The Colas c16.s1: I.—Chronology and Order of Succession
The Cōḷas c16.s1
- Language
- English
I.—Chronology and Order of Succession
c16.s1.1
The relatively short interval of about thirty years from the death of Parāntaka to the accession of Rājarāja I is one of the most difficult passages of Cōḷa history. The evidence is confusing, and no two scholars are agreed in its proper interpretation. It is hardly possible to put forward a scheme of succession without a more or less full discussion of possible alternatives.
c16.s1.2
We may begin by setting forth in some detail the nature of the evidence at hand. Stone inscriptions constitute the chief source of our knowledge, and there are several among these that unmistakably belong to this period.¹ The inscriptions of Kannaradēva (Kṛṣṇa III) bearing regnal years higher than twenty-three, and found in the districts of North Arcot and Chingleput constitute the first group of these records; the latest year in them is twenty-eight, and this would take us to A.D. 965.² Of the remaining stone inscriptions of the period, several belong to Maduraikoṇḍa Rājakēsari and bear dates in regnal
c16.s1.3
years ranging from five to seventeen. Four inscriptions belong to Sundara Cōḷa with titles Madhurāntaka and Pāṇḍiyanaic-curam-iṟakkina; two of
c16.s1.4
these are dated in years five and seven, while the dates in the other two are lost. A considerable number with years ranging from two to five are those of Parakēsari who had the title (Vīra)
c16.s1.5
Pāṇḍiyan-talai-koṇḍa; and a still larger number belong to a Pārthivēndra-varman with the same title and with other variants to his name like Vēndrādivarman, Pārthivēndrādhipativarman and so on; these records give dates up to the thirteenth regnal year.³ Lastly, we have a number of Parakēsari Uttama Cōḷa's inscriptions with regnal years from two to sixteen; two of them are precisely dated and fix the limits of his reign.
c16.s1.6
The title Parakēsari, and the dates in his inscriptions⁴ which couple Kali year 4083 (A.D. 981-2) with his regnal year thirteen, fix Uttama Cōḷa's place as the immediate predecessor of Rājakēsari Rājarāja I. It should also be observed that some of the numerous records
c16.s1.7
which give no more detailed specification of the ruling sovereign than to call him Rājakēsari or Parakēsari will, no doubt, on any scheme of chronology, be found to fall in this period; but these may, for the most part be left on one side in the present discussion.
c16.s1.8
Besides the stone inscriptions, we have the evidence of copper-plate grants. The only grant that falls in this period is that in the Anbil plates of the fourth year of Sundara Cōḷa. Unfortunately for us, Mādhava Bhaṭṭa, the composer of the Sanskrit *praśasti* in these plates, was less anxious to record facts which must have been very well known to him than to display his capacity for *alaṅkāras* in his verse. As it is, even for the fact that Sundara Cōḷa was a Rājakēsari we have to depend on the opening of the Tamil part of the record. The Tiruvālaṅgāḍu plates imply, and the Leyden grant expressly affirms, that after the death of Rājāditya the succession took place in the following order: Gaṇḍarāditya, Arindama, Parāntaka, Āditya and Madhurāntaka. The Karandai plates of Rajendra I as also the Kanyākumārī inscription and the Chārāla plates of Vīrarājēndra give only the names of Arindama and Parāntaka II between Parāntaka I and Rājarāja, apparently because the authors of the *Praśastis* had no other object in view than to trace the descent of the reigning king in the direct line from Vijayālaya. The genealogy of the Cōḷas from Parāntaka I to Rājarāja as it is given in these records may be set down before proceed-ing further:
c16.s1.10
One thing is clear and it is necessary to grasp this firmly; that we cannot treat the regnal years gathered from the stone records as those of kings whose reigns succeeded one another in regular order in the manner implied by the copper-plate grants. For in that case the period would, if we omit Pārthivēndra-varman from the reckoning and add the highest regnal years known of Maduraikoṇḍa Rājakēsari, Sundara, Āditya and Uttama, extend to 45 years, a period much too long for the interval between Parāntaka I and Rājarāja.
