in the second syllable would even be according to the Prakrit tendency marked by S. N. Majumdar1 himself and once before invoked by him in the course of the Parirhda-controversy: vakra-titvata num. With "Palae" as a prefix, the compound "Palaesimundus" would correspond to the Indian Pāre-simhadvīpa and would have a meaning on the analogy of the "Palaeogoni" of Mêgasthenes which is best explained2 in the manner of Goldstiicker by taking "goni" for janās ("a people") so that the whole expression "Palaeogoni" would stand for Pāre-janās and mean "a people on the other side (of the waters)". "Palaesimundus" would similarly stand for "the island of Siriiha on the other side (of the waters)". The compound Pāre-simhadvīpa is not found in Indian literature; but neither is Pāre-janās which is the only reasonable interpretation of the word in Mêgasthenes.
If the above line of thought is correct, we have "Simhala" already in use in the time of Sandrocottus, the father of Samudragupta. And Ptolemy adds "Salike" (="Saline") without knowing "Simhala" 's secret presence in his own "palae Simoundou".
But the correctness of this line of thought is not indispensable. Even if "Palaesimoundou" be born of a mixture of Pāre-samudra and Pāre-sindhu and even if the Greeks after Mêgasthenes knew only "Taprobanè", it does not follow that "Simhala" or its Pali equivalent "Sīhala"3 was not in use from antiquity. As Jacqueline Pirenne4 suggests, Ptolemy, instead of saying "now" should have said "also" when he wrote "Salike". Pirenne5 indicates the correct position in the naming of Ceylon by referring to the book Christian Topography (XI) written in the 6th century A.D. by Cosmas
1.Op. cit., p. 301.
2.Ibid., p. 253.
3.Ibid., p. 252.
4."Un Probleme-clef pour la Chronologie de 1'Orient: Le Date du 'Periple de la mer Erythree', " Journal Asiatique, Tome CCXLIX, Fascicule No. 4,1961, p. 453.
5.Ibid., pp. 452-3.
Page 420
Indicopleustes: "the island is called by the Indians Sielediba, by the Greeks Taprobanè."
This, of course, should not be understood as meaning that the Indians did not employ a designation corresponding to "Taprobanè". They did and they employed still another term, as we shall see. What Cosmas Indicopleustes conveys is simply that a designation corresponding to his "Sielediba" was always there side by side with "Taprobanè" 's Indian equivalent. Just as, even in the 6th century A.D. when he lived, the Greeks still used "Taprobanè", so also even in the early days when "Taprobanè" came into use among the Greeks the Indians employed the native equivalent of "Sielediba". That is the sense we should, with Pirenne, read in the statement of Christian Topography about Ceylon.
The full accurate position in the matter may be learnt from the Pali Mahāvamsa (VII. 7,41,42) which applies three names to Ceylon: "Sīhala", "Lanka" and "Tambapannī". In connection with Vijaya, the traditional founder of the Aryan colony in Ceylon on the day of Buddha's death, the Mahāvamsa mentions the origin of the appellation "Sīhala" for the island. Aiyangar1 sums up its story: "From Simhavahu or Simhala, the father of Vijaya, the island received the name Simhala-dvīpa, the Arabic Sirendib, the Portuguese Ceilao, and the modern Ceylon." According to this account, "Simhala" is an ancient designation at least as old as 543 B.C., the traditional Ceylonese date of the Nirvāna of Buddha - and we should hardly wonder on finding it in an inscription of the 3rd century B.C.
From various quarters suggestions converge about the antiquity of Samudragupta's "Simhala". And even the supposed indication from the Periplus that such a term was not in vogue in 70-80 A.D. would lose all relevance if the unorthodox thesis of Pirenne,2 set forth with considerable acumen, proved to be correct - namely, that the Periplus is to be dated to c. 225 A.D. For, then, the Geography would be anterior to the Periplus and we should have the fact that the latter has still no mention of anything answering to the former's "Salikè". The non-mention would lack significance altogether. May not the same lack of significance apply even if we accept the current idea of the Periplus preceding the Geography?
1."Southern India and Ceylon", The Age of Imperial Unity, p. 235.
2.Op. cit., pp. 441-459.
Page 421
The Ceylonese Chronicles' Mix-up of Samudragupta and Aśoka
We may now turn back without misgiving to our contention that Aśoka has nothing to do with Ceylon and that the real link is between Samudragupta and Devānampiyatissa to whose island the missionary monk Mahendra went during the former's reign in India.
But how did the Ceylonese Chronicles commit the error of foisting on Aśoka the events that touched Ceylon during Samudragupta's reign in the 3rd century B.C.? One broad psychological clue is in Barua's remark:1 "The essence of Aśoka's Dharma-vijaya may be shown to have been recorded in the Gupta coin-legends: sucharitair svargam jayati, 'He conquers even the heavens by virtues'. " Another clue is: among the Mauryas, Aśoka is the next important king after a Chandragupta and similarly among the Guptas the undeniably important Samudragupta succeeds a Chandragupta. A third clue is the Chronicles' allotting to Aśoka a reign-period fairly close to the one that would be Samudragupta's by the Purānic chronology. A fourth is from documentary sources.
Is it just a happy coincidence that Samudragupta should have his inscription engraved on the Aśoka Pillar at Allāhābād? The inscription itself calls forth by the character of its description a memory of the Buddhist emperor. As Majumdar2 tells us: "It describes the political condition of India and the achievement and personality of Samudragupta with such fullness of details as is not to be found in the record of any other king of Northern India, with the single exception of Aśoka."
Mankad3 carries us one step forward in the comparison when he reports: "Indeed, it has been suggested with considerable force and reason that Samudragupta had actually and intentionally tried to mould his career after the example of Aśoka Maurya. Dr. S. K. Aiyangar says (Ancient India, pp. 247-8): 'The whole series of these conquests (the conquests of Samudragupta) as detailed in one elaborate inscription which has come down to us of this great ruler, had for their object nothing more than the bringing under the control and influence of one suzerain monarch the whole territory included in the area, which in the best of its days
1.Op. cit., p. 287.
2."The Foundation of the Gupta Empire". A New History. .., p. 136.
3.Purānic Chronology, pp. 273-74, 276-77.
Page 422
constituted the empire of Aśoka. It must be borne in mind that this record of Samudragupta is indited on a pillar which carries on it an Aśokan inscription as well. Did Samudragupta then emulate the exploits of Aśoka? Did he, in fact, know the history of Aśoka or the extent of his empire and could he have read the document on the pillar of Aśoka? The answer to this question may be given in the affirmative for certain reasons.' In this passage and in what follows in the next two or three pages, by means of able and lucid reasoning a modern scholar proves that both in his temporal and spiritual outlook Samudragupta had followed in the footsteps of Aśoka Maurya.... Aiyangar, with keen insight, has used an argument to show that Samudragupta had read the inscriptions of Aśoka Maurya. He has said that, according to Harisena, Samudragupta has the epithet of Kavirāja, and Kavirāja, according to Rajasekhara, was a title to be conferred on one who was able to make compositions not only in Sanskrit but also in the various dialects of the country. Samudragupta, possessing the title Kavirāja, was thus a good linguist. As such, Aiyangar says, he could have read and understood the inscriptions of Aśoka Maurya."
Samudragupta seems to have had an Aśokan aura. Perhaps the Ceylonese received a Buddhist mission during his reign and later confused him with the Buddhist Aśoka who had become a legend to them.
Who knows if the confusion was not further helped by a curious fact which emerges from the Buddhist narratives in Sanskrit and of which Samudragupta too may have been aware. It is a fact showing an "Aśoka-Samudra" association. Barua1 tells us that, in a legend transmitted by the Divyāvadāna, Aśoka's first connection with Buddhism was not with the Sangha as a whole but with an individual representative named the Venerable Samudra. Addressing Samudra, Aśoka said: "I take refuge in (thee), the sage, and also in Buddha, the supreme embodiment of qualities as well as the Doctrine taught by the elect." So, in the sense of "Protected by Samudra", Aśoka could have been called "Samudra-gupta".
All in all, it is likely that the Ceylonese Chronicles blended the third Maurya with the second Gupta.
Page 423
Problems of the Allāhābād Pillar Inscription
Face to face with Samudragupta's Daivaputra-Shāhi-Shāhānushdhi-Śaka-Murunda of the Allahabad Pillar Inscription, and even with the constellation of nine obeisance-offering tribes named in it , we are bound to be told :
"Can you put in the 3rd century B.C., in post-Alexandrine India, such an ensemble as is formed by the tribes of the Mālavas, Ārjunāyanas, Yaudhēyas, Madrakas, Ābhīras, Prārjunas, Sana-kānīkas, Kākas and Kharaparikas?
"Can you doubt that when foreigners at the frontier are mentioned, they are the Kushānas and the Śakas? But both these peoples are known to have come to India round about the beginning of the Christian Era. 78 A.D. is generally believed to mark the beginning of the reign of Kanishka, the greatest Kushāna emperor. And surely it is the Kushānas alone who bore the resounding title Daivaputra-Shāhi-Shahānushāhi. It continued right up to their later representatives, the 'Little Kushānas', the chief of whom was Kidāra, a contemporary of Samudragupta by the current dating of the latter.
"The Śakas of Western India - often called the Western Satraps or Kshatrapas and Mahākshatrapas - lasted on through Samudragupta's usually dated reign and ended their rule when Chandragupta II conquered them - Chandragupta II who, after his conquest of them, issued his silver coins in imitation of those of the Satraps. Now, these Śakas date themselves in the era of 78 A.D., which has come to be designated the Śaka Era. How then can Samudragupta's Śaka Murunda exist in the earlier half of the 3rd century B.C. where Samudragupta will have to be pushed up if he is to substitute Aśoka? There he will be a contemporary of the successors of Alexander the Great - Greek kings, some of whom had diplomatic connections with India. Not only Sandrocottus but also Amitrachates, his son, received an envoy from the Greek world. There was even a third envoy, either to him or his father or his son . You have identified Amitrachates with Samudragupta. So . far as the meaning of the name or even the name itself is concerned, your identification has some plausibility; but, with regard to the Allāhābād Pillar Inscription, the new date involved leads to a fantastic contradiction of all ascertained history."
Page 424
The Nine Tribes and Post-Alexandrine India
Our general answer is: "Ascertained history and insecure hypothesis are badly mixed up everywhere in this argument, and the conclusions do not inevitably follow."
Let us begin with the tribe-names. Just because all the names do not openly chime with the ones familiar to us from Alexander's historians we do not need to read quite different milieux for them and for those others. We must probe a'little below the surface. Even in the case of the Mālavas who at once recall the Malloi and the stiff resistance Alexander received from them in the Punjāb where in his time they "lived in the land lying north of the confluence of the Rāvī and the Chenab",1 we get to the full truth only by linking them with the tribes succeeding in the list. For, while they are known in later history as peoples of Western India, their immediate companions in the inscription can be taken to show them in the north-west in Samudragupta's time - located exactly as they should be in the 3rd century B.C., in post-Alexandrine India.
In fact, these companions are not only north-western but also the very tribes we can trace in relation to Alexander's campaign. The Yaudhēyas are mentioned by as old a writer as Pānini2 and "the heart of the Yaudhēya territory may have been the eastern Punjāb":3 the Brihat-Samhitā "locates them in the northern or north-western division of India along with the Ārjunāyanas".4 Jayaswal5 writes: "The find-spot of Yaudhēya coins suggests that an unnamed state beyond the Beas reached by Alexander was theirs." Altekar6 agrees with Jayaswal: "The great republic beyond the Beas, of which Alexander's army heard alarming reports in 326 B.C., was most probably the Republic of the Yaudhēyas." Cunningham prefers to identify them with the Ossadioi or Asso-dioi with whom Alexander fought: the latter name he connects with the name Ajudhiya which is a variant for Yaudhēya.7
1.The Age of Imperial Unity, p. 163.
2.Ibid., p. 165.
3.Ibid., p. 166.
4.Ibid., p. 167.
5.Hindu Polity, I, p. 67.
6.The Journal of the Numismatic Society, XI, p. 50.
7.Mookerji, Hindu Civilization, II p. 340, fn. 2.
Page 425
Their neighbours the Ārjunāyanas who "have been assigned to the region lying west of Agra and Mathura about the Bharatpur and Alwar States of Rājputana"1 may have been, according to some scholars,2 the Agalassoi who fought Alexander after the submission of the Sibae (Sivis) near the Jhelum in the Punjāb and who are called by the Greeks Agesinae or Argesinae, names which McCrindle thought equivalent to "Arjunayana."3
"The Madrakas or Madras... had their capital at Śākala (modern Sialkot in the Punjāb)," writes Sircar,4 adding that they "were probably subjects of the younger Porus (Paurava king) in the days of Alexander".
As for the Ābhīras, "the most important section of the people", says Sircar,5 " lived in the northern Konkan and the adjoining part of the Maratha country", but about the Ābhīras in Samudragupta's inscription he6 says: "it is uncertain whether the reference here is to the Abhira kingdom in the north-western Deccan." He7 inclines to place them in central or western India - believing, as he does, that Sarnudragupta flourished in the 4th century A.D .'The moment we put this Gupta emperor in the 3rd century B.C. we have to pay attention to other facts noted by Sircar8 himself: "In one context the Mahabharata places the Ābhīras in Aparānta; but in another it associates the people with the Sudras, and assigns both the tribes to the land near vinaśena where the Sarasvatī lost itself in the sands of the Rajputana desert. The Ābhīras are also found in association with the Sudras in Patanjali's Mahābhāshya." Patanjali takes us to the pre-Christian era and the association he and the author of the Mahābhārata make for the tribe joins up with the one noticed in the Greek historians and suggests a new locale. For, as Sircar9 mentions: "In the second half of the fourth century B.C. when Alexander invaded north-western India , the
1.The Age of Imperial Unity, p. 162.
2.Ibid., pp. 50-1.
3.Mookerji, op. cit., p. 339, fn. 4.
4."Northern India after the Kushānas", The Age of Imperial Unity, p. 160, fn. 2.
5.Ibid., fn. 3.
6."The Deccan after the Sātavāhanas", ibid., p. 223.
7."Northern India after the Kushānas", ibid., p. 160, fn. 3.
8."The Deccan after the Sātavāhanas", ibid., p. 221.
9.Ibid., fn. 1.
Page 426
Śūdras, called Sodroi by the Greeks, lived in northern Sind to the south of the Punjāb rivers."
As regards the remaining tribes in the same list - the Prārjunas, Sanakānīkas, Kākas, Kharaparikas - Raychaudhuri1 says that their territories lay probably in Mālwā and the Central Provinces: "The Prārjunas," he adds, "are mentioned in the Arthaśāstra attributed to Kautilya [p. 1194] and are located by Smith [JRAS, 1897, p. 872] in the Narsinhapur District of the Central Provinces. A clue to the locality of the Sanakānīkas is given by one of the Udayagiri inscriptions of Chandra-Gupta II discovered in Eastern Mālwa. The Kākas find mention in the Mahābhārata [VI.9.64] - Rishika Vidabhāh Kākas Tanganāte-Paratariganāh". There is geographically nothing about them to make us doubt the location we have given to the others. Nor is there chronologically anything to demand a revision of our placing Samudragupta a little after Alexander's time.
The Correct Outlook on "Daivaputra-Shāhi-Shāhānushāhi"
When we come to the frontier foreigners of the inscription, the terrain seems to grow difficult, but a bit of keen scrutiny can carry us across. The first point to mark in Samudragupta's compound is the one made by D. R. Bhandarkar:2 "... the initial word is not Devaputra but Daivaputra, a Taddhita form which shows that the term cannot stand by itself but must be taken with what follows. If this is a correct view, Daivaputra had better be taken along not only with Shāhi but also Shāhānushāhi..." Samudragupta is referring to one single potentate rather than three in a row. The second point is to ask: "Does any Kushāna carry such a full royal title?"
Only the early members of the group - known as the "Great Kushānas" - are found to have styled themselves "Devaputra": Kujula-kara Kadphises whose territories were conquered by the famous Kanishka I,3 Kanishka I himself who is usually dated to the last quarter of the 1st century A.D. ,4 a nameless ruler of the Taxila Inscription of A.D. 79,5 and Huvishka, whose dates range between
1.The Political History of Ancient India, pp. 545-46.
2.Quoted by Mahājan, Ancient India, p. 401.
3.The Age of Imperial Unity, pp. 140, 141.
4.Ibid., p. 141.
5.Ibid., p. 140.
Page 427
A.D. 106 and 138 and who is alluded to in a Mathurā epigraph.1 Another epigraph speaks of "Shāhi Huvishka".2 Nobody except Kanishka I is styled both "Devaputra" and "Shaonanoshao" ("Shāhānushāhi"),3 though the latter title's Indian translation along with "Devaputra" is applied to others - e.g. Huvishka and the Taxila Inscription's unknown ruler - but at the end4 instead of the beginning as in Samudragupta's epigraph. Kanishka I, however, does not seem to have had "Shāhi" to accompany his two honorifics. The later members of the "Great Kushānas" whose "empire broke up some time after A.D. 230 or 240"5-assumed the title "Shaonanoshao" but not "Devaputra", nor were they called "Shāhi".6 The "Little Kushānas" in the 3rd quarter of the 4th century A.D. are still poorer in honorifics. Even the greatest of them - Kidāra - who is supposed to have been meant by Samudragupta bore only the title "shā".7 Neither "Devaputra" nor any equivalent of "Shāhānushāhi" is found on his coins. Thus Samudragupta's triple designation of a monarch whom our historians take for a Kushāna has no precise parallel in Kushāna history, and - what is crucial to our inquiry - two of the three components of that designation hang in the air if the king in question lived between 350 and 375 A.D.
Further, the Kushānas do not have the monopoly of the title whose Indian echo is "Shāhānushāhi". This echo occurs also in a mediaeval Jain book, Kālakāchārya-kathā, which speaks of Śaka invaders from the west of the Sindhu as "Shāhis" and their overlord as "Shāhānushāhi".8 As for the title in another form than the Indian, Rapson9 tells us about the Śaka and Pahlava rulers who preceded the Kushānas: "their normal style is 'Great King of Kings', a title which is distinctively Persian ; It has a long history from the Kshāyathiyānām Kshāyathiya of the inscriptions of Darius down to the Shāhān Shāh of the present day. " Rapson adds
1.Ibid., p. 150.
4.Ibid., pp. 150, 14-.
5.The Classical Age, p. 54.
6.Ibid., p. 51.
7.Ibid., p. 56.
8.The Age of Imperial Unity, p. 155 and fn. 3.
9."The Scythian and Parthian Invaders", The Cambridge History of India, I, p. 567.
Page 428
that in the Parthian empire the title was probably first assumed by Mithridates II the Great (123-88 B.C.) in imitation of his predecessors the Achaemenids. Rapson uses the word "probably" because, as he says in a footnote, "some numismatists attribute the Parthian coins on which the title first occurs to Mithridates I (171-138 B.C.)."
We may add an observation of Sircar1 about the Indo-Greek king Hermaeus "who flourished in the first half of the first century A.D." Sircar says that on some coins Hermaeus "is found in association with the Kushāna chief Kujula Kadphises". "The Kushāna chief is endowed with dignified royal titles in the legends of some of his later coins", but on the other coins he is mentioned in the legend without any royal title. This, according to Sircar, "no doubt points to his subsidiary position to the Greek 'king of kings' of Kābul during the early part of his life". Nor, we may finally remark, were the Indo-Greeks the sole Greeks to be considered "king of kings". The earlier Bactrian Greeks also enjoyed the same distinction. Eucratides,2 who ascended the throne almost at the same time as the first Mithridates mentioned above, has a coin with the legend Maharajasa rajatirajasa Evukratidasa.3
So we see that in the first few centuries before Christ the equivalent of "Shāhānushāhi" was in common use on the coins of various kings in the wake of the dissolved Achaemenid empire which had set the model for the title. The very title which the Guptas themselves bore - Mahārājadhirāja ("Supreme King of great kings") - is a concentration of the formulas prevalent in that age. And it is interesting to mark that often in that age the formulas figure in Greek. Thus the later coins of Maues, the Śaka (c . 20. B.C.-22 A.D:), have Basilêôs Basileôn along with the Kharosthi Rajadirajasa:4 Another Śaka, Ranjuvala (before 15 A.D.), carries the corrupt Greek legend Basilêi Basilêôs Sôteros, "King of kings, the Saviour".5 "A large number of coins," writes Sircar,6 "found all over the Punjāb as well as in Kandahār and in the Kabul valley bears a particular symbol and a corrupt Greek
1."The Yavanas", The Age of Imperial Unity, pp. 118, 138.
2.Ibid., p. 108.
3.Corpus Inscriptionum Indicarum, II, i. xxix n.
4.The Age of Imperial Unity, p. 126, fn. 1.
5.Ibid., pp. 134-35.
6."The Kushānas", ibid., p. 140.
Page 429
legend Basilêus Basileuôn Sôter Mêgas, 'the King of kings, the Great Saviour', sometimes with the Kharoshthī legend Mahārājasa rājadirājasa Mahātasa tratarasa, 'of the Great King, the King of kings, the Great Saviour.' The issue of these coins "is associated with Wema Kadphises by the use of the nominative instead of the more usual .genitive in the Greek inscription, and of the title Sôter Mêgas, as well as by the similarity in form between certain letters both in the Greek and Kharoshthī inscriptions." About the coins of Kanishka, the successor of Wema Kadphises, and about his family, Sircar1 informs us: "A striking feature of the coins of Kanishka and his successors... is that, unlike the issue of the earlier foreign rulers of India, they have no Kharoshthī legend on the reverse. Although the script of the legends is a corrupt form of Greek, the language is sometimes Persian; cf. the title Shaonano Shao..."
The use of the Greek language by Śakas and Kushānas or of the Greek script even when the language is different appears to associate pointedly the equivalents of ShāhānuShāhi, like the one on a coin of Eucratides between 171 and 150 B.C., with the presence of Greek rule in India and in her neighbourhood. This is but natural because Alexander and his military and political heirs were the immediate successors of the Achaemenids who had brought into fashion the formula "King of kings". Especially if an heir of Alexander ruled over the old Persian empire he would bulk in the Indian mind as the "Shāhānushāhi" par excellence. Such was Seleucus Nicator.
After the battle of Gaza in 312 B.C. he not only recovered Babylon but brought all the eastern provinces of Alexander's empire under his authority as far as India. Even after losing Alexander's Indian conquests to Sandrocottus and ceding to him in 305 B.C. the Greek satrapies west of the Indus, he remained king over all Persia right up to the north-eastern limit of the Jaxartes. Such too was the son of Seleucus, Antiochus I Sôter (born 324 or 323 B.C.) whom his father made king of the eastern provinces in 293 B.C. and who during this part sovereignty of Seleucus's empire as well as during his own rule of this empire from 281 to 262 B.C. held together those provinces. Antiochus I would be the contemporary of Samudragupta if the latter were identified with Amitrachates, the son of Sandrocottus. Master of
1. Ibid., pp. 147-48.
Page 430
all Persia, he would answer precisely to the designation "Shāhānushāhi" no less than "Shāhi", the pair of words that combines two of the several formulas we meet with in the epigraphs of Darius and Xerxes: "the King" and "King of kings".1 True, we have no coins or inscriptions of Antiochus I with this sort of double designation, but we are here concerned with how he was looked upon in India of the first quarter of the 3rd century B.C. and how his own ambassador Daimachus to Amitrachates represented him as ruler of the Persian kingdom bordering on the Indian monarch's territory. A further fact of great relevance is: "Antiochus I Sôter (324 or 323-262) was half a Persian, his mother Apame being one of those eastern princesses whom Alexander had given as wives to his generals in 324."2 No heir to Alexander's eastern conquests deserved more to be called "Shāhi-ShāhānuShāhi" by the contemporary Indian court.
Perhaps there will be a rather peevish objection: "You drew attention to the fact that none of the Kushānas combined 'Shāhi' and 'ShāhānuShāhi' but used only one or the other title. Now you are ready to accept Antiochus I as being indicated by the combination even though the double title was never associated with him. As in his case, why cannot some Kushāna be looked upon by an Indian king as both 'Shāhi' and 'Shāhānushāhi'?"
The objection forgets the highly relevant background of Persia and its monarchy for Antiochus I. The absence of the concerned combination is counterbalanced by the naturalness with which this background would legitimize such a double title for him on the lips of an Indian king. No Kushāna could claim to be the inheritor of the Persian crown which was distinguished by that combination, not to mention that none of the Kushānas had Antiochus I's special claim of being half a Persian.
What about "Devaputra"? Is not the appellation applicable exclusively to the Kushānas? Certainly not. First of all, it is a natural Indian word. With the meaning "God-son" it is found as early as the Rigveda (X.62,4). It also occurs in an entirely religious context in the Gupta period itself. Now the sense is dissimilar but the form is basically the same. An inscription of Budhagupta dated year 157 "of the Guptas" on a stone image of Buddha made by the
1.Olmstead, A History of the Persian Empire, pp. 175, 231.
2.The Encyclopaedia Britannica (1920), Vol. 20, p. 304.
Page 431
monk Abhayamitra speaks of him as seated with the gods as his sons (Devaputravato).1
Then there is a Brāhmī inscription from Mathurā, partly damaged, with no overt relationship to the Kushānas. The extant portion of it has been translated somewhat differently by V. S. Agrawala and D. C. Sircar.2 What is common to both the renderings is the expression: "Devaputra Magha." Sircar's general comment3 is: "In the present state of our knowledge, it is difficult to be definite regarding the identity of Devaputra Magha, although his title Devaputra as well as the palaeography of the inscription under review would point to his membership of Kanishka's house. In the epigraphic and numismatic records of the Kusana kings of India, we often find the epithet Devaputra associated with other royal titles. Curiously enough it has been used singly in the inscription under review. The absence of any regal title may suggest that Magha was a prince or subordinate ruler.... Elsewhere (The Age of Imperial Unity, p. 176) we have suggested the possibility of the Magha dynasty of Kausambi being 'founded by a lieutenant of the Kusanas (of Kanishka's house) named Magha'. Whether Devaputra Magha, mentioned in the Mathura inscription, had anything to do with the Maghas of Kausambi can hardly be determined in the present state of our knowledge." Sircar is honest in not forcing any inference about who this Magha was, but, while palaeo-graphically placing him in the Kushāna period, he is inclined to link him with the Kushāna Kanishka's house on the strength of the term "Devaputra". But in view of the fact that among the Kushānas it never stands alone we cannot unreservedly endorse the linking. Even the association of other titles need not Kushānize their bearer if a different identification is possible: a single title should definitely give one to think. It should suggest the likelihood of non-Kushānas being Devaputras.
By the way, we may point out in the palaeographical context that in The Age of Imperial Unity (pp. 175-77) Sircar chooses to date the Magha rulers of Kauśāmbī by the Śaka Era of 78 A.D ., which he attributes to Kanishka and by which some of the Magha dates are quite away from the currently believed Gupta times,
1.Mookerji, The Gupta Empire, p. 113.
2.Sircar, "A Brāhmī Inscription from Mathura", Bharatiya Vidya, Vol. XLV, 1953, pp. 6, 8.
Page 432
being as early as A.D. 129, 130, 159, 164 and 165 - although Altekar1 who is of the same opinion about the era has yet noted: "there is no doubt that the characters of the Magha inscriptions are almost the Gupta characters." So Sircar's palaeographical observation cannot be binding on our understanding of "Devaputra".
Finally, a lop-sided outlook on the appellation can result from concentrating not only on the Kushānas who have used it but also on its affinity with the old Chinese "T'ien-tzu" meaning "Son of Heaven". A new prospect opens before us as soon as we listen to Raychaudhuri's remarks:2 "The Greeks used titles like 'the divine king', 'the godlike queen', etc... Greek and Chinese influence is clearly discernible in the title of Devaputra." Yes, and the Greeks who used such titles were no other than Alexander and his successors. It may be worth while dwelling a little on the theme.
We read in the Encyclopaedia Britannica:3 "The Egyptians had ascribed deity to their kings, and were ready enough to add Alexander to the list. It is also not unlikely that the Persian kings had received some form of divine worship (L. Taylor, Journ. Hellenic Studies, 1927, pp. 153ff.). From the Greeks he certainly received such honours.... The offering of divine honours to the king which was begun under Alexander became stereotyped in the institutions of the succeeding Hellenistic kingdoms. The successors themselves received divine honours. Such worship might be organized and maintained by royal authority. The first proved instance of a cult of the latter kind is that instituted at Alexandria by the second Ptolemy for his father soon after the latter's death in 283-282, in which later, 279-278, he associated his mother Berenice also, the two being worshipped together as theoi Sôteres. Antiochus I followed the Ptolemaic precedent by instituting at Seleucia-in-Pieria a cult for his father as Seleucus Zeus Nicator. So far we can point to no instance of a cult of the living sovereign (though the cities might institute such locally) being established by the court for the realm. This step was taken in Egypt after the death of Arsinoe Philadelphus (271), when she and her still-living brother-husband, Ptolemy II, began to be worshipped together as
1."New Indian States in Rajputana and Madhyadesa", A New History..., p. 41, fn. 2.
2.An Advanced History of India, pp. 125, 126.
3.Vol. 16, pp. 568, 571.
Page 433
theoi adelphoi. After this, the cult of the reigning king and queen was regularly maintained in Greek Egypt, side by side with that of the dead Ptolemies. Under Antiochus II (261-246) a document shows a cult of the reigning king in full working for the Seleucid realm with a high priest in each province, appointed by the king himself; the document declares that the Queen Laodice is now to be associated with the king. The official surname of Antiochus II, Theos, suggests that he himself had here been the innovator. Thenceforward, in the Hellenistic kingdoms of the east, the worship of the living sovereign became the rule..."
We may observe that Ptolemy II Philadelphus (285-246 B.C.) is also said to have sent an envoy, Dionysius, to India. He may have come either to the court of Amitrachates or that of his son. But he must have carried to it the royal title, theos adelphos, for the king, signifying literally "Divine Brother" and generally "Divine Son" or "Son of God', Devaputra.
One more quotation on the theme - from the Encyclopaedia of Religion and Ethics:1 "Both Seleucus and his son Antiochus were worshipped by the Athenian colonists in Lemnos... It is probable that all through the epoch the dynasties of Seleucus and Ptolemy ruled in Asia and Egypt respectively, the Greek cities which were subject to them, and some which were merely allied, expressed their loyalty in a cult. We find at Ilion a priest of Antiochus I soon after his accession...; and a cult of the same king celebrated by Bargylia after his death and by the Ionian Confederacy during his lifetime."
Lastly, we have to determine when exactly the word "Devaputra" as referring to a foreign king was heard on Indian soil. Arrian (VII.2) writes about Alexander in India: "In Taxila, once, he met some members of the Indian Sect of Wise Men whose practice it is to go naked, and he so much admired their powers of endurance that the fancy took him to have one of them in his personal train. The oldest man among them, whose name was Dandamis (the others were his pupils), refused either to join Alexander himself or to permit any of his pupils to do so. 'If you, my lord,' he is said to have replied, 'are the son of God, why - so am I. I want nothing from you, for what I have suffices..."2
1.Edited by James Hastings (T & T. Clark, Edinburgh, 1935), Vol. 4. p. 527.
2.Arrian's Life of Alexander the Great, translated by Aubrey de Selincourt (The Penguin Classics, Harmondsworth, 1958), p. 226.
Page 434
This was evidently in answer to a command translated to Dan-damis (Dandi-Swāmī?) to meet Alexander, "Son of God". In fact, another account1 preserves the words of the command: "The son of the mighty god Zeus, King Alexander, who is the sovereign lord of all men, asks you to go to him..."
Thus in 326 B.C. Alexander's interpreter used the expression "Devaputra" to Dandamis who in his answer repeated it. India of Sandrocottus knew of Alexander as "Devaputra" and it is extremely probable that the practice of deification among Alexander and his successors led the successor of Sandrocottus, Amitrachates, to use "Devaputra" for both Antiochus I Sôter and Ptolemy II Phila-delphus, both of whom despatched envoys to the Indian court as respectful allies.
