Pramāṇa · Search · About · API

The Pandyan Kingdom from the Earliest Times to the Sixteenth Century c5.s1: Sources

The Pāṇḍyan Kingdom from the Earliest Times to the Sixteenth Century c5.s1

Language
English

Sources

c5.s1.1

The sources of Pāṇḍyan history may now be considered. The value of indigenous literature for the historian of India has generally been somewhat underrated. There are, it is true, few professedly historical works of a contemporary character and for the most part we have to rely on casual references to historical events in works of general literature or to winnow a large mass of legend in search of a grain of fact. But with patience and care it may be found that many useful suggestions are derived from these works. The local legends centering round Madura exist in three versions of which the earliest dating from the ninth or the tenth century A.D. is the *Tiruvilaiyāḍal Purāṇam* of Perumbarrappuliyūr Nambi, the other versions being another and much later work of the same name by Paranjōtimunivar and the Sanskrit *Sthalapurāṇa* known as the *Halāsyamahātmya*. The two later works give a list of seventy-three or seventy-four kings forming one continuous line of rulers while the earliest version only mentions, and that very casually, seven kings of the family. The set lists of the later *Purāṇas* were very early recognized to be worthless for purposes of history;¹ in fact most of the names are, on the face of it, inventions of later times. But several of these stories have a quaint interest for the general student who sees here almost the same tendencies at work as produced the legends of Regal Rome. The older *Tiruvilaiyāḍal* thus gives a story (No. 12) in explanation of the name Māḍakkulakīlmadurai which we find in

c5.s1.2

¹ For the lists see Sewell’s *Antiquities*, vol. ii and Elliot, *Coins of Southern India*, pp. 128-9; also p. 121 for a short critique of the list. Nelson, *Madura Country*, part iii—contains an English version of the stories following the Sanskrit *Purāṇa* mentioned in the text. Other local *Purāṇas* like the *Kadambavanapurāṇa* have not been noticed.

c5.s1.3

Antiquary. Also Mr. P.T. Srinivasa Aiyangar, ‘ Pre-Aryan Tamil Culture ’ in the Journal of Indian History, vol. vii.

c5.s1.4

many Pāṇḍyan inscriptions as the name of the capital. Another story (No. 36) refers the name Madura to the fact that Śiva converted into sweet nectar the river of poison emitted by a giant cobra set upon the city by the magic of the Jainas. On another occasion these heretics sent an elephant against the city (No. 26) and the Lord petrified this beast in the form of the Ānamalai Hill, and as he used a lion-faced arrow on the occasion, the Pāṇḍyan king made a temple for Narasimha on the hill. We shall have occasion later to notice the true history of the temple. Apart from such stories, which need not be further retailed here, these Purāṇas,—especially the earliest version, which has been engaging our attention more than the others,—may be found occasionally to contain hints of great importance. Thus the life of Māṇikkavāsagar is treated in such detail as to explain the occasions on which he sang particular hymns of his Tiruvāśagam and his life is narrated before that of Gnānasambandar; and in this order, this version is followed by the later versions also. It must also be noted that some of these ‘sacred sports’ are referred to in the works of early Tamil literature and it is significant that none of these early references shows any sign of hostility to Jains or Buddhists.¹ These local and traditional Purāṇas are often very well supplemented by references in quasi-historical and religious works like the Periyapurāṇam, while the few direct references to historical persons and occurrences furnished by the saints and poets of the Tevāram and the Tiruvāymoḷi are of inestimable value. Then, we have the Sangam works which form a class by themselves and contain much valuable information which still awaits systematic and scientific treatment by the historian; the task has been begun in the works of the late Mr. Kanakasabhai and Dr. Ś. K. Aiyangar and other scholars. The ‘ Age of the Sangam ’ however is one of the debated problems of South Indian History and will engage our attention presently.

c5.s1.5

¹ See e.g. Silappadikāram, canto 11, ll. 23-30.

c5.s1.6

Turning to foreign literary sources, some of the earliest references are furnished by the Old Testament of the Bible and the Greek and Roman writers like Megasthenes, Pliny, the author of the Periplus and Ptolemy. These references have been discussed out of all proportion to their intrinsic importance and much ingenuity has been spent in trying to press into service texts which are very obscure in themselves.¹ Then, there are occasional references in old Sanskrit works composed in North India, besides very passing references in the records of Chinese pilgrims which are not very helpful to our present purpose. The Ceylonese chronicle Mahāvamsa contains several references to the Pāndyan kingdom and its affairs, but these must be carefully used as the chronology of the Mahāvamsa still rests on insecure foundations.² For mediæval Pāndya history we get some help from the Muhammadan historians of the time as also from Marco Polo.

