The Epigraphical Legacy at Gangaikondacholapuram: Problems and Possibilities

The Epigraphical Legacy at Gangaikondacholapuram: Problems and

Possibilities†

Daud Ali

Gangaikondacholapuram served as an important centre of Chola activity throughout the eleventh and twelfth centuries, yet our knowledge

of its history remains in many respects uncertain and fragmentary, despite the impressive efforts recently exerted by the École Française d’Extrême-Orient in Pondicherry.1 This paper will point out some peculiar features in the surviving epigraphic corpus at Gangaikondacholapuram, the exploration of which may shed relevant light on its early development and historical importance in Chola and post-Chola history. Integral to the approach taken in this essay will be a close attention to the spatial and chronological distribution of the surviving inscriptions. Leslie Orr has recently drawn our attention to the fact that while the geographical fi nd spots of inscriptions have long been recorded and noted by scholars, we have often been much less sensitive to the other aspects of their physical character, including their spatial locations and relative placement on monuments and within architectural programmes.2 Gangaikondacholapuram, particularly when compared to its ‘parent’ monument at Tanjavur, offers considerable scope for such methods. The goal of this paper is to make some preliminary observations regarding the epigraphic profi le of the temple and to offer some suggestions as to their historical implications. Many of these suggestions must remain provisional, however, until further research is conducted.3

The medieval inscriptions of Gangaikondacholapuram are located on and around the famous Śiva temple known in contemporary records as Gaṅgaikoṇḍacoḻīśvara.4 The temple’s name, like that of the city itself, commemorated the famous expedition to the Ganges by the Chola ruler Rājendra I (r. 1012–1044 AD) in 1022–1023 AD. The temple complex has yielded a fair number of inscriptions, many of them damaged or fragmentary, that have been copied over the years by surveyors and epigraphists. The fi rst collection of material at the site was carried out sometime between 1806 and 1818 by surveyors under the direction of Colonel Colin Mackenzie. Some fourteen epigraphic records were hand-copied and relayed to Madras where clean transcripts were prepared on paper, and fi led with the large number of

Epigraphic Enigmas.indd 1Epigraphic Enigmas.indd 1 2/10/2012 8:14:53 PM2/10/2012 8:14:53 PM

similar documents from across southern India. These transcripts eventually made their way into the Government Oriental Manuscripts Library and were later edited by T.N. Subramaniam in his South Indian Temple Inscriptions, published in 1953.5 By this time, of course, the Archaeological Survey of India (ASI) had long begun its own work of collecting inscriptions – with notices published in the Annual Reports on South Indian Epigraphy and full texts in Epigraphia Indica and South Indian Inscriptions. Surveyors from the ASI had visited Gangaikondacholapuram in 1892 and 1908, and texts recorded on the former of these visits were published in 1923. Subramaniam was of course aware of such developments, and in publishing over 1000 texts of the Mackenzie transcripts, tried wherever possible to correlate them with the numbered notices from the Annual Reports or the published texts of South Indian Inscriptions.6 But in the case of the fourteen inscriptions copied by Mackenzie’s assistants at Gangaikondacholapuram, Subramaniam was not able to corroborate even a single one with the efforts of the ASI.7 This is a remarkable disparity, one hardly addressed by any of the more recent studies of the temple. It is as if Mackenzie’s surveyors at the beginning of the century and the ASI epigraphers at its end had visited two entirely different sites! While this discrepancy requires further investigation, for the purposes of this essay the records in Subramaniam’s collection will be treated with only secondary importance, and will not be reckoned as authenticated epigraphs at the site.