Hence, although "Shāhi-Shāhānushāhi" applied with certainty to Antiochus, can we say for sure whether "Devaputra" referred to him or to Ptolemy? There are a number of points which induce us to choose the former.
Antiochus I as Both "Devaputra" and "Shāhi-Shāhānushāhi"
(1)We cannot be definite that the ambassador of Ptolemy II came during the reign of Amitrachates rather than that of his son -and except through the ambassador we have no evidence of contact between Amitrachates and Ptolemy.
(2)Antiochus I had not only his Asiatic kingdom bordering on that of Amitrachates and an official relation with him through his ambassador but also a personal one through letters. Athenaeus (XIV, 67, 652F and 635A)2 gives us the anecdote of how Amitrachates wrote to Antiochus I, asking him to buy and have conveyed to him some sweet wine, some figs and a sophist to teach him to argue. Antiochus, forwarding the figs and the wine, explained that sophists were not a marketable commodity among the Greeks. The double relation, on top of his proximity, tilts the Bālānce in favour of Antiochus as against Ptolemy.
(3)Both the personal relation and the official were a continuation of the close contact established between Sandrocottus and Seleucus in 305 B.C. when they met at the Indus (Strabo, XV.724;
1.McCrindle, India as Described by Mêgasthenes and Arrian, p. 124. See also The Age of Imperial Unity, p. 554.
2.The Cambridge History of India, I, pp. 432-33.
Page 435
Appian, Syr., 55) - a contact furthered by frequent visits to the Indian court by Seleucus's ambassador Mêgasthenes (Arrian, V.6, 2) and by personal exchanges between the two kings (Athenaeus, O, 32Y, 18D).1 So it is not for us a choice merely between a newly formed friendship of Antiochus with Amitrachates and a similar friendship of Ptolemy with him. The contact of the former two is a natural development and sequel of an old family-friendship which is more likely than a recent one to figure in an important inscription assigned by us to Amitrachates.
(4) This old family-friendship has certain features which eminently qualify it to figure in that inscription's category of rulers or states or peoples who, although independent, were in a sense homage-bearing. Let us hear Mookerji2 tell how the friendship arose and what form it took: "While Chandragupta was busy laying the foundations of his empire in India, the Greek king Seleucus, who had succeeded Alexander in the eastern part of his empire, was moving towards India to recover the lost possessions of the late emperor. But while Alexander had to fight against a divided India, split up into a multitude of states, his successor had to face a united and a much stronger India organised by an able leader. Seleucus reached the Sindhu about 305 B.C. The Greek writers do not give the details of his conflict with Chandragupta, but merely record the result. Seleucus had to purchase peace by ceding to Chandragupta territories then known as Aria, Aracho-sia, and Paropanisadae (the capitals of which were respectively the cities now known as Herāt, Kandahār and Kābul), and probably also a part of Gedrosia (Balūchistān). In return Chandragupta presented him with 500 war elephants. The terms of the peace leaves no doubt that the Greek ruler fared badly at the hands of Chandragupta... The peace was ratified by a matrimonial alliance between the rival parties. This has been generally taken to mean that Chandragupta married a daughter of Seleucus, but this is not warranted by known facts. Henceforth Seleucus maintained friendly relations..."
We may add that Seleucus did not only transfer to Sandrocottus the Greek satrapies west of the Indus: he also acquiesced in the Indian's de facto sovereignty over all the country once held by
2."Chandragupta and the Maurya Empire", The Age of Imperial Unity, pp. 60-61..
Page 436
Alexander's prefects to the east of the Indus.1 Of course the transfer was the more important event. But Seleucus had come with the intention of recovering what Sandrocottus had wrested from Alexander's governors in India itself. And it is when we remember this that the transfer of Greek satrapies west of the Indus in addition to acceptance of the Indian's hold over the provinces on the other side takes on its proper colour and shows the homage-bearing attitude on the part of Seleucus. This attitude is hinted at in a remark of George Macdonald:2 "There may even have been a nominal and unmeaning acknowledgement of suzerainty.." The "matrimonial alliance" itself - the jus connubium, the right of marriage between the families of Seleucus and Sandrocottus - over and above all the territorial waiving of rights by the former is a sign of the acknowledged nominal suzerainty. We at once recall the phrase in Samudragupta's inscription about the offering of princesses to him by independent rulers to prove their respectful alliance with him. It is as if Antiochus I, among others, sent such an offering to Samudragupta through his ambassador -repeating what his father Seleucus had done in dealing with Sandrocottus, Chandragupta I.
The very despatch of an ambassador could fall within one of the four types of feudatory homage listed by Samudragupta. The first of them reads: "personal service" or "personal attendance".3 It would be performed by proxy when Daimachus came from Antiochus I to Amitrachates. And the homage-gesture would be accentuated by a striking fact repeated now from the time of Amitrachates's father. Just as Sandrocottus received Mêgasthenes from Seleucus without himself sending any representative in return, so too his son got Daimachus from Seleucus's son without a reciprocating move on his own part - a one-way transaction on the diplomatic level which could bear broadly the hue of .a vassal's relationship.
Antiochus I, therefore, fits perfectly into the context of the Allāhābād Pillar. Ptolemy does not bring the necessary background or relevant circumstance.-
1.George Macdonald, "The Hellenic Kingdoms of Syria, Bactria and Parthia", The Cambridge History of India, I, p. 431.
3.Raychaudhuri, The Political History,... p. 546 and Majumdar, A New History..., pp. 148, 149.
Page 437
(5)Although many of Alexander's generals shared in his empire after his death, it is only the family of Seleucus who stood in the very shoes of Alexander vis-à-vis India. Seleucus alone was concerned with the fate of the Indian province of his master and came, as it were, as a second Alexander. Consequently, when he and his son began to be deified and worshipped just like Alexander, they in particular would inherit in India the title by which Alexander made himself known on Indian soil: "Son of God", "Devaputra".
(6)Numismatics also renders the title especially appropriate to the Seleucids. We learn from the Encyclopaedia Britannica:1 "After his (Alexander's) death his deified portrait appeared on the coins of Lysimachus in Thrace and on the early coins of Ptolemy I in Egypt. It is not till 306 [B.C.] that we have a portrait of a living king on his coins when Ptolemy I appears, still as a god with the aegis of Zeus. Seleucus I similarly puts himself on his coins as Dionysus..." Now, Dionysus is well-known as the Son of Zeus by Semele, and the etymology of the very name is supposed to be: "dio - name of the Thraco-Phrygian sky god resembling Zeus; nys-possibly akin to Lat. nurus. Gr. nyos, and may mean 'child' or 'son'; hence perhaps 'son of god' ".2 It is very likely that Seleucus's coins came to the Indian court. Historians have noted that Antiochus minted some of his coins according to Indian measures in order to facilitate trade with India. And Seleucus's numismatic identification of himself with Dionysus, "Son of God", may easily have made his deified family carry at the Indian court the title "Devaputra" in direct and prominent continuation of Alexander's self-deification with the same term in India.
Furthermore, in connection with Antiochus I, there is here an item of nomenclature worth noting. We have mentioned Bhanr darkar's observation that the term "Daivaputra" instead of "Devaputra" is a Sanskrit Taddhita form. Now we may cite Raychaudhuri's remark:3 "As to the form Daiva see Achaemenian inscriptions of Xerxes, and forms like Bhaimarathi (instead of Bhīmarathī)." The first part of the remark points straight to Persia, where Daiva rather than Deva was the natural vocable, and, pointing to Persia, this part focuses our attention not on any Kushāna but on the half-Persian son of Seleucus.
1.Vol. 16, p. 619.
3.The Political History..., p. 546, fn. 7.
Page 438
So the equation is complete: Antiochus I is Samudragupta's "Daivaputra-Shāhi-ShāhānuShāhi".
Who were Earlier - the Kushānas or the Guptas?
Our next concern in strict relation to the Allāhābād Inscription is Samudragupta's phrase: "Śaka Murunda." But a small sidetrack is in order since we have put aside the usual association of the preceding phrase with the little Kushānas. These members of the Yueh-chi tribe were the successors of those Yueh-chis who first ruled in India, the Great Kushānas whom the current time-scheme sets from the beginning of the 1st century A.D . to the end of the 2nd: in other words, the Great Kushānas fairly precede the age of the Guptas as at present calculated from the era marking in our opinion their end and distantly succeed the Gupta Era we have computed to be their beginning. What is their proper place?
From a comparative study of the two dynasties' coins an impressive opening argument can be built for the priority of the Kushānas. We have to meet this argument squarely if we are to situate the Great Guptas earlier. Altekar1 indicates the case for the Kushānas very well:
"In the realm of numismatics, conservation was a very strong force in ancient times; the early gold issues of the Guptas, therefore, closely resemble those of the later Kushānas, to whose imperial position they succeeded in the north. The obverse of the Kushāna prototype, the king standing and offering incense at the altar, is very common in the earlier stages of the Gupta coinage. The Hindu king is also seen wearing the Kushāna overcoat and trousers. His name is written perpendicularly under the arm, as on the Kushāna prototype. The reverse is again a close copy of the Kushāna type, which has [the goddess] Ardoksho seated on a high-backed throne. The Kushāna monograms also reappear with only slight variations (cf. Kācha-type).
"There was, however, a definite move to Hinduise the type soon afterwards. The Greek legend on the Kushāna prototype was replaced by the Brāhmī one from the very outset. The peaked Kushāna cap was never put upon the head of the Gupta emperor. Ardoksho on the reverse was Hinduised usually by transforming her into the goddess Lakshmī and seating her on a lotus. Lakshmī,
1. "The Coinage", A New History..., pp. 302-03.
Page 439
that was thus introduced on the reverse, was destined to figure on the gold coinage for more than a millennium; we find her not only on the coins of the Chedis and Gāhadavālas, but also on some coins of Mahmudbin Sam.
"The Kushāna prototype, king offering incense at the altar, lingered on for a few decades, but side by side with it, the Gupta mint masters introduced a number of original artistic types. Samudra-gupta's new types were the Archer type, the Battle-axe type, the Conch type, the Tiger-slayer type and the Aśvamedha (Horse-sacrifice) type. Chandra-gupta II added the Lion-slayer, the Horseman and the Chhatra (Royal umbrella) types. The coins of most of these types are quite original in their conception and show no foreign influence whatsoever..."
R. K. Mookerji1 has a good summary formula along the same lines: "The degree of Indianization of the Gupta coins is a key to their chronology." But Mookerji next raises a question which is destined to prove most disturbing; He continues: "From this point of view, what are known as 'Chandra Gupta coins'... cannot be attributed to Chandra Gupta I because...they show a degree of independence of Kushān models which makes them later than several other types of coins issued by his successor, Samudra Gupta." Among the latter's coinage Mookerji picks out the "Standard type" and says: "This is the commonest type of Samudra Gupta's coins, the closest copy of Kushān coins, and, therefore, the earliest type of Gupta coins." He2 adds: "The Rev. is a downright copy of the late Kushān Ardochsho Rev. Ardochsho is seen seated, facing on a high-backed throne, holding cornucopia in 1. arm and fillet in outstretched r. hand."
Reverting to the "Chandra Gupta coins", Mookerji3 comments:
"If Chandra Gupta I had issued any coins, they should have been, as the earliest Gupta coins, of the Standard type... No such coins of Chandra Gupta I have been discovered. On the other hand, the Chandra Gupta type is more Indianized than the Standard type of Samudra Gupta, as shown (1) in the figure of the queen added on the obv. and (2) in the lion taking the place of the throne, though its dependence on Kushān technique is seen in traces of the back of the meaningless throne being still kept up.
1.The Gupta Empire, p. 31.
2.Ibid., p. 32.
3.Ibid., p. 33.
Page 440
"Further, the goddess seated on a lion first appears on the reverse of the Lion-slayer type of Chandra Gupta II coinage. Therefore, Chandra Gupta I type should be considered as the immediate predecessor of this type of coins of Chandra Gupta II."
Referring to the alliance made by Chandragupta I with the Lichchhavis whose princess Kumāradevī he married, an alliance "of which even his illustrious progeny was so proud, and constantly making mention in their inscriptions," Mookerji1 declares: "The importance of the alliance...is also celebrated by the issue by Samudra Gupta of special coins commemorating the event..." The "Chandra Gupta coins" are ascribed to the son of Chandragupta I and Kumāradevī and this celebration of his Pārents' marriage is dated to the end of his reign, just before his own son, Chandragupta II, ascends the throne. A rather odd timing, one would think. But how is it possible to avoid it?
Now comes the surprise of the whole situation - from no other historian than Altekar himself who has laid down the same principles of Gupta coin-chronology as Mookerji. He2 tells us:
"With the assumption of the imperial title Mahārājadhirāja Chandra-gupta I started his gold coinage. The view of Allan3 that the coins bearing the figures of Chandra-gupta and his queen Kumāradevī on the obverse are commemorative medals struck by Samudra-gupta is untenable. Chandra-gupta I owed his imperial status, to a considerable extent, to the valuable assistance that he had received from the Lichchhavi relations of his wife; it was, therefore, but natural for him to issue throughout his reign a joint coinage, whose type would be acknowledging the Lichchhavi help in a suitable manner. Had Samudra-gupta issued these coins as commemorative medals, the name of the commemorator would naturally have appeared somewhere upon them, as it does on the admittedly commemorative issues of Agathokles and Eukratides. As it is, Samudra-gupta's own name is conspicuous by its absence on these coins.
"On the obverse of the coins of Chandra-gupta I, we see the king and his queen Kumāradevī standing and facing each other; the king is probably giving the marriage ring to his consort. The names of both are engraved by their sides. On the reverse there is
1.Ibid., p. 15.
2.Op. cit., pp. 301-02.
3.Catalogue of Coins of the Gupta Dynasties...(London, 1914), pp. Ixiv-lxviii.
Page 441
Durgā seated on a lion. The legend is Lichchhavayah, which is probably in acknowledgement of the help the Guptas had received from their Lichchhavi relations."
Majumdar1 lends his weight to Altekar's thesis: "Mr. Allan's contention that these [coins] were struck by Samudra-gupta to commemorate the marriage of his Pārents is hardly convincing. The view, held by the old numismatists, that these coins were issued by Chandra-gupta jointly with the Lichchhavis and their princess Kumāradevī, who was his consort, has been very ably defended by Dr. Aiyangar and Dr. Altekar.2"
Raychaudhuri3 chooses to strike a Bālānce: "There is a difference of opinion among scholars regarding the attribution of these coins... It is difficult to come to any final conclusion till the discovery of coins whose attribution to Chandra Gupta I is beyond doubt."
Obviously the argument à la Mookerji has not completely satisfied Raychaudhuri; but he has not explained how one would attribute coins to Chandragupta I in an indubitable way. Mookerji4 begs the question: the coins have to be "the closest copy of the Kushān original". He refuses the possibility of any independence at the start of the Guptas. Evidently, Raychaudhuri does not agree: else he would not hold the matter to be still undecided. What would be the alternative to Mookerji's demand? Could Raychaudhuri mean coins with Chandragupta's name yet without the Kumāradevl-Lichchhavi motif and the lion-poised Durgā? But how would this motif and the goddess Simhavāhanā in themselves render any coin dubious? Unless one makes up one's mind beforehand that the slightest departure from the Kushāna model signifies a late period, there is no conceivable criterion. Raychaudhuri's hesitation to come down on one side or the other implies lack of critical thought on the subject and must involve at the back of his mind a recognition of Chandragupta's right somehow to differ from the so-called "Kushān original".
As we have hinted, to date - as Mookerji does - the "Chandra
1."The Rise of the Guptas", A New History..., p. 128.
2.Numismatic Supplement to the Journal and Proceedings of the Asiatic Society of Bengal New Series, Calcutta, XLVII, pp. 105 ff; Journal of Indian History, Madras, VI, Suppl. pp. 101 ff.
3.The Political History of Ancient India, Fifth Ed., 1950, p. 530, fn. 4.
4.Op. cit., p. 32.
Page 442
Gupta coins" as "the immediate predecessor" of "the Lion-slayer type of Chandra Gupta II coinage" where appears "the goddess seated on a lion" is a queer reading of Samudragupta's intention. Samudragupta is made to celebrate his Pārents' marriage at the close of his own life and after he had issued his own various coin-types, instead of at the very beginning when as the son of his father's Lichchhavi wife rather than of any other of less important standing he would justify the event recorded in his Allāhābād Pillar Inscription, namely, his father's selection of him as successor in preference to other princes who are said to have resented it. The extremely late dating is absurd in the context of Samudragupta's life. A less pointless occasion cries out to be discerned.
Raychaudhuri's ambiguous neutrality' diminishes no whit the strength of Altekar's and Majumdar's position and a realistic assessment of historical circumstances strengthens it. Nor is Aiyangar the sole other supporter. Vincent Smith,1 ascribing the coins to Chandragupta I as a sign of his recognition of dependence on his wife's people, gives the footnote: "That seems to me the natural interpretation of the coin legends. Mr. Allan, of the British Museum, regards the coins as having been struck by Samudragupta in honour of his parents, a view which I cannot accept." A. L. Basham, who was entrusted with revising a part of Smith in the 1970 edition of the latter's Oxford History of India published after his death, leaves Smith's opinion intact and unanhotated unlike his procedure at some places where he comments in the light of new research. This tacitly makes Basham a concurring party. Even in 1954 Majumdar2 could boldly inform us: "The view that the coins were issued by Samudra-gupta.. .is no longer held by any scholar."
But if the anti-Mookerji team is right, the entire theory that the Guptas imitated the Kushāna coins and progressively Hinduized them suffers a shock. Altekar does not seem to realize how his brief for Chandragupta I lands him in serious self-contradiction. It appears impossible for him to give up his powerful criticism of Allan: it should be equally so for us to accept his own expose of what he terms conservatism in the realm of numismatics and to acquiesce in Mookerji's criterion of chronology for the Gupta coins. Consequently, the view that the Guptas followed the
1.The Oxford History of India, Third Ed., 1970, edited by Pert, al Spear, p. 165, fn.
2."The Rise of the Guptas", The Classical Age, p. 4, fn. 2.
Page 443
Kushānas in time is, to say the least, flung into the melting-pot.
This conclusion cannot be averted by arguing that the position of Kumāradevī to her husband's right on the coins goes against all suggestion of the husband as their issuer. True, A. K. Narain1 has shown by quoting various authorities that Indian custom demands the wife to be invariably to the husband's left - except at marriage. But what Narain should lead us to infer is simply that we cannot interpret the coins as a normal joint issue of Chandragupta I and Kumāradevī indicative of a joint rule. Provided we do not so interpret them, there is nothing to make us carry them over to any point in Samudragupta's reign. Obviously they are commemorative of the occasion of marriage. As such they can be ascribed to the first Gupta and our conclusion in relation to the time of the Kushānas can stand.
Neither can the reference to a remnant of Ardoksho's portrayal - "the back of the meaningless throne" - help a return from the melting pot to the solid state. What looks like a fragmentary back may not be one at all: it may be some additional feature we cannot explain. If we compare it to what we have in Samudragupta's Standard type which has Ardoksho in full - we can perceive that the mint-master has an appreciably different detail: he shows four vertical poles instead of two short as in the standard type. Only our preconception may conceal from us some subtle meaning here. Moreover, is the high back itself characteristically Kushāna? Mookerji2 has said: "Both the Tiger and Lyrist types are the most Indian of Samudra Gupta's coins." Yet about the Lyrist type he3 writes: "The Obv. shows King seated... on a high-backed couch..."
An extra point of sufficient importance to be made - when we keep in sight the so-called full Ardoksho of the Standard type - is that Altekar4 is wrong to use the adjective "Roman" in speaking of "the Roman goddess, seated Ardoksho" figuring in the coins of Kanishka III (c. 180-210 A.D. on the assumption that Kanishka I flourished around 78 A.D., the conventional Śaka Era). Perhaps the Romanizing of Ardoksho is by sympathy with the figure of "Roma" on the reverse of the coins attributed to a successor of
1.The Journal of the Numismatic Society, Vol. XIV, 1957, Part II, pp. 135-44.
2.Op. cit., p. 36.
3.Ibid., p. 35.
4."The Punjāb, Sindh and Afghanistan", A New History..., p. 15.
Page 444
Kanishka I: Huvishka (between 106 and 138 A.D.)?1 As the Roman influence in India cannot go too far back in the pre-Christian centuries and is certainly later than the immediate post-Alexandrine age, a Roman Ardoksho would militate against placing Chandragupta I and his successor as early as we have done. But actually it is a mistake to Romanize this goddess.
Richard N. Frye,2 the specialist in matters Persian, writes of Kushāna coins "where Hellenistic or Roman deities, such as Heracles, Hephaestos, Serapis, are pictured as well as Irānian gods and goddesses such as Mithra, Ardoxsho, Atar and Verethragna". Another expert, J. Duchesne-Guillemin,3 dealing with the same subject, lists "Ardoxsho" strictly among the Irānian divinities that came to the fore with the general Irānianization of coin-legends in the later part of Kanishka's rule. At another place,4 this expert informs us: "Some have sought to draw important conclusions from what appears to be a female Mithra [on a Kushāna coin]. The name is certain. But as the type is exactly the same as that of Ardoxsho, a woman carrying a cornucopia, Cumont concluded this as an engraver's mistake. However, Bussagli and Gnoli would rather see in this a tendency, of which there are other instances, to merge the two deities." If Ardoksho is definitely not Roman and if out of four lines two lingering as though from the back of her throne are the sole sign suspected of her vaguely haunting from the Kushāna time the coinage of Chandragupta I, this coinage which throws into confusion the hypothesis of the derivation of Gupta numismatics from the Kushānas may be taken to prove that the Guptas need not have been posterior to the Kushānas. They may be considered as affected by Persian ideas in general for a time in the wake of several centuries of Persian sovereignty to the immediate west of the Indus before Alexander's invasion and then by the Graeco-Persian cultural influence coming with that invasion and increasing with the contact - attested by a number of Classical reports - between Sandrocottus's dynasty and the Persia-possessing semi-Persianized royal family of Seleucus Nicator.
1.Sircar, "The Kushānas", The Age of Imperial Unity, p. 150.
2.The Heritage of Persia, p. 230.
3.The Religion of Ancient Irān, translated by K. M. JamaspAsa from the French (Bombay, 1973), p. 165.
4.Ibid., p. 166.
Page 445
Vincent Smith1 cites one instance of Persia's cultural influence from Mêgasthenes's account of Sandrocottus: "We are told... that the ceremonial washing of the king's hair was made the occasion of a splendid festival when the courtiers offered rich presents to the king. That observance recalls the Persian hair-washing ceremony on the sovereign's birthday, described by Herodotus." In the same context, Smith, holding that though the Arthaśāstra attributed to Kautilya belongs as a whole to a date later than Megasthenes's, some parts of it point to the Greek ambassador's period, remarks that the Arthaśāstra's rule about the king's consulting physicians and ascetics while being "seated 'in the room where the sacred fire has been kept' seems to be an indication that Magian ritual was honoured..." If he is right, we have an independent confirmation of Persia's cultural influence which would be deducible from the incense-fed fire-altar on the Gupta coins. But here we may discern a mingling of the foreign influence with India's own practice of the fire-cult. In any case there is no call to dub this feature of the coins a relic from the Kushānas.
The coat-and-trousers can certainly be acounted for by Persian influence in the immediate post-Alexandrine epoch. Besides, to regard them as nothing else than a blind copy from Kushāna numismatics would be absurd. Would the Guptas show themselves to their subjects in coat-and-trousers on their coins if they never wore them at any time? Altekar2 naturally infers "from the effigies of the Gupta emperors on some of their coins" that "overcoats, coats and trousers were often worn by the Indian kings". Furthermore, the coins do not always picture the kings as peace-time figures who would wear dhoti and turban but frequently as warriors and conquerors. In dealing with the time of Sandrocottus Mookerji3 has drawn our attention to "a nearly life-size figure of an infantry soldier armed as described by Megasthenes" which "appears among the sculptures at Bhārhut which are generally taken to date from the age of Aśoka": that is, the age in which Aśoka is conventionally put - c. 269-232 B.C. - but which for us would be part of Samudragupta's reign. The soldier4 wears what
1.The Oxford History of India, 3rd Ed., p. 103.
2."Social and Economic Conditions", A New History..., p. 353.
3.Chandragupta Maurya and His Times, p. 276.
4.The Age of Imperial Unity, Plate XVI, Fig. 37.
Page 446
Mookerji1 in his book on the Guptas calls a "close-fitting cap" such as we see on the heads of both Chandragupta I and Samudragupta in the early coins, and there is armour on the legs and no trace of anything like a dhoti. In his earlier book Mookerji2 goes on to quote Cunningham's report of military figures sculptured at Sānchī, usually dated to the 1st century A.D. "In one of them," says Cunningham, "... the soldiers wear a tight-fitting dress and kilt... The swords are short and broad and tally with the description of Megasthenes." As for the Kushāna dress, Smith,3 employing the Kushāna tribe-name, says: "The Yueh-chi princes... dressed in long-skirted coats..." The coat of Kanishka, the greatest Kushāna, as seen in his coins as well as in his famous headless statue,4 comes down below the knees. But, in the Chandragupta I coins and in Samudragupta's Standard or Battle-axe type, the coat-front stops at the waist.5
A wholesale imitation of the Kushānas is nowhere to be marked even in the early Gupta coinage. The peaked Kushāna cap is absent everywhere. In the Standard type itself, which Mookerji deems the closest to the Kushāna model, he6 has observed not only a quite non-Kushāna "close-fitting cap" but also the Garud? (Eagle) which is unmistakably a Gupta symbol and he has called the altar "Tulasīvrindavana, a completely Indian feature". In "some specimens" he notes "the king wearing shorts and full socks", and in all specimens "the jewellery worn by the king is Indian".7 The "standard" itself is not otiose. Mookerji8 writes: "... the Standard indicates the conquered territories where the flag of victory was planted. It is also appropriately associated with the legend Parākramah ['Mighty']."
Finally, is it indeed Ardoksho on the Gupta coins? Mookerji et al. use that name because the Kushāna coins have it for the same or similar figure. The Gupta figure, which Mookerji9 identifies variously as Durgā, Lakshmī or Sarasvatī on account of suggestive
1.The Gupta Empire, pp. 31, 33.
2.Chandragupta Maurya..., p. 276.
3.Op. cit., p. 155.
4.Ibid., Plate ii facing p. 182, Plate 8 facing p. 155.
5.The Gupta Empire, Plate I facing p. 24, Plate II facing p. 25.
6.Ibid., p. 31.
7.Ibid., p. 32.
8.Ibid.
9.Ibid., pp. 31-36.
Page 447
signs, need not be designated Ardoksho at all. The only item which relates her to Ardoksho is the cornucopia in her right hand. Because of the cornucopia Sircar,1 noting Ardoksho on the reverse of Huvishka's coin-types, terms her "the goddess of abundance". Mookerji2 particularizes the cornucopia as "the horn of plenty, the horn of the goat Amalthea by which Zeus was suckled" and which "is represented in art as overflowing with flowers, fruit and corn" and "thus... is a pre-eminently foreign feature" on the Gupta coins as well as on the coins of the Kushānas. Undoubtedly, the cornucopia is a Greek touch, but in the immediate post-Alexandrine age it would be very natural, as much so as Indo-Persian ones. Hence the cornucopia could easily go with some Persian goddess who could be fused with Indian female divinities and who could be a pre-Kushāna counterpart of the Kushāna period's Ardoksho.
Duchesne-Guillemin3 indicates Ardoksho to be the the same as the Avestan goddess Asi Vanuhi of "good fortune" mentioned in Yasht 17 as bringing "riches... to the houses of the devout". There is also, as Duchesne-Guillemin4 tells us, the Avestan Pārendī, "goddess of abundance", who corresponds to "the Vedic Pūramdhi". Whatever superhuman power the name "Ardoksho" stood for in - to quote Duchesne-Guillemin5 again - "Bactriana of the Kushāns", where she "enjoyed great popularity", is covered precisely by those two feminine deities of the Avesta. It is they who, in combination with the Greek cornucopia-motif and with the traits of Indian goddesses, take on Gupta coins the form which reappears as Ardoksho on Kushāna issues and which Mookerji, guided by the present-day chronology, dubs "a downright copy of the late Kushān Ardochsho". The moment we turn the chronology the other way round and look about us with eyes sensitive to the manifold cultural currents mingling in an earlier epoch, the picture totally changes.
All in all, on the strength of the coin-depictions there is no reason to make the main initial Guptas start after the Great Kushānas and flourish on the heels of the Little Kushānas. We may well regard the Kushāna dynasty as imitating in a number of
1."The Kushānas", The Age of Imperial Unity, p. 150.
2.Op. cit., p. 32.
3.Op. cit., p. 28, 141.
4.Ibid., p. 128.
5.Ibid., p. 141.
Page 448
respects the Guptas on some of their coins, rather than vice versa.
The weights of the coins have little significance in themselves apart from the depictions. About the Kushāna coinage we have statements like Smith's1 that Kadphises II struck "gold coins agreeing closely with the Caesarean aurei in weight", or Sircar's:2 "Some scholars believe that Kanishka's gold coinage was suggested by the Roman solidus..." The impression we get is not of any inevitability of influence: the Kushānas might have adopted on their own the weight they did. The choice of weight depends on convenience and capacity and mere resemblance does not argue straight copying. In fact, the aureus weighs 124 grains or 8.035 grammes,3 while the Kushāna coins "weigh 120 grains on the average".4 The early Gupta coins have the same weight as the latter,5 but this too need not be mere imitation and the tendency towards originality manifests itself in a marked difference from the Kushāna practice over the years. Altekar6 observes: "The weight standard of the later Kushāna and Scythian coins is the same as that of the earlier Imperial Kushānas." In contrast we may study Altekar's account7 of the weight of the Gupta gold coins: "... in the reign of Chandra-gupta II it was raised to 124 gr. Some types of Kumāra-gupta weigh as much as 132 gr. Skanda-gupta raised the weight to 144 grains and thus transformed his gold coins into suvarnas of the traditional Indian standard of eighty ratis." In the light of this trend of change the initial near-similarity to the Kushāna weight is of little moment.
But the straight leap of Kumāragupta's coins from 124 to 132 grains cannot but be meaningful since it immediately joins them to ancient Persian coinage, "the famous gold darks with the figures of kneeling archer on them".8 Mookerji9 relates: "The standard gold coin of ancient Persia was the Daric, weighing about 130 grains, probably first minted by Darius [522-486 B.C.]." And it is curious that the largest variety of coin-legend among Kumāra-
1.The Oxford History of India, Third Edition, p. 158.
2."The Kushānas", The Age of Imperial Unity, p. 144, fn. 2.
3.Ibid., p. 139.
4.Altekar, "The Coinage", A New History..., p. 296.
5.Ibid., p. 304.
6.Ibid., p. 296.
7.Ibid., p. 304.
8.Frye, op. cit., p. 140.
9.Chandragupta Maurya and His Times, p. 350.
Page 449
gupta's gold issues comes in the Archer type, 6 kinds as against 5 in the Horseman type and 4 in the Lion-slayer type.1 The Daric affinity should not be surprising. Of all the Persian kings after Cyrus, who were connected with the Indus region, Darius I had most to do with it and there is extant even a Daric struck in India in c. 337 B.C.2 But the Daric affinity whisks us away from the post-Kushāna period to a substantially pre-Kushāna one.
What about the Dīnāra?
The only point that can cause some concern is the use, by the Guptas, of the coin named the dīnāra. We may be told: "The dīnāra is based on the Roman denarius aureus and the latter could not have been known in India earlier than about the last quarter of the 1st century B.C. because the earliest denarii found by archaeologists in India are of Augustus whose reign started in 31 B.C."