c5.s1.7

¹ See Kanakasabhai, Tamils, p. 54, connecting Megasthenes' story about Pandaia and Pliny's Pandæ who were ruled by women with Silappadikāram, canto 23, ll. 11-13 which does not bear the interpretation given there. These classical references are collected together and briefly discussed by Caldwell, Comparative Grammar, especially pp. 88 ff. See also his Tinnevelly, pp. 17-22, on Korkai, Comorin, Paumben, etc.

c5.s1.8

² Hultzsch discusses these in J.R.A.S., 1913 pp. 517 ff

c5.s1.9

The archæological evidence bearing on the subject of our study is very extensive. Epigraphy is the most fruitful source of ancient history. In the variety, wealth and occasional length of both its stone and copper inscriptions, South India is remarkably fortunate. The number of Pāndyan records registered in the Epigraphical reports of Madras and Travancore are now a few thousands; but not many of these can be referred to dates prior to A.D. 1000. There are no doubt many more still awaiting discovery and registration. It may be noted in passing that, since the rejection by government of Dr. Hultzsch's suggestion to prohibit the renovation of temples till the inscriptions in them have been copied, ‘a more vigorous attempt had to be made to secure impressions of the inscriptions thus threatened with destruction.’¹ ‘It was the practice in ancient times, whenever a temple had to be rebuilt, to copy the lithic records found on its walls into a book and then re-engrave them again on the new walls’² and it would be well if this practice were followed by the renovators of temples in our own day.

c5.s1.10

¹ H. Krishna Sastri, Introduction to *S. I. I. (Texts)* vol. iv, see also *A. R. E. 1902.* ‘What the Mussulmans did not destroy is being demolished by pious Hindus!’—Hultzsch. Mr. Krishna Sastri has remarked elsewhere (*A.R.E.*, 1913, part ii, para 41), ‘Some intelligent engravers on the stone helped by the members of the Archæological staff must, in my opinion, be enough to carry out this old scheme of preserving ancient records from complete ruin.’

c5.s1.11

² *A.S.I.*, 1909-10, pp. 128-9.

c5.s1.12

The bulk of the early inscriptions employ the script known as *Vaṭṭeḻuttu* in the Tamil parts and the grantha in the Sanskrit parts; *Vaṭṭeḻuttu* gave way to the present Tamil script about the time of the Chola conquest of the Pāndya country, say about the end of tenth century A.D.³ or the middle of the eleventh. It may also be noted that several of the later inscriptions are in excellent verse while the longer records of the early Pāndyas attain to great literary merit as prose compositions.

c5.s1.13

³ See *A.R.E.*, 1905, p. 43; also *Travancore Archæological Series*, vol. i, p. 286.

c5.s1.15

Sometimes the set forms of the historical introductions in the inscriptions of particular kings help the historian in identifying the records and fixing their age.

c5.s1.16

Considerable light has been shed in particular on the history of the Pāṇḍyas of the First Empire of the seventh to the tenth centuries A.D. by several important documents brought to light since 1906. These are the dated stone inscriptions from Ānamalai and Aivarmalai; the Trichinopoly and Ambāsamudram inscriptions of Vara-guṇa; and the copper plate records known as the Śinna-manūr plates (two sets) and the Veḷvikkuḍi grant. The Madras Museum plates of Jaṭilavarman are now better understood than they were when they were published in 1893. All these records (except the Museum plates) are still new, and there is much room for difference as to their import at several points. Pāṇḍyan affairs often derive elucidation from the records of the contemporary Cholas and among these the Tiruvālaṅgāḍu plates and the Leiden grant of Rājēndra Chola deserve special mention. The Pallava grants are also occasionally very helpful.