To date, epigraphers from the ASI have made fi ve visits to the site, copying a total of thirty-nine inscriptions.8 As previously mentioned, the fi rst visit was in 1892, when nine inscriptions were copied from what the Annual Reports on Epigraphy call ‘the enclosure wall’ of the temple, but as we shall see, these inscriptions are actually found on the central shrine. Sixteen years later in 1908 another visit was made to the site, leading to the collection of six further inscriptions, three from the main shrine and three from the adjoining maṇḍapa in front of it. A brief visit to the site in 1962 revealed an inscription incised on a stone fi xed into the wall of the north entrance stairs to the maṇḍapa. A visit two years later revealed a further seventeen inscriptions: one inscription on the central shrine, four fragmentary inscriptions built into the tiers and steps of the platform abutting the central shrine, six short and mostly fragmentary inscriptions built into the base of the platform of the large mahāmaṇḍapa in front of the temple, a fragmentary inscription built into the subsidiary Durgā shrine, two fragments built into the sidewalls of the east gopuram of the temple, and a fi nal three damaged records near or part of the executive offi ce for the temple. A fi nal visit in 1993 revealed six more fragments on loose stones lying in the temple precincts. This list may not be entirely complete. During a site visit conducted in 2008 with Dr A. Murugaiyan and

2 New Dimensions in Tamil Epigraphy:


Dr S. Rajavelu, an attempt was made to corroborate all inscriptions noticed by the ASI, correcting inaccurate locations where necessary (see Appendix and Figure 4). Although some important errors have been corrected, most notably the misdescription of the location of inscriptions copied in 1894, not all inscriptions noted by the ASI could be located, and at least three new inscriptional fragments were noticed, in addition to what seems to be a Tamil-Telugu fragment built into a large lion-well, indicating its construction by a zamindar of Udaiyarpalayam.9

All commentators have noted that the structure of Gaṅgaikoṇḍacoḻīśvara seems to have been largely modelled on the Rājarājeśvara temple at Tanjavur, founded by Rājendra’s father. Pichard’s exhaustive study, however, has argued that it was neither a mere replica nor an architectural ‘rival’, but was a ‘matter of elaborating and developing the architectural experiment’ represented by Rājarājeśvara.10 A key difference at Gangaikondacholapuram is the lack of a second, outer enclosure wall as at Tanjavur. Instead we fi nd only a single prakāra and gateway, on the east side. The central shrine of each temple, crowned with a large vimāna, is preceded by a covered entry front-porch (mukhamaṇḍapa) and a larger covered entry hall (mahāmaṇḍapa). As at Tanjavur, there are a number of subsidiary shrines within the temple enclosure at Gangaikondacholapuram. On either side of the main shrine, to the north and south, are two Chola-period temples which must have originally contained liṅgas now known as the northern and southern ‘Kailāsa’. These shrines seem to have been identical in construction, with front porches and entry halls, though the northern Kailāsa (currently occupied by the goddess Bṛhannāyakī), is better preserved. To the northeast of the central shrine is a small temple containing Caṇḍikeśvara, steward of Śiva, which may be dated to the original construction of the temple, and to the southwest is a small shrine to Gaṇapati, the original structure of which has been dated by Nagaswamy to the thirteenth century on stylistic grounds.11 Later structures within the temple precincts include a temple to Durgā, assigned to the fourteenth or fi fteenth centuries,12

and a lion-well already mentioned. The statue of Nandi facing the main shrine was found to be constructed from stone fragments and stucco. It is not known whether the original was composite or monolithic.

In addition the original structure seems to have undergone a number of later repairs the most notable of which were those to the maṇḍapas in front of the main temple. The pillars, walls, ceilings and stairs of these structures are probably later additions. Pichard has argued that because the socle of these fore-structures is at the same level as that of the main tower we may postulate a single stage of construction for the platform, and thus the presence of an original maṇḍapa which had at some point collapsed.13