But is it quite certain that the Indian dīnāra comes from the Roman denarius? F. Spiegel3 tried to derive, from the Greek form "denarion" of the Roman coin, the term "danare" found in the Avesta for a dry measure or weight. But surely "danare" is older than "denarion"? We may suggest that the Indian dīnāra originally comes not from the Roman denarius but from the Avestan danare. From danare we can derive a modified form daenara,4 then the Old Persian dlnnar, the Indian dīnāra and the modern Persian dinar. If we commence the Guptas from 315 B.C., there should be no historical incongruity in this etymology. And from the fact that the Gupta epigraphs speak of the dīnāra only in the time of Chandragupta II (e.g., No. 7 of Fleet's Gupta Inscriptions5) we may guess danare to have denoted at the start the weight to which this emperor raised the initial 120 grains: 124 grains. As we can gather through Altekar's account, the dīnāra-value did not stay constant in the hands of the Guptas and touched the daric-value after a while until ultimately, as Fleet's No. 64 proves,6 the
1.Mookerji, The Gupta Empire, pp. 86-88.
2.Smith, op. cit., PL 5, fig. 1.
3.Commemoration Volume I, p. 363.
4.Geiger, Civilisation of the Eastern Irānians in Ancient Times, translated by Darab Dastur Peshotan Sanjana from the German, Vol. II, p. 137.
5.Mookerji, op. cit., p. 51.
Page 450
Irānian-sounding dīnāra became interchangeable with the Hindu-named suvarna, a gold coin of "eighty ratis" - quite a quantity away from the Roman denarius aureus of a fixed 124 grains.
In the period of the Kushānas when the gold coins they struck were of a weight almost that of the denarius aureus, which by then had become fairly current in the East, the word "dīnāra" got associated with the foreign currency. The association does not disqualify danare, through daenara, from being the real Pārent of the Gupta dīnāra.
Rather, we should say "one of the real parents". For, danare brought explicitly the element of weight along with its name. The material aspect - both stuff and size - of the dīnāra together with the general sound of the coin's name may be traced to a gold ornament or a gold seal referred to in works like the Unādi-sutra as "dīnāra". According to the author of the Unādi-sutra, the word is derived from "dlna" which means "poor", and "ri", indicating "to go", the whole connoting "what goes to the poor". By the conventional chronology, Brāhmanic literature in the style of the Sutras is roughly dated between 600 and 200 B.C. With the Guptas starting in 315 B.C. the date would naturally go further into antiquity. No matter how much of a late-comer, this "dīnāra" could serve as part origin of their coin.
Even if we granted the coin's derivation from the Roman denarius of 124 grains and said that its weight underwent change in the course of time, we would not need to say that the Roman currency reached India only as late as the reign of Augustus. The heavy denarius is a considerably old coin. At first it was minted of silver in 268 B.C., and later the gold denarius was of the same weight as the silver.1 No doubt, the earliest gold denarii we have discovered in India are of Augustus, but it is not true that no Roman coins earlier than those of Imperial Rome have been found here. As M. S. C. Vidyabhusana2 admits, "several Roman coins of the Consulate period have been discovered in the Manikjāla stūpas and in the Hazara district of the Punjāb". These coins are said to have been "very probably brought to India by traders several years after they had been prepared in Rome". Vidyabhūsana postulates "traders" because, according to him, "it is almost
1.Webster's New International Dictionary of the English Language (1956), p. 697, col. 2.
2.Journal of the Royal Asiatic Society, 1906, p. 1.
Page 451
certain that Rome did not attempt to spread eastwards till the later years of the Consulate" and "it was only in the reign of Augustus that the conquest of Asia by Rome began". But there is no reason to think that no one except Roman traders could have brought the coins to India or that stray Roman traders could not have done it much before the reign of Augustus. Through such agents or those non-Romans who were at the same time in contact with Rome and with India, the Consulate coins could very well have come. Against the background of the Manikjāla stūpas and the Hazara district of the Punjāb no convincing evidence exists that an influx of pre-Augustan denarii before the 1st century B.C. was not possible.
In fact, the possibility of sporadic trade-contact or else occasional contact of culture between Rome and India is from 264 B.C. onward, for it is in 264 B.C. that the first Punic War took the Romans to Africa and it is at about the same time that Ptolemy II Philadelphus of Egypt is known to have sent an ambassador to India and to have had cultural communication with this country through Indian books in the Alexandria Library.1 We know,2 further, that Indian processions used to go through Ptolemy's streets and Ptolemy's successors too were in touch with things Indian. Again, a Greek colony, with multi-national elements, flourished in Arachosia (Kandahār region) at this time. Since the earliest Roman denarii go back to 268 B.C., the above circumstances would allow the arrival of a few of these coins into India through various channels and their imitation by the Guptas who, if Sandrocottus was Chandragupta I, may be expected to have been greatly interested in the Western world.
The earliest Gupta mention of the dīnāra seems to have been in the Gadhwa Stone Inscription of Gupta year 88 (Fleet's Gupta Inscriptions, No. 7).3 By our Gupta Era of 315 B.C., the date resolves into 227 B.C., a full 41 years later than the striking of the first silver denarius of 124 grains in Rome and sufficiently subsequent also to the various favourable circumstances we have already listed. So, even if the dīnāra derives from the denarius, we do not have to wait till Augustus's reign for it to appear in India in some form or other of rich metal, and the Guptas do not have to
1.Bhandarkar, Aśoka, p. 163.
2.The Age of Imperial Unity, p. 616.
3.Mookerji, op. cit., p. 51.
Page 452
be posterior to the Kushānas for the opportunity of imitating it.
The Archaeological Question
What remains is the archaeological question. Do the Kushāna levels at excavated places always underlie the Gupta ones and what are the signs which are assumed to prove the posteriority in time? If it is merely the presence of coins, the case is not established, since coins could be hoards from an earlier age. If it is other artifacts, we must first know for sure what are Gupta artifacts. Evidently, scholars do not take very seriously the bare archaeological precedence. Although "it has been pointed out... that at the excavations at Taxila, coins of the Kanishka group of Kushāna kings were found in the upper (i.e. later) strata of the earth and those of the Kadphises group in the lower (i.e. earlier) strata", "according to some scholars, notably Fleet, Kanishka was a predecessor of Kadphises I and Kadphises II and was the founder of the era of 58 B.C. which ultimately came to be known as Vikram-Samvat".1 At present the majority of scholars do not favour Fleet and company in this respect, but archaeology as such is hardly a decisive factor against them.
Moreover, in the context of our stand, we must bear in mind that our run of the Gupta dynasty - with some temporary breaks in certain localities such as those where the Kushāna power arose - is from 315 B.C. to 320 A.D., so that some Gupta material would be prior and some posterior to the Kushānas.
The archaeological question does not admit of a simple answer. Besides, archaeology itself is at times a complicated affair. The strata have to be clean-cut and exclude the possibility of a mix-up. Instances are there of radical reversals of archaeological reading. The case of "the walls of Jericho" in relation to the period of the Israelite conquest of Canaan comes from the books of Kathleen Kenyon as a great lesson in complete chronological somersault.2
A Chronological Caution in Passing
Although not directly relevant to our discussion, a certain point
1.Sircar, "The Kushānas", The Age of Imperial Unity, p. 144 and fn. 2.
2.See especially Digging Up Jericho (Ernest Benn Ltd., London, 1957), pp. 44-46, 261-62, 271.
Page 453
is worth making in relation to the Great Kushānas because rarely do historians take note of it when adjusting their chronology. One must not overstrain oneself to establish their time-bracket - from the beginning of the 1st century A.D. to the end of the 2nd. Thus we may be told: "The title Kaisera (Caesar) adopted by Kanishka of the Āra inscription of the year 41 points to a date considerably later than Augustus Caesar who died in A.D. 14".1 But we may answer in the words of S. K. Dikshit:2
"As to the title of this Kanishka (II), both Prof. Luders and Dr. Sten Konow read it as '[Kajisara', of which the first letter is supposed to be doubtful and can be acutally only conjecturally read. In this connection Mr. N. G. Majumdar states: 'An examination of the stone, which I undertook in February 1924, left no doubt as to the reading of the last four letters "isarasa". We are afraid that it is just possible that these Western scholars have introduced into this reading a conception of their own, and presumed that this late Kusana ruler assumed in imitation - which is the best way of flattery - the Roman title of Caesar. It is strange that such a late emperor should be introducing such an original title when none of his predecessors did it, or anything like it. We believe that there is as little reason for presuming that this ruler assumed the Roman title, as there is for presuming that the Indians of the times of the Kusana emperors used not Indian months, nor even Chinese months, but Macedonian months.' There is probably no 'k' preceding the letters, clearly verified by Mr. N. G. Majumdar, and these letters may therefore stand for nothing else but 'Iśvarasya'."
Dikshit gives the footnote: "Cf. MBh (Bori), V.33.2 where Iśvara appears as a descriptive title (?) of a certain Mahārāja." We may add a fact from the records of the Kushānas themselves: "The Kharoshthī legend on Wema Kadphises' coins has the expressions sarva-loga-iśvarasa-mahiśvarasa.3 The ground sought for the chronology of the Ara inscription's Kanishka by a backlook at Augustus Caesar seems an utterly misguided ingenuity.
1.Sircar, op. cit., p. 144.
2."The Problem of the Kusanas and the Origin of the Vikrani Samvat", Annals of the Bhandarkar Oriental Research Institute, XXXVIII. 1957. pp. 100-01.
3.The Age of Imperial Unity, p. 140.
Page 454
The Correct Outlook on "Śaka-Murunda"
Now we shall attend to Samudragupta's "Śaka-Murunda". How shall we relate it to the India of the 3rd century B.C.?
It is seldom realized that to India's north-west, just beyond the Indus, there were Indianized Śakas living in the time of Alexander and even before him. A clue to them comes in the river-name Maśakāvatī which the grammarian Patanjali mentions in his Mahābhāsya (11.287) and which the Kāśika connects with the considerably earlier Pānini along with other river-names like Pus-karāvatī (IV.2.85; VI.1.219; VI.3.119). Maśakāvatī reminds us, in the first place, of the Massagetai or Massagetae whom the geographer Ptolemy1 (c. 140 A.D.) lists among the Śakai and puts along the range of Askangka, and whom Herodotus (1.201) too relates to the Śakas (Scythians). Maśakāvatī reminds us, in the second place, of the locality Maśakā which the Mahābhārata, as Agrawala2 informs us, includes in its Śakadvīpa together with several other geographical entities like Chaksu (Oxus) and Kumud (Komedia of Herodotus, a mountain in the Śaka country). The Mahābhārata's Maśakā in its turn takes us at once to the time of Alexander, which we consider to have immediately preceded Samudragupta's. We hear of the Assakenoi (Asvakayana) who fought fiercely with the Macedonian invader of their capital which the Greeks (Arrian, Anabasis, IV.26) called Massaga. Agrawala3 equates the Greeks' Massaga with Pānini's Maśakāvatī, just as Mookerji4 does. And when we bear in mind S. N. Majumdar's observation5 - "The Massagetes were the Maśakās described in Sanskrit Literature as the warrior tribe of the Śakas" - we realize that the people on India's frontiers whose capital was Massaga in the late 4th century B.C. and therefore also in the early 3rd, where we have put Samudragupta, were a section of the Śaka Massagetae.
A proof of this is the way they fought Alexander. According to J. H. Haskins,6 scholars, while fully accepting the Massagetae as
1.McCrindle, Ancient India as Described by Ptolemy, edited by S. N. Majum-dar,1927, pp. 284, 318.
2.Op. cit., p. 68.
3.Ibid., p. 73.
4."Foreign Invasions", The Age of Imperial Unity, p. 45.
5.McCrindle, Ptolemy, p. 397.
6."The Royal Scythians", Natural History (New York), October 1960. p. 16.
Page 455
Scythians, distinguish them in one particular from other Śakas such as the Western tribe whom Herodotus called the Royal Scythians and reported to have proceeded westwards after their defeat in Asia by the Massagetae. This particular is the position of women among them. The women of the Massagetae stood very high. Haskins1 writes: "We are explicitly told [Herodotus, I. 214] that they were led in battle by a queen." Now let us glance at the battle given by the Assakenoi army of Massaga to the Macedonians. Mookerji2 relates: "This army was led by the late King's mother, queen Cleophis... The example of the queen-commander leading the struggle for freedom in person brought the entire womenhood of the locality into the fighting..."
Evidently, Samudragupta's Śakas can be linked to the Massagetae of the hilly regions in the valley of the Kumār, Panjkore and Swat.
The expression "Murunda" may be taken, as by many scholars,3 to be only a Śaka word meaning "lord". Then a separate definition of it would not be required. But we should prefer to regard it, with many other scholars,4 as the name of a distinct tribe. The Chinese5 have spoken of the capital of Meou-lun (a word equated with Murunda) situated 7000 li from the mouth of the Great River, undoubtedly the Ganges. And Ptolemy's Geography6 also speaks of the Maroundai as living a little above the Ganges-delta. But how shall we think of a Murunda-tribe in north-western terms when we know of its existence in the east?
McCrindle7 opines that originally the Murundas may have been "a people of Lampāka (Lamghān) at the foot of the Hindu-Koh" who brought gifts to Samudragupta, but he immediately pulls himself back to condemn, with the modern chronology in mind, the theory that Ptolemy's Maroundai were people who, driven from the valley by enemies, had crossed the Indus and advanced southwards till they had established themselves on the Ganges. He tells us that Saint-Martin had disproved this theory "since... Samudragupta... reigned subsequently to the time of Ptolemy, and
3.A New History..., p. 148.
4.Ibid., p. 147.
6.McCrindle, op. cit., p. 213.
7.Ibid.
Page 456
they could not therefore have left their ancestral seats before he wrote". Here is a dilemma.
Looking at Ptolemy's Maroundai as well as the Chinese Meoulun, we cannot help considering with historians like Raychaudhuri1 Samudragupta's Murunda to be connected with them. And yet modern chronology renders it impossible to explain the connection satisfactorily: how could a tribe which was earlier on the Ganges be later in Lamghan? As soon as we desist from placing Ptolemy nearly two hunderd years before Samudragupta instead of nearly-four hundred and forty years after him, the whole difficulty vanishes. We can then consistently connect the people of Ptolemy with the tribe of Samudragupta and think of the former as finally settling on the Upper Ganges after a forced migration into India from Lampāka where they had lived as neighbours to the Śakas of Massaga in the 3rd century B.C. to which we have dated Samudragupta. In the matter of the Murundas, both Samudragupta and Ptolemy are best explained by means of the Purānic chronology.
As for the presence of the Murundas for Alexander's historians, they may be understood as dubbing this tribe no less than the Śakas of the trans Indus area the Assakenoi, because of the racial affinity between the two tribes and their common Indian-Irānian culture.
Yes, Samudragupta's "Śaka-Murunda" can be defined for the India of the early 3rd century B.C. in north-western terms. And if we did not have to look beyond Samudragupta we might stop here . But these terms cannot exhaust his category . His category must cover the Western Satraps also - the Śakas of Western India whose rule was abolished by Samudragupta's son Chandragupta II. If, as claimed , the Western Satraps dated themselves by the Śaka Era of 78 A.D., how can we get Samudragupta into any relation with them in the period before 260 B.C.?
Samudragupta, Chandragupta II and the Western Satraps
Majumdar2 says: "It is a significant fact that the long series of coins testifying to the almost unbroken rule of the Western Kshatrapas for more than three hundred years comes to an end between A.D. 388 and 397 and is replaced by coins of similar design issued
1.The Political History of Ancient India, p. 547.
2."The Vikrama Sarhvat and Śakabda", A New History..., p. 167.
Page 457
by Chandragupta II." As Chandragupta's dates1 show that he reigned from the 56th year of the Gupta Era to the 94th, his reign on Fleet's theory would fall between 376 and 414 A.D., precisely covering the last dates of the Śakas of Western India.
In the face of such a conjunction, seemingly devastating to our theory, our counter-attack may be made in Majumdar's own words2 elsewhere about the Śaka Era of 78 A.D.: "... the era is not associated with the Śakas for the first five hundred years or more when it is simply called Varsha." This clearly means that the Western Satraps during their career of more than 300 years dated themselves in an era which they have not specified at all. Historians have simply assumed that .these Śakas have used the so-called Śaka Era of 78 A.D. So far as the Śakas' own dating goes we are not pinned down to this era.There is no objection inherent in their own dating to our carrying them where our chronology would situate Chandragupta II. And then they could be in relation to Samudragupta in the period before 259 B.C. and be included in his "Śaka-Murunda".
Perhaps the idea that the Western Satraps were already almost three hundred years old in India when Changragupta II came to the throne in 259 B.C. seems shockingly to contradict accepted history's teaching:3 "The earliest independent Scythian king of Indo-Scythia seems to have been Maues (Moa, Moga)... (c. 20 B.C.-A.D. 22)..." About him and all who followed him in time it is said:4 "There is no doubt that the Śaka occupation of the western part of Northern India was principally the work of the Śakas of eastern Irān ... The nomenclature of the early Śakas in India show an admixture of Scythian, Parthian and Irānian elements. This no doubt suggests that the Śakas, before their entry into India, lived for a considerable period of time in the Irānian Śakastān under Parthian rulers when they must have also received a good deal of admixture of blood." As regards the Śaka settlement in eastern Iran , "some scholars believe that, after the dispersal. of the Śaka tribes from the Oxus valley of the Yueh-chi , their main movement, checked by the Greek kingdom of Kābul , went westwards in the direction of Herāt and thence southward to Seistān, It is further
1.Ibid., pp. 166, 172.
2."The Vikrama Sarhvat and Śakabda", The Age of Imperial Unity, p. 157.
3.Sircar, "The Śakas and the Pahlavas", Ibid., pp. 124, 125.
4.Ibid., p. 121.
Page 458
pointed out that the tide must have been actually flowing from the time when the Śakas had been displaced by the Yueh-chi from their home beyond the Jaxartes in the second century B.C."1 How the Śakas who poured into Eastern Irān came thence to India is explained:2 "The Parthian emperors who were then in possession of eastern Irān struggled hard with the Scythians, and Phraates II (138-128 B.C.) and Artabanus I (128-123 B.C.) lost their lives in the strife. It was Mithridates II (123-88 B.C.) who finally subdued the Śakas. The tide of Scythian movement, thus checked by the Parthians, ultimately flowed towards the valley of the Sindhu."
No doubt, all this has historical substance at its core; but the history of Śaka entry into India is not exhausted by the fact of an influx from a Parthianized eastern Irān, and that history has several mysterious points. The account of the displacement of the Śakas from beyond the Jaxartes by the Yueh-chi we owe to the Chinese3 and, according to the Chinese, "the Śakas were... successively driven by the Yueh-chi from the valley of the Jaxartes to that of the Oxus, and then to Ki-pin."4 Ki-pin is either Kāpisa (Kāffiristān) or Kāshmir or a combination of parts of both.5 The earliest Śaka ruler of Ki-pin, by the Chinese chronology, dates to a little before 73 B.C.6 But, says Sircar,7 "the relations of the Śakas of the Ki-pin country... with India proper as well as with the Śakas of eastern Irān, who occupied wide regions of western and northwestern India, are unknown." May it not be that we have no knowledge of the relations because the Śakas of Ki-pin and those of western and north-western India belonged to widely separated epochs?
Another point of mystery is created by the phrase Śaka- Yavana in Patanjali, who is at present dated to 175 B.C.8 As Patanjali speaks of the Śakas as having been, like the Yavanas, absorbed into Indian society, Bhandarkar9 notes that by 175 B.C. the Śakas must have established themselves in the north-western portion of
1.Ibid., pp. 120-21.
2.Ibid., p. 121.
3.Ibid., p. 122.
1.Ibid., p. 123.
8.Ibid., pp. 121-22, fn. to p. 121; Bhandarkar, Indian Culture, I, pp. 279-80.
9.Op cit.
Page 459
India if not within Āryavarta proper. So neither Maues nor any possible scion of the Ki-pin Śakas could have been the first of their kind in India - that is to say, east of the Indus . Patanjali's Śakas may have been part of the Aryan tribes whose degradation is reported by Indian tradition to have occurred in ancient times or else they may have been sufficiently settled descendants of an early invasion.
Still more of an eye-opener - the most dazzling point of mystery, so to speak - is what Agrawala1 suggests when he draws attention to the Śaka origin of the Sanskrit word "kantha" for "town". All scholars know that it stems from the Śaka " kand" 'or "kanda" as found in the place-names in the Oxus region most associated with the Śakas in old times, place-names like Samarkand, Khaqand, Chinkand, Tashkent, Panjkand, Yarkand. But Pānini, according to Agrawala, gives a string of kanthā-ending place-names in his own Uśīnara country in the heart of the Punjāb: Chihanākanthā, Madarakanthā, Vitulakanthā, Patatkakanthā, Vaidālikarnakanthā, Kukkutakanthā, Chitkanakanthā, the first one in sūtra VI , 2.125 and the rest in Gana. "How such names could be in India," comments Agrawala, "is an unexplained problem. It points to an event associated with Śaka history even before Pānini , probably an. intrusion which left its relics in place-names long before the Śaka contact with India in the second century B.C." And Agrawala's surmise is supported by J. Przyluski's conclusion on an independent inquiry:2 "In India there was a Scythian (Śaka) immigration long before the time of Panini which caused a tribal welter in the Punjāb."
History is full of complexities undreamt-of by historians wedded to a pet theory, in the light of which they want to simplify everything. There is no a priori objection to thinking of a Śaka entry into India before Maues. And in the period preceding the 4th century B.C., we have the preconditions for such an entry.
First, we do not need to restrict Śaka occupation of eastern Iran to the time of the Parthian rule there from Mithridates I (c. 150 B.C.) onward. The inscription of the Achaemenid emperor Darius (522-486 B.C.) at Naksh-i-Rustam in c. 515 B.C. refers to no less than three settlements of the Śakas who were his subjects: the
1.India as Known to Pānini, p. 68 f.
2."Nouveaux Aspects de l'histoire des Scythes" Revue de I'Universitaire de Bruxelles, Vol. 42 (1937), pp. 209, 218-223.
Page 460
Śaka Tigrakhauda (Śakas with pointed helmets), Śaka Hauma-varga and Śaka Taradaraya (Śakas over, or beyond, the sea).1 The first-named are probably also mentioned in Achaemenid records "as those who lived beyond Suguda or Sogdiana (modern Bukhara region) and in the works of Herodotus as the neighbours of the Bactrians."2 The second group "has been identified by Thomas with the Scythian settlers of Drangiana in the Helmund valley, which afterwards came to be known as Śakastari (the land of the Śakas; Sanskrit Śakasthana; medieval Sijistān; modern Seistāns)"3 The third kind , inhabitants "of the land beyond the sea are usually believed to have been those who dwelt in the Russian steppes to the north of the Black Sea".4
The Helmund valley is eastern Irān, which is almost next door to India. So there were Śakas nearby already in 515 B.C. And it is these Śakas who have to be taken as overflowing towards the valley of the Sindhu and coming to our notice as the Kshaharāta and Kārdamaka families. And there is absolutely nothing in these families' nomenclature that bears any inevitable Parthian stamp forcing us to put them in the Parthian period of eastern Irān. The only problem for us is to fix as precisely as possible their earlier date and see whether it falls before the Naksh-i-Rustam inscription and, if it does, whether they fit any of the three Śaka categories listed there and, if they do fit, what circumstances are responsible for the fitting.
At the very outset we may note that it is unconvincing to equate the Śakas "of the land beyond the sea" to the Scythians of the Russian steppes north of the Black Sea. Herodotus (IV. 1-145) in his detailed narrative of Darius's invasion of Scythia to the north of the Black Sea tells us how foiled and frustrated the Persian army was by the Scythians' tricks and how Darius for the safety of his army had to leave this territory without really conquering it. So these Śakas could not be his subjects and we may legitimately ask if the Kshaharātas and Kārdamakas could be his Śaka Taradarayā. All depends on where these families flourished. Sircar5 cogently
1. The Age of Imperial Unity, p. 120.
2. Ibid.
3. Ibid.
4. Ibid.
5. "The Śaka Satraps of Western India", The Age of Imperial Unity, pp. 179, 181. 182, 184.
Page 461
infers from the archaeological data available that they ruled not only in Sind but also in Malwā, Gujarāt, Kāthiāwār and western Rājputāna. The inscriptions of the Kārdamaka chief Chashtana have been found at Andhau in Kachchha.2 Mankad3 has rightly opined that the Śakas living in Kachchha and Kathiawar "could very appropriately be described as Śakas across the ocean, i.e. the Indian Ocean".
But can Darius be considered the overlord of these Satraps? Herodotus (IV.44) tells of a naval expedition despatched by Darius in 517 B.C. under Scylax of Caryanda to explore the Indus down to the sea, and then sail to Egypt. Herodotus continues: "After these persons 'had sailed round, Darius subdued the Indians and frequented the sea ." Mankad1 thinks the account is not properly worded here. In correction he cites The Cambridge History of India (I, p. 336) on Scylax's expedition: " .. .it seems much more likely that Darius must previously have won by force of arms a firm hold over the territory traversed from the headwaters of the Indus to the ocean, in order to have been able to carry out such an expedition." Mankad agrees with the Cambridge History that Darius was already in possession of the Indian country up to the mouth of the Indus, i.e. the whole of Sind. What , however, he adds is that the words of Herodotus - "Darius subdued the Indians" - should not merely be considered misplaced. His argument is that, though the entire Indus could not have been explored without previously subduing the Indians all about the region, Darius carried out a further subdual and this conquestis expressed by Herodotus while the other is left implicit. Mankad suggests that the Indian regions Darius conquered after possessing the whole of Sind must have been Kachchha and Kāthiāwār. That Darius did not go to the east of Sind is clear to Mankad from the statement of Herodotus (III. 94-98) that to the east was a desert (the Rājputāna desert). Therefore, the Indian regions Darius subdued should be to the south of Sind: Kachchha and Kāthiāwār - the countries where the Western Śakas had their main seat. So Darius from 517 B.C. could have been their overlord
Mankad believes they actually came to India in the wake of
1.Ibid., p. 183.
2.Op. cit., p. 185.
3.Ibid., pp. 184-85.
Page 462
Darius's conquest - that is, the conquest he made of what the inscription at Persepolis (c. 518-515 B.C.) and the Hamadan Gold and Silver Tablet inscription as well as the inscription at Naksh-i-Rustam call Hi(n)du.1 Darius's earlier records, the Behistun inscription (520-518 B.C.) and the Susa palace inscription, mention Ga(n)dāra but not Hi(n)du as one of his provinces.2 Hi(n)du is usually understood to have been the northern Punjāb,3 but the narrative of Herodotus directs us to Sind or at least leads us to include Sind in that term. Perhaps Herzfeld4 is right in suggesting that the Punjāb is shown by another province-name of Darius -Sattagydia - which Herzfeld derives from Sapta Sindhava, so that Hi(n)du would be Sind, the region of the lower Indus river.
It is tempting to concur with Mankad and have quite a neat picture. But certain chronological considerations militate against it. First is the starting-point of the era the Western Satraps follow. Its initial date may be roughly guessed by putting together three facts we have marked in Part One of our book. (1) The coins of the Satraps end in the year 310 of their era. (2) The reign of Chandragupta II ends in the 94th year of the Gupta Era. (3) The coins of Chandragupta II replacing those of the Satraps were issued in the 90's of the Gupta Era. Evidently, the Satrapal coins ended a little before the years 90-94 of the Gupta Era. If the founder of the Guptas mounted the throne in 315 B.C., those coins ended a little before 225-221 B.C. Then, since they ended in the year 310 of the era of the Satraps, this era must have started somewhat prior to the period reached by adding 310 to 225-221 B.C.: that is, somewhat prior to 535-531 B.C.
Now, at the time Mankad wrote, the earliest date known of the Kshaharātas was the year 41 of an era and of the Kārdamakas the year 52. With these dates as sole evidence of the Satraps' presence in India we could easily take them to have arrived in India on the heels of Darius's addition of Hi(n)du to his empire. C. 517 B.C. could be their starting-point in the new country. And, if their era commenced a little before 535-531 B.C., the earliest known year -41 - of it could be in India after c, 517 B.C. But not long ago a new inscription of Chashtana was unearthed at Andhau and it is dated
1.The Age of Imperial Unity, P.41.
4.Archaeologische Mitteilungen aus Irān, I (Berlin, 1929), p. 99.
Page 463
the year 11.1 That is indeed a far cry from the year 52. With their era definitely preceding 535-531 B.C., the Satraps are thus proved to have been in India before the reign of Darius. They must have arrived during the reign of one of the two predecessors of Darius, either Cambyses (530-522 B.C.) or Cyrus (558-530 B.C.). As to their era, its initial year may have begun in their home-land to celebrate some event or it may have celebrated their advent in the new country.
Here we may be asked: "Is there any corroboration from outside the Kshaharātas and Kārdamakas that a Śaka Era was counted from some such year?" In history, certain dynasties may follow eras to which we cannot give any name from outside sources. Thus some scholars take the Maghas of Kauśāmbī to have followed an unknown local era.2 An outside corroboration is unnecessary for the historical reality of any era. But, of course, if we can obtain that corroboration, the historical reality acquires a brighter and clearer look.
Unbelievable though it may seem, we have not only a general corroboration independently of the Kshaharātas and Kārdamakas but also, from an outside source, a precise specification of the initial year of their Śaka Era. We have only to revert to a topic already treated when we refuted the claim that the traditional Indian chronology is internally inconsistent and that the dating of the Kaliyuga and the Bharata War does riot come to us in a single coherent voice. We dealt with a pronouncement of Varāhamihira, the author of the famous astronomical works, Panchasiddhāntikā and Brihatsamhitā.
Varāhamihira is the first Indian astronomer to use the phrase "Śaka Era". He has been understood as alluding to 78 A.D. But an entirely new vision dawns on us when we carefully consider it in the context of the most often quoted sentences in the Brihatsamhitā (XIII, 1-4) which refer to the earlier astronomer Vriddha Garga, the traditional cycle of the Seven Rishis and the "period" of King Yudhishthira. Varāhamihira's Śaka Era proves to be not 78 A.D. but 551 B.C. which is eminently compatible with our
1."Andhau Inscription of Castana, Śaka 11" by Shobhana Gokhale, Journal of Ancient Indian History, Vol. 11, parts 1-2, 1968-69 (University of Calcutta), pp. 104-115.
2.The Age of Imperial Unity, p. 175, fn 3.
Page 464
calculation that the era of the Kshaharātas and Kārdamakas should be a little before 535-532 B.C.
The result is astonishing and caps, crowns and clinches all we have said about the time of the Western Satraps. What remains is to tie up a few loose ends about them. Here we shall deal in some greater detail with a topic we have treated before in the wake of our treatment of Varāhamihira.
Commencing the era of the Western Satraps in 551 B.C., we can conclude from Chashtana's inscription of the year 11 that by (551-11=) 540 B.C. they were already in India, and we know from this date that their entry into the new land was during the reign of Cyrus - 558-530 B.C. Did it have any connection with his activities?
Frye1 considers it not unlikely that Cyrus, before "he turned his attention to Lydia, probably in 547 B.C.", was "engaged in expeditions in eastern Irān" in order "to consolidate his power". Xenophon, on the other hand, says in his Cyropaedia (VIII, 6, 20-21) that "Cyrus, after reducing Babylon, 'started on the campaigns in which he is reported to have brought into subjection all the nations from Syria to the Erythraean Sea' (i.e. the Indian Ocean)".2 Babylon, as Frye3 tells us, was taken in 539 B.C. Here Xenophon contradicts himself, for it is evidently on a certain statement of his own that Frye bases his supposition of pre-547 B.C. expeditions. Xenophon (VI.2, 1-11) has the account of an embassy sent to Cyrus by an Indian king. "This embassy conveyed a sum of money for which the Persian king had asked, and which ultimately served him in a delicate matter of espionage before the war against Croesus [of Lydia] and the campaigns in Asia Minor."4 We must aver that Cyrus made two expeditions to the east, one before the conquest of Lydia and the other after. But Xenophon's chronology for the second is incorrect and Stems from a misunderstanding of Herodotus.