c16.s1.11
Then there is the possibility that Gaṇḍarāditya and Ariñjaya also ruled as kings. A Rājakēsari inscription possibly of Rājaraja I's reign⁵ mentions the second year of Gaṇḍarādityadēva *alias* Mummuḍi-cōḷa-dēva, while taking stock of the gifts made by his queen Sembiyan Mahādēvi at various times to the temple of Tiru-veṇkāḍu. Records of the eighth year of Rājakēsari⁶ show that Arikuḷakēsari, the son of Parāntaka, stood in the relation of Piḷḷaiyār or Āḻvār⁷ (terms often applied to junior members of the royal family) to the ruling king who, being a Rājakēsari, could only have been Gaṇḍarāditya.
c16.s1.12
As for Ariñjaya, besides the statements in the copperplates regarding his rule, there is a Rājakēsari record dated in the twelfth year⁸ which mentions two queens of ‘Ariñjigaivarman who died at Āṟṟūr’; and inscriptions from Mēlpāḍi dated late in the reign of Rājarāja I⁹ mention the construction by him of the Cōḷēśvara temple as a memorial shrine (*paḷḷipaḍai*) to Āṟṟūrttuñjina-dēva. These references render it probable that Ariñjaya lived long enough to rule in his own right, though, perhaps, only for a short time. Taking all these facts into consideration, we see clearly that there must have been a great deal of overlapping of the regnal years of the different kings quoted in the lithic records.
c16.s1.13
Another preliminary question that must be considered is the exact significance of the phrase '*talai-koṇḍa*' which is of some importance to the history of this period; because Vīra Pāṇḍya claims to have done this to a Cōḷa king and others claim to have done the same thing to Vīra Pāṇḍya himself. The phrase has been generally understood to mean 'having cut off the head,' and the identity of the Cōḷa king who died at the hands of the Pāṇḍya has been much discussed.¹⁰ It appears, however, that the true meaning of the phrase is that the vanquished king had to acknowledge his defeat by humbling himself before the conqueror in a particular manner, as it were placing his head at the disposal of the conqueror. Hultzsch has pointed out¹¹ that in the inscriptions of Kulōttunga III, the phrase '*Pāṇḍiyanai muḍittalai koṇḍaruḷiya*' employed in some records is explained by another phrase: '*avan muḍi mēl aḍi vaittu*' of other inscriptions of the same reign; so that the process of 'taking the head' consisted in the vanquished king bowing and touching with his head the feet of the conqueror seated in open darbar. The celebrated Kṛṣṇa Dēva Rāya of Vijayanagar only varied this traditional procedure slightly when he demanded that, as the price of peace, Adil Shah of Bijapur should visit him and kiss his foot.¹² The *Guruparamparai*, a work which often gives a true account of the practices of Cōḷa times, mentioned that the king from Gaṅgaikoṇḍa-śōḻapuram mounted his state elephant by placing his foot on the head of his *sāmanta*, or feudatory.¹³ The above interpretation of *talai-koṇḍa* has a bearing on the discussion of this period, because it follows that a king whose 'head was taken' by another need not be assumed to have died at the time this happened to him. It is, of course, another matter where we have clear statements of a person being killed. Applying this to the cases arising in this period, we see that the Tiruvālangāḍu plates make an express declaration that Āditya II killed Vīra Pāṇḍya in battle and brought his severed head to the Cōḷa capital.¹⁴ Even here the lateness of the testimony throws suspicion on the event. There is no clear evidence, however, that a Cōḷa sovereign lost his life at the hands of Vīra Pāṇḍya, and apparently Vīra Pāṇḍya's boastful title meant no more than that he inflicted a tempo- rary humiliation on a Cōḷa king or prince.