Daud Ali 3


A good number of fragmentary inscriptions have been incorporated into the later repair of these structures. When colonial offi cers fi rst documented the site in the nineteenth century, much of the temple was overgrown and in disrepair, with the enclosure wall and gopuram partly collapsed and missing. This damage was likely originally sustained in the eighteenth century during struggles between the English-supported Nawab and the local Poligar of Udaiyarpalayam.14 Large parts of the temple were in ruins by the time Mackenzie’s surveyors visited the site between 1806 and 1818.15 But the greatest damage to the site occurred in 1836, when large amounts of fallen debris from these structures as well as large parts of the enclosure wall that remained were taken away by state offi cers for use in the construction of a nearby dam. As a colonial visitor poignantly remarked in 1855, ‘it must not be omitted that when the lower Kollerun anikat was built the structure was dismantled of a large part of the splendid granite sculptures which had adorned it and the enclosing wall was almost wholly destroyed in order to obtain materials for the work’.16 While we may only guess as to the other material taken from the inner precincts of the temple at this time, the large number of inscriptional fragments that have come to light over the years in and around the temple (see Appendix) suggests that this event sadly makes any complete and certain understanding of the temple’s history impossible.

The epigraphic legacy of the temple at Gangaikondacholapuram refl ects the monumental vicissitudes of the site. More than half of the surviving epigraphic records are fragmentary – either on stray stones recovered from within the temple complex or incorporated into later repairs or new monuments. Moreover, a substantial number of the extant in situ records, covering all walls of the central shrine and several of its adjoining buildings (see Figure 4) have been damaged or mutilated in some way and are thus incomplete. This state of affairs makes any certain and comprehensive reconstruction of the inscriptional profi le of the temple problematic, if not impossible. Apart from the fi nal removal of the enclosure wall, we have no clear idea when the damage was sustained at the temple or when repairs or additions were carried out. This prevents any certain conclusions regarding the original location of the epigraphic fl otsam at the site. The descriptive details of the records collected by Mackenzie’s surveyors as they have fi ltered down through Subramaniam’s South Indian Temple Inscriptions, however, do not seem to offer any clear light on either the condition of the complex at the beginning of the eighteenth century or the distribution of its epigraphs.17

Despite such limitations, some educated guesses may be proposed and a number of epigraphic features of the site are notable.



The town of Gangaikondacholapuram, as is well known, was named by Rājendra I to commemorate the victorious military expedition of his armies to northern India and his retrieval of water from the Ganges river for his kingdom. The historical introductions (meykkīrttis) of Rājendra I’s Tamil inscriptions begin to mention the event only from his eleventh and twelfth regnal years, so we may assume that the expedition was completed by 1023 AD.18 But it is not until the seventeenth year of his reign (1029 AD) that the town of Gangaikondacholapuram comes to be mentioned in his inscriptions.19 A typical example may be cited from the Rājarājeśvara temple at Tanjavur in which Rājendra claims to issue a royal order ‘while in the gallery which surrounds the king’s fl ower garden (āram) on the northern side of the royal hall known as Muṭikoṇṭacōḻaṉ within the palace (kōyil) at Gaṅgaikoṇṭacōḻapuram’.20 Given this six-year interim, it would be logical to conclude that the town was either constructed or given its new name at some point during this period – between the completion of the Gangetic expedition concluded sometime just before 1023 AD and the fi rst instance of a reference to the site in 1029 AD.

Site explorations and excavations at Gangaikondacholapuram have revealed that the town was surrounded by a large rectangular enclosure wall extending from east to west (c. 1900 by 1350 meters) and abutted on its perimeter by a moat. There are almost no remains of the interior of the town, except a large foundation which must have been the royal palace.21 Notably, the east-facing Gaṅgaikoṇḍacoḻīśvara temple was constructed outside the walls of the town complex in a northeasterly direction. The presumption has been that the temple was constructed at more or less the same time as the city, but to date no inscription has been found at the temple from Rājendra’s reign.22 A later record refers to arrangements for the temple only from his twenty-third year (1035), some six years after our fi rst reference to the city as an administrative centre. Three hundred meters to the west of the town was a vast reservoir, fed by rains and a small river, known in Chola epigraphy as the Cōḻagaṅga, or the ‘Ganga of the Cholas’. Pichard has argued that this layout suggests that the larger urban population existed outside the walled inner city, which was reserved for the palace and military/administrative elites.23 Unfortunately, we cannot be certain of the condition of the settlement prior to Rājendra’s foundation there – whether, for example, it was a previously inhabited settlement or was founded by imperial fi at.24 The reigning historical consensus, however, is that Gangaikondacholapuram, situated some sixty kilometers to the northeast of Tanjavur on dry and under-populated land – what Pichard has called a ‘human desert’ – was founded through deliberate political policy rather than as the result of an organically developed local