What Herodotus (1.153) says is that Cyrus, after organising Persian rule in Lydia, "started eastward on his march to Ecbatana, taking Croesus [whom he had captured but spared] with him. He
1.The Heritage of Persia, p. 105.
2.A. V. Williams Jackson, 'M he Persian Dominions in Northern India down to the Time of Alexander's Invasion", The Cambridge History of India, I, p. 330.
3.Op. cit., p. 106.
4.The Cambridge History of India, I, p. 331.
Page 465
did not think the Ionians important enough to constitute a primary objective, for his mind was on Babylon and the Bactrians and the Sacae and the Egyptians, against whom he intended to lead an expedition in person."1 Herodotus recounts how Cyrus's representative, Harpagus, dealt with the Ionians, and then (1.177) adds: "While Harpagus was turning upside-down the lower, or western part of Asia, Cyrus was engaged with the north and east, bringing into subjection every nation without exception. Most of his minor conquests I will say nothing about, but mention only those of his campaigns which gave him the greatest trouble and are in themselves the most interesting."2 Actually Herodotus recounts only two campaigns: that against Babylon and the one against the Śakas of the Massagetai tribes, living eastward of the Caspian Sea. But it is illuminating to note two statements of his.
He (1.190) writes that after conquering Assyria, which means the reducing of Babylon, "Cyrus's next desire was to subdue the Massagetae".3 The campaign against these Śaka was the last in the life of Cyrus , for , according to Herodotus (1.214), he lost his life in it. Thus the reduction of Babylon and the unsuccessful fight with the Śaka east of the Caspian Sea occupied the years 539-530 . B.C. Whatever else Cyrus accomplished was, therefore , done before 539 B.C. And that is precisely the testimony of Herodotus (1.178): "Having subdued the rest of the continent, he turned his attention to Assyria ..."4 Herodotus did not deem the pre-Babylon conquests worth telling us about , but he is quite clear that they occurred between the Lydian-Ionian engagement (547 B.C.) and the reduction of Babylon (539 B.C.). The whole eastern campaign in the direction of India - subsequent to the subjugation of Lydiais to be accommodated within these 8 years.
What were his achievements here? The Greek writers on the campaign of Alexander the Great hold that Cyrus never reached India - meaning that he never crossed the Indus.5 But nobody denies his drang nach Osten. Pliny refers to Cyrus's conquest of Kapiśa (in the Ghorbund valley); and there is the assertion of
1.The Histories, translated by Aubrey de Selincourt (The Penguin Classics, Harmondsworth, 1960), pp. 75-76.
2.Ibid., p. 85.
3.Ibid., p. 95.
4.Ibid., p. 85.
5.The Age of Imperial Unity, p. 40.
Page 466
Arrian (Indica I, 1-3) that the Indians between the rivers Indus and Cophēs (Kābul) "were in ancient times subject to the Assyrians, the Medes, and, finally, to the Persians under Cyrus to whom they paid the tribute he imposed upon them".1 From the Behistun inscription (c. 520-518 B.C.), the earliest record of Darius the successor of Cyrus's son Cambysses who is credited with no campaign in India's neighbourhood, we may infer that the mention of "Gadāra" (Gandhāra) as one of the provinces of his empire indicates his inheriting this province from Cyrus.2 "According to Xenophon (Cyropaedia, I. 1.4), Cyrus 'brought under his rule Bactrians and Indians' and extended his sway up to the Erythraean sea."3
As the Indian Ocean (Erythraean Sea), in Xenophon's further words, "bounded the empire of Cyrus on the east",4 we may well surmise that the lower Indus valley no less than more northward areas skirting the Indus were part of his dominion. No doubt, the lower Indus valley was lost not long after, for Darius had to conquer it afresh; yet, while Cyrus's sway lasted, assuredly covering the period up to 539 B.C., since he started the Assyrian war after "having subdued the rest of the continent", the Kshaharātas and Kārdamakas must have passed into Sind and then into Kachchha and set up as Kshatrapas and Mahākshatrapas. These designations they considered equivalent to Rājan, a title which, whether they called themselves Kshatrapa or Mahākshatrapa, they invariably used,5 and it really denoted their essential status. For all practical purposes they were independent rulers.
If they came into India with the help of Cyrus, they must have nominally acknowledged him, during his hold on Sind, as their overlord. Similarly, when Darius recovered that province and pushed on southward they must have looked upon themselves as his representatives.6 But their practical independence is proved not only by the term Rājan but also by what Sircar further tells us
2.Ibid., p. 41.
3.Ibid., p. 39.
4.The Cambridge History of India, I, p. 330.
5.The Age of Imperia Unity, pp. 180, 182.
6.An inscription of Nahapāna (Rapson's Catalogue, pp. lviii, clxxxv) of the period 41-46 (=515-510 B.C. according to our Śaka Era), which refers to a gold currency, must be alluding to the darics of Darius. Nāhapāna's own coins were in silver and copper (The Age of Imperial Unity, p. 180).
Page 467
about them. Sircar,1 who would like to show them as feudatories to the Kushānas, has yet admitted that no record of the chief Kshaharāta figure, Nahapāna, refers to his overlord. Nor is there any evidence of who the overlord of the Kārdamakas was. Writing on Rudradāman, Chastana's grandson, Sircar2 notes "the absence of any reference to the Kushānas in the records of his time as well as his claim that he himself assumed the title of Mahākshatrapa".
The two families must both have hailed from the same place, the Helmund valley in eastern Irān, first established themselves in the valley of the lower Indus and then penetrated into Kachchha and Kāthiāwār, reckoning their new history in terms of a common era of 551 B.C. which they had either brought from their homeland or started on entering India. What the era meant and how it happened to be common is perhaps best explained by a hypothesis of Mankad's. Mankad3 writes:
"From the coins and inscriptions of these Ksatrapas two families have been traced (i) Bhūmaka-Nahapāna and (ii) Zāmotika-Castana-Jayadaman-Rudradāman, etc. Some scholars consider these two families as unconnected with one another, but Lévi, Sten Konow and some other scholars have suggested that the names Bhūmaka and Zāmotika [also read as Ysāmotika] are identical and refer to one and the same person. They think that ysam or zam of Ysāmotika or Zāmotika is originally the Irānian-Scythian word meaning earth and that Bhūmaka is its Sanskritised form."
Nahapāna and Chashtana would then be brothers and their common era would mark the rise to power of their father under Cyrus during the Achaemenid emperor's first expedition to eastern Irān and India's borderlands before 547 B.C. We have done enough to show the legitimacy of our new scheme of the Śakas . There are , however, four significant challenges to be met. One concerns the contact of the Kshaharāta Nahapāna and the Kardamaka Rudradāman with the Āndhra-Sātavāhana dynasty.
1."The Śaka Satraps of Western India," The Age of Imperial Unity, p. 181.
2.Ibid.,
3.Purānic Chronology, p. 182.
Page 468
12
The Śakas and Gautamīputra Sātakarni
Nāhapāna's last inscriptional year is 46 of the era his family followed.1 We are brought down to (551-46)505 B.C. in our time scheme. There is no inscription or coin of any Kshaharāta after this date. The Kshaharāta coins which Gautamīputra Sātakarni of the Sātavāhana dynasty restruck, as the Jogalthembi hoard proves, are of Nāhapāna and no other later ruler' - a fact which justifies the claim made for Gautamīputra that he " uprooted.the Kshaharāta dynasty".3 Gautamīputra's fight with the Kshaharāta dynasty is evidenced by hisinscription from Nāsik in his 18th regnal year.' The inscription "records the grant of some land that is stated to have been in possession of Rishabhadatta..."5 Rishabhadatta was a Hinduised Śaka chief - originally Ushavadāta - who "was the husband of Dakshamitrā, daughter of Nāhapāna, and was the viceroy in the southern province of his father-in-law's dominions".6 "It is interesting that the order regarding the grant was issued from a 'victorious camp of the army that was gaining success', and that the Sātavāhana king is represented as stationed at the time at a place called Banākataka (probably situated on a river called Benā) in the Govardhana (Nasik) district. Apparently Gautamīputra's presence in that region at the head of an army was connected with his campaign against the Kshaharātas, which led to the liberation not only of the Upper Deccan but also of considerable parts of western and central India. :"
The usual inference drawn from all these facts is that "in the latter part of, or shortly after, the Śaka year 46" Nāhapāna 'seems to have been defeated and killed"! and that the defeat and slaughter were effected by Gautarniputra "in or shortly after the eighteenth regnal year of the Sātavāhana king".9 The inference does not follow. Gautamīputra's son Pulumāvi's last inscription is in his 24th year: this does not prevent historians from making him live for 5 more years." Nāhapāna could very well have died after Gautamīputra's 18th regnal year if that year fell later than the
1. The Age of Imperial Unity, pp. 180, 182.
2. Ibid., p.80, fn. 1.
3.Ibid. 7. Ibid., p. 201.
4.Ibid., p. 201. 8. Ibid., p. 182.
5.Ibid. 9. Ibid., p. 201.
6.Ibid., p. 181. 10. Ibid., p. 204.
Page 469
Śaka's inscription of the year 46. There need be no practical coincidence of the two dates.
It may also be that Nāhapāna died in the year 46 without having been defeated and killed by Gautarniputra and that his natural death without any male heir left his son-in-law in charge of his empire. Gautamīputra may have extirpated Rishabhadatta and by destroying the last male member connected with Nāhapāna brought aboutthe end of the Kshaharāta dynasty. It is to be noted that Gautamīputra does not mention Nāhapāna: he mentions only Rishabhadatta.
In the light of the true situation, all that we have to see is whether our Purānic dating brings Gautamīputra into just the right relation needed for his contact with Nāhapāna.
The general consensus of the Purānas makes Gautamīputra the 23rd Āndhra-Sātavāhana and after him the consensus is for 7 more kings, thus running the total to 30.' In the course of ascertaining which Purānic number out of several for the full dynastic period of the Āndhras agreed with the Sapta Rishi calculation about them, we had occasion to count the reign-lengths of these terminal 7 according to Pargiter:2 28+7+3+29+6+10+7==90 years. As we have ended the Āndhras in 390 B.C. after commencing them in 802 B.C., Gautamīputra's reign should close in (390+90==) 480 B.C. With a reign of 24 years, a number given by one version of the Matsya Purāna and accepted by modern scholars, he came to the throne in 504 B.C. Ruling in 504-480 B.C., he must be a contemporary of Nāhapāna whose last inscription was in 505 B.C. Gautamīputra's 18th regnal year would be (504-18==) 486 B.C. Nāhapāna would either be killed in the wake of the ouster of his son-in-law Rishabhadatta in that year or he would have died issueless - of old age before it and Rishabhadatta's overthrow would spell the end of the Kshaharāta dynasty.
We may even bring Gautamīputra's victory over Rishabhadatta closer to Nāhapāna's last inscription by a slight readjustment which modern research demands and the Puranic chronology mostly permits. Pargiter3 admits that his 3 years for the 26th Āndhra (Śivaskandha) is more or less conjectural. Sircar4 counts 7
1.Pargiter, The Purāna Text of the Dynasties of the Kali Age, pp. 36-7.
2.Ibid., pp. 71-2.
3.Ibid., p. 25, Note 7; p. 71, note 19.
4."The Sātavāhanas and the Chedis", The Age of Imperial Unity, p. 205.
Page 470
years, which seem to be a variant found in the Purānas themselves. It is possible also to add one year to Vāsishthīputra Pulumāvi who succeeds Gautamīputra. Sircar1 says that according to the Purānas he."ruled for twenty-eight or twenty-nine years". The Pulumāvi who is the last Āndhra can also have at least 1 more year, since an inscription of his, discovered at Myakadoni in the Bellary District, is dated in his 8th regnal year.2 So we can date Gautamīputra to 510-486 B.C., shift the Nāsik epigraph to 492 and reduce by 6 years the time between it and 505 B.C. Either way, our chronology proves itself adequate.
What about the other Śaka family, the Kārdamakas? The third of them, Rudradāman, grandson of Chashtana, is the author of the famous Junāgarh inscription dated the year 72. By our Śaka Era the year resolves into (551-72=) 479 B.C. Rudradāman "claims to have twice defeated Sātakarni, lord of Dakshināpatha, whom he did not destroy as he was a near relative".3 Sircar4 opines: "This Sātakarni seems to be no other than Gautamīputra. The closeness of relation between the two rulers is explained by the Kānheri inscription which refers to a Kārdamaka princess as the daughter of Mahākshatrapa Ru(dra) who is generally identified with Rudradāman, and as the wife of Vāsishthīputra Pulumāvi and a son of Gautamīputra." It appears certain that a son of Gautamīputra was Rudradāman's son-in-law, for the Purānas list as Pulumāvi's immediate successor a Sivaśrī Sātakarni who may be identified with a Vāsishthīputra Śivaśrī Sātakarni of coins discovered in the Krishna and Godāvari Districts.5 But this son cannot himself be the "near relative" whom Rudradāman defeated yet spared. If he were, he would straight away be called "son-in-law" and not "near relative". Besides, when Rudradāman was Mahākshatrapa, Kānheri -where the inscription speaking of Rudradāman's daughter was found - was under him, as the Junāgarh inscription lists among his dominions Aparānta, in which Kānheri is situated; so during his lifetime there could be no "lord of Dakshināpatha" operating from Aparanta.6 Nor can Vāsishthīputra Śivaśrī's brother Pulumāvi
1.Ibid., p. 204.
2.Ibid., p. 206.
3.Ibid., p. 183.
5.Ibid., p. 205.
6.Ibid., p. 205.
Page 471
be such a "lord" from any other province, for Rudradāman names "Sātakarni" and "the two names, Sātakarni and Pulumāvi, are never known to have been borne by a single individual".1 Finally, we have independent evidence of Gautamīputra Sātakarni at war with the Kārdamakas: Bhagwanlal Indrajit2 found a coin of his that had been executed in the style of Rudradāman's father Jayadaman whom he must have defeated, a style which was never imitated by any other Sātavāhana king. Everything converges upon Gautamīputra, and our dating of the Junagarh inscription - 479 B.C. - is precisely such as to eliminate all other candidates. This date is only 7 years later than the last of Gautamīputra's reign, 486 B.C., and falls in the 7th year of the reign of Pulumāvi who in any case is a non-starter in this question because of his name. The reign of Vasishthīputra Śivaśrī Śātakarni is over two decades still in the future (457-450 B.C.).
We may justly consider the first challenge, which has to do with the Śakas and the Āndhra-Sātavāhanas, to have been fully met.
The Hathigumpha Inscription and a Sātakarni
The next challenge is the Hāthigumphā Inscription of the Chedi king Khāravela, ruler of Kalinga. This inscription is dated either to the early part of the 2nd century B.C. or to the later part of the 1st century B.C. In both cases, it is said to refer to a Sātakarni whom Khāravela defied in his 2nd regnal year. It is also said to mention the Chedi monarch's enlarging in his 5th regnal year a canal excavated by a Nanda rāja and he used the phrase ti-vasa-sata which, as Sircar3 tells us, translators interpret as either 300 or 103 years earlier. About this Nanda rāja we get a further phrase which makes him take away to Magadha, obviously at some time in the past, Jain images from Kalinga, which Khāravela succeeds in recovering.
As the Purānas distinctly mention Mahāpadma Nanda of the pre-Mauryan Nanda dynasty of Magadha as a conqueror of Kalinga, Khāravela's Nanda rāja cannot be anybody except Mahāpadma.
1.Ibid., p. 203. fn. 1.
2.Quoted in Shobhana Gokhale's "Andhau Inscription of Castana, Śaka 11", op.cit., p. 110.
3."The Sātavāhanas and the Chedis", The Age of Imperial Unity, pp. 214, 215, 100.
Page 472
If Khāravela can be dated to the later part of the 1st century B.C., and if Mahāpadma may be put 300 years before, we get this monarch just where modern historians place him - in the 4th century B.C., a little prior to Chandragupta Maurya alias Sandrocottus. And the defied Sātakarni will also be where historians, like Sircar, position the third member of the Sātavāhana line: Sātakarni I. Then our chronology of both the Guptas and the Sātavāhanas will have to be thrown overboard. Along with it will have to go our time-scheme for the Western Śakas.
Actually, the Hāthigumphā Inscription poses no danger to us. If Khāravela's Sātakarni belonged to the Sātavāhanas, we have to reckon with three attacks on the dominions of Sātakarni and consider their plausibility face to face with his position as gathered from other evidence. Not only in the 2nd regnal year but also in the 4th and 10th Khāravela attacked territories which were included in the Sātavāhana empire if that empire existed in his day; and each time it was a different territory. Under these circumstances, which show no setback anywhere for the Chedi king, how could Sātakarni I, the third member of the Sātavāhanas, be described in the Nānāghāt inscription of his wife Nāgārnika as Śūrah, Virah. and Daksināpathapati and as the performer of two Aśwamēdha sacrifices and one Rājasūya? Surely , he could not be the Lord of Dakshināpatha after having so many different parts of his kingdom conquered successively by Kharavela? We must remember that the Purānas accord him a reign of merely 10 years: they would all be covered by Khāravela's three expeditions, leaving him no chance to reassert supremacy over the Deccan. Obviously, Khāravela's Sātakarni was the petty ruler of only a small section of the country west of Kalinga, through which the Chedi king, "without even thinking of Sātakarni" , sent in his 2nd regnal year a large army en route-us the banks of the " Krishnabanā" (the river Krshnā) to threaten the city called Rishika- or Mushikanagara. To equate this negligible Sātakarni with the . powerful Sātavāhana monarch whose sway was far-flung and unabated is illogical.
A Sātakarni other than a Sātavāhana or else belonging to an offshoot of^that family is quite conceivable in Khāravela's time. There is a plethora of Sātakarnis known to history, outside the several among the 30 kings of the great line:1 Vishnukada Chutu-
1. Ibid., p. 209.
Page 473
kulānda Sātakarni, Skanda Sātakarni, Vijaya Sātakarni, Kumbha Sātakarni, Krishna Sātakarni. The first named was about 100 years later than the great line in the current time-scheme. Another Sātakarni, not in the above list but figuring in theTalagunda Stone inscription, is by the present reckoning over 200 years later, "apPārently a king of the Chutu family of Kuntala"1 - Kuntala which, was in the heart of the Kanarese country, practically the same as the district round modern Banavāsi. In lieu of Sātakarni I, some such Sātakarni has to be thought of for Khāravela.
As for the Nanda rāja, 300 years after whom Khāravela is said to come, nothing compels us in the inscription to think of 300 or even 103 years as "earlier" in connection with a canal excavated by a Nanda rāja. There is no "earlier" in the inscription: it is an interpretative word supplied by the reader. We can very well construe the passage as does Sten Konow:2 "And now in the fifth year he has the aqueduct, which was shut (or opened) in the year 103 (during the reign of) the Nanda King, conducted into the town from Tanasuliya Vata." Konow makes 103 years stand not for an interval reaching down to Khāravela after a Nanda but for a period counted in some era and hence for an interval between a Nanda rāja and some event prior to the Chedi king's day. Knowing Khāravela to have been a Jain, Konow surmises the computation to be in what is called the Mahāvira Era; but there is no inevitability in the notion. Why should Khāravela refer to the Mahāvira Era when speaking of a Nanda rāja and, despite his evident fervour for Jainism, not once refer to this era when speaking of the various events in his own life? If he did associate it with the Nandas, he would be acting rather strangely under the circumstances. The most natural thing for us is to ask: "Cannot the computation be in the years of the Nanda dynasty itself?"
The Ceylonese Chronicles assign only 22 years to this dynasty, the Purānas 40 according to some versions and 100 according to others. So the Purānic account no less than the Ceylonese suggests a negative answer. But the relevant account in the case of a Jain king would be the Jain account. V. Smith3 tells us that the Jains extend the duration of the dynasty to 155 years. He adds that the number does still greater violence to the reason than the Purānic
1.Sircar, Select Inscriptions, p. 450.
2.Acta Orientate, Vol. I, pp. 14-26; also Raychaudhuri, op. cit., p. 256, fn. 3.
3.The Early History of India, p. 42.
Page 474
100; but the point is neither here nor there. We are concerned with what Khāravela is likely to have accepted. And it seems significant that a Jain king should speak of the year 103 in connection with a Nanda and that the Jain accounts should make room for such a number by assigning 155 years to the dynasty.
If one objects that the Jain book tells us of no Nanda Era, we may reply: "The omission of it is hardly in itself a serious argument. No book mentions the so-called Kalachuri or Chedi Era which our historians take as starting from A.D. 248-49.1 It is inferred only from inscriptions. We have no inscriptions or coins of the Nanda dynasty to decide whether it had an era or not. If the Hāthigumpha inscription suggests an era of theirs, why should we not believe it? The Nanda dynasty is the earliest known conqueror of Kalinga; so it is quite possible that, even if the Nandas themselves did not establish any official era; the people of Kalinga might remember or be made to remember the years of this dynasty either from the time of its inception or from the time of its sovereignty over their country and keep a running count of its career. But, strictly speaking, no Nanda Era need be involved by Khāravela. It is highly probable that the number 103 applies to the reign of only the first Nanda whom the Purānas call Mahāpadma and consider the conqueror of Kalinga. The Purānas, with a total of 100 years for all the Nandas, give 88 to him: the Jain tradition, with a total of 155, may very well have led Khāravela to give him 103 or more.2 The ancient mind was not averse to believing in rare cases of extraordinarily long reigns."
Thus the Hāthigumphā inscription turns out to be no obstacle to our Śaka Era of 551 B.C. and to the link we forge between the Sātavāhanas and the Western Satraps in terms of it.
The "Periplus" and Ptolemy on the Śakas and Sātavāhanas
The third challenge founds itself on certain references in that navigational record, the Periplus of the Erythraean Sea generally
1.The Age of Imperial Unity, p. 222.
2.Smith's source for 155 must have been the Apāpāpurikalpa or Pāvāpurikalpa. In the Tttthogoli Painannaya we find 150 years, while Meritunga gives 158 (vide Mankad's Purāhic Chronology, pp. 178, 188, 189). But the differences are of no moment for us, since all the figures go beyond 103 and also make it probable that this number applies to one reign only.
Page 475
accepted as written between 70 and 80 A.D. by an Egyptian Greek who was a merchant in active trade and personally made a voyage to India - and some statements in a book by the geographer Ptolemy from Alexandria, the famous Geography written in about 140 A.D. on the basis of information collected by Marinus of Tyre a few years earlier.
The Periplus (Cap. 52) speaks of the country called "Dachina-vades" and mentions "Calliena which in the time of the elder Saraganus became a lawful market-town; but since it came into the possession of Sandares the port is much obstructed, and Greek ships landing there may chance to be taken to Barygaza under guard." Dachinavades has been identified with Dakshinapatha, Calliena with Kalyāna near Thānā, Barygaza with Broach, Saraganus with some Sātakarni and Sandares with Sundara Sātakarni who is placed in the Purānic list a few generations earlier than Gautamīputra Sātakarni. Thus the Āndhra dynasty seems to be precisely in the period where it should be if the Mauryas were established in Alexander's time and followed by the Śungas and Kānvas who preceded the Āndhras.
Ptolemy (VII, i, 63, 82) writes of Ozéné as the capital of Tiastenes and Baithana as that of Siroptolemaios. Ozéné appears to be Ujjayinī and Tiastenes to be Chashtana, the grandfather of Rudradāman, who, according to inscriptions at Andhau, was Mahākshatrapa in the year 52, which, if counted from 78 A.D., the commonly named Śaka Era, yields 130 A.D., just the right time for the Geography's information. Baithana seems to be Pratishthāna and Siroptolemaios to be Śrī Pulumāvi, Gautamīputra's son, who, like other kings, has "Siri" before his name in inscriptions. As modern historians date Gautamīputra's death to about 130 A.D., his son would really be the contemporary of Chastana for a few years and reign in the period upon which the Geography bears. So, once again, should we not discredit the Purānic chronology?
The True Interpretation of the "Periplus" and the "Geography"
The best way to begin our answer is to show that the Periplus, which is understood to prePāre historically the right background for the statements contained in the Geography, is actually a poor preparation.
Page 476
The Elder Saraganus (Sātakarni) implies a Younger Saraganus. If Sandares is equated with Sundara Sātakarni, we have in him a Saraganus all right, but the Purānic Sundara is not preceded by any Sātakarni, in relation to whom he can be a Younger Saraganus: he is preceded by Purīndrasena, Mantalaka and Hala, none of them specifically called Sātakarni.1 The Elder Saraganus is missing. In the eyes of some scholars, "Sandares" is not the right name of his successor: the name is "Sandanes", and this potentate appears to have belonged to a different family from that of the Elder Saraganus. McCrindle,2 referring to the information to be gained from the Periplus about the place Ariaké mentioned by Ptolemy, writes: "we learn that Sandanes after having made himself master of Kalliena (now Kalyāna), which had formerly belonged to the house of Saraganus the elder, subjected its trade to the severest restrictions..." So the identifications, highly questionable in themselves, with an earlier and a later ruler of the well-known Āndhra dynasty are rendered more debatable. Though one may not be debarred from imagining that successors of the dynasty might have lingered in parts of Mahārashtra after the main family had become extinct, all talk of Sundara Sātakarni and of a Sātakarni predecessor of his can be considered practically irrelevant.
Besides, the historical information in the Periplus is connected with the name of a powerful king Mambarus whose capital was Min-nagara and whose dominion seems to have comprised Kathiawar, Gujarat and parts of Rājputāna3 but who is utterly an unknown quantity. The identification usually suggested - viz. that his name is a mistake for Nambanus which again is a Greek corruption of the name Nāhapāna - is unwarranted if we accept the modern chronology, according to which Nāhapāna comes in the period 119-125 A.D., nearly fifty years after the time of the Periplus, 70-80 A.D.4 However, Sircar5 has the impression that Mambarus (with a capital named as the early Scythians used to name their chief city) is a Scythian (Śaka) ruler and then it strikes
1.The Age of Imperial Unity, pp. 705-6.
2.Ancient-India as described by Ptolemy (Ed. 1927), p. 40.
3.The Age of Imperial Unity, p. 178.
4.Ibid., p. 179.
Page 477
him as "not improbable that Sandares was the viceroy of Mambarus in Aparānta".
The suggestion of Professor Aiyangar in his Beginnings of South Indian History1 to take Mambarus as Lambodara (Prākrit Lamoboara), one of the Āndhra kings, appeals to S. N. Majumdar2 as philologically sounder than taking him to be Nāhapāna. But historically it is equally unsound. Lambodara, as a glance at the Puranic list3 will tell us, is the 7th Āndhra king , more than a century prior to Sundara Sātakarni who should be contemporaneous with the Periplus if Pulumāvi and Chashtana are contemporaneous with the Geography. Sircar's view is the best , but by removing Sundara Sātakarni from the scene it spoils the background required for the Geography .
In that case the seemingly recognisable figures of the Geography grow obscure. An alternative to the equations Tiastenes-Chashtana and Siroptolemaios-Siri Pulumāvi becomes tenable. Even independently of the failure of the Periplus to be evidential in favour of the present chronology, there can be ground for an alternative. But the ground turns into absolute terra firma now. Arid we may confidently offer the possibility that those two names are not personal but dynastic terms. We may recall the Matsya Purāna (271.39):
Paulomāstu tathāndhrāstu mahāpadmāntare punah:
antaram tac chatānyastau sat-trimsat-tu samā stathā.
"Further, the Pulomas (lit., the Puloma-offspring) and the Āndhras at an interval from Mahāpadrna - that interval was 836 years. " Here the Āndhras and the Pulomas stand for the same dynasty: "Puloma" is clearly a dynastic term just as "Āndhra" is. Conse- : quently, it would be quite reasonable to believe that "Siroptolemaios" is a dynastic label for a late scion of the Puloma alias Āndhra line. A great dynasty does not always fade out the moment its central branch is supplanted by another or has gone extinct. Raychaudhuri4 writes : "Petty Maurya kings continued to rule in Western India as well as Magadha long after the extinction
1.p. 116.
2.Notes to McCrindle, op. cit., p. 345.
3.The Age of Imperial Unity, pp. 706-7.
4.The Political History of Ancient India, p. 240.
Page 478
of the Imperial line. King Dhavala of the Maurya dynasty is referred to in the Kanawa inscriptions of A.D. 738... Maurya chiefs of the Konkan and Khandesh are referred to in the early Chalukya and Yādava epigraphs. A Maurya prince of Magadha named Purnavarman is mentioned by Hiuen Tsang." The Āndhra line is itself said by the Brahmānda and the Vāyu Purānas (170-71; 99, 357) to have split into 12 branches after the death of Puloma III, the last of the great Āndhras. We should hardly be surprised if petty Āndhras continuing after him were entitled Pulomas with all the more appropriateness. We have already mentioned later Sātakarnis The continuation of the Āndhra Satavahana family is attested also by Kalha.na's Rāiatarangini (IV , 367; VII, 1283, 173-2), where Sātavāhana-kula is given as the name of the Lohara dynasty of Kashmir.
The Śaka family of the Kārdamakas may also have continued in a series of small chiefs and been known as the Chashtanas because of Chashtana who was the first in his own line to be called rāja: they may have reigned as local potentates in the province of their illustrious ancestor. At least a Jain work, TrailokyaPrajnapati,1 mentions immediately after Naravāhana who is evidently Nāhapāna, a line of kings continuing for 242 years and entitled Bhachchatthana which is clearly a corrupt variant of Chashtana as a dynastic term. The number of years given by the Jain work to the Chastanas is too small for our need. But the usage we want is there to render credible our idea that in Ptolemy's day some descendant of the Kārdamaka Mahākshatrapa of the Andhau inscriptions of the years 11 and 52 - that is, 540 and 499 B.C. - might be holding power in Ujjayinī while a "Puloma" was in authority at Pratisthāna.
The credibility of our idea increases on remembering the testimony of Pliny, based on Mêgasthenes, regarding a powerful Āndhra king, with a large dominion, in about 300 B.C. We stressed that if the Āndhra dynasty had begun later, there could never have been such a king at this time and we took Pliny's statement as a pointer towards the Āndhra chronology deduced by us from the Purānas as well as from Varāhamihira. Now we may stress another aspect: the statement indicates the continuation of the Purānic "Pulomas and Āndhras" beyond the strict termination of thejnain dynastic line with Siri Pulumāvi III prior to the advent
1. Vide Mankad, Purānic Chronology, p. 198.
Page 479
of Sandrocottus. We have no reason to believe that they ceased before the epoch of the Geography and that the "Siroptolemaios" of this book was not a dynastic label.
Possibly we shall be reminded: "Is it not curious that, whatever the lack of support from the Periplus, the Geography should chronologically place Tiastenes and Siroptolemaios just where modern historians place Chashtana and Siri Pulumāvi?
We have to ask in our turn: "Does the Geography really do so? Does it indeed give the state of affairs in its own time: 130-140 A.D.? Have its sources genuine directness and immediacy? We can easily show that the answer cannot be 'Yes'."
S. N. Majumdar1 has indicated the sources of the Geography: "Ptolemy had to rely (and specially so in the case of India of which he had not the slightest personal observation) upon second-hand information - reports of travellers, navigators, works of previous writers and certain Indian sources." McCrindle,2 appreciating the value of Ptolemy's account as an antiquarian record, has yet stated: "Ptolemy's information concerning many parts of the earth, whether owing to their remoteness or the conflicting accounts of travellers regarding them, was imperfect in the extreme." Of course the above remarks of both Majumdar and McCrindle refer primarily to the geographical and not the historical part of the book. But the same imperfection of information that was responsible for the erroneous geography is bound to be responsible for a haziness in whatever history got associated with the geographical treatment.