c16.s1.14
Evidence has been cited above showing that Gaṇḍarāditya was a Rājakēsari and that he ruled for at least eight years. His rule might have com- menced from the death of Rājāditya, whose place he must have taken as heir apparent in Parāntaka's lifetime. A suggestion has sometimes been made that Gaṇḍarāditya was a Parakēsari and that Arindama, placed immediately after him by the Tiruvālangāḍu plates, was identical with Maduraikoṇḍa Rājakēsari.¹⁵ This seems to rest really on two assumptions; that Rājāditya ruled after Parāntaka's death, and that as he was a Rāja- kēsari, Gaṇḍarāditya who succeeded him must have been a Parakēsari. But it has been shown that the first of these assumptions is not true. It is likely enough that Rājāditya assumed the title Rājakēsari when he was made heir apparent; but considering that he predeceased his father, Gaṇḍarāditya who took his place must have assum- ed the same title so that the sovereign ruler next after Parān- taka Parakēsari might be a Rājakēsari. In fact the general opinion now prevailing is that Gaṇḍarāditya was a Rājakē- sari,¹⁶ though the implication that Rājāditya predeceased his father does not seem to have been so clearly grasped. For we shall see at a later stage of this discussion that an appli- cation of the rule that every heir apparent, whether one or more, adopted the title Rājakēsari or Parakēsari, according as the ruling sovereign was Parakēsari or Rājakēsari, furnishes a neat solution of the difficulties that could not otherwise be adequately met.
c16.s1.15
The Rājakēsari title may, therefore, be admitted for Gaṇ- ḍarāditya. But the suggestion, first made by Venkayya,¹⁷ that Gaṇḍarāditya was no other than Maduraikoṇḍa Rājakēsari, appears to have been accepted without sufficient examination of the evidence.¹⁸ The title Maduraikoṇḍa was taken to imply that the king was the son of Maduraikoṇḍa Parakēsari, Parāntaka I. Hence when Gaṇḍarāditya was believed to be a Parakēsari, Maduraikoṇḍa was identified with Ariñjaya Rājakēsari.¹⁹ Later when Gaṇḍarāditya became a Rājakēsari himself, he was said to be also Maduraikoṇḍa Rājakēsari. But this asumption is by no means an argument, and one must be prepared to relinquish it if reason is shown for doing so. Now, let us turn to the provenance of the records of Maduraikoṇḍa Rājakēsari. All of them, except three records of the fourteenth and seventeenth years, come from the North Arcot and Chingleput districts. Three of these are dated in his fifth year, and a fourth in the seventh. One of the records of the fifth year belongs to a place called Karikkal, near Shōlingur, in the Walajapet taluq of North Arcot. Its date falls about A.D. 954. At the latest, that is, assuming that Parāntaka I died in 955 and that Gaṇḍarāditya did not begin his rule till after that event, it might be A.D. 960. It is difficult to see how this can be reconciled with the fact that Kṛṣṇa III was still at Mēlpaḍi in A.D. 959 distributing conquered territory among his followers, and that his inscriptions are found in Toṇḍaimaṇḍalam up to about A.D. 965. These objections to identifying Gaṇḍarāditya with Maduraikoṇḍa Rājakēsari are weighty, and should prevail even if no other identification were possible from the later inscriptions of the time. But fortunately there is open to us an easy, and doubtless the correct, solution of the question which would be readily accepted, once the idea is given up that Maduraikoṇḍa must have been a son of another Maduraikoṇḍa, that is, Parāntaka I. A single record from Koḍumbāḷūr,²⁰ from which the date has been lost, opens by mentioning ‘*uḍaiyār madurāntakan Sundara-śōḻan.*’ This unique record disposes of the notion that the title Madurāntaka (Madiraikoṇḍa) in the records of this period must have been borne only by a son of Parāntaka I, for we know of no Sundara-śōḻan who was his son. This inscription offers the clue to the correct identity of Maduraikoṇḍa Rājakēsari. Sundara Cōḷa, the son of Ariñjaya, was a Rājakēsari, as the Anbil plates tell us, and he had also the title ‘*Madurāntakan.*’ He has yet another title which gives evidence of his Pāṇḍyan war, namely, *Pāṇḍiyanaic-curam-irakkina.*²¹ These facts seem to point definitely to the conclusion that Sundara Cōḷa Parāntaka II was the king who, in some of his records, is called Madi(u)raikoṇḍa Rājakēsari.