Daud Ali 5


urbanism, a hypothesis which at the very least explains the site’s gradual but pronounced contraction into insignifi cance in post-Chola times.25

Rājendra’s reasons for establishing the new palace-city are not entirely understood. Nagaswamy has argued that the place became celebrated as the commencement point of the Ganges campaign, and therefore was the most suitable spot for its commemoration.26 There exists, however, no clear epigraphic evidence that this was the case. Other scholars have suggested that the location had military strategic advantage in defending the kingdom against dynastic rivals to the north.27 This line of argument, however, has been thoroughly and effectively dismissed by Pichard, who has argued that the city’s location on the north side of the Kollidam makes more immediate tactical sense for a southward defense, but only if the deltaic area was to be conceded, hardly a strategic advantage of the site. Moreover, as Pichard points out, it is diffi cult to imagine such a defensive mentality as motivation at a time when Chola imperial power was at its greatest.28 Pichard has instead suggested that the city may have been founded precisely because of its remoteness and distance from the old city of Tanjavur. This would have potentially extricated Rājendra from factional politics at the capital, and given him a tabula rasa to assert his authority anew. Such decampments of royal cities are not unknown in both South and Southeast Asia.29

Of the thirty-nine extant inscriptions recorded at Gaṅgaikoṇḍacoḻīśvara by the ASI, twenty-three may be associated with the Chola period, (but only about twenty with absolute certainty), and, as mentioned above, among these not a single surviving epigraph dates to the reign of Rājendra I himself. For long it was thought that no inscription of Rājendra I even referred to the construction of the temple, but the important Essalam copper plate charter, published by Nagaswamy, dated in the twenty-fourth year of Rājendra I’s reign (c. 1036 AD) mentions not only that the king established Gangaikondacholapuram but that he built a great temple to Maheśvara there.30 This suggests that the temple was at least under construction but perhaps completed by this year, which makes the absence of any of his own records at the site all the more curious. Aside from a fragment in the temple precincts discovered in 1994 of the Chola king Rājādhirāja dated in the twenty-seventh year of his reign (1045 AD),31 the earliest extant dated Chola inscription at the site is that of Rājendra I’s third son Vīrarājendra (r. 1063–1069 AD), with another short fragmentary inscription which may be dated to the reign of Adhirājendra (possibly the son of Vīrarājendra). There are four inscriptions at the site clearly associated with the Chola king Kulōttuṅka I, two of which are dated to the forty-fi rst and forty-ninth years of his reign (though the latter refers to a gift at another temple). Three fragments found within the temple



precincts have been dated to the reign of Chola king Vikrama (r. 1018–1035 AD). An undated fragmentary inscription includes the eulogistic introduction (meykkīrtti) of Kulōttuṅka II (r. 1133–1150 AD), and six inscriptions may be dated in the reign of Kulōttuṅka III (r. 1178–1218 AD), including three fully dated inscriptions with meykkīrttis and three fragments, one of which refers to a land grant to another temple. The remaining Chola period inscriptions are undated and fragmentary, being built into various structures surrounding the central shrine.