c5.s1.17

One difficulty that is common to an interpretation of all Pāṇḍyan records arises from the way in which they mention regnal years in double dates x years opposite y years. Several suggestions have been made but none of them is quite satisfactory, and the usual procedure is to treat the date as equivalent to x + y years,¹ and calculate the date of accession accordingly. One instance which may go to justify this practice is found in the larger Sinnamanūr grant where the regnal year இரண்டாவதின் எதிர் பதினொன்றாவது is rendered in the Sanskrit part of the grant by Ṣoḍaśe Rājyavarṣe. Again, some inscriptions give the regnal year and the number of days since the commencement of the reign or since the commencement of the current regnal year. And when we get to the numerous epigraphs of the mediæval and later Pāndyas, such difficulties increase enormously. Almost invariably the records of the Pāndyas who ruled in Tinnevelly in the period of the decline of the Pāndya power, i.e. in the fifteenth century and later are dated in the Saka era; on the other hand Saka dates are the exception in the inscriptions of the mediæval Pāndyas or the Pāndyas of the Second Empire as we may call them. But several records contain astronomical data which yield often strange and perplexing results. Many kings have been made and unmade by hasty calculations and equally hasty corrections and the student of history who is not a specialist in astronomy has great difficulty with the astronomers.¹ And when it is remembered that the texts of the bulk of these inscriptions still await

c5.s1.18

¹ The history of this question is very interesting and the curious reader is referred to the following :—Burgess and Natesa Sastri, *Tamil and Sanskrit Inscriptions*, p. 30, n. 4 ; Hultzsch : *I. A.*, vol. xx, pp. 288-9 ; Kanakasabhai, *Tamils*, pp. 59-60 and notes. More recently, Mr. T. A. Gopinatha Rao suggested (*Sen Tamil*, vol. iv, p. 114) that in a date ‘இரண்டாவதின் எதிர் நாற்பத்தாண்டு’ the first figure referred to the actual regnal year of the ruling prince and the second gave the date counting from the coronation of the previous ruler, and that this method was usually continued till the ruling prince was crowned. But this explanation fails in a

c5.s1.19

¹ These difficulties will call for more attention later. But a few samples may be noted here. Inscr. No. 422 of 1917 is referred to A.D. 1357 in p. 112 and to A.D. 1445 in p. 113 of A. R. E. 1917-18. At p. 89, A. R. E. 1923-24 we find Nos. 327 and 334 of 1923 with calculated dates A.D. 1278 and A.D. 1417 ascribed to the same king.

c5.s1.20

date like 'பதிமூன்றாவதின் எதிர் பன்னிரண்டாவது' I. A., vol. xx, p. 288. And Mr. V. Venkayya could only say, 'The second figure in these double dates which are frequently met with in Pāndya inscriptions has been taken to refer to the actual reign of the king and the first either to his appointment as heir-apparent or to some other event prior to his coronation'. A. S. I. 1903-4, p. 272 n). Mr. C. V. Narayana Aiyar (Journ. Ind. His., vol. vii, part 2) assumes that the date opposite to which other dates follow must be constant in the case of the same king. Even this is not so. See, e.g., Nos. 548, 624 and 625 of 1926 which belong to the same king; also Nos. 159-63 of 1894 and No. IV in Trav. Arch. Series i, pp. 99 ff., 1. 50-54.

c5.s1.21

publication one gets some idea of the conditions under which this part of the subject has to be studied.

c5.s1.22

As pointed out in the *Epigraphical Report, 1913* (p. 85) : ‘The subject-matter of the majority of the well-preserved inscriptions is, a gift made to a temple either of land or of money, for maintaining daily worship, special festivals, lamps, flower gardens and repairs; for feeding Brahmanas and providing jewels; or, it may be for supplying ghee of sheep and cows, to burn perpetual lamps in the temple. Lands were presented or, sometimes, sold to the temple by private persons and village communities. In the latter case, the sale amount was recovered from the temple treasury through the god Chandēśvara (the supposed manager of Śiva temple) and through temple trustees (*sthānattār*) (in the case of Vishnu temples). All land-gifts, whether sold or presented, were made entirely tax-free, the parties selling them invariably agreeing to meet the *irai* on such lands from their own pocket. Lands thus owned by the temple (*devadāna*) were in turn leased out to be permanently enjoyed as *kāṇi*, to select tenants or to the donors themselves (if cultivators), in consideration of a fixed amount of coin paid, or grain measured, at the temple treasury, regularly every year. Money gifts made to the temple were deposited with village assemblies and private individuals on permanent interest (*nilai-poliyūttu*) from which alone the temple had to meet the expenses specified by the donor. If the interest was not paid in any one year, the depositories agreed to pay it with the amount due for the following year together with a fine (*daṇḍam*) fixed for the default period by the officer (*dharmāsana* or *dharmāsanabhaṭṭa*) who was one of the members of the temple establishment. A curious condition was that the man who came to collect the arrears thus due was to be fed twice every day till, perhaps, the amount was fully paid up. Sheep and cows granted to a temple for maintaining lamps were received by the shepherds (*manrādis*) and the prescribed measure or measures of ghee supplied without fail. The cattle were, it may be noted, considered "never to die or never to grow old" for the apparent reason that they multiplied and increased in number.