If modern historians are right, the Ozéné of Tiastenes must be within what is called Indo-Scythia - that is, the part of northwestern India which is taken to have been under Śaka rule.3 But Surendranath Roy4 remarks that Larike, in which Ptolemy (Sections 62-63) situates Ozéné, is put by Ptolemy clearly outside of Indo-Scythia. Therefore his Ozéné cannot have been in Śaka hands in the 4th decade of the 2nd century A.D. and must have been the capital of Tiastenes (however we may interpret this name) at some time in the past. If Lariké lay, as B. C. Law5 also
1.Ancient India as described by Ptolemy, p. XXI.
2.Ibid., p. 3.
3.S. K. Majumdar, op. cit., p. 369.
4.Proceedings of the Indian Historical Congress, 1939, p. 344.
5.Indian Culture, Vol. III. p. 736.
Page 480
states, to the east of Indo-Scythia, why should we think of the family of Chashtana, the Śaka Satraps, to have been flourishing in the period in which modern historians set them? Ptolemy's reference may be to a period before A.D. 78 which they regard as the Śaka Era. The Periplus which belongs to the period of this Śaka Era says (Section 48): "Eastward from Barygaza is a city called Ozéné, formerly the capital where the king resided." Even in the time of the Periplus this city was divested of its old glory. McCrindle1 opines that the king mentioned may have been the traditional Vikramāditya of Ujjayinī. Or he may have been of the line of "Tiastenes", who had returned to Ozéné after the passing of Vikramāditya but had been dispossessed in the time of the Periplus and of the era of 78 A.D. At any rate, Roy's remark spotlights the probability that "Tiastenes" indicates no Śaka occupation of Ozéné in 130-40 A.D. Ptolemy's Tiastenes and Ozéné do not coincide chronologically with the Chastana and Ujjayinī of modern historians.
Only if we take them as we have done we can bring them into touch with the Śaka Chashtana and his Ujjayinī; for then we are not compelled to fit the latter two into 130-40 A.D. A late scion of our Chashtana of 540-499 B.C. might be master of Ujjayinī at whatever time it would be convenient for us to give him. Trusting the Periplus about Ozéné, we should place him somewhere between 57 B.C., the traditional Vikrama Samvat; and 78 A.D., the conventional Śaka Era. We must understand Ptolemy as alluding to matters not of 130-40 A.D. but of a century or so earlier. And with his Tiastenes his Siroptolemaios will be dislodged from where our historians date their Siri Pulumāvi.
Nor would Ptolemy's allusion here to another epoch than his own be quite an exception in his Geography. If with modern historians we put Khāravela in the last quarter of the 1st century B.C. - as we certainly can - Majumdar's words,2 apropos of Ptolemy's mention of the city Pityndra which has been identified with the Pithuda of Khāravela's Hāthigumphā inscription, become significant: "We cannot expect to find Pityndra, for it was destroyed by Khāravela a few centuries before Ptolemy who seems to have mentioned it on the authority of his old materials. (The Indian Antiquary,1926, pp. 145-46)."
1. Op. cit., p. 155.
2. Ibid., p. 387.
Page 481
The fate of Ptolemy's history is exactly what we should expect from our interpretation of the historical data supplied by the Periplus. If these data, which are likely to be contemporary with the book's author, cannot support the modern chronology, neither can Ptolemy's history. A new constellation of events must swim into our ken.
Greek on Śaka Coins
Fourth and final is the challenge: "How is it that Śaka Satraps from eastern Irān commenced using in India in the late 6th and early 5th century B.C. a legend in Greek on their coins over and above Kharoshthī and Brāhmī writing? Greek on Indian coins in the time of Cyrus (558-530 B.C.) is an impossibility. Surely, then, these Satraps could not be dating themselves by an era of 551 B.C.?"
We have to note what sort of Greek is here and in what historical context the coins were issued. The Greek is corrupt and "later degenerated into a sort of ornament",1 "a sort of ornamental fringe around the obverse of the coins".2 Evidently the use of Greek characters is rather in an elementary stage and hardly an established habit: contact with Greece is proved to be merely incipient and not at all fraught with crucial meaning. And, if we consider the historical context in which, by our dating in the time of Cyrus the entry of the Western Satraps into India, we put these coins, we shall realise that the coins are most likely to be what they actually are like.
We have already shown that Cyrus's campaigns in the Indian borderlands both preceded and followed his conquest of Lydia in 547 B.C. and certainly preceded his reduction of Babylon in 539 B.C. With Chashtana's first available inscription dated in the year 11 of the era of 551, we should assume Cyrus's rule over the territories which extended from Gandhāra to the Indian Ocean and even involved tribute from an Indian king from east of the Indus, to have taken place between 547 and 541 B.C. The point of importance here is that the second phase of the rule came in the wake of Cyrus's possession of Lydia, a Greek state whose king, whom he defeated but spared, was Croesus. "The Lydians," says
1.The Age of Imperial Unity, p. 180.
2.Ibid., p. 184.
Page 482
Herodotus (194), "were the first people we know of to use gold and silver coinage..."1 Frye,2 like all scholars, confirms Herodotus: "...the Lydians are generally credited with the introduction of coinage on a state scale." Greek coinage thus was given a gate of entrance into Cyrus's Irān. We may also remember that Croesus had the reputation of being the richest man of his time and, once he was allowed to live, he became a close friend of Cyrus, accompanying him everywhere. Herodotus (1.153) informs us that Cyrus "started eastward on his march to Ecbatana, taking Croesus with him."3 It would seem that Croesus was inseparable from Cyrus, for, even on the last expedition of his life - the one against the Śakas east of the Caspian Sea - Croesus was present and Herodotus (1.207), describing a crucial meeting of Persian officers to decide the strategy of the forthcoming battle, reports "Croesus the Lydian" as saying: "My lord, I have already told you that since God has made me your servant I will do all I can to avert any danger which I may see threatening your house..."4 Obviously Croesus went wherever Cyrus took his army and we must expect him to have been with his master when the latter, as Xenophon (Cyropaedia, 1.1,4) writes, "brought under his rule Bactrians and Indians" and extended his sway up to the Erythraean Sea, i.e. the Indian Ocean.5 We have also to note from Herodotus (1.153) that Cyrus got all the treasures belonging to Croesus collected and conveyed to himself.6 The influence of Croesus and his coinage in some form or other on whoever might be minded to strike coins in Cyrus's time was inevitable. Croesus's coins have no Greek writing but are only stamped on one side with the facing heads of a lion and a bull.7 But already stray coins - that is, coins not issued on a state-basis as by Croesus - in Greece had begun to have inscriptions on them. One specimen of them has come down to us: "The coin with a stag (unique, British Museum Collection) is remarkable for the Greek inscription reading: 'I am the badge of Phanes,' and was perhaps struck at Ephesus"8 in c. 650 B.C. Among the
1.The Histories, p. 52.
2.The Heritage of Persia, p. 140.
3.The Histories, p. 75.
4.Ibid., p. 97.
5.The Age of Imperial Unity, pp. 39-40.
6.The Histories, p. 75.
7.The Encyclopaedia Britannica, 1973, Vol. 16, p. 763, col. 2.
8.The Encyclopedia Americana (New York, 1966), Vol. 20, p. 545, col. 1.
Page 483
treasures of Croesus, there must have been coins of various kinds and, although he never used any writing on his own coinage, coins like the one from Ephesus may have been recommended. It should not be a matter of great wonder to find on the satrapal silver and copper issues what Sircar1 calls "traces of Greek legend".
Perhaps we may wonder why Cyrus himself was not encouraged to issue coins. Here we are somewhat in the dark.- Darius who came to the Persian throne 8 years after Cyrus is well known for his gold darks and silver shekels (Greek sigloi). Probably he was the first Persian to strike coins, but the intimacy between Cyrus and one so coin-minded as Croesus raises a doubt. Just as Darius introduced a reform in the tax-system (Herodotus, III.89) he may have improved the monetary system and enlarged it rather than innovated it. Besides, do we not hear from Xenophon (Cyro-paedia, VI.2.1-11) that Cyrus received a sum of money as tribute from an Indian king even before he embarked on the conquest of Lydia? It seems best to think that at least after that conquest Cyrus had some sort of coinage whose remnants we have not yet chanced upon.
In any case, we have a historical context in which the Satrapal issues, with all their characteristics, would not be misfits. The tendency to use Greek legends on coins seems no unnatural one among the Satraps of the Achaemenid empire, along with the tendency first exemplified by the Kshaharātas and the Kārdamakas to put their own busts upon their coins. It is to "a number of Persian satraps" who "struck coins in their own names" that "belong a number of the earliest and finest portraits on coins".2 And the portrait tendency as well as the tendency towards Greek legends reaches its climax,- as it were, in the coins of Sophytes, over which a controversy once raged. Sophytes was once taken to be a Hinduised Greek from some pre-Alexandrine Greek colony in India or an Indian subordinate of Alexander. Now we know that Sophytes was an Irānian satrap ruling in north-eastern Irān under the Achaemenids and that he became a Governor perhaps in 335 or 340 B.C., asserted his independence at the fall of the Persian empire and got reinstated in his office by Alexander.3 The point
1."The Śaka Satraps of Western India", The Age of Imperial Unity, p. 179.
2.The Encyclopaedia Britannica, Vol. 16, p. 780, col. 2.
3.Journal of the Numismatic Society of India, Vol. XI, pp. 93-99.
Page 484
significant for us in general is his coins: they are Irānian issues in pre-Alexandrine days with the issuer's head on them and a workmanship highly imitative of Greek types and an inscription on them in clear Greek characters. Coins of Śaka Mahākshatrapas, showing in late 6th and early 5th centuries B.C. the issuer's portraits and a rudimentary Greek legend which soon became a mere ornament, could be a very natural beginning of such coinage in Indo-Irānian terms.
Thus the last of the challenges may be taken as met within reasonable limits. And Samudragupta, so far as his relations with foreign powers are involved by the Allāhābād Pillar inscription, can stand in time with the successors of Alexander, just as he can with Devānampiyatissa of Ceylon.
13
We have dealt with quite a number of supposed hurdles to our shifting the era of the Imperial Guptas to 315 B.C., from Fleet's epoch for it: 320 A.D. Against our move two other "hurdles" may be set up.
The Challenge of the Testimony of "Gupta Feudatories"
The first hurdle may be seen thus: "The inscriptions and coins of certain Indian rulers become chronologically intelligible only if the Gupta Era mentioned by them or the unspecified era used by them is assumed to be the epoch of Fleet."
Obviously, this objection is too general. We do not deny the existence of the era of 320 A.D., but to us it is the end of the Imperial Guptas, not their beginning - an era followed by those who rejoiced in their end. The objection would be valid only if in the records of any kings regarded at present as feudatory to the Imperial Guptas and as using their era we could find the mention of an Imperial Gupta as overlord.
The Imperial Guptas and the Later Guptas
None of these kings - the Valabhī monarchs, the Maukharis, the rulers of Kāmarūpa and Orissa - have left us the name of any Imperial Gupta in connection with the Gupta Era used or with
Page 485
reference to overlordship.1 Among the chiefs of Bengal, only one seems to have used the name of some Gupta. Majumdar2 writes: "The name of a Gupta ruler is invoked as suzerain in a grant of N. Bengal in A.D. 543. Unfortunately the first part of the name is lost, but it might well have been 'Vishnu' and refer to the last Gupta ruler. We do not know how and when the Guptas lost this last stronghold. A land-grant found in the Gayā District in the very heart of Magadha was issued in A.D. 551-2 by Nandana who is called Kumārāmātya Mahārāja. As there is no reference to any Gupta ruler in this record we may conclude that by A.D. 550 the Guptas had ceased to exercise effective authority over the greater part of Magadha. Nandana's title Kumārāmātya, however, shows that like the Viziers of Oudh in the eighteenth century he still dared not throw off nominal allegiance to the Guptas." Majumdar's inferences are legitimate in the context of his beliefs, but once we conceive of the Gupta Era à la Albērūnī everything changes. Instead of thinking of Vishnugupta of the Imperial line we may think of one of those whom our historians distinguish from that line and designate the Later Guptas, and it is most interesting to see that their history satisfies all demands arising from the facts stated by Majumdar.
Majumdar3 himself has told us this history:
"An inscription found at Aphsad near Gayā4 gives the following genealogy of the early kings of this dynasty:
1.Krishna-gupta
2.Harsha-gupta
3.Jlvita-gupta
4.Kumāra-gupta
5.Dāmodara-gupta
6.Mahāsena-gupta
7.Mādhava-gupta
8.Āditya-sena.
Although no royal title is given to any of these, Krishna-gupta is called nrpa (king) and similar epithets are applied to his successors.
1.The Classical Age, pp. 61, 62, 67, 68, 70, 90, 91, 92, 93.
2."The Fall of the Gupta Empire", ibid., p. 44.
3."Northern India after the Break-up of the Gupta Empire", pp. 72-76; "Northern India during A.D. 650-750", pp. 126-28, ibid.
4.Corpus Inscriptionum Indicarum, III, 200.
Page 486
The Aphsad inscription describes in very general and conventional terms the military achievements of the first three kings. The third king is said to have carried his arms to the Himalaya mountains as well as to the sea. But there is nothing to show whether these campaigns were undertaken by the Later Gupta rulers as feudatories on behalf of their suzerains or as independent chiefs...
"More details are available about the next king Kumāra-gupta. He defeated the Maukhari king Iśāna-varman who is described as 'a very moon of kings'... That the success attained by Kumāra-gupta was both great and permanent is proved by the facts recorded in the Aphsad inscription that he had advanced up to Prayāga where he died and that his son Dāmodara-gupta again defeated the Maukharis, though he was probably killed or was seriously wounded in the battle. There is no reason to discredit these claims, particularly as the Maukhari records do not claim any victory over their opponents...
"Mahāsena-gupta, the son of Dāmodara-gupta, carried his victorious arms as far as the Lauhitya or Brāhmaputra, and defeated Susthita-varman, the king of Kamarupa or Assam... Mahāsenāgupta... soon fell on evil days...His two sons Kumāra-gupta and Mādhava-gupta found shelter at the court of king Prabhakara-vardhana of Thaneswar, whose mother Mahāsena-gupta, as the name shows, was probably a sister of king Mahāsenā-gupta. The two young princes become attendants of Rājya-vardhana and Harsha-vardhana, the two sons of Prabhākara-vardhana.... Shortly after the death of Harsha... either Mādhava-gupta or his brother seized the opportunity to make himself master of Magadha... Mādhava-gupta must have been fairly advanced in age when he ascended the throne, and his reign was probably a short one. He was succeeded by his son Āditya-sena... Āditya-sena assumed the imperial title of Mahārājadhirāja. There is hardly any doubt that he ascended the throne in the third quarter of the seventh century A.D. We know the names of three successors of Āditya-sena, viz. Deva-gupta, Vishnu-gupta and Jīvita-gupta. They all continued the imperial titles and were evidently rulers of some power..."
Page 487
The true Interpretation of the Grants of North Bengal and Gayā District
Some light is thrown on the chronology of the kings preceding Āditya-sena by the fact noted by Majumdar1 that Iśāna-varman, whom the fourth Later Gupta Kumāra-gupta defeated, flourished in c. 550-576 A.D. With this fact before us we may very well take the name of the unknown Gupta ruler in the grant of N. Bengal in 543 A.D. to have been of Kumāra-gupta himself or his predecessor Jīvita-gupta. As for the title Kumārāmātya Mahārāja assumed by Nandana in the Gayā-District land-grant of 551-2 A.D., one would be tempted to interpret the first half of the title as not only being a component of a familiar designation of an official appointed by a higher authority - a designation essentially meaning a civil governor - but as also being a covert pointer to Kumāra-gupta whose reign overlapping with the time of Iśāna-varman may quite credibly include the date of the land-grant. Even if it does not, another of the Later Guptas can cover it with his reign: the needed candidate would again be Jīvita-gupta.
The sole proviso for these possibilities is that the Later Guptas be proved to have dominated Magadha and other parts of eastern India at the time. On this matter we may quote Majumdar2 again : "Āditya-sena, the grandson of Mahāsena-gupta, ruled in Magadha, and so did all his successors. The Aphsad inscription of Ādityasena gives a continuous account of the whole dynasty from the very beginning up to his reign without indicating in any way that it had migrated from a different place. It may , therefore, be presumed that Magadha was the kingdom over which the dynasty ruled from the beginning." Then Majumdar considers some objections and the answers to them and he remarks: "Thus although it is impossible to come to any definite conclusion it seems very reasonable to regard the Later Guptas as rulers of Gauda and Magadha with suzerainty over Mālava." This means that there is no evidence for the Imperial Guptas rather than the Later Guptas, in the two grants mentioned by Majumdar.
1."Northern India after the Break-up...", The Classical Age, p. 70.
2.Ibid., pp. 75-76.
Page 488
The Vākātakas and the Imperial Guptas
The one family that mentions the Imperial Guptas are the Vākātakas, but they do not employ the Gupta Era. What they do is noted by Sircar:1 "The records of the Vākātaka kings are dated in their regnal year and not in any era." They provide us with no clue to placing them side by side with the Valabhl kings, the kings of Bengal, of Kamarupa and Orissa, and the Maukharis. They can be placed in the same centuries before Christ as the Imperial Guptas.
An attempt is made at times to connect the end of the Vākātakas with rulers who are known to have existed in the 6th century A.D., but no success attends it. We can only swim in conjectures. Sircar2 admits: "The actual events leading to the fall of the Vākātakas are unknown. They are not mentioned amongst the powers that stood in the way of the Chālukya occupation of the Deccan in the latter half of the sixth century. The early Chālukya monarchs had to subdue the Nālas of the southern Madhya Pradesh and adjoining regions, the Mauryas of the Konkan and the Kalachuris of northern Māharāshtra and the adjacent countries. It is not impossible that the major part of the dominions of both the Vākātaka houses had passed to the Nalas before the middle of the sixth century..." On one nuance of the same point Sircar3 writes again "Whether the Nālas were responsible for the fall of the Vākātakas of Vatsagulma... cannot be determined in the present state of insufficient knowledge. But the suggestion is not altogether improbable." How can it be anything except "altogether improbable" when we are utterly in the dark, and the only source of light - the Chalukya records - has not even a hint about the Vākātakas?
"Gupta-Feudatories" and the Gupta Era of 320 A.D.
Here we may go on to remark that the idea of Gupta-feudatories using the Gupta Era of 320 A.D. is left unsupported also by the fact that certain monarchs' clear independence of any overlord
1. "The Deccan after the Sātavāhanas", The Age of Imperial Unity, p. 219. For a couple of instances, see The Classical Age, pp. 181, 185.
2.Deccan in the Gupta Age", The Classical Age, p. 187.
3. lbid, p. 190.
Page 489
in the period under consideration is not linked organically with their use or non-use of this era. Thus the Later Guptas, who are taken to be feudatories of the Imperial line and whom we may expect because of their name to continue with the Gupta Era even after independence, have left no inscription dated in it. A dated inscription of one of them, Āditya sena, an independent monarch, gives the number 66 which has been referred by our historians to the Harsha Era of 606 A.D., 1 about which Majumdar2 says: "Although in not a single instance has the era been expressly associated with the name Harsha, its existence has been inferred from certain remarks of Alberuni [Sachau's tr., Vol. II p. 5]." On the other hand, Iśāna-varman Maukhari, whose three predecessors are said to have been Gupta-feudatories and who is considered the first Maukhari to set up an independent kingdom,3 is understood by our historians to be still using the Gupta Era on his coins; and even his two successors Sarva-varrnan and Avantivarman, who are allotted the reign periods c. 576-580 A.D. and 580-600 A.D. respectively, are assumed to have used the same era.4 Much more pointed is Śaśānka's Gangem plate of the Gupta Era year 300 (620 A.D.) when no Imperial Gupta is even conceivable on Fleet's epoch. Or , better still, take the rock inscription at Tezpur5 of Harja-varman, the Kāmarūpa king, of the Gupta Era year 510, which would take us to 830 A.D. It seems we do not really know why the Gupta Era was dropped or continued: the feudatory-hypothesis does not cover the situation adequately and hence we cannot affirm that the Imperial Guptas were in power after 320 A.D.
Some Debatable Inscriptions
There are just a few inscriptions which may seem to provide ground for the belief that the Imperial Guptas reigned from 320 to at least 569 A.D. One is the Sumandala copper-plate6 from Orissa.
1."Northern India during A.D. 650-750", Ibid., p. 127.
2."Harsha-Vardhana and His Time", Ibid., p. 115.
3."Northern India after the Break-up...", Ibid., p. 68.
4.Ibid., p. 70.
5.Mookerji, The Gupta Empire, p. 16.
6.Epigraphia Indica, XXVIII, pp. 79-85. The Classical Age, pp. 42 &92; p. 42 has the fn.: "An account of this inscription has been published in the Indian Historical Quarterly XXVI, p. 75."
Page 490
Here a chief named Dharmarāja acknowledges the suzerainty of a monarch named Prithvī-vigraha who is stated to have been ruling the Kalinga rāshtra, apparently as a viceroy of the Gupta emperors. The phrase in which the date is indicated runs: "varttamāna-Gupta-rājye varshaśata-dvaye panchāśad-uttara." This has to be translated: "In the year 250 of the current sovereignty of the Guptas." If the inscription is dated according to the Gupta Era of 320 A.D., the Imperial Guptas were ruling in (320+250=) 570 A.D. But there is a detail that goes against 320 A.D., a detail based on three other inscriptions.
The first is another copper-plate inscription, discovered in a locality in Puri District, of the Vigraha dynasty1 to which Dharma-rāja's suzerain belonged. It mentions a monarch named Loka-vigraha and gives the year 280 and shows the monarch to have been holding sway over both Dakshina-Tosall and Uttara-Tosali in evident independence of Gupta sovereignty.
However, two copper-plate inscriptions2 of a Mahārāja named Śambhuyaśas belonging to the Mudgala-gotra and the Māna-family show him to be master of not only Uttara-Tosali in 579 A.D. but also Dakshina-Tosall in 602 A.D. on the very likely assumption that the years 260 and 283 are of the Gupta Era. If the copperplate of Loka-vigraha of the year 280 is also of the Gupta Era of 320 A.D., as it should be in consonance with the belief that the year 250 of Prithvī-vigraha belongs to this era, we have in him and Śambhuyaśas two simultaneous sovereigns each in complete control of the two Tosalis.
The paradox would be resolved on shifting either Loka-vigraha or Śambhuyaśas to an entirely different period. As the former joins up with Prithvī-vigraha and as the inscription mentioning this predecessor of his is the sole one in our possession to speak unequivocally of the era of a contemporary Gupta rājya, Loka-vigraha may be disconnected from the Gupta Era of 320 A.D. This means that Prithvī-vigraha may be referred to another Gupta Era, one which counts the years of actual Imperial Gupta sovereignty and not the years from the end of that sovereignty. Thus Prithvī-vigraha may be considered as existing in the year 250 after 315 B.C. i.e., 65 B.C.
1.Epigraphia Indica, XXVIII, pp. 328ff.
2.Ibid., XXIII, pp. 301f and IX, pp. 287f.
Page 491
Sircar's suggestion1 that Loka-vigraha was misstating things and that his claim is to be deemed "an echo of an earlier period when the Vigrahas were actually ruling over the Gonjan-Puri-Balasore region" appears to be an ad hoc device to get out of a compromising position. The natural conclusion would be that the Sumandala inscription of Dharmarāja does not bring the Imperial Guptas into a period which would give the lie to the Purānic chronology of them.
Even if we accept Sircar's rather arbitrary solution, are we compelled to keep Prithvī-vigraha in relationship to the overlords our historians affirm for him? Cannot we take the Guptas involved to have been the Later ones instead of the Imperial dynasty? Admittedly, there is no explicit pointer to the former's sway over Orissa as over Mālwā and further north, yet it would not be illogical to assume it. In the Aphsad Inscription, as Majumdar2 recounts, their "third king is said to have carried his arms to the Himalaya mountains as well as to the sea". The concluding phrase is explained by another which mentions "the Lauhitya" or Brāhmaputra as the limit of his triumph in the seaward direction which "leaves little doubt" that by the time of Mahāsena-gupta "the Later Guptas had... strongly established themselves both in Mālwā as well as in the eastern parts of the Gupta empire".3 We have also Majumdar's reference4 to Mahāsena-gupta's success "in bringing under his sway, at least for some time, the extensive dominions from Mālvā to Bengal". He5 has also observed that "the Later Guptas came into immediate possession of a large part of the dominions of the Imperial Guptas". Kaliiiga in Orissa could easily have been taken by the Later Guptas in their eastward stride. Of course this is an inference, but, surprisingly enough, Majumdar himself carries us no further than inference about Orissa as part of the Imperial Guptas' kingdom. He6 says: "hardly anything is known of the history of Orissa during the Gupta period. It would be interesting to speculate why Samudra-gupta advanced to the eastern coast of the Deccan through the hilly region of Kosala
1."Facts of Early Orissan History", The Journal of Indian History, Vol. XXXIV, Part III, p. 271, Dec. 1956.
2."Northern India after the Break-up...", The Classical Age, p. 72.
3.Ibid., p. 73. 5. Ibid., p. 76.
4.Ibid.6. Ibid., p. 92.
Page 492
(Chattisgarh), and avoided the more direct and easier route through West Bengal and Orissa. This would be all the more inexplicable if Orissa formed a part of the empire; and it is difficult to believe that he would have proceeded so far south without subjugating Orissa. In any case no royal dynasty ruling over Orissa during the fourth and fifth centuries A.D. is known to us, and we may well believe that it was included in the territory directly administered by the Gupta emperors." So there is as little or as much logic in assigning Orissa to the Later Guptas as to the Imperial ones. On any hypothesis, the Sumandala inscription does not discredit the Purānic chronology for the latter.
Much less do the inscriptions of the Parivrājaka Mahārāja Hastin and his son Samksobha. Fleet argued for a Gupta Era in them, but, really speaking, what they imply is the existence of the Guptas at the time they were made; the era is not definitely shown to be of the Guptas. Even Fleet,1 though taking the Gupta Era to be most probably there, had to admit: "There is nothing in the expression tending to give the era the name of Gupta era." The expression stands as follows: Śatpānchaśottare abdaśata Gupta nrpa-rājya bhuktau" - " In the year156 while the Gupta kings were ruling." D. N. Mookerji' has remarked: "If the Gupta era was really intended, the expression could have been written as Guptam: pa rāiya bhuktau śatpānchaśottare abdaśata," All that we can say is that Hastinand his son were contemporaries of.the Guptasand using some era which cannot be identified. If there is no conclusive evidence to place them in the A .D. period, they may be placed in the B.C. period where we have put the Imperial Guptas. If there is, we have to ask whether, while using some unidentified era , they cannot be minor contemporaries of the Later Guptas who were definitely in the A.D. period. Seen all round, their inscriptions do not bear out the non-Puranic chronology for the Imperial line.
The Current Chronology Impossible
On the strength of all that we have studied so far we may go to the extent of asking whether the Imperial Guptas can even possibly be placed after 320 A.D. up to nearly the end of the third quarter of the 6th century. What most prompts the question is the
1.Gupta Inscriptions, p. 95.
2.Bharatiya Vidya, Vol. V, Miscellany, p. 108.
Page 493
information already got from Majumdar that the Later Guptas were ruling over Gauda, Magadha and Mālvā from the very beginning of their dynasty. Is it not incompatible with this information to posit the Imperial Guptas as rulers over the same provinces? If Āditya-sena, the 8th of the other line, came to the throne, as said by Majumdar, in the last quarter of the 7th century A.D., and if the 4th ruler, Kumāra-gupta who defeated Isana-varman, was on the throne round about 550 A.D., we have about 125 years for four members of the dynasty. So we may with reason put the first of the Later Guptas, Krishna-gupta, another 125 years earlier, in about 425 A.D. The Later Guptas may not have been uninterrupted masters of Magadha, Gauda and Mālwā: they may have had ups and downs - some of which we know of - but they are shown to have been masters in general of dominions which our historians give to the Imperial Guptas. 414 or 415 A.D. is the accession-date allotted to Chandragupta II's son, Kumāragupta I who reigned for 40 years2 and was followed by Skandagupta, Kumāragupta II, Purugupta, Budhagupta, Narasirhhagupta, Kumāragupta III and Vishnugupta, the last two being allotted the years between 535 and 570 A.D.3 We are told that the Later Guptas ruled over a powerful principality within Magadha, but this is a supposition made in order to accommodate the modern version of history. We are quite free, on the facts in our possession, to discard it and give them all Magadha. The Imperial Guptas seem not only unnecessary but out of place in the period ascribed to them.
Any Chronological Connection between the Two Gupta Families?
Perhaps a Parthian shaft will be tried on chronological lines: "Does not the run of the Later Guptas where chronologically you with modern historians place them imply for their Imperial namesakes a close connection with them in time?" An answer negating in no way our new chronology for these namesakes can be framed with the help of modern historians themselves.
Majumdar4 writes: "The name or surname Gupta is not abso-
1.The Classical Age, pp. 73-75.
2.Ibid., p. 23.
3.Ibid., p. 43.
4."The Rise of the Guptas", ibid., p. 1,
Page 494
lutely unknown in ancient Indian history. Names of officials ending in Gupta and queens of the Gupta clan or family are mentioned in old records, particularly of the Śunga and the Sātavāhana period. But whether all of them were connected with a Pārent Gupta clan or there were different families, without any such connection, who assumed the name Gupta, cannot be determined, though the latter view seems more probable. In any case, the particular Gupta family, ruling in the fourth century A.D., cannot be affiliated to any ancient family or clan." Here the central point driven home directly or indirectly is that there were several separate Gupta families and they could be as far apart in time as the Śungas and the Imperial Guptas who were divided by at least 400 years.
The next point to note comes out in Raychaudhuri's statement:1 "Petty Gupta dynasties, apPārently connected with the Imperial Line [which ended in about the middle of the sixth century A.D.], ruled in the Kanarese district during the twelfth and thirteenth centuries A.D., and are frequently mentioned in inscriptions." Here we see that in the eyes of modern historians the time-factor need not be crucial for relationship. The converse of the fact that related Guptas could be widely separate in time is that unrelated Guptas could be chronologically near as well as that the placing of a Gupta family in a certain period does not necessarily entail a close link in time with those who happen to be known as Guptas, whether they be related or unrelated.
Finally, were it said that if there is a relationship the chances of time-proximity are liable to be more than if there is none, we should reply in the words again of Majumdar to cut the very ground from under this plea. Writing against the background of the modern chronology he2 passes on to us the following comment on the hypothesis "that the Later Guptas were descended from the Imperial Guptas": "The similarity of the name-endings, the common names like Kumāra-gupta and Deva-gupta, and the fact that the Later Guptas came into immediate possession of a large part of the dominions of the Imperial Guptas, no doubt favour such a supposition, and it has even been suggested that Krishna-gupta, the founder of the dynasty, was identical with Govinda-gupta, a son of Chandragupta II, whom we know from a Basārh seal and an
1. The Political History of Ancient India, p. 412.
2. Op. cit., p. 76.
Page 495
inscription. But there is not enough evidence in support of this identification. On the other hand, we should remember that not even the slightest hint of any such relationship is given in the records of the Later Guptas. It is difficult to believe that the court-poets of the Later Guptas would have missed such a splendid opportunity of glorifying their patrons if they had the least claim to such an illustrious lineage." No relationship exists to render possible any argument for more chances of time-proximity, even apart from the inherent inconclusiveness of the argument. The Parthian shaft is from every angle totally wide of the mark.
14
The second of the two further "hurdles" to the Purānic chronology of the Imperial Guptas is the famous Mandasor Inscription No. 52 in which a Kumāragupta is said to have been "the ruler of the earth" in Mālava Era 493.1 On the assumption that the era of the Imperial Guptas is from 320 A.D., the reign of Kumāragupta I can be placed in 415-455 A.D. And on the assumption, common to all modern historians, that the Mālava Era is the same as the era of 57 B.C., associated with the traditional king Vikramāditya of the Mālava province, whose capital was Ujjayinī, the year 493 added tp 57 B.C. gives us 436 A.D. - a date fitting into that reign-period.