The most important of the surviving inscriptions at Gaṅgaikoṇḍacoḻīśvara is dated to the reign of Rājendra I’s third son, Vīrarājendra. This long and complex record (Figure 1), mentioning many offi cials and villages, though described as ‘on the enclosure wall’ by the ARE reports and South Indian Inscriptions, actually commences from the northeastern corner of the central shrine. Vīrarājendra’s orders were issued from the bathing hall at the Chola palace in Kanchi during the fi fth year of his reign (1068 AD). The inscription is particularly valuable because it supplements or completes a number of orders, no longer extant, made during the reigns of earlier kings, most notably those of Vīrarājendra’s father Rājendra I (referred to as aiyyar in the inscription) and his elder brother Rājādhirāja (referred to as aṇṇar ).32 These earlier orders are dated in the twenty-third and twenty-fourth years of Rājendra’s reign (1035 and 1036) and the twenty-sixth and thirtieth years of Rājādhirāja’s reign (1044 and 1048), and though no longer extant, together constitute the earliest royal activities at the temple. They thus are of great value in refi ning a chronology for the site as a whole. We may recall that the town as a seat of Chola power begins in Chola inscriptions from the seventeenth year of Rājendra’s reign (1029), and the temple from the twenty-fourth year (1036) in Rājendra’s Essālam plates. The fact that our earliest reference to an order relating to the temple occurs in 1035, six years later than our fi rst reference to the city as a whole, suggests a number of things. At the very least, that the temple was not completed until this date, some years after the town had enjoyed the prestige of being a political centre. Its construction no doubt had commenced earlier, probably as early as the foundation of the walled city itself. But as at Tanjavur, where the Rājarājeśvara temple was constructed over a number of years, Gaṅgaikoṇḍacoḻīśvara was surely not constructed overnight, and the gap in time between our fi rst references to the city as a seat of political power and our fi rst reference to the endowment of the temple represent at least part of the time it took to build such a massive monument. The nature of Rājendra I’s record is also signifi cant. Vīrarajendra’s donations apparently supplemented a number of large land grants to the temple made in the twenty-third and twenty-fourth years of Rājendra’s reign. As Nagaswamy

Daud Ali 7


has noted, a number of the villages named in this inscription had in fact earlier been donated by Rājendra I’s father Rājarāja I (r. 985-1014) to his temple at Tanjavur.33 This apparently deliberate divestment of resources from Rājarājeśvara to Gaṅgaikoṇḍacoḻīśvara, Pichard has argued, suggests a policy of premeditated ‘severance’ from Tanjavur.34

The fate of Rājendra I’s original orders relating to the temple is unknown. It has been suggested that they may never have in fact been inscribed on the temple because it was not fi nished in Rājendra’s lifetime – and that they were entered into the royal register, but only engraved posthumously, during the reign of Vīrarājendra.35 This situation would then bear some resemblance to Rājarājeśvara in Tanjavur, which was completed only after Rājarāja’s death. But whereas the inscriptions of Rājarājeśvara were ‘back-dated’ to the occasion of their issuance during the king’s reign, the inscriptions at Gangaikondacholapuram are dated in the reign of Vīrarājendra. Moreover, while Vīrarājendra’s inscription ‘completes’ Rājādhirāja’s order, and may thus be perhaps considered an indictment of a previously unincised record, its relationship to Rājendra I’s order is apparently only referential – it cites Rājendra’s orders by way of supplementing them with further gifts of land from the same villages, rather than completing them. This would seem to imply, at the very least, that Rājendra I’s orders were in fully completed form, whether in palm leaf or stone by the time of Vīrarājendra. As the central shrine has survived intact, and bears no inscriptions of Rājendra himself, beginning instead, on its north face with the inscription of Vīrarājendra, it seems quite unlikely that Rājendra’s orders were ever


Figure 1: Vīrarājendra’s order, northwest corner of the main shrine.


inscribed on the temple. Yet given the fact that we have reference to such orders, the decision not to include them on the main shrine, even posthumously, remains a mystery.