c16.s1.16
There is one record of Maduraikoṇḍa Rājakēsari which, at first sight, seems to render it more likely that Gaṇḍarāditiya rather than Sundara Cōḷa was the king of that record. This inscription²² from Tiruvorṟiyūr is dated in the fifth year and registers the endowment of a lamp by one of the nobles of Uḍaiyār Śrī Uttama Cōḷa-dēva who accompanied him to the temple. Krishna Sastri, in editing this record, says: 'A reasonable doubt may arise why Uttama Cōḷa is given here the title of a ruling king and not that of a prince. It was perhaps because he was the chosen successor of Gaṇḍarāditya at the time. We know, however, that he actually came to the Cōḷa throne only after one or two other kings had reigned subsequent to his father's death.' But if this was so, if Uttama Cōḷa was in the fifth year of his father's reign old enough to be chosen heir apparent, to adopt regal style and to visit temples with his (*perundaram*) nobles, and if his father continued to rule for twelve years thereafter, it is hard to see why he did not succeed his father immediately on his death and had to wait until most probably Ariñjaya and Sundara, possibly also Āditya II, had finished their rule. Not only does Krishna Sastri not explain this, but he says elsewhere:²³ 'At the time of Gaṇḍarāditya's death, Uttama Cōḷa must have been a young boy, as he was set aside in the order of succession till three kings after Gaṇḍarāditya had ruled and died.' Surely it is not easy to reconcile the two positions that the king of the Tiruvorṟiyūr record was Gaṇḍarāditya, and that at his death, his son was a young boy, so young that he had to wait through the reigns of three successors of Gaṇḍarādityā before he could himself rule. There is a more serious discrepancy. Even assuming that Gaṇḍarāditya counted his regnal years from the death of Rājāditya (a.d. 949), if we identify him with Maduraikoṇḍa Rājakēsari, who ruled for at least seventeen years, his reign would extend up to 966. And Madhurāntaka Uttama Cōḷa began to rule in 969-70. The interval of three years is too short to take in the reign of even Sundara Cōḷa whose records unmistakably give him a rule of not less than seven years,²⁴ much less those of three kings. The Tiruvorṟiyūr record cited at the beginning of this paragraph cannot, therefore, be assigned to Gaṇḍarāditya. It must be admitted, however, that even if we take it to be a record of Sundara Cōḷa, the difficulty still remains of explaining the regal title of Uttama Cōḷa in it. Seeing that Sundara had a son Āditya who was ably assisting him in his warlike enterprises, it seems unlikely that he recognised, and that so early in his reign, a prince from a collateral, although senior, branch of his family as heir apparent. The only suggestion that offers itself, and for which support may be found in analogous cases from Cōḷa records, is that though the gift recorded in the inscription was made in the fifth year of Sundara Cōḷa when Uttama might have been old enough to have a retinue of his own with which he went about touring the country, it was not actually engraved on stone until Uttama Cōḷa had begun to reign in his own right, which he did, as we shall see, immediately after Sundara Cōḷa.²⁵
c16.s1.17
There remain two further questions to be discussed—the position of Vīra-Pāṇḍiyan-talai-koṇḍa Parakēsari and the identity of Pārthivēndravarman who bears the same title. The former may certainly be identified with Āditya, the son of Sundara Cōḷa, who according to the Tiruvālaṅgāḍu plates and the Leyden grant fought against Vīra Pāṇḍya.²⁶ But Āditya II was followed by another Parakēsari, viz., Madhurāntaka Uttama Cōḷa. This appears, at first sight, to violate the normal rule of the Rājakēsari alternating with the Parakēsari title among the ruling sovereigns of the Cōḷa dynasty. Krishna Sastri says of Uttama Cōḷa:²⁷ 'Contrary to the usual order, according to which he ought to have been a Rājakēsarivarman, his predecessor Āditya II being Parakēsarivarman, he too was called a Parakēsarivarman, evidently because he was the son of a Rājakēsarivarman and succeeded to the throne not by the right he possessed, but at the request of his cousin's son Rājarāja I who was the chosen successor.' Two remarks may be made on this explanation: first, it seems hardly consistent with the position held by Krishna Sastri, not by us, that Uttama Cōḷa was chosen successor in Gaṇḍarāditya's reign, and that on account of his tender age, 'the claims of Gaṇḍarāditya's chosen successor were temporarily set aside and postponed.'²⁸ For, if this was so, how can it be said that he succeeded to the throne not in his own right, but at the request of Rājarāja? Then again the implication that the son of a Rājakēsari must be a Parakēsari is contradicted by the example of Rājarāja I, who was a Rājakēsari himself and the son of a Rājakēsari. It may also be stated here that Ariñjaya, who must have ruled a short while as Parakēsari, was also the son of a Parakēsari. The true explanation of two Parakēsaris coming one after another seems to be that the earlier Parakēsari, the chosen heir apparent, Āditya II, died in the lifetime of his father Sundara Cōḷa, and the prince chosen next for the throne also took the title of Parakēsari in order that Rājakēsari Sundara might be followed on the throne by a Parakēsari after his death. At the death of Āditya II,²⁹ Sundara Cōḷa had to choose Uttama Cōḷa and not his younger son Rājarāja, either because Uttama Cōḷa forced the choice by threatening civil war, or because Rājarāja of his own will preferred to wait. The verses in the Tiruvālangāḍu plates, which are the only direct source of our information, can support either interpretation. They declare on the one hand that Uttama Cōḷa was eager to rule and on the other that Rājarāja was too good a kṣatriya to dream of the throne for himself while his father's cousin wanted it. And we shall see that Uttama Cōḷa made clear his eagerness to rule by, possibly, instigating a political murder.
c16.s1.18
The identity of Pārthivēndravarman, whose records, are found in the districts of North and South Arcot, and Chingleput, is involved in much obscurity. The suggestion³⁰ that he was the same as Pṛthivīpati II, the Gaṅga feudatory of Parāntaka, is the result of a very dubious inference from the resemblance in meaning between the names Pṛthivīpati and Pārthivēndra. It is also partly the result of a confusion between the Gaṅga king and a feudatory of Kṛṣṇa III who, though he had the name Kannaradēva-Pṛthivigaṅgariyar, was quite different from the Gaṅga.³¹ After examining the records of Āditya II and Pārthivēndravarman, Krishna Sastri has reached the following conclusions:³² 'Both these kings claim the epithet, "who took the head of Pāṇḍya or Vīra-Pāṇḍya"—evidently the same Pāṇḍya king who was at war with Sundara Cōḷa Parāntaka II—and the title Parakēsarivarman. Inscriptions of the former are very few and found only in the south, the latest regnal year being the 5th. Of the latter, there are many in Toṇḍai-maṇḍalam and the latest regnal year is the 13th. Pārthivēndra Ādityavarman may have been a prince of the royal family and viceroy of Toṇḍai-maṇḍalam. Āditya Karikāla appears to have been the actual successor.' The inscriptions of Pāṇḍiyan-talai-koṇḍa Parakēsari are not so very few after all; and what is more important, they are not confined to 'the south,’ if by that is meant the country to the south of and outside Toṇḍai-maṇḍalam. There are at least five inscriptions of his in North Arcot and more in South Arcot. On examination we find that the Pārthivēndravarman records differ from those of Āditya Parakēsari in their provenance only in so far as the former are found in Chingleput district also, and are not found south of Toṇḍai-maṇḍalam. We also find the following features which have a cumulative significance in the consideration of his identity. He is a Parakēsari;³³ he has the regal title and calls himself ‘Kōvirājamārāyar’;³⁴ in an early record of the third year he is even called Pārthivēndra Ādittaparumar;³⁵ his queens have also the full regal style—*uḍaiyār dēviyār Villavan mahādēviyār*,³⁶ *Perumā-naḍigaḷ dēviyārt-tanmapponnār-āgiya Trailōkya mahādēviyār*.³⁷ It seems clear that, far from being a feudatory of the Cōḷa kings, the ruler who can lay claim to so much distinction must himself be a Cōḷa monarch, and the name Āditya and the title Parakēsari clearly suggest his identity with Āditya Karikāla Parakēsari. ‘Pārthivēndra Ādityavarman’ and its variant forms occurring in his records show that he took the title ‘Pārthivēndra.’ The Cōḷa kings were very fond of high-sounding *birudas*, and almost every one of them had many such titles. As the latest recorded regnal year in his inscriptions seems to be the thirteenth year,³⁸ it follows that he must have been chosen co-regent very soon after his father Sundara Cōḷa’s accession. This looks very probable considering that the Leyden grant gives him credit for having successfully attacked Vīra Pāṇḍya as a young boy. Afterwards he must have been deputed to rule the northern part of the Cōḷa dominion. Apparently he died in the life-time of his father and was succeeded in the place of heir apparent by Parakēsari Uttama Cōḷa.
c16.s1.19
The results of the foregoing discussion may thus be summarised before the history of the period is taken up:
c16.s1.20
| Rājakēsari Gaṇḍarāditya | A.D. 949/50³⁹–957 |
| Parakēsari Ariñjaya | c. A.D. 956–957 |
| Rājakēsari Sundara Cōḷa (Madurai-koṇḍa) | c. A.D. 956–973 |
| [Parakēsari Āditya II Pārthivēndra Karikāla | c. A.D. 956–969] |
c16.s1.21
There is just one record⁴⁰ of Pārthivēndravarman from the Chingleput district which is dated in the fifteenth year. Though the inscription is damaged, a careful examination of the impression of the record establishes two things: the date of the record is clearly 15; and though the chiselling of the letters is by no means good, and the stone is apparently much decayed, there is no reason to question the genuineness of the record which exhibits most of the characteristic palaeographic features of the period. This date must, if accepted, upset the scheme of chronology given above which is based on the identification of Āditya II with Pārthivēndra, for we cannot possibly find room for the fifteen years before the accession of Uttama Cōḷa and within the reign of Sundara. The period of thirteen years takes us almost to the limit and involves the assumption, in itself by no means unlikely, that Sundara associated his son Āditya in his administration very soon after his accession to the throne. On the other hand, this inscription of the fifteenth year from Parandūr is the only record so far known of Pārthivēndravarman with a regnal year higher than thirteen. We have several records of the thirteenth year, none whatever of the fourteenth, and only this one of the fifteenth. And it is extremely difficult to see any other solution to the riddle presented by his records than the one offered above; for if Pārthivēndra was not Āditya himself, how are we to explain the numerous coincidences to which attention has been drawn above, and which cannot all of them be accidental? The suggestion may therefore be made that in the regnal year in the Parandūr record, the second figure which looks like the ordinary *i* (இ) contains some mistake on the part of the engraver. This record need not be taken into account, therefore, unless other records with higher regnal years than 13 are discovered in future years.
c16.s1.22
One final consideration remains to be urged in addition to those so far dealt with. If Āditya and Pārthivēndravarman were not identical, the highest regnal year for Āditya would be that found in the records of Parakēsari who took the head of Vīra Pāṇḍya, namely five. We must naturally assume these five years to have been those immediately preceding Uttama Cōḷa’s accession in a.d. 969-70; this would lead to Āditya’s initial year falling somewhere about a.d. 964/5, which would be in the 8th or 9th year of Sundara Cōḷa according to the chronology suggested above. The fact that Sundara Cōḷa is clearly described as having come out successful in the Pāṇḍyan war before his seventh year (a.d. 963), together with the statement of the Leyden grant that Āditya took part in the fight at Śēvūr when still young, may lend some support to the view that Āditya had become sub-king earlier in Sundara’s reign. This argument cannot be pressed far, as it is possible that Āditya fought without being formally associated in the government; or a second fight with the Pāṇḍyas might have come about later, about the year a.d. 964/5