The Correct View of the Mālava Era
The whole argument apropos of the Mandasor Inscription is vitiated by the common folly of all modern historians in identifying the Malaya Era with the Vikrama Era. There is no intrinsic reason for the identification. What we are justified in holding is that the Mālava Era is identical with the Krita Era found in various inscriptions but not that either of these eras is identical with the Vikrama Era. The Mandasor Inscription of Naravarman2 dated 461 and the Negari inscription of some Vaishyas3 dated 482 speak of both the Mālava and the Krita Eras together. But no epigraphic proof exists for equating them with the Vikrama Era. Nowhere do we see the latter mentioned as another name for either of those
1.A New History..., pp. 181-82.
2.Sircar, Select Inscriptions, p. 377.
3.Bhandarkar, List of Inscriptions of Northern India, No. 5.
Page 496
two. The threefold equation is made merely because it is convenient for the modern chronology of the Guptas.
The Correct View of the Mandasor Inscription
In the second place the particular chronological situation posed by the inscription is confused by our historians. What the situation is may be gathered from Majumdar's account,1 though it identifies the Mālava Era with the Vikrama:
"After referring to Kumāra-gupta as the ruler of the earth, [the inscription] mentions king Viśva-varman and his son Bandhu-varman. While this Bandhu-varman was ruling over Daśapura a temple of the Sun-god was built by the guild of silk-cloth weavers in the Mālava year 493 (=436 A.D.). In course of time, under other kings, part of this temple fell into disrepair and so in the year 529 (=472 A.D.) the same guild repaired the temple.
"The main object of the inscription was thus to record the repair of the temple in 472 A.D., and it must have been composed in or shortly after that date. Under ordinary rules of construction, Kumāra-gupta should be understood to have been the overlord at the time the record was set up i.e., in 472 A.D., but most of the scholars have taken the references to Kumāra-gupta in connection with the original construction of the temple."
Among the dissenters who connect Kumāra-gupta with the time of the repair, Majumdar2 names two historians and, in consideration of the fact that Kumāra-gupta I who reigned in 415-455 A.D. by Fleet's epoch is ruled out, he adds that these historians "naturally take this Kumāra-gupta to be Kumāra-gupta II".
Anybody approaching the inscription and its object dispassionately must admit that the Kumāragupta concerned has to be connected with the year of the repair. However, by Fleet's epoch it cannot be Kumāragupta II, for the date of his accession is the year 154 of the Gupta Era,3 which, by Fleet's epoch, comes to (320+154=) 474 A.D., a date wide of the mark for our inscription by 2 years. We have somehow to fall back on the earlier Kumāragupta and the Mālava year 493, yet not at the expense of the ordinary rules of construing the inscription's references to "the
1."The Expansion of the Gupta Empire", A New History..., pp. 181-82.
2.Ibid., p. 182, fn. 1.
3."The Imperial Crisis", The Classical Age, p. 29.
Page 497
ruler of the earth". In other words, the same Kumāragupta should be taken as such a ruler both in the Mālava year 493 and in the Mālava year 529, during the careers of several local kings of Dasapura. Surely, this is the correct view of the inscription: both the extremes of the inscription's time-count must have the contemporary sovereign's mention. As only one sovereign is brought in, he must cover the two extremes. But how will Kumāragupta I do it? By Fleet's epoch he can cover the year 493, for, by equating the Mālava Era with the Vikrama we get 436 A.D., which falls within Kumāragupta I's 415-455 A.D. The year 529 escapes him, reducing itself as it does by the Vikrama Era to 472 A.D.
Thus the particular chronological situation posed by the inscription excludes either Kumāragupta on the basis of Fleet and consequently excludes that basis as well as the equation of the Mālava with the Vikrama Era. In the context of the modern historical perspective our inscription hangs in the air.
We have to seek another perspective. And we may start by a look at the Mālavas and their emergence into history.
The Mālavas
The Mālavas, who ultimately settled in Central and Western India after a stay in Rājputāna, are an old people. Their republic is mentioned by the historians of Alexander's invasion of India: they were then at the confluence of the Ravi and the Chenab. They were in the same or another region of the Punjāb as far back as the time of Pānini who names them (IV.2,45) among the "warrior communities" of "the land of rivers" (V. 3, 114). The very word "gana" which occurs several times in the inscriptions of the Mālava Era - Mālavanām ganasthityā, Mālava-gana sthitivasāt, Mālavaganānnāta - serves for Pānini to classify them (III. 3, 86) as republics (Samghas). Even according to the modern dating of Pānini, the Mālava Era could easily be as old as any date upward of 500-400 B.C. which is the lowest limit possible for him in the time-scheme favoured by our historians. This era can certainly antedate by a number of centuries the Vikrama Era and may be taken to mark some important event connected with the great love of liberty and independence which the Mālavas are reported by Alexander's historians1 to have expressed to the Macedonians.
1. McCrindle, The Invasion of India by Alexander the Great, p. 154.
Page 498
Perhaps a leader named Krita figures in that event and so lends the Mālava Era its other designation.
The True Starting-point of the Mālava Era
Going by the failure of the modern historical perspective for the Mandasor Inscription, we may resort to the sole alternative left -namely, the one that identifies Sandrocottus with Chandragupta I of the Imperial Guptas instead of Chandragupta Maurya. And here our guiding light should be the realisation that Kumāragupta I, with his reign of 40 years from Gupta Era 95 to 135, should compass the 36 years from Mālava Era 493 to 529 of our inscription.
Counting from 315 B.C. as the accession-date of Chandragupta I, we get for the years 95-135 of Kumāragupta I the reign-period 220-180 B.C. Any of the first 4 years of the 40 years could suit our inscription's 36, but it would be in the fitness of things for us to frame these 36 years in the centre of this reign-period, making them run from 218 to 182 B.C. Then, to reach the starting-point of the Mālava Era, we may count either 493 years backward from 218 B.C. or 529 from 182 B.C. We arrive at 711 B.C.
A Test for the New Starting-point
We may check 711 B.C. for the Mālava Era by scrutinising another Mandasor inscription about which Majumdar1 writes: "An inscription found at Mandasor2 records some constructions by Dattabhata, Commander-in-chief of the forces of king Prabha-kara, in the Mālava Samvat 535 (=467-68 A.D.). The inscription mentions emperor Chandra-gupta II and his son Govinda-gupta, and we are told that Dattabhata's father Vayurakshita was the general (senādhipa) of Govinda-gupta. The date of the record places it during or immediately after the reign of Skanda-gupta and presumably Prabhakara was his governor. But it raises several interesting problems. First, as to the position of Govinda-gupta. He is also known to us from a clay-seal at Vaisall which records the name of 'Mahādevi Śrī Dhurvāsvamini, wife of the Mahārājādhirāja Śrī Chandra-gupta, mother of the Mahārāja Śrī Govinda-
1. "The Expansion of the Gupta Empire", A New History..., p. 180.
2. Bhandarkar's List No. 7.
Page 499
gupta'. It has been suggested that Govinda-gupta was the governor of Vaiśālī during his father's reign. In that case it is not very likely that he was alive in the year 467-68 A.D., far less that he held any important position in that year. How are we then to explain the reference to him as a great ruler in the Mandasor record which does not even mention the name of the emperor Skanda-gupta? Dr. Bhandarkar has pointed out that 'as Indra is represented as being suspicious of Govinda-gupta's power, the latter seems to have been a supreme ruler'. This would mean that he had rebelled, either against his brother Kumāra-gupta or the latter's son Skanda-gupta..."
The feature we are concerned with in this story arises from the fact that Dattabhata, while speaking of his father in the past tense, speaks of Govindagupta as if he were alive at the time of the inscription. If that were not so, Majumdar would not weigh the problem whether Govindagupta were "alive in the year 467-68 A.D." Majumdar decides that in this year, "the last-known date of Skanda-gupta",1 Govindagupta's having been alive "is not very likely". How theft is it that Dattabhata suggests the opposite? The plain answer is: the inscription, dated by the Vikrama Era of 57 B.C., fits ill into the present historical scheme, but, as this scheme has no other era to date by and yet cannot separate Dattabhata from the Imperial Guptas, the scheme itself which puts Kumāragupta and Skandagupta in the 5th century A.D. becomes suspect.
Operating with our Mālava Era of 711 B.C. and our Gupta Era of 315 B.C., we get for the inscription the date: (711-524=) 187 B.C. and find that this year comes 7 years before the end of the reign of Kumāragupta I. The possibility of Govindagupta's being alive at this date is certainly much greater than at the end of Skandagupta's reign about 20 years later in or even after Gupta Era 148 or, in our time-scheme, 167 B.C. The inscription of Dattabhata clearly supports this time-scheme rather than the current chronology vis-à-vis the Mālava Era and confirms the starting-point we have given to the latter.
(Any of the other 3 years - 220, 219, 217 B.C. - of Kumāra-gupta's reign open to us for reaching back to the starting-point would serve as well.)
1. Op. cit., loc. cit.
Page 500
Yaśodharman as Vikramāditya of Ujjayinī
711 B.C. for the Mālava Era yields also a most illuminating relation with Mālava history as recorded in two other Mandasor inscriptions, one of which is dated the year 589 of the Mālava Era and both of which refer to the magnificent military triumphs of a king of Mālava named Yaśodharman.
Yaśodharman, the Aulikara chief, seeming originally to belong to a family of feudatories under an unknown overlord, rose into prominence by his crushing victory over the formidable king Mihirakula and became the sovereign of an empire which extended from the Himālayas to the Western Ocean, from the river Brahmaputra to the Mahendra mountain and included "those countries which were not enjoyed (even) by the Gupta lords and which the command of the chiefs of the Hūnas failed to penetrate".1 Now, if 711 B.C. is the commencement of the Mālava Era, the year 589 brings us to (711-589=) 122 B.C. We are not far from the time of the traditional Vikramāditya of Ujjayinī and the era of 57 B.C. associated with his name. Could we say that the legend of Vikramāditya took its rise from the achievements of Yaśodharman?
R. B. Pandey,2 who has no thought of identifying the two figures, is struck by the fact that the empire of Vikramāditya as described in the Brihatkathāmanjari and the Kathāsaritsāgara should be paralleled almost uniquely by that of an actual Mālava king. But how shall we relate Yaśodharman to the era of 57 B.C.?
If we could say that he ruled - though not necessarily always over the same extent of empire - from c. 122 B.C. to the year of this era - having a reign of 65 years, not an impossibility for one who may have risen to power very young - and that the era was established by his subjects at his death in grateful memory of his ridding Mālava of its enemies, there would be little difficulty in conceiving an equation between the real master of Mandasor (ancient Daśapura) and the legendary lord of Ujjayinī, both places neighbouring ones in Mālava.
We may even submit a reason for the switch from Daśapura to Ujjayinī in the popular mind. Modern historians seek to explain several aspects of the Vikramāditya-legend by an appeal to some
1.ibid, p. 203.
2.Vikramāditya of Ujjayinī, p. 127.
Page 501
attested facts of Gupta history. Bhandarkar1 was the first to emphasize the significance of the third Gupta emperor, Chandra-gupta II, in the controversy over Vikramāditya. It is well established that Chandragupta II took the title Vikramāditya as well as extirpated the satrap Śakas of Western India. He is also known to have made Ujjayinī a secondary capital after his victory over the Śakas. If Yaśodharman was the original of the legendary Vikramāditya, his Daśapura could easily become that figure's Ujjayinī under the influence of Chandragupta II Vikramāditya's bringing Ujjayinī into eminence towards the end of his reign as computed by us: 259-221 B.C.
As for Yaśodharman's reign of 65 years, we may refer to Merutuhga's Therāvali,2 telling the story of the legendary Vikramāditya. It assigns to him a reign of 60 years, a number close enough to ours. But Merutuhga shares the popular belief which attributes the Vikrama Era to the king himself. Popular belief, however, is not backed by all tradition. As we noted earlier, many Jain works3 consider the era as marking the death of Vikramāditya. Majumdar, who has quoted Merutuhga, quotes another Jain source too where we learn: "135 years after Vikrama having passed, again the Śakas expelled Vikramaputra (Vikrama's son or descendant) and conquered the kingdom."4 Here the reference is to the so-called Śaka Era of 78 A.D., 135 years prior to which would be 57 B.C., marking the year of "Vikrama having passed", the year of his death. Again, as we observed before, several inscriptions no less than literary compositions count 57 B.C. as Vikramāditya's death-year.5 So it is extremely probable that Vikramāditya was Yaśodharman and that the celebrity he won for destroying the Śakas was due to his triumph over Mihirakula who, as we shall presently argue, must have been a Śaka and not, as commonly supposed, a Hūna.
Yaśodharman is generally taken to be a destroyer of the Hūnas, and he may have come into some conflict with them, but actually
1.The Journal of the Bombay Branch of the Royal Asiatic Society, 1900, p. 398.
2.The Age of Imperial Unity, p. 155.
3.The Journal of the University of Nagpur, December 1940, pp. 52-3.
4."The Vikrama Sarhvat and Śakabda", The Age of Imperial Unity, p. 155.
5.Epigraphia Indica, XIX, Appendix No. 109; XX, Nos. 383, 455; "Inscriptions of Kathiawad", New Indian Antiquary, June 1940, p. 112; Bhāratiya-Prāchina-Lipi-Mālā, p. 170.
Page 502
his great inscription1 mentions no conquest over the Hūnas and merely says that he penetrated where neither "the Gupta lords" nor "the chiefs of the Hūnas" had gone. It gives prominence solely to his victory over Mihirakula who is nowhere in it called a Hūna and whose defeat is mentioned quite separately from the allusion to the Hūnas. Majumdar2 has remarked: "In the Mandasor inscription reference is made both to Mihirakula and to the Hūnas, but in a manner which, far from connecting the two, might even suggest a definite distinction between them." Evidently, if an hypothesis alternative to the current one could be elaborated, the distinguishing attitude of the inscription would be more logically explained.
The Hūnas, the Śakas and Vikramāditya
Before we go further, a small point may be disposed of. We may be told: "Vikramāditya, unlike Yaśodharman, is associated only with the Śakas, never with the Hūnas, and never thought of as in possible conflict with them." The answer is that this is a wrong assertion. The Brihatkathāmanjari,3 in recounting Vikramāditya's conquests, includes the defeat, at his hands, of "the mean Hūnas with barbarous hordes". What is true is not the non-association with the Hūnas but their minor role as compared to that of the Śakas in his life. And this is precisely the truth we are prompted to feel in regard to Yaśodharman by the attitude of the inscription.
Mihirakula: Hūna or Śaka?
What are the facts, if any, outside the inscription to favour the Hūna hypothesis about Mihirakula rather than the alternative we propose? Our historians hold him to be the son of a chief from North India who goes by the name of Toramāna and who is mostly regarded as a Hūna; but, while inclining to agree, Majumdar4 honestly tells us: "There is no conclusive evidence to this effect... We possess little definite information about him." Equally little ground do we have to take Mihirakula for a Hūna. None of his
1.Corpus Inscriptionum Indicarum, III, pp. 147-48.
2.The Imperial Crisis", A New History..., p. 197.
3.X, i, 285-86, quoted by Pandey, op. cit., p. 114.
4."The Disintegration of the Empire", The Classical Age, p. 35.
Page 503
inscriptions or Toramāna's, says Majumdar,1 "calls these kings Huns or contains any reference to that people." Majumdar2 further admits there is a question whether Toramāna and Mihirakula were Hūna or Kushāna. His final verdict3 goes still further: "... there is no conclusive evidence as to their nationality."
Since the time this was written some new information has come our way. Two seals have been discovered in Kauśāmbi during the excavation of the monastery at Ghositārāma.4 One of them is counter-struck with the letters To Ra Ma Na, the other with the legend Hūna-rāja. The inference is likely to be drawn that the reference in either seal is to the same individual. But would such an inference be binding any more than the one commonly drawn that the mention of Mihirakula as well as of the chiefs of the Hūnas by Yaśodharman in the same inscription proves Mihirakula to be a Hūna chief? Just as Majumdar has pointed out the distinct lack of connection between the two mentions, we may emphasise the fact that neither of the two seals connects "Toramāna" and "Hūna-rāja". From Yaśodharman we know that the chiefs of the Hūnas no less than the Gupta lords were contemporaries of Mihirakula. They would naturally be contemporaries of Toramāna, too. It is therefore quite conceivable that the two Kauśāmbi seals speak of two different kings. Conclusive evidence as to the nationality of Toramāna and his son is still wanting.
Nothing prevents us from looking on them as neither Hūna nor Kushāna but as Śaka. Majumdar5 has cited an "inscription found at Kura (Salt Range, the Punjāb), referring to Rājādhirāja Mahārāja Toramāna-Shāhi-Jau(bla)." Inscriptions of Mihirakula found in Utuzgen (central Afghānistān) also bear the titles "Saho Zobol" or "Saho Jabula", which A. D. A. Bivar6 suggests to have been the official title of the dynasty. Now, "Shāhi" is not only a Kushāna designation. Majumdar7 himself, in his summary of Kālā-kachārya-kathā's story of the Śaka conquest of Ujjayinī, speaks of Kālākachārya going to the west of the Sindhu and bringing from
1."The Imperial Crisis", A New History..., p. 196.
2."Minor States in North India during the Gupta Empire", The Classical Age, p. 59.
3."The Imperial Crisis", A New History..., p. 194.
4.G. R. Sharma, The Excavations at Kauśāmbi (1957-59), pp. 15-16.
5."The Imperial Crisis", A New History..., p. 196.
6.Journal of the Royal Asiatic Society, 1954, pp. 112 ff, 115 ff.
7."The Vikrama Sarhvat and Śakabda", The Age of Imperial Unity, p. 155, fn.
Page 504
there 96 "Shāhi" chiefs who belonged to "Sagakula". Indeed, if we put Mihirakula's defeat by Yaśodharman in 122 B.C. and accept the view of all historians1 that Mihirakula was in possession of Kāshmir, he fits accurately into the picture which the well-known Sinologist P. C. Bagchi has drawn of the Śaka influx into India a little earlier as reported by Chinese historians.
Scrutinising the Chinese account regarding the movement of the Śakas (whom, by the way, the Chinese call "Saī"2) from the region of Ta-hia to the region of Ki-pin and disputing the identification of Ki-pin with Kābul-Kapiśa, Bagchi3 says: "The route to Ki-pin which the Śaka king followed is clearly stated. He passed the Hien-tu or the hanging pass while moving towards Ki-pin from the original seat in Ta-hia. This route was recognised by Chavannes and Sir Aurel Stein as the Bolor route through the Yasin valley. This was the route which was usually followed by the ancient travellers from the region of Wakhan to the Indus valley and to Kāshmir and Udyana. If we follow the Chinese account literally we are driven to two conclusions: the first that the Śakas who were turned out of Ta-hia by the Yue-chis entered the Bblor route and second that Ki-pin which they conquered was Kāshmir.... This must have taken place before 128 B.C.... The identification of Ki-pin with Kābul-Kapiśa is an impossibility. Lévi and Chavannes were the first to propose the identification of Ki-pin of the Chinese annals with Kāshmir. They pointed out that in a number of Chinese translations of Buddhist texts the translators use Ki-pin for translating the name Kāshmir up to 581 A.D. Since 581 A.D. the Buddhist translations as well as the Chinese documents use the name Ki-pin to denote Kapiśa and not Kāshmir."
Bagchi then shows that the geographical details given in the Chinese account point to Kāshmir, and he adds: "The name Ki-pin itself seems to suggest the same. In Han pronounciation, the first word Ki was definitely a Ka followed by some consonant which might have been an s. Pin was pronounced almost certainly in early time pir or wir. Hence Ki-pin clearly stood for Ka(a)-pir or Ka(s)-wir. This form of the name is also found in the early Greek records in which Kāshmir is either Caspiri or Kasperia. Ki-pin was thus a correct phonetic transcription of the old name of Kāshmir.
1. The Classical Age, p. 35.
2. The Age of Imperical Unity, p. 122.
3. Presidential Address, The Indian History Congress, 1943, pp. 33-35.
Page 505
The Śakas of Ta-hia could not have come to Kāshmir via Seistān and the lower Indus valley - they must have come there by the shorter route, i.e., the Bolor route from Ta-hia."
Thus, if Mihirakula's defeat is put in 122 B.C., there is nothing incongruous in considering this king of Kāshmir a Śaka successor to the Śakas who conquered Kāshmir before 128 B.C.
Against this may be urged the remarks of S. N. Sen1 about the Chinese pilgrim Sung-yun: "Early in 520 A.D. Sung-yun entered the country then under Mihirakula's rule", and he says also: "The contemporary account of Sung-yun of that ill-mannered ill-favoured barbarian warrior is not without interest." From this we may incline to think that Mihirakula could not have lived in the 2nd century B.C. But the fact is that his contemporaneity with Sung-yun is merely an inference inspired by the current chronology. Sung-yun gives no name to any Hūna king of his own day, and if we accept the current chronology and believe that Sung-yun met a Hūna king, all we can say of Mihirakula is, in Majumdar's words:2 "It is not... unlikely that he is the king whom Sung-yun met in Gandhāra." If we do not accept this chronology, nothing obliges us to put Mihirakula in Sung-yun's time. Further, the description Sung-yun gives seems to be of a king whom he places not in his own time but two generations prior to himself,3 so that it cannot in any case be applied to Mihirakula.
The contemporaneity of Sung-yun and Mihirakula or, in general, the current dating of the latter is also not binding on us from the account left by Cosmas, surnamed Indicapleustes (Indian navigator), an Alexandrine Greek, in his Christian Topography, a book probably begun in A.D. 535 but not put in its final form till A.D. 547. Cosmas in one place speaks of a White Hun King named Gollas as the Lord of India and an oppressor of the people. Majumdar4 comments: "It is generally believed that king Gollas in the above account refers to Mihirakula whose name is also written as Mihiragul." Here a certain semi-echo in the name can be granted; yet the identification hardly forces itself. The characteristic component of the name is surely "Mihira", not "gul". And
1.India through Chinese Eyes (University of Mādras, 1956), pp. 163-64.
2."The Disintegration of the Empire", The Classical Age, p. 36.
3.S. Beal, Si-yu-ki: Buddhist Records of the Western World, Vol. I, Introduction, pp. 99-100.
4.Op. cit., p. 36.
Page 506
there is a significant fact which works against the conjectured identification. "It must be noted," observes Majumdar,' "that, whereas the chief seat of Hūna power, according to both Sung-yun and Cosmas, was to the west of the Indus, Mihirakula's capital, according to Hiuen-Tsang, was at Sakāla (Sialkot) and that of Toramāna, according to the Jaina book [Kuvalaya-mālā], on the river Chenāb."
Even the name "Mihirakula" has nothing exclusively Hūna about it. It only means: "Of the family of the Mihiras." "Mihira" is the same as "Mer" or "Meher" and signifies "Sun". The Hūnas were indeed sun-worshippers, but the Śakas too were most certainly the same, as we know from the Purānas2 where Śaka-dvīpa is the land of the Maga-dvijas who worship the sun-god. The cult of Shiva which Mihirakula, if we may judge from his coins, seems to have practised is also known to be characteristic of the Śakas: the Mahābhārata (VIII. II, 8-38) regards Śaka-dvīpa as a centre of Shaivism.
What makes it still more probable that Mihirakula was of the Śaka group ("Sagakula", as the Kālakāchārya-kathā names it) that came to India and ruled over Kāshmir a little prior to the time of Vikramāditya and that Yaśodharman who defeated him was this very Vikramāditya is that Toramāna whom our historians look upon as Mihirakula's father is put by Kalhana's history of Kāshmir, Rājatarahginī, just one generation before Vikramāditya. Kalhana separates Mihirakula from Toramāna by a good number of generations and hence far beyond any such date as either our historians or we ourselves give to Yaśodharman. But if, as is almost certain, Toramāna is Mihirakula's father, the Śaka-destroyer and era-inspirer Vikramāditya of Rājatarahginī (III) must belong to the same time as Yaśodharman.
In another way too Kalhana leads to the conclusion that the time of Yaśodharman was the same as Vikramāditya's. Hiuen Tsang3 has described king Bālāditya's fight with Mihirakula. Kalhana (III) suggests the contemporaneity of Bālāditya and Toramāna. He states that Toramāna forbade the currency of the coins with the name "Bāla" within his territory and largely circulated the dīnāra
1."The Imperial Crisis", A New History..., p. 198.
2.Agni, 119-21; Brāhmanda, XX.71; Kurma, 1.48. 36-37.
3.Sen, op. cit., pp. 135-39.
Page 507
coined by himself. D. N. Mookerji1 remarks: "It is evident therefore that Toramāna ruled for a few years contemporaneously with Bālāditya." And hence not only can Toramāna's son Mihirakula be considered a contemporary of Bālāditya but also Bālāditya can be considered a contemporary of Vikramāditya whom Toramāna preceded by a generation. And since Bālāditya and Yaśodharman were contemporaries the time of Vikramāditya must be the same as Yaśodharman's.
Hiuen Tsang, whom we have mentioned above, has himself indicated the time of the defeat of Mihirakula by Bālāditya. Majumdar2 informs us: "The most important point to be noted is that Hiuen Tsang places the defeat 'some centuries ago', i.e., several hundred years before c. 633 A.D. when he visited Śakala. This is hardly compatible with the view that the incident referred to by him took place about 530 A.D. As Watters has pointed out, other Chinese authorities also seem to place Mihirakula long before that date." Majumdar's own reaction is: "This naturally casts grave doubts on the credibility of Hiuen Tsang's story about Mihirakula." But Hiuen Tsang's chronology, supported by other Chinese authorities, is just what it should be by our hypothesis.
Hiuen Tsang's Chronology of the Imperial Guptas
And we may draw attention to the fact that the Chinese pilgrim is not only in general agreement with Our criticism of the modern view but also indicates in particular the correctness of our own by his hint on the age of Bālāditya. Sen3 says that in Hiuen Tsang's account Bālāditya was the great-grandson of Sakrāditya who is credited by the pilgrim to have built the earliest monastery of Nālandā shortly after Buddha's death, and in between Sakrāditya and Bālāditya we have the mention of Sakrāditya's son Buddha-gupta and of Buddhagupta's son Tathagatagupta. Many of our historians4 identify Hiuen Tsang's Bālāditya with one of the kings of the Gupta dynasty, Narasimhagupta surnamed Bālāditya. For, the pilgrim's Buddhagupta, grandson of Sakrāditya, must be identified with the historical Budhagupta, grandson of Kumāragupta I,
1.Bhāratiya Vidyā, Vol. V, Miscellany, p. 113.
2.Op. cit., pp. 196-97.
3.Op. cit., pp. 135-39, 162.
4.The Classical Age, pp. 37, 43.
Page 508
especially as Kumāragupta's title Mahendrāditya includes like Sakrāditya a name of Indra. So we may conclude that Hiuen Tsang believed the Imperial Guptas to have flourished shortly after the Nirvāna of Buddha. Of course, the belief is mistaken: both the Indian time-scheme and the modern can- demonstrate its shortcoming, but the mistake should drive home with some precision Hiuen Tsang's sense of the antiquity of the Guptas.
On Fleet's theory, Narasimhagupta who has to be placed just before the year 215 of the Gupta Era1 would be just previous to (320+215=) 535 A.D., fairly close to Hiuen Tsang's own time. If Hiuen Tsang had thought his Bālāditya to be pretty close and yet affirmed Sakrāditya to have existed shortly after Buddha's death we should absurdly have had to consider Hiuen Tsang as dating Buddha in the early years of the Christian Era. But he does nothing of the sort although he gives four different views prevalent in his time about the number of years elapsed since Buddha's Nirvāna: (1) 1500 years, (2) 1300 years, (3) 1200 years, (4) 900 years.2 As Hiuen Tsang's own time in India is c. 630-643 A.D., none of the views puts Buddha's death after Christ: the latest date is (900-643 A.D.=) 257 B.C. We do not know which of the four dates was favoured by Hiuen Tsang himself, but, judging from even the latest, we can be certain that he did not regard the Guptas as having been near his own day: they were to him sufficiently far to go considerably beyond Fleet's epoch and, by even the latest of his own dates for Buddha's death, they were for him ancient enough to be previous to Christ. By the modern chronology the last Gupta was in c. 569 A.D., just a little over 60 years before the Chinese pilgrim's visit to India.
Clearly, Hiuen Tsang can be taken to support our chronology of the Guptas. And his expression, "some centuries ago", in connection with Mihirakula can be no slip on his part. So far as he is concerned, we are free, if not actually encouraged, to put Narasirhhagupta Bālāditya where our Gupta Era of 315 B.C. would want him - namely, just before (315-215=) 100 B.C., a time which bears out our dating of Yaśodharman, since Yaśodharman fought Mihirakula before Bālāditya did so.3 His inscription says
1.The Classical Age, p. 43.
2.Watters, II, p. 28.
3.A New History..., p. 201, fn. 1, the first 6 and a half lines.
Page 509
that Mihirakula's head was never before bent to anybody.1 Thus, through Bālāditya, Hiuen Tsang allows for Yaśodharman the period of the traditional Vikramāditya and sanctions broadly our identification of the one with the other. He also lends support, by his chronology, to our proposition that Mihirakula was a Śaka king whom Vikramāditya defeated and by whose overthrow he won the title "Śakāri".
Yaśodharman, the Guptas and the Ephthalite Huns
Perhaps a last volley will be fired at us in the following terms: "A very cogent argument, against transposing Yaśodharman to 122 B.C. and identifying him with Vikramāditya as well as against setting the Guptas in the pre-Christian period, is our knowledge that the Hūnas who broke into India were the Ephthalite Huns and that these Hūnas started moving east from the Oxus valley in about the middle of the 5th century A.D. and entered India in the time which on Fleet's Gupta Era would be the reign of Skanda-gupta who himself in the Bhitari Seal Inscription refers to his victory over them. Facing these facts, how can we admit the wholesale chronological revolution you propose?"
Quite a good answer can be given. We have already disconnected Mihirakula and Toramāna from the Ephthalite Huns. There is no evidence, either, that the invaders of India during the reign of Skandagupta were these tribes whose campaigns our historians date in the middle of the 5th century A.D. - unless we take it for granted that Skandagupta lived where Fleet would place him. Nor are the Hūnas of Yaśodharman's inscription distinguished by anything to render them identifiable with the Ephthalites. Indian chronology has nothing to make the Hūnas (Chinese "Hiung-nu" or "Hsung-nu") impossible in India in the pre-Christian period. The Purānas put them no less than the Śakas and Tushāras in just this period. And we know that the Huns were active on the borders of China for over 300 years before Christ. We may cast a glance at their activities.2
In 318 B.C. they were troubling the Chinese empire. In 215 they were again engaged in war and suffered a defeat. Round about 212
2.M. Grenet, La Civilisation Chinoise, pp. 51,112, 118, 119,122,128, 132, 145, 156.
Page 510
the Emperor Shi Huang-ti was completing the Great Wall to keep them out. Foiled by the Great Wall they might easily have sought other fields of adventure to the south-east. And indeed it was while the Great Wall was being completed that they formed for the first time a nation united and strong, capable of just such an excursion. On our computation Skandagupta's reign in Gupta Era 135-147 falls in 180-168 B.C., precisely the time when the Huns, on a south-eastern war-path, could have struck at India in the way our historians have pictured from the Bhitari Seal inscription. No a priori ground exists against a Hūna invasion in the 2nd century B.C. Whether it actually occurred or not can be judged only according as we date Skandagupta and Yaśodharman. Our chronological revolution stands intact.
Some Chronological Relations
What remains to be done is to take a few other inscriptions into account and fix two chronological relations properly. The first relation is between Toramāna and Mihirakula. The second is between Yaśodharman, who is known to have dealt with Mihirakula, and Bhānugupta who, according to the inscription, must have been in conflict with the same king.