But if this is the case, it also gives rise to other questions as to the overall epigraphic programme at the site. It is interesting to compare the spatial distribution of inscriptions at Gaṅgaikoṇḍacoḻīśvara with Rājarājeśvara in Tanjavur. At Tanjavur, the placement of epigraphs reveals a carefully planned programme. The inaugural inscription on the monument, dated in the twenty-sixth year of the king’s reign (1011 AD; but likely inscribed after his death) and located on the north wall of the central shrine (vimāna), commences ‘Let the gifts made by us, by our elder sister, by our wives, by other donors to the lord of the sacred stone temple called Śrirājarājeśvara, which we caused to be built in Tanjavur…be engraved on stone on the sacred vimāna.’36 A clear precedence is set in this order which was largely adhered to in the engraving of the over eighty of Chola inscriptions at the site. Among the majority of these inscriptions, dated in the twenty-ninth year of Rājarāja I’s reign and the early years of Rājendra I’s reign, those appearing on the vimāna pertain exclusively to gifts to the main deity of the temple, while those on the enclosure walls and gateways record the establishment and support of lesser and generalised temple functions –including the establishment and endowment of moveable metal images of various subsidiary deities and saints, the establishment of subsidiary shrines, and the arrangements for general temple personnel (watchmen, dancers, etc…) and supplies (most notably lamp-oil).

At Gaṅgaikoṇḍacoḻīśvara there seems to be no key ‘inaugural’ inscription as at Tanjavur – a fact which perhaps should not surprise us, given that most other imperial temples lack such apparently premeditated epigraphic programmes. Despite the substantial number of Chola inscriptions at the site, only three are found on the central shrine. And it is notable that the earliest inscription on the central shrine, and the fi rst Chola record at the site which is more than a fragment, commencing on the kumuda from the eastern corner of the north wall (see Figures 2 and 4) belongs already to the fi fth year of Vīrarājendra (1068). Beyond this lengthy and important order, there are two inscriptions on the central shrine of the reign of Kulōttuṅka I: one dated to his forty-fi rst year (1111 AD) and another to his forty-sixth year (1116 AD). Of these, the former seems to have been without any donative or operative portion and the latter seems to relate to another temple. These three inscriptions on the main shrine would seem to present a very different picture to Tanjavur. They do not give the picture of a carefully orchestrated ritual and political order being incised in stone

Daud Ali 9


as at Rājarājeśvara. The distribution of inscriptions on the central shrine at Gangaikondacholapuram differs from Tanjavur partly as a result of each monument’s different treatment of the mouldings on the adhiṣṭhāna. It is clear that the fl at façade of the jagati at Tanjavur provided a readier template for inscriptional elaboration (though the kumuda was also used) than at Gangaikondacholapuram, where no extensive fl at surface seems to have been part of the original moulding plans. This of course did not stop the engravers from making full use of the kumuda, padma and even the small ‘toe’ jagati at the base of the adhiṣṭhāna at Gangaikondacholapuram. But the architect’s apparent indifference to epigraphic and proclamatory potential in his design of the shrine’s moulding suggests that the imperial court, to the extent it was involved in the planning of this imperial temple, had no unifi ed agenda or programme for the ‘textualization’ of the monument. Such a hypothesis is strengthened by the generally miscellaneous and fragmentary nature of the existing epigraphy at the site.

The actual patronage and support provided to the temple by the Chola court is likewise problematic. Not only is the main land grant, though referring to lands donated to Rājarājeśvara, missing, but the peripheral and fragmentary inscriptions within the precincts of the temple, while they do suggest the royal court’s activity at the temple in the eleventh century, 37 mention no substantial amount of moveable wealth (images, ornaments) or services (temple personnel or temple supplies) donated to the temple in anywhere near the scale detailed


Figure 2: Location of inscriptions on the moulded base of the central shrine at Tanjavur (