c5.s1.23

A very small percentage of the inscriptions treat of subjects other than the gifts specified above. Sale or exchange of land among private parties, inquiries into temple management made by officers (*adhikāri*) appointed by the king with a view to collect the outstanding arrears of a temple, assignment of taxes by kings or of tolls by merchants for the benefit of a temple, dedication of hereditary servants (men or women), settlements of disputes, specification of caste or communal privileges, memorials to heroes who died either in cattle raids or on battlefields and other public charities, such as the construction of a tank, the planting of a grove, the gift of a water-trough, etc., are also, sometimes, permanently recorded on stone.' It may be added that these records often yield information of value relating to land tenure, public revenues, village administration, and generally the state of social and political life, affairs and activities.

c5.s1.24

But here a warning is necessary. The passage just quoted from the epigraphical report furnishes a comparatively harmless instance of a tendency to combine information from diverse sources, separated widely in time and space, and so to form a general picture of the social or political life of the country. This tendency has particularly unfortunate results in the study of institutions as it is likely to produce an appearance of flat uniformity and absence of change; it will also increase the difficulty of detecting the presence or otherwise of any signs of change or growth. It is only by carefully limiting our observations to the time and place indicated by the sources of our information that truth can be served in the present, and perhaps future work rendered easier.¹

c5.s1.25

¹ Cf. similar remarks of Prof. K. V. Rangaswami Aiyangar in another connection, *Some Aspects of Ancient Indian Polity*, pp. 31-2.

c5.s1.26

We do not derive much help from numismatics for the study of Pāṇdyan history. Coins definitely attributable to the early Pāṇdyas are very rare. A few gold specimens are known² and these bear only the Pāṇdyan figure of the fish. It is very interesting that the name Kūna known only to tradition and not to epigraphy is borne on a copper coin figured by Sir Walter Elliot (No. 140). It is well known that Roman coins belonging to different periods have been found in several places in South India and the abundance of Roman copper of the fourth and fifth centuries A.D. in and near Madura has sometimes led to the supposition that a local mint issued these pieces for daily use in a settlement of foreign merchants. Many Roman coins of the Early Empire have been found in and near Madura.³ By far the largest number of coins belong to the mediæval Pāṇdyas and bear legends substantiated by inscriptions. These coins often show the influence of the Chola conquest by the presence of a tiger design or of Ceylonese influence indicated by the presence of ‘a rude human figure, standing on the obverse, and seated on the reverse’ (Elliot, p. 108). The earliest coins of the Ceylon type date from the eleventh century; ‘it came into use in Drāvida only, at the time the Chola-Pāṇdyan dynasty were masters of the whole of it’ (Elliot, p. 109). It should also be observed that ‘the constant warfare which raged between Chola and Pāndyan (rulers) not only renders it well nigh impossible at any particular time to fix the exact boundaries of their respective territories, but also causes considerable uncertainty in the identity of a large number of their coins.’¹

c5.s1.27

² See Elliot, *Coins of Southern India*, p. 121.

c5.s1.28

³ See Sewell in *J.R.A.S.*, 1904, pp. 595 and 600-15.

c5.s1.29

¹ Tufnell, *Hints to Coin Collectors in Southern India*, pp. 11-12. The most interesting of the Pāndya coins known so far have been described in this publication and Elliot, as also in the papers of Sir T. Desikachari in the *Tamilian Antiquary* and that of Hultzsch in *I. A.*, vol. xxi, pp. 323-6. The Roman coins are discussed by Sewell in the *J.R.A.S.*, 1904. It may be noted that Sir T. Desikachari mentions that gold coins with the fish design were found in South Canara; cf., in this connection, the observations of Prof. D. R. Bhandarkar, *Asoka*, p. 40.

c5.s1.30

This review of the sources indicates that there is a large mass of material for the history of the Pāndyas which awaits critical discussion and cautious summing up. Much good work has been done already; but more still remains to be done and it will be our endeavour in the following chapters to do something in this way.

Linked texts (8)