Vainyagupta and Bhānugupta appear to have "ruled at the same time respectively over the western and eastern parts of the empire".1 The western part would include Eran where several of our inscriptions have been found. The inscription, dated in Gupta Era 191, "records how a feudatory chief named Goparāja accompanied 'the mighty king, the glorious Bhānu-gupta, the bravest man on earth', and fought a famous battle. Goparāja died in this battle..."2 Majumdar3 opines that the battle "at Eran was probably fought against Toramāna. For at a date which cannot be long removed... we find Toramāna as the overlord of Eran". The turn of Majumdar's phrase implies that we do not know the date at which Toramāna achieved his overlordship. And such is indeed the fact. Another Eran inscription simply records "the construction of a temple by Dhanyavishnu, after the death of his brother Mātri-vishnu, 'in the first year while the Mahārājadhirāja, the glorious
1.Majumdar, "The Imperial Crisis", A New History..., p. 190.
Page 511
Toramāna of great fame, is ruling the earth'.'1 All we can be sure of is that Toramāna was not master of Eran till the Gupta year 165. For, an earlier inscription of Mātrivishnu and Dhanyavishnu shows Eran still under Budhagupta,2 the father of Vainyagupta and Bhānugupta. The last absolutely certain year of Budhagupta is Gupta Era 175.3 Any year between 175 and the year of Bhānu-gupta's battle - 191 - could be Toramāna's "first year".
We may assume that, on Budhagupta's death, when internal dissensions broke out and the empire was split in two, Toramāna swooped down on Eran. We may date his conquest, therefore, in the 175th year of the Gupta Era. This gives us (315-175=) 140 B.C. Toramāna could not have been king of Eran for long. Apart from the inscription in his 1st regnal year there, we have no epigraph of his anywhere except the undated one from the Punjāb where he is called "Toramāna-Shāhi-Jau(bla)". For the length of his Eran-period we must turn to the epigraphs of Mihirakula and Yaśodharman. Majumdar4 informs us: "...there is an inscription at Gwalior dated in the 15th regnal year of Mihirakula... The name of the father is mentioned, but only the first two letters 'Tora' can be read, the rest being utterly lost. It has been restored as Toramāna." So we must allot at least 15 years to Mihirakula. Before Yaśodharman defeated him in 122 B.C., he must have reigned for that length of time: that is, from 137 B.C. Then Toramāna, as would seem appropriate, can be alloted no more than 3 years in Eran - from 140 to 137 B.C.
From such a chronological set-up it should be clear that the battle joined by Bhānugupta in the year 191 G.E. - or (315-191=) 124 B.C. - could not have been against Toramāna. It must have been a valiant effort to oust Mihirakula from Mālava. But the effort was in vain. "Unfortunately, the Eran inscription of Goparāja is silent about the result of the battle."5 Majumdar6 rightly remarks: "...it is somewhat strange that if [Bhānugupta] had really achieved such a great victory it should not have been expressly stated in the record while referring to him." Yaśodharman-Vikramāditya
1.Ibid., p. 191.
3.Ibid., p. 189, with fn. 1.
4.Ibid., pp. 195-6.
5.Ibid., p. 191.
Page 512
was destined to do within 2 years what Bhānugupta had failed to accomplish.
Did Skandagupta fight the Hūnas?
We have argued on the assumption that the Hūnas are connected with Skandagupta no less than with Yaśodharman. But in the interests of historical accuracy we should add that no strict proof exists to connect the Hūnas with the Gupta king.
The latter's Junāgarh inscription of the Gupta Era 136-8 refers "to a struggle with hostile kings including those against the Mlech-chhas".1 Majumdar2 comments: "The war with the Mlechchhas probably refers to his fight with the Hūnas which is specifically referred to in the Bhitari Pillar Inscription. But whether the Mlechchhas are the same as Hūnas, or were a different tribe, both the records claim that Skanda-gupta completely defeated these enemies. The verse describing the conflict with the Hūnas, though mutilated, leaves no doubt that it was a severe one." Surely, the mutilation Majumdar speaks of must render some things uncertain. That there was a severe conflict and that Skandagupta scored a victory is quite clear. What is unclear must then be the mention of the Hūnas. And this is precisely the fact.
Fleet originally read the inscription and noted some gaps. In one place his reading suggested that the king of a tribe called the Pushyamitras was the Gupta emperor's adversary. But, while construing the expression "Pushyamitrārhs cha", he observed that the second syllable of the name was damaged.3 H. R. Divakar proposes to read the whole thing as "Yudhyamitrarhs cha",4 which makes the expression allude to the hostile chiefs simply as enemies. Fleet's reading the Hūnas is also a conjecture. The word admits of no clear decipherment. Sircar, in some letters5 to the present writer, has frankly said so, even though wondering what else could possibly be read. So we have no sound reason to introduce the Hūnas into Skandagupta's reign period of 180-168 B.C. Unknown tribals from the west of the Indus seem to have
1."The Expansion of the Gupta Empire", Ibid., p. 177.
3.Ibid., p- 174, fn. 2.
5.Dated 3.8.1957, 6.9.1957, 17.9.1957.
Page 513
irrupted into India - a mixed horde which could be, termed Mlechchha.
Gradually the distinction among them of Śaka and Hūna must have taken place. The first clear mention of the latter is in the Yaśodharman-inscription's phrase: "the chiefs of the Hūnas." But here there is no indication of war between them and Yaśodharman any more than the reference to "the Gupta lords" is to a war with the Gupta family. Although some importance is undoubtedly attached to "the chiefs of the Hūnas" just as to "the Gupta lords", Mihirakula who stands by himself separately from the former as much as from the latter is the real cynosure of the epigraph's attention.
But, of course, whether or not Skandagupta really fought the Hūnas and whether the Bhitari Pillar named them or not makes no odds to our dating of the Imperial Guptas and of Yaśodharman, and to our view of Mihirakula's and Toramāna's nationality.
Who Founded the Era of 78 A.D.?
To round off this section a glance needs to be cast at the problem of the Śaka Era of 78 A.D. and see how it may figure in our new perspective as a sequel to the Vikrama Era of 57 B.C. Here there is a peculiarity seldom brought to notice. Repeatedly it is 78 A.D. and not 57 B.C. that is associated with Vikramāditya and his destruction of the Śakas.
Perhaps the earliest instance of this peculiarity is in Bhattotpala. Commenting on the terms "Śakendrakāla" and "Śakabhupakāla" in Varāhamihira's Brihatsamhitā (VIII. 20 and 21 respectively), he writes that the kings of the Śakas who were Mlechchhas were killed.by Vikramāditya and hence the era was called Śaka and began from the date of the slaughter of the Śaka kings.1 In reference to Varāhamihira there is a paradox here. The terms Varāhamihira has employed do not have the negative sense Bhattotpala has expressed. They suggest the coming into power of a Śaka king or kings rather than the killing of the kings of the Śakas: the sense is positive and out of accord with the comment. This is enough to make the comment rather eccentric.
1. The Benares edition, p. 193 gives the original Sanskrit. Transliterated by Shrava, op. cit., p. 47, it runs: "Śaka nāma mlechchha-jātayo-rājānaste yasmin-kāle Vikramāditya-devena vyāpaditāh sa kalo loke Śaka iti prasiddhah."
Page 514
The impression gains strength from our argument that Varaha-mihira's Śaka-kāla was of 551 B.C. and that his own date was in the first century B.C., contemporaneous with the legendary Vikramāditya himself: so the era of 78 A.D. would be beyond his ken. We have also argued that Bhattotpala subscribes to the ancient Śaka-kāla as well; but his mention of the era of 78 A.D. is not in conflict with his date which, according to us, is around 337 A.D. Only in regard to Varāhamihira the mention would be anomalous.
However, the comment found in Bhattotpala, for all its eccentric look both by being off the point of the older astronomer's terms and by appearing to be chronologically inapplicable, serves to stress that the name of Vikramāditya was at one time associated not with the era of 57 B.C. but with that of 78 A.D., the Śaka-kala which Shrava1 has shown to have been called "Śaka-nripa-kdlatita-samvatsara or the era at the expiry of the time of the Śaka king or kings". Nor is it a solitary freak. While the inscriptions do not bring in Vikramāditya in connection with such an era, literary allusions confirm the substance of that Bhattopala-eccentricity.
Āmarāja, whom we have quoted for the date of Vikramāditya's death and who was the commentator on Brahmagupta's Khanda-khādiyaka, states in language somewhat reminiscent of this strange gloss that when Śaka kings were killed by Vikramāditya the Śaka Era began.2 Then there is the locus classicus of the same information with a new point raised, which sets the Vikrama Era in the right relation vis-à-vis the Śaka Era: the passage from Alberum,3 part of which we culled when discussing his'pronouncement on the Gupta Era:
"The epoch of the era of Śaka or Śakakāla falls 135 years later than that of Vikramāditya. The here-mentioned Śaka tyrannised over their country between the river Sindh and the ocean, after he had made Āryāvarta, in the midst of this realm, his dwelling-place. He interdicted the Hindus from considering and representing themselves as anything but Śakas. Some maintain that he was a Sūdra from the city of Almansūra; others maintain that he was not a Hindu at all, and that he had come to India from the west. The
1.Op. cit., p. 49.
2.The Sanskrit reads: "Śaka-nāma mlechchhā rājānaste yasmin kāla Vikramā-dityena vyāpāditāh sa Śaka-sambandhi kālah Śaka ityuchyate" (Khandakhādyaka, with Vāsanā-Bhāshya by Āmarāja, p. 2, Calcutta Edition, 1925).
3.Op. cit., II. p. 6.
Page 515
Hindus had much to surfer from him, till at last they received help from the east, when Vikramāditya marched against him, put him to flight and killed him in the region of Karūr, between Multan and the castle of Lonī. Now this date became famous, as people rejoiced in the news of the death of the tyrant, and was used as the epoch of an era, especially by the astronomers. They honour the conqueror by adding Śri to his name, so as to say Śri Vikramāditya. Since there is a long interval between the era which is called the era of Vikramāditya and the killing of Śaka, we think that Vikramāditya from whom the era has got its name is not identical with that one who killed Śaka, but only a namesake of his."
What is outstanding in Albērūnī's pronouncement is not only that 78 A.D. is linked with the Śaka-killing exploit of Vikramāditya but also that there must be two Vikramādityas, since 57 B.C. which is 135 years earlier is linked too with the same name. The reason for this belief in 1031 AD., Albērūnī's date, seems to be the spectacular short-term rise of Yaśodharman, as we have shown, in the period allotted traditionally to the figure connected with 57 B.C. Since Yaśodharman too tackled a Śaka chief - Mihirakula -of great power, a parallal to the legend of the Vikramāditya of 78 A.D. arose for 57 B.C. - a parallel evidently not yet developed in the time of Alberum. This cause is additional to the one we touched upon earlier: the memory of the Śaka-destruction by Chandragupta II Vikramāditya in the third century B.C. in the traditional-Purānic scheme of chronology. However, the question remains: "Who was the Śaka-destroyer in 78 B.C., bearing the name Vikramāditya?"
We have ruled out Kanishka the Kushāna as the starter of the Śaka-kāla of 78 A.D. For one thing, India always distinguished the Kushāna-Kanishka-line as Tukhara or Tushara from the line of the Śakas. For another, Kanishka's own date of accession, even if found possible to put in the first century A.D. against the argument for the second, need not be fixed exactly at 78 once the Western Śaka satraps whose era was said to have commenced in that year are removed to a far earlier epoch. Actually, as Shrava points out, there would be at any period a verbal difference between their era and Kanishka's inasmuch as they counted from it in terms of varshe in contrast to the use of samvatsara in all the Kushāna inscriptions, copper plates, scrolls, etc.
With Kanishka out of the picture, whom shall we posit as the
Page 516
Vikramāditya of Āmarāja, Alberum and the odd text in Bhattot-pala's commentary? As with Yaśodharman the original for the legend of 57 B.C., we may take the name "Vikramāditya" as an attributive or descriptive term and not the personal label for the hero of 78 A.D. Someone performing certain actions came to bear this appellation. The appellation itself derived from a floating memory mainly of Chandragupta II (259-221 B.C. by our chronology) who was styled "Vikramāditya" and subsidiarily from other members of the Gupta line who took the same title: among these the chief was Skandagupta (181-169 B.C.), grandson of Chandragupta II. It is the grandfather who is remembered in the later honorific "Śakri" (Śaka-slayer), but very strikingly the name of the legendary Vikramāditya's father in both the books of about the eleventh century A.D., that celebrate him - the Brihatkathāmanjari and the Kathāsarit-sāgara - is Mahendraditya,1 the very name borne by Skandagupta Vikramāditya's father Kumāragupta.2 The two books are talking about the king whom the era of 57 B.C. celebrates: we have no details about the later warrior beyond what Alberum has to tell.
What we may call a negative detail of some help is the absence of "Ujjayinī" in Albērūnī's account of Vikramāditya. A pointer of positive guidance may be drawn from the Jain tradition about the earlier warrior. In Majumdar's recital3 of the Kālakāchārya-kathā's narrative of the Śaka conquest of Ujjayinī which was ruled by King Gardabhilla, the ravisher of Kālakāchārya's sister and thereby the prompter of the enraged brother to summon the Śakas from beyond the Sindhu to cross Sind and Gujerat and attack Ujjayinī, we read at the end: "After 17 years Vikramāditya, son of Gardabhilla, regained his kingdom by expelling the Śakas. Kālakāchārya, after defeating Gardabhilla and releasing his sister, went to the court of king Sātavāhana at Pratishthāna."
What stands out for our purpose, directing us to the future when after 135 years the Śakas have come again and expelled Vikrama-putra from Ujjayinī, is the Sātavāhana dynasty with Pratisthāna as its capital. The capital is the very one that figures in the Indian legends associated with King Sālivāhana as the founder of the era of 78 A-D. Even the dynasty is indicated in the legends, for the
1. Pandey, op. cit., pp. 15, 21.
2. The Classical Age, p. 23.
3. "The Vikrama Samvat and Śakabda", The Age of Imperial Unity, p. 155.
Page 517
name "Sālivāhana" of the king of Pratishthāna derives, as is well known,1 from "Salavahana" which in its turn comes from "Sātavāhana". And we must observe with Shrava2 that side by side with the Vikrama Era of 57 B.C. there was already in the Śaka year 981 a connection of the name "Sālivāhana" with 78 A.D. No. 134 among the inscriptions of Northern India collected by D. R. Bhandarkar is found equating this Śaka year with the Vikrama year 1116.3 So we have (78+981=) 1059 A.D. as the first instance to hand of the Sālivāhana legend. Here we are only 28 years later than 1031 A.D., the time of Alberum - and it is a fact hardly remembered that the Arab scientist is not devoid of an inkling of the name in question. Fleet" reports; "...it is perhaps worth noting that Alberum (A.D. 1031), in a very short abstract of the story in the Kathāsaritsāgara, has the curious form Samalvāhana, i.e., in the classical language Salavahana." We may add, again from Fleet,5 that the form in the classical language existed not only in legends: it was-also a name in actual use, however rarely, for there is "a copper plate record from the Chamba State, of about the middle of the eleventh century which mentions a King Salavahana as the father of the then reigning King Somavarman" - and there is, too, "a Rohtāsgarh inscription of A.D. 1631, which mentions a Tomara prince Sālivāhana who flourished at Gwalior in or just before that year; see Kielhorn's Northern List, No. 318." The first of these two epigraphs dating to c. 1050 A.D., is contemporary with the earliest reference - 1059 A.D. - we possess in epigraphy to the Śaka-Sālivāhana Era.
Fleet himself denies the reality of any King Sālivāhana as the founder of the Śaka Era. His conclusion6 is: "The name of the supposed king Sālivāhana was introduced in connection with the Śaka era in imitation of the association of the name of the supposed King Vikrama with the era of 58." But he7 grants the
1.Cf. ibid., p. 197.
2.Op. cit., p. 43.
3.Ibid., fn. 4. The inscription runs: Ekādasa-śata-varshāmga tadadhikam shodaśam cha vikramendresharh, Samvat nava-sata-ekāsaiti-Śaka-gata śālivāhana cha nrpadhisa-sake.
4."Sālivāhana and the Śaka Era", Journal of the Royal Asiatic Society, London, 1916, p. 818, fn. continued from p. 617.
5.Ibid., p. 615, fn. 1.
6.Ibid., p. 820.
7.Ibid., pp. 816, 820.
Page 518
source of "Sālivāhana" to be the family name "Sātavāhana" of the kings who had Pratishthana as one of their capitals. His employment of the epithet "supposed" is due to an analogy from the opinion that Kanishka and no such king as Vikramāditya established the era of 57 B.C. All post-Fleet historians have discarded the idea of Kanishka's responsibility for that era. And now that we have shown the implausibility of Kanishka's having established the Śaka-kāla of 78 A.D., we should be entitled not only to look, as we have done, for the real Vikramāditya of 57 B.C. but also for the real Sālivāhana of the Śaka-kāla.
The suggestion from his name and from his capital that he belonged to the Sātavāhana family is increased by an observation of Sircar's:1 "A tradition recorded in the Dvātrimśat-puttalikā represents Sālivāhana...as of mixed Brāhmana and Naga origin. The association of the Sātavāhana kings with the Nagas and also with foreigners like the Śakas is proved by epigraphic evidence." While strengthening the suggestion we have spoken of, the last part of Sircar's words may seem odd when we are confronting Sālivāhana and the Śakas. But occasional intermarriages between the Śakas and the Sātavāhanas did not cancel their more frequent hostilities. Gautamīputra Sātakarni is known to have extirpated the Kshaharata dynasty of the Śakas.2 The fourth ruler - YajnaŚri Sātakarni - after Gautamīputra was no less a Śaka-extirpator. Sircar3 informs us about his silver coins: "this series has the head of the king on the obverse, while the reverse type is a combination of the Ujjain and hill symbols with the rayed sun. It is well known that the king's head and the rayed sun associated with a hill are constant features of the coins of the Śakas of Western India. Yajna was therefore a great king who ousted Scythian [Śaka] rule not only from Aparānta, but probably also from parts of Western India and the Narmadā valley."
With the era of 78 A.D. going a-begging for any identifiable figure in the current historical accounts we may postulate on the authority of various legends that a scion of the Sātavāhana family in the third quarter of the first century after Christ celebrated his destruction of Śaka power by founding an era. The existence of Śaka power and of Sātavāhana kingship in this period is historically
1. "The Sātavāhanas and the Chedis", The Age of Imperial Unity, p. 197.
2. Ibid., p. 201.
3. Ibid., pp. 205-06.
Page 519
justifiable in the light of our discussion of the Periplus dating tp 70-80 A.D. and of Ptolemy's Geography written in 130-40 A.D. The Periplus implies that Ozéné (Ujjayinī), where its king is said to have formerly resided, had lost its glory. This can mean that the Śakas, whose capital it was, were dispossessed of it in 78 A.D. We saw how the Geography, which is known to have been based on information gathered by Marinus of Tyre some time earlier, proves to be historically no less than geographically quite inaccurate for its own period and how its mention of a Śaka ruler of Ozéné and of a Sātavāhana monarch of Baithana (Pratishthāna) can be taken to refer, against the background of the Periplus's date, to the period between 57 B.C. and 78 A.D. The legendary Sālivāhana, connected with the era of 78 A.D. which marks the expiry of Śaka rule, could very well be a successor of that Sātavāhana monarch.
True, we have no epigraph or coin of this personage nor any cognizance of his relationship with the Kushānas of his period. Before Kanishka's extension of Kushāna sovereignty to further north and south, our exterminator of Śaka rule in Ujjayinī in 78 A.D. did not need to fit into their picture. Epigraphs and coins are not always required for historians to believe in the reality of a king. Sircar1 has written: "the existence of a line of Sātavāhana kings holding sway over South Kosala (modern Chhattisgarh and the adjoining region) is suggested by a tradition recorded, among others, by the Chinese pilgrim Hiuen Tsang. According to this tradition, when Nāgārjunā (c. second century A.D.) lived in a monastery built by Aśoka to the south of the capital city of Dakshina-Kosala, the king of the land was So-to-p'o-ho, i.e., Sātavāhana. That Dakshina Kosala in the.eastern M.P. was once under the Sātavāhana kings is possibly supported by the discovery of a copper coin of king ŚivaŚri Āpīlaka in Chhattisgarth." Stray coins are never a sign of its issuer's presence in a place or of his connection with it, for coins like seals and other portable objects can travel far from their proper provenance. So we have really nothing else than Hiuen Tsang's record of a tradition to go by. If its suggestion is worth noting as a probable pointer to a fact, we should make no bones about the "So-to-po-ho" we have proposed for the Śaka-nripa-kālātīta-samvatsara.
1. "The Sātavāhanas and the Chedis", The Age of Imperial Unity, p. 210.
Page 520
15
A Look Back at the Yavanas
Our reconstruction of Aśokan history and chronology took as its starting-point a reconsideration of the term "Yona" or "Yavana". We argued that neither in Aśoka's edicts nor in ancient Indian literature could this term have stood for the Greeks. We said further that, in consonance with the Persian word "Yauna" for all Greeks, it denoted them not before the Bactrian Greek kings and the Indo-Greek kings had come with Persian associations. According to us, the new significance was helped by the disappearance of the non-Greek tribe of Yonas who, originally Indian, had become a frontier people Irānianized and converted to the Mazdean religion.
We may add that the last indubitable sign of their presence in an inscription is connected with the pre-Gupta Sātavāhana king, Gautamīputra Sātakarni, who is said by his son to have been Śaka-Yavana-pahlava-nisūdana1 "the destroyer of the Śakas, the Yavanas and the Pahlavas." What exactly is meant by the destruction of the Śakas we have already seen: "the extirpation of the Ksha-harata dynasty" to which the Śaka chief Nahapāna belonged. Who the Pahlavas were as a tribe in Gautamīputra's time we cannot say. They are not epigraphically heard of again as a tribe; and our historians designate as Pahlavas the Parthian kings who ruled over Eastern Irān from the middle of the 2nd century B.C. and extended their sway to Afghānistān until the middle of the 1st century A.D. The Yavanas do not figure with certainty in any post-Sātavāhana inscription as a distinct people. The Guptas have no mention of them. If we start the Guptas in 315 B.C., the first epigraphic reference after them to a Yavana is in the Besnagar Inscription of the last quarter of the 2nd century B.C., where "Yavana" denotes the Greek Heliodorus who was ambassador from the Indo-Greek king Antialcidas.
Some literary references are there, in late Buddhist books dealing with the time of the Bactrian Greek and Indo-Greek kings as well as in the dynastic sections of the Purānas2 where 9 Yavanas along with 18 Śakas, 13 Murundas, 14 Tushāras and other tribal
1. The Age of Imperial Unity, p. 200.
2. Pargiter, The Purāna Text of the Dynasties of the Kali Age, p. 455.
Page 521
chiefs are said to have ruled, presumbly in the Punjāb and Western India, before the Guptas arrived on the scene. A strange people called Kilakila Yavanas also make their appearance in the same period; the chief of them is declared in the Vishnu Purāna1 to be Vindhyaśakti who is familiar to students of history2 "as a brāhmana belonging to the Vishnuvriddha gotra" and as "the founder of the Vākātaka dynasty", one who cannot by any chance be taken for a Greek. Hence even the slightly earlier "8 Yavanas" need not have been Greeks, either. By the currently held chronology itself they cannot be equated to the Greeks. We are authoritatively told:3 "Prima facie, it appears strange that the Purānas should have described one of these dynasties as Yavana; Greek rule had become extinct long ago. It is however likely that the term may be referring to the tribe Jouan-Jouan of the Chinese writers, who were threatening the Kushāns in c. 320 A.D. and who may have eventually penetrated to the Punjāb during the confusion subsequent to the overthrow of the Kushāns." Our historians here are swimming in surmises. We have no cause to be put out of countenance merely because we are ourselves also uncertain.
In the Brihatsamhitā (XIV, 17-18) Varāhamihira mentions the Yavanas as a tribe at the end of a long list including the Pahlavas and Kambojas. He is commonly dated to the 6th century A.D.; we have placed most of his life in the 1st century B.C., in accordance with his own Śaka Era of 551 B.C. as well as with Indian tradition. But, in any case, we have to note with Cunningham:4 "Dr. Kern, in his preface to the 'Brihatsamhita' of Varāhamihira, p. 32, states that Varāha's chapter on Geography is taken almost intact but changed in form from the 'Parāsara-tantra', and must, therefore, be considered as representing the Geography of Parāsara, or perhaps yet more ancient works, and not as the actual map of India in Varāhamihira's time." So we may stick to our view of the Yavana-tribe.
But, if the Yavanas as a tribe did not stand out in our Gupta times and if the Greeks were not yet known as Yavanas, what
1.The Vishnu Purāna: A System of Hindu Mythology and Tradition, translated by H. H. Wilson (London, 1940), p. 477.
2.Sircar, "The Deccan after the Sātavāhanas", The Age of Imperial Unity, p. 219.
3.A New History..., p. 21.
4.The Ancient Geography of India (1924), p. 6, fn.
Page 522
was India's name for the Greeks in the 4th century B.C. when Chandragupta I, as Alexander's contemporary Sandrocottus, became king of Magadha? We may not expect the Greeks to be mentioned as a people by Samudragupta (Amitrachates of the Greeks); for their rule round the Indus-region in the immediate wake of Alexander had already been ended by Sandrocottus, though as a somewhat fluid cultural entity they continued in Arachosia. But it should be different with the records, if any, left by Chandragupta I.
Such a record would also solve an old problem: "How is it that Alexander's campaign and its aftermath have found no allusion in Indian historical writing of the age?" It would provide the needed allusion.
The Meherauli Pillar Inscription of "Chandra"
A number of gold coins stand in the name of Chandragupta I, but there is only one record which may be ascribed to him: the famous Meherauli Iron Pillar Inscription of King "Chandra", engraved in a Brāhmī script similar to that of the Allāhābād Pillar epigraph of Samudragupta.' Of course, palaeographical resemblance is not always a reliable criterion, yet it can be a good additional argument if other circumstances of direct history point in the same direction. Such circumstances seem to be present in this "one record". To be more precise we should say "one considerable record". For, an inscription of "Mahārājādhirāja Śri Chandra" has been discovered on a Jain image at Vaibhara hill,2 which, because of the phrase "Mahārājādhirāja Śri Chandragupta" in several Gupta epigraphs, could very well be of the first Imperial Gupta, but it is a brief minor record provoking little heat of dispute. Controversy has long raged over the other, and many scholars choose Chandragupta II for its author; but, as Majumdar3 admits, "there is no convincing proof" of their position. We believe, with some scholars, that a better case can be made for Chandragupta I. The more commonly covered grounds for this case, as well as the answer to the usual objections, we shall set forth later. At the moment we shall touch on a point which
1.Mookerji, The Gupta Empire, p. 69.
2.The Archaeological Survey of India, Annual Report, 1926, p. 125.
3."The Expansion and Consolidation of the Empire", The Classical Age, p. 20.
Page 523
becomes uniquely significant when we identify Chandragupta I with Sandrocottus.
Chandragupta II is not independently known to have had anything to do with military action round about the Indus, whereas Sandrocottus is twice connected by the Greek and Latin historians with such an action, and Chandra of the Meherauli Pillar is credited, in Mookerji's words,1 with "a running fight across the seven mouths of the Sindhu". The question then would be not of choice between the two Chandraguptas but of choice between the two military actions of Sandrocottus in the Indus-context: "Which of these actions is relevant to Chandra's fight?"
The first is the war of liberation Sandrocottus waged against the prefects of Alexander soon after Alexander's death: the result was the conquest of the Greek satrapies in India. The second is the confrontation in c. 305 between Sandrocottus and one of Alexander's successors, Seleucus Nicator, when he, after subjugating the Bactrians, "passed over into India". In the references to the former exploit there is no direct mention of the Indus. In those to the latter we have it not only by implication in Justin's "passed over into India", which suggests a crossing of the river. We have it also explicitly in the words of Appian (early 2nd century A.D.)2 on Seleucus: "He crossed the Indus and waged war on Sandrocottus, king of the Indians who dwelt about it..." That Sandrocottus in turn should cross the Indus in driving out the invader would be quite natural. So, in the mere matter of Indus-crossing, Chandra's "fight" joins up better with Sandrocottus's opposition to Seleucus than with his campaign against Alexander's prefects.
There is, in addition, the fact that as a result of opposing Seleucus Sandrocottus obtained from him the Greek satrapies lying to the west of the Indus. These satrapies would come into the picture logically in relation to a passing over by Sandrocottus outside India. The campaign against the prefects of Alexander resulted only in Sandrocottus's obtaining the Greek satrapies lying on the Indian side of the river. There was no going across. Hence the Indus-adventure of Chandra links up again better with the war between Sandrocottus and Seleucus.
However, we have to face the clear expression: "across the
1.Op. cit., p. 69.
2.Syr., 55: McCrindle, Ancient India as Described by Megasthenes and Arrian, pp. 9-10.
Page 524
seven mouths of the Sindhu." Can we conceive Seleucus to have attacked India in the locality of the Indus's seven mouths rather than in the north-west where earlier Alexander had crossed the river? No scholar so far has thought of anything except the northwest. But apropos of Sandrocottus's war with the prefects we may remember that the treaty of Triparadisus in 321 B.C. among Alexander's generals refers to the removal of Peithon to the provinces between the Indus and the Paropanisus from the province of which he was the governor: we know that he was the governor of Sind. Do we not have here a hint that Sind, including the seven mouths of the Indus, was the province first lost by the prefects?
Peithon certainly brings in a coincidence which appears meaningful. But what the terms of the Triparadisus treaty throw into relief is, as Mookerji1 has said, that no part of India to the east of the Indus was included as a part of Alexander's empire. We are still not carried beyond the river's mouth westward. And there is no sign that Peithon was forcibly driven out: he is as if removed by Greek orders in view of events taking place on the Indus's eastern side. Thus the entry of Sind into the picture is meaningful more in appearance than in reality: there is no true connection with Chandra's battle.
On the other hand, while we are not told that Seleucus invaded India at the Indus's mouths we are not told either that he did so in the north-west. We simply do not know from our sources where exactly he crossed the Indus. And, if we may judge from the territories he ceded to Sandrocottus at the end of the war, we get a vision of the whole of the Greek possessions to the Indus's west as the area where the war may have been waged. For, as J. C. Powell-Price2 informs us: "The Indus valley and Afghānistān and Balūchistān were given up..." Mookerji3 also tells us that Sandrocottus was ceded by Seleucus "the Satrapies of Arachosia (Kandahār) and the Paropanisadae (Kābul), together with portions of Aria (Herāt) and Gedrosia (Balūchistān)". In consequence, to quote Mookerji4 once more, Sandrocottus "extended his empire beyond the frontiers of India up to the borders of Persia". In other
1.Chandragupta Maurya and His Times, p. 50.
2.A History of India (Thomas Nelson, London, 1955), p. 38.
3.Op. cit., pp. 59-60.
4. The Gupta Empire, p. 69.
Page 525
words, the war with Seleucus won for him the territories lying westward along the entire length of the Indus, from Kābul down to the mouths. It is immediately to the west of these mouths that the "portions of... Gedrosia (Balūchistān)", ceded to Sandrocottus, lie. Seleucus may have carried out his penetration there just as well as further up the river.
The Indus-delta was no terra incognita to Seleucus. Alexander had already explored it and, before sending Nearchus on his voyage from there to the Tigris in the Persian Gulf, he had himself touched the sea. We read in Arrian's Anabasis (Vl.xix):1 "Then having gone beyond the mouths of the river Indus, he sailed out into the open sea..." Seleucus may have found all the rest of the Indus well guarded on the other side and so proceeded southward, watching for a weak spot. He may have remembered what had happened when Alexander had begun "to sail down the right branch of the river" towards the mouth of it. Arrian (IV.xviii) records: "The Indians of that region had fled..." Judging the delta to be the required weak spot, Seleucus may have attempted his invasion over it, partly succeeded but finally got driven off by the resolution and skill and resources of a military expert who was more than his match.
We may reconstruct the sequel as follows. Seleucus was pushed back northward along the way he had come. Then he made overtures of peace. We hear of their results from both Justin and Appian: a friendly exchange took place between the erstwhile enemies. Perhaps the Greek was called over to negotiate where Chandra, to celebrate the various victories rounded off by this last one, set up - as Mookerji2 says - "his pillar in honour of Lord Vishnu on the hill known as Vishnupada" - not far from the Kurukshetra and the Beās.3
With the road open to us to opt for the Indus-delta, we may scrutinise the remainder of the inscription for helpful pointers in our direction. If the epigraph records the triumphs of Sandrocottus, any reference to the conflict with Alexander's prefects should come first in the series of achievements; for, as we have seen, he became king originally as a result of it. But actually the Indus-battle stands last in the series. The very first is, in Mookerji's
1.McCrindle's translation.
2.Op. cit., p. 69.
3.Sircar, Select Inscriptions, p. 277. n.5.
Page 526
phrases:1 "Conquest of the Vanga countries (Vangeshu) by his battling alone against a confederacy of enemies united against him (Śatrun-sametyagatān)." Vanga is central Bengal: its conquest takes us to eastern India. According to our reading of Justin's passage on Sandrocottus's military deeds, there would have been no action there at the start of his career. If the Vangā-war preceded the one on the Indus, the latter could not be the campaign against the Macedonian prefects.
Moreover, to date the inscription to so early a period in the life of Sandrocottus must raise the query: "What about his accession to the throne of inland India?" Whether the conquest of Vanga preceded or succeeded the action in the Indus-region against those prefects, it could not have come before he held the sovereignty of inland India. Where in the inscription is any allusion to his achievement of it? Prominent in inland India at the time of Alexander's march up to the Hyphasis (Beās) was a king named Xandrames. Surely, Sandrocottus took over the lands belonging to Xandrames not long after - or else, if we are mistaken, a little before - the struggle with the prefects. How is it that an inscription made at so early a point in Sandrocottus's life has not the slightest hint of this take-over?
Both by omission and commission the Meherauli Pillar puts itself in a time to which the early events of the life of Sandrocottus are somewhat remote. C. 305 and some years preceding it are here the theme. Most probably the early events figure in another inscription which remains to be found. And, in the theme developed in the present epigraph, the focus of attention for us is the military encounter we have traced between Sandrocottus and Seleucus. But the moment we look at the terms in which this encounter is described we have already to our hand the name we are seeking for the Greeks in the 4th century B.C. For, the full phrase concerned goes: "Conquest of the Vāhlīkas in a running fight across the seven mouths of the Sindhu."
What were the Greeks called in the 4th Century B. C. ?
Vāhlīkas or Bāhlīkas - that is what we may take Sandrocottus alias Chandragupta I to be calling the Greeks with whom he came into contact in c. 305 B.C. It must have been the current designation
1. Op. cit., p. 69.
Page 527
ever since Alexander invaded India. But we still have to ask whether there is a more likely identification of them for the Meherauli Pillar.
S. N. Majumdar1 writes in a note to Ptolemy's Geography: "Reference in Sanskrit Literature clearly indicates the Bāhlīkas, or rather the two tribes of them, as dwelling in the Punjāb. One of these tribes was closely connected with the Mādras, for Śalya, a prince of Mādra (with its capital at modern Sialkot), is also called lord of the Vāhlīkas (Mahābhārata, Ādi CXIII, 4425-40; LXVII, 2642) and his sister Mādri is called Vāhlīki also (Adi, CXXV, 4886). The other people of this name appear to have been connected with the Daradas (Dards of Dardistān) (Bhisma, CXVII, 5484, etc.). Thus one Bāhlīka was situated in the plains of the Punjāb, alongside the Mādradeśa, i.e. between the rivers Chenāb and Sutlej; and the other among the lower slopes of the Himalayas, very possibly between the Chenāb and Beās."
The people living round Balkh or Bactria are also known as the Vāhlīkas. Then there is the dynasty of three Bāhlīka rulers placed by the Purānas in the region of Mahisamati on the Narmada.
The last can be rejected immediately since these rulers are to the east of the Indus and do not involve its crossing. The people living round Balkh are to be rejected too, because, as Allan2 correctly points out, the inscription cannot mean that "Chandra's arms penetrated to Balkh, the route to which would not be across the mouths of the Indus".
An attempt is sometimes made to get round this difficulty by saying3 that Chandra who mentions other things besides these mouths is merely referring to the four limits of the territory traversed by him in the course of his digvijaya: the eastern limit is Vanga, the southern the southern sea, the western the seven mouths of the Indus and the northern the Vāhlīkas. In support of this it is sought to be proved4 from numerous inscriptions and literary works that Chandra's claim of having conquered the chakravartikshetra is the conventional claim of a Digvijain and would call for just these four limits. But all such arguments
1.Ancient India as Described by Ptolemy, p. 395.
2.Catalogue of Coins, the Gupta Dynasty, etc., Introduction, p. xxxvi.
3.Sircar, Select inscriptions, p. 276, n. 4.
4.The Journal of the Royal Asiatic Society of Bengal, Letters, Vol. V, pp. 407-15.
Page 528
completely overlook the syntax of the sentence concerned. Even Majumdar,1 who is for Balkh, is forced to translate the sentence thus: "he defeated a conspiracy of hostile chiefs in Vanga and, having crossed in warfare the seven mouths of the Sindhu, conquered the Vāhlīkas."
First of all, a particular incident or a specific complex of circumstances is mentioned in regard to Vanga: there is no generality as one may expect if a mere eastern limit were intended. Secondly, the crossing of the Sindhu's seven mouths is organically related to the conquering of the Vāhlīkas, as if the latter could never have been defeated without the unique circumstances of a running fight across those mouths. Even if a limit is indicated, it is a single one, the western - and, with the southern sea taken as another limit, we shall have the northern limit missing. We may also observe that Majumdar does not bring in the southern sea as part of the detailed statement. The southern sea's mention has the air of a generality and it differs palpably from the manner of the earlier declaration. Mookerji2 recounts it as: "Spread of his fame, as a conqueror, up to the southern seas." It is indeed difficult to think in terms of a Digvijain's four limits. The interpretation "Balkh" is quite gratuitous.
Even the Bāhlīkas of north-western India, the two people in the upper Indus-region, cannot serve: they too do not need to be reached across the mouths of the river. The eneiny is undoubtedly to the west of the Indus and must be one with whom the fight had to begin at the mouths and get essentially decided there, even if a pursuit had to be maintained subsequently along the whole west bank and officially terminated when the celebrating Pillar was mounted on the Vishnupada hill.
An unconventional meaning has to be found for the term "Vāhlīkas". Of course, the possibility is there that, as S. M. Majumdar3 conjectures, the indigenous Vāhlīkas themselves migrated southwards near to the mouths of the Indus. But nowhere in Sanskrit literature do we have the slightest sign of such a migration. Many tribes are put in various divisions of India or various parts of the same division. Thus4 not only are the Kiratas placed in the east as well as in the north-west: the Trigartas too are
1."The Expansion and Consolidation of the Empire," The Classical Age, p. 20.
2.Op. cit., p. 69. 3. Op. cit., p. 395.
4. Barua, Aśoka and His Inscriptions, Part I. pp. 101-2.
Page 529
placed by the Mahābhārata in the upper no less than in the lower part of the north. The Śūdras and the Ramathas (known also as the Ramanas) whom the Epic locates in the southern part are located by the Markandeya Purāna in the northern.1 Many tribes are widely diffused. The Vāhlīkas are not among them. A few tribes, some located in the north and some in the south, are at times listed in a mixture of northern and southern tribes. Thus the Parades, whom the Markandeya Purāna places in the northwestern division along with the Aparantas, Gandhāras, Yavanas, and Kambojas, occur in the Vāyu Purāna and in the Harivarmśa along with the Śakas, Yavanas, Kambojas, Pahlavas, Khaśas, Mahishikas, Cholas and Keralas, some of whom belonged to the Uttarāpatha and some to the Deccan.2 Similarly, in the Matsya Purāna, the Pulindas are associated with the Āndhras, Śakas, Chulikas and Yavanas.3 The Vāhlīkas do not come in any such hotchpotch even. They are always with the northerners. No less than the Mahābhārata and the Purānas, the Brihatsamhitā of Varāhamihira mentions them with northern peoples.
Raychaudhuri4 claims to find a people that can be called Vāhlīka just where it is wanted: he says, "The Vāhlīkas beyond 'the seven mouths of the Indus' are apparently the Baktrioi occupying the country near Arachosia in the time of the geographer Ptolemy {Indian Antiquary, 1884, p. 408)." There is a mistake here: Ptolemy does not state, as Raychaudhuri believes, that the Bactrioi are between the mouth of the Indus and Arachosia. What comes between is the country of Gedrosia (modern Balūchistān) and here in the maritime parts immediately adjoining the Indus the Arbitai are put by Ptolemy;5 in the north of this country, not far from the Indus, are the Rhamnai (the Indian Ramanas). Then we have Arachosia itself to the north-west of Gedrosia. The tribes of south Gedrosia who would be nearest to - though distant enough from - the lower Indus are the Rhoploutai and the Eoritai.6 And only between Arachosia and Drangiana (Seistan) to the west of Arachosia are the Bactrioi.7 It is impossible to identify
4.The Political History of Ancient India, p. 535, fn. 2, end.
5.Op. cit., p. 320.
6.Ibid., p. 316.
7.Ibid., p. 313.
Page 530
them with the Vāhlīkas of the Meherauli Pillar. Besides, they are obviously a small and insignificant tribe and are absolutely unknown to history and it is hardly conceivable that they should have constituted a serious menace to Chandra's empire, comparable to the conspiracy of hostile chiefs in Vanga and deserving to be mentioned in a record claiming all-India fame for its hero.
The sole alternative left among the suggestions made by scholars is the one of Allan. He1 opines that in the time of Chandra "the name Vāhlīka had acquired a more general significance and was used like Pahlava, Yavana, etc., of a body of foreign invaders of India." Allan, of course, was speaking of the Vāhlīkas against a background of the 4th century A.D. So his comparison with the Pahlavas and Yavanas cannot be pressed into our service. In our period - the 4th century B.C. - we have no proof of "a more general significance" for the names "Pahlava" and "Yavana". But his idea of "Vāhlīka" as denoting "a body of foreign invaders" rather than any of the known Vāhlīkas inappositely offered by other scholars is none the less brilliant and the only logical course. It is supported also by the fact that the very word "Vāhlīka" or "Bāhlīka" etymologically suggests "outsiders".2 And it can hold for the 4th century B.C. just as well as for the 4th century A.D. It can hold even better, for in the 4th century A.D. we have no foreign invaders of any kind near the mouths of the Indus. Even if we put Chandragupta I and Chandragupta II in that century we come across no sign of such invaders: it is only towards the end of the reign of Kumāragupta I and the beginning of that of Skandagupta that we can think of invading foreigners, but then too not in the Indus-delta. In contrast, the 4th century B.C. shows us two foreign invasions, Alexander's and Seleucus's. One of the invaders left India for Gedrosia from the mouths of the Indus, the other is likely to have tried entering the country from Gedrosia over the same mouths. And both were of the same race. Chandra's Vāhlīkas can be none else than the Greeks.
Even apart from the suggestion of "outsiders", how apt the name "Vāhlīka" is for the Greek invaders of the 4th century B.C. will be apparent the moment we ask: "By what route did they swoop down on India?" Mookerji3 writes of Alexander: "At the
1.Op. cit., p. xxxvi.
2.The Vedic Age, p. 258.
3."Foreign Invasions", The Age of Imperial Unity, pp. 43-4.
Page 531
beginning of 327 B.C. he had completed the conquest of Eastern Irān beyond the Hindu Kush by overrunning the regions now known as Bokhara as far as the Syr Daria (Jaxartes). In May 327 B.C. he advanced towards India. He returned through the Kushān Pass..." Clearly the descent through the Kushān Pass which lies in the Hindu Kush was from Bactria. So the general name "Vāhlīka" or "Bāhlīka" for a body of foreign invaders suited his men down to the ground. In the Anabasis (IV.xxii)1 Arrian has the specific phrase that Alexander "advanced from Bactra towards India".
In connection with Seleucus we may hark back to Justin's words immediately preceding those about Seleucus's passing over into India. Justin writes: "He first took Babylon, and then with his forces augmented by victory subjected the Bactrians." Here also the Greeks moved from Vāhlīka into India, after conquering that province and establishing their power in it.
Thus the combination of India-ward Greeks and Vāhlīka was strengthened. Need we be surprised if in an inscription Sandrocottus should call Seleucus and his invading army the Vāhlīkas?
For the sake of interest we may observe in addition that Antio-chus III, one of the successors of Seleucus, gave up in 206 B.C. a two-year siege of the capital of Bactria held by the Greek Euthydemus, soon after "crossed the Hindu Kush, marched down the Valley of the Kābul river, and reached the country ruled by Sophagasenus (probably Sanskrit Subhagasena), king of the Indians..."2 Once more there is the India-ward movement from Vāhlīka.
Vāhlīka is all along bound up with the Greek approach to India. Even the Greeks who were definitely called Yavanas were those who were Bactrians - themselves Vāhlīkas in the strict sense. Agrawala,3 quoting the Brāhmanda Purāna (16.18), makes the statement: "Originally the Yavanas or Greeks who had settled in Bactria were designated in Purānic tradition as Bāhlīka Yavanas." Bāhlīka or Vāhlīka, as the first name for the Greeks before the term "Yavana" caught on, seems a certainty from every standpoint, and the Meherauli Pillar of King Chandra carries the epigraphic proof
1.McCrindle, The Invasion of India by Alexander the Great, p. 1.
2.The Age of Imperial Unity, pp. 104-5.
3."Bhuvana Kosa Jariapadas of Bharatvarsha", Purāna (Varanasi), Vol, V. No. 1, p. 167.
Page 532
from the hands of Sandrocottus if we can demonstrate definitively its author as being Chandragupta I.
Chandragupta I as "Chandra" of the Meherauli Pillar
It is not only Chandragupta II who has the abbreviation "Chandra" on some coins. It is also Chandragupta I who has this brief legend on the single type of coins known to be his - the special ones commemorating his marriage with the Lichchhavi princess Kumāradevī." That they were issued by himself and not by anybody on his behalf seems now certain: Majumdar2 tells us, "The view that the coins were issued by Samudragupta is no longer held by any scholar."
However, in no inscription of Chandragupta II is the short form "Chandra" used, so that the abbreviation in the Meherauli Pillar epigraph where space for the longer form is available creates some surprise. As we have no authenticated inscription of Chandragupta I to compare this epigraph with, we can hardly be surprised if "Chandra" is employed in his case. On the contrary it is what we may consider possible: although by his own example he must have set for his successors the rule of using the family-name "Gupta", the necessity of sticking to that name as a dynastic title could not have been felt so much by him as by his successors after his designation "Chandragupta" had carried a renown wanting to his ancestors and giving him the right to call himself "Mahārājadhirāja" ("Supreme King of great kings") for the first time in the family in contrast to the name "Mahārāja" of his father and grandfather. His father Ghatotkacha bore no "Gupta"-ending though Ghatotkacha's own father was Gupta. In spite of establishing a general precedent for those who came after him, Chandragupta himself had no particular reason to attach importance at all times to the terminal component of his own name.
We may further note that he is the first Indian king to assume the title of "Mahārājadhirāja": hence not only in his time but in all Indian history up to him he was the sole supreme sovereign. This circumstance is perhaps echoed in what Mookerji3 lists as one of the distinctive declarations of the Meherauli Pillar about
1.Mookerji, The Gupta Empire, p. 15.
3.Op. cit., p. 69.
Page 533
"Chandra": "Achievement of sole supreme sovereignty in the world (aikādhirājyam) by the prowess of his arms."
Even the term used for "prowess" argues for Chandragupta I rather than his namesake grandson. For, it is not the term characteristic of Chandragupta II, namely, "Vikram", but the term "Virya". This term seems to get connected with the grandfather of Chandragupta II through his father Samudragupta's use of it in the Allāhābād Pillar. Again, the religious expression - "having in faith fixed his mind on the god Vishnu" - is similar in spirit yet significantly different in form from the one repeated stereotypically by Chandragupta II: Paramabhāgavata. The successors of Chandragupta II followed the usage started by him, but Samudragupta, although a Vaishnava who adopted the emblem of garudadhvaja, appears to have differed doctrinally from his paramabhāgavata successors.1 So the different form of the Vaishnavite expression on the Meherauli Pillar strikes us - through Samudragupta - as linking up with his father Chandragupta I.
As for a career of conquest like "Chandra"'s, we can say of Chandragupta I, as of his grandson, that independent sources lead us to infer such a career for him. The assumption of the unique title Mahārājādhirāja should be sufficient proof of territorial expansion. Mookerji2 argues: "The title of 'King of Kings' must have been acquired by his conquests by which he was able to rule over an extensive territory." But it has often been held that the range of his expansion is summed up in the Purānic statement:3 "(Kings) born of the Gupta race will enjoy all these territories, namely, along the Ganges, Prayāga, Saketa and the Magadhas." A dissenting note, however, is struck by Majumdar:4 "Apart from the uncertainty of the reading and interpretation of the passage, we cannot even be certain that it refers to the period of Chandragupta I." Majumdar5 remarks about the territory indicated: "Some have even taken it as the Gupta dominions in the period of decline after Skanda-Gupta." Majumdar's own verdict,6 based on the word Gupta-vamśa-jāh is in the passage: "the Guptas, in plural number,
1.The Classical Age, p. 414.
2.The Gupta Empire, p. 12.
3.Pargiter, op. cit., p. 73.
4."The Rise of the Guptas", The Classical Age, p. 4.
5."The Rise of the Guptas", Appendix, A New History..., p. 134.
6.Ibid., pp. 134-35.
Page 534
are described as the rulers. It is prima facie unreasonable, therefore, to think that the author of the passage had specifically the kingdom of Chandragupta I in view." The plural should indicate that these territories formed the central block over which the Guptas exercised direct and immediate sovereignty, the area which roughly remained proper to them in the great days of their dynasty for more than 200 years after Chandragupta I had acquired it. Expansion beyond this area is thus not ruled out for any Gupta, including Chandragupta I himself.
In support of not restricting Chandragupta's domain to the Purānic limits, one may quote Samudragupta's Allāhābād Pillar Inscription. It is quite possible to take Samudragupta's sway over the 5 frontier kingdoms to the east and over the 9 tribal republican states to the west and north-west as inherited from his father's conquests: there is no direct indications that it was established by him as there is that he fought against certain kings of Northern and Southern India. Nor is Chandragupta's domain necessarily restricted to the Purānic limits even if the Allāhābād inscription shows a series of conquests by Samudragupta roughly beyond the region marked by Prayāga (Allāhābād) and Sāketa (Oudh) on the west and the Magadhas on the east. Mookerji1 draws our attention to a problem connected with the eastern conquests. Three countries are mentioned: Samatata, Davāka, Kāmarūpa. The last two are covered by Assam. Samatata, says Mookerji, denoted "certain remote parts of Bengal" and "was probably to the east of Tamra-lipti and bordered on the sea, as stated by Hiuen Tsang." If Samudragupta's campaign in Bengal was concerned with the outlying districts to the east of Tāmralipti, who could have conquered Bengal proper, the central parts, down to "the southern seas" (the Bay of Bengal) mentioned on the Meherauli Pillar? We know for certain that Bengal proper was within the empire of his son Chandragupta II, because, under his grandson Kumāragupta I, its northern portion was called Pundravardhana.2 Did Chandragupta II annex Bengal proper or did Chandragupta I do so? Mookerji concludes that in the absence of any definite evidence on the subject the conqueror could have been either the one or the other. And he grants that if Samudragupta did denote Bengal's outlying portions the central ones could be inferred to have been annexed
2.A New History..., p. 174.
Page 535
by a predecessor, and Samudragupta could be considered to have completed the work of his father, and Chandragupta I might be identified with Chandra who proceeded to form the central block of Gupta dominion outlined in the Purānas and then secured his rear by carrying out the first achievement described in the Meherauli Pillar: "Conquest of the Vanga countries (Vangeshu) by his battling alone against a confederacy of enemies united against him (śatrun-sametyāgatan)."1 But, according to Mookerji, the difficulty of the identification is, so far as the conquests are involved, that "by no stretch of imagination can Chandra Gupta I figure as a conqueror of territories in the Punjāb and North-west which Samudra Gupta was the first of the Guptas to deal with".
Is the difficulty genuine? The imagination can be prevented from stretching only if an irrefutable reason is found from the Allāhābād Pillar to keep Chandragupta I within the geographical limits mentioned in the Purānas. If Mookerji himself has considered it possible that Chandragupta did break these limits in the direction of the Vanga countries, the sheer ban is removed and it should be equally possible to imagine him breaking beyond them in the direction of the Punjāb and the North-west. The only proviso would be that we show why Samudragupta would have to deal again with those regions.
We may ask a counter-question: "If Chandra is identified with Chandragupta II, why should the latter have to deal again with the Punjāb and the North-west after Samudragupta had dealt with them?" Mookerji2 believes that, though Samudragupta had exercised sway over the northern Śakas, Chandragupta II had also to confront them and that he carried out campaigns in the Indus-region and consolidated his father's conquests. Logically, we can reverse the role of Samudragupta and say that instead of his conquests being consolidated by his son he consolidated the conquests of his own father Chandragupta I. The rationality and the probability of the two statements are exactly the same, provided the first Chandragupta is as likely as the second to have proceeded to the Punjāb and the North-west. The sole thing certain about Chandragupta II is that he conquered the Western Śakas, the Mahākshatrapas who ruled in Mālwa and Kāthiāwār. Although we may be sure that he had more conquests to his credit, "we have,"
1.Mookerji, op. cit., p. 69.
2.Ibid., p. 70.
Page 536
writes Majumdar,1 "no definite information regarding the nature and result of these other campaigns." Fact to fact, there is nothing to choose between the two Chandraguptas in the matter of the Punjāb and the North-west. Consequently it is just as reasonable to make Samudragupta the consolidator of his father's achievements in those regions as to make Chandragupta II consolidate those of his father.
Even with regard to Bengal the same reverse process is possible. Majumdar2 does not quite see a difference between the Vanga countries and the regions annexed by Samudragupta. He writes: "Vanga denotes Eastern Bengal, nearly the same country as Samatata which is included in the tributary states of Samudragupta." And by imagining either a revolt there or else a desire by the new king to fix more firmly his sovereignty he explains why, if Chandra was Chandragupta II, the latter had once more to fight Vanga. Can we not with equal logic say that, if Chandra was Chandragupta I, Samudragupta had to reconquer Vanga because of the same reasons - most probably a revolt?
The plausibility of a revolt requiring Samudragupta to go over a considerable portion if not the whole of the terrain of his father's dominion in a new digvijaya is suggested by certain attitudes and circumstances discoverable in Samudragupta's inscription.
There is the phrase - already quoted by us once - in which he is called the cause of "the destruction of evil". This attitude to him renders all his "uprooting" and "extermination" of various kings punitive, as if they had done wrong, as if he had fought them for treason to and rebellion against rightful authority.
Then there is the seventh verse, a part of which is lost. "Mention is first made," comments Majumdar,3 "of his complete victory over two rulers named Achyuta and Nāgasena, and a third, belonging to the family of the Kotas. This is followed by a statement that he took his pleasure at the city Pushpa. The lacunas, caused by the peeling off of the surface of the stone in this part of the record, leave us in the dark about the connection between these events, but the construction of the sentence makes it very likely that Samudragupta's victory over the kings, at least the third one, is closely connected with his visit to the city called
1."The Foundation and Consolidation of the Empire", The Classical Age, p. 20.
3."The Foundation of the Gupta Empire", A New History..., p. 139.
Page 537
Pushpa. In other words, it was probably the victory over one or all of them, that enabled him to take possession of the city."
As "Pushpapura was... a well-known name of Pātaliputra"1 and as the kingdom of Chandragupta I, after his marriage with the Lichchhavi princess Kumāradevī, "almost certainly included the greater part of Bihar and also very probably a portion of U.P. and Bengal",2 Pātaliputra which was the chief city of Bihar must have been involved in the campaign of the seventh verse. That is to say, Samudragupta's capital was at stake. Of course, as Majumdar3 conjectures, "it is just possible that the poet refers to the triumphal entry of Samudra-gupta into his own capital after his brilliant military campaign"; but, if we may justly hold that Samudragupta could take his pleasure at the Gupta's own capital city Pātaliputra only by recovering it from enemies on whom he inflicted an absolute defeat, this must mean a revolt at the very heart of his empire. And in concord with it there were most probably uprisings in other provinces than Magadha. Samudragupta can thus be seen as winning back Pātaliputra and then going forth to crush the various rebellions whether on the east or on the west and carrying his arms as far as he could, even beyond the limits of his father's previous victories.
Further, there is the fact known from the Allāhābād Pillar that Chandragupta chose his son as his successor in the midst of a public assembly, exclaiming with exultation, "Thou art worthy, rule this whole world." This appointment of a son to the throne while the father was yet alive was an extraordinary event and must bear a special significance. Some have held that we have a parallel in the case of Samudragupta's own son. "But this view," says Majumdar,4 "rests on the doubtful interpretation of an expression which cannot be regarded as certain." The Allāhābād Pillar leaves us in no uncertainty. What is more, it implies a greater gesture of recognition than the mere appointment of Samudragupta. Majumdar5 has well remarked: "The royal declaration is usually taken to mean that Chandra-gupta I publicly announced Samudragupta as the heir apparent to the throne. The words put in his
1. Ibid.,
2.Ibid., p. 132.
3.Ibid., p. 140.
4."The Expansion and Consolidation of the Empire", The Classical Age, p. 18.
5."The Foundation of the Gupta Empire", Ibid., p. 7.
Page 538
mouth, however, taken literally, mean that Chandra-gupta I formally abdicated in favour of his son." And Majumdar adds in a footnote: "Dr. Chhabra has come to the same conclusion after a critical discussion of the whole passage... (Indian Culture, XIV. 141)."
The extremely unusual character of the event must signify Chandragupta's deep gratefulness for distinguished services rendered by his son, and the services may very reasonably have been a rush to rescue the father from whatever loss had been suffered, less demanding at the start, precisely the campaign in which Samudragupta "uprooted" Achyuta, Nagasena and the scion of the Kota family and achieved possession and pleasure of Pātaliputra. Rewarded for this campaign with the crown and invested thereby with full power, he may have set out on a vast digvijaya to restore and expand his father's kingdom.
In this way our suggestion as to why his military expeditions were undertaken in spite of his father's wide conquests receive indirect support and our identification of Chandra with Chandragupta I is made easier.
If it is asked whether we have any clue to the cause for a revolt, we may cite Justin (XV.iv) on Sandrocottus: "India... after Alexander's death, as if the yoke of servitude had been shaken off from its neck, had put his prefects to death. Sandrocottus was the leader who achieved their freedom, but after his victory he forfeited by his tyranny all title to the name of liberator, for he oppressed with servitude the very people whom he had emancipated from foreign thraldom." The same anti-thraldom impulse which worked in Sandrocottus against the Macedonian prefects seems to have driven his dependents to raise the standard of rebellion against him towards the closing part of his iron-rod rule.
Two Objections to our Case Answered
Our case is pretty strong now, but it has two objections to meet. Mookerji takes up the word suchirān occurring on the Meherauli Pillar and meaning "a long reign". He1 says that it cannot hold true for Chandragupta I. We have already shown that if, as is very reasonable, we may assume that Chandragupta's founding the Gupta Era is linked with his marriage to the Lichchhavi princess
1. Op. cit., p. 10.
Page 539
Kumāradevī who may have brought him Pātaliputra as dowry, the gift of his throne to Samudragupta for unusual prowess and ability must involve two things: (1) Samudragupta must have attained the age of at least twenty-five to thirty and (2) Chandragupta must have reigned twenty-five to thirty years at the minimum till then. As Chandragupta II ascended the throne in the 56th year of the Gupta Era, we have quite a long period to divide between Chandragupta I and Samudragupta for their reigns. We may well allot to the former a reign as long as his grandson's: about 33 years. Mookerji's objection carries no weight.
A more serious one could be based on Greek sources. The Meherauli Pillar mentions seven mouths of the Indus, whereas Arrian (Indica, II), recounting the history of Alexander numbers only two. If Chandragupta I was Sandrocottus, why this discrepancy? We must try to understand what Arrian intends by "mouths".
His words in the Indica are: "these mouths are not close to each 'other, like the five mouths of the Ister (Danube), but diverge like those of the Nile, by which the Egyptian delta is formed. The Indus in like manner makes an Indian delta, which is not inferior in area to the Egyptian, and is called in the Indian tongue Pattala."1 Some elucidation of his reference comes from his Anabasis of Alexander (Ch. XVIII), where he tells us of the Indus dividing near the city Patala into "large rivers both of which retain the name of Indus as far as the sea."2 By "mouths" Arrian means not precisely the places of immediate entry into the sea, for he speaks of the divergence of the mouths, thus implying the high-up common point of parting at the city Patala, and he speaks of the large delta to which this city gives the name Pattala and which is formed by the mouths, thus implying the whole triangle made by that point and the points of entry into the ocean. Arrian's "mouths" are simply the two main oceanward branches of the lower course of the river from the city Patala sufficiently distant from the sea and they do not preclude offshoots from the branches within the delta which also empty into the sea.
We become sure of the non-preclusion when we read in the geographer Ptolemy exactly the same information about the two branches. Referring to the delta made by them he3 writes in almost
1.McCrindle, Ancient India as Described by Megasthenes and Arrian, p. 183.
2.Chinnock's translation, p. 321.
3.McCrindle, Ancient India as Described by Ptolemy, p. 136.
Page 540
a paraphrase of Arrian: "The insular portion formed by the bifurcation of the river towards its mouth is Patalene." This information, however, does not prevent Ptolemy1 from giving the mouths of the river as seven. Why should its occurrence in Arrian be taken to limit the mouths to just two in Alexander's day? Even the author2 of the Periplus (c. 70-80 A.D), nearly a 100 years before Ptolemy, gives the same number as that later writer. It seems the most natural thing that, in c. 326-300 B.C. when Arrian's chief sources here wrote their books, the number should not be different.
The reason for Arrian's concentration on merely the bifurcation at the city Patala is not difficult to find. The records about Alexander used by him in the Anabasis were concerned only to describe Alexander's exploration of the seaward route via both the main branches of the river from this city and did not need to mention the smaller offshoots which Alexander neglected and left unexplored. In the Indica, at the end of chapter XVII which closes the section based principally on Megasthenes, Arrian declares: "Since my design in drawing up the present narrative was not to describe the manners and customs of the Indians, but to relate how Alexander conveyed his army from India to Persia, let this be taken as a mere episode." What led to his omission of the seven mouths in the one book can be legitimately thought to have worked also in the other to the same effect. The seven mouths are a detail beyond his purpose and outside his course of history.
Nothing vitiates our thesis and its implication that in the 4th century B.C. The Greeks were designated Vāhlīkas, not Yavanas.
And, in King Chandra's vivid record of his victory over the Vāhlīkas on the westward side of the Indus's seven mouths, we have Indian history speaking through the founder of the Imperial Guptas and taking notice of the last of the events connected with the Greek invasion of our country in the final quarter of the 4th century B.C.
Last Words
The Meherauli Pillar may be said also to mark in conclusion the victory not only of independent logic working on the very data
1.Ibid., p. 83.
Page 541
offered by modern historians but also of the traditional Indian chronology, which is found both in the Purānas and in Megasthenes, over the modern time-scheme in the matter of identifying Sandrocottus. As the historical point from which everything is to be traced backward and forward, Chandragupta I instead of Chandragupta Maurya stands solidly where Megasthenes and other Classical writers have placed Sandrocottus.
Page 542
Home
Disciples
Amal Kiran
Books
Share your feedback. Help us improve. Or ask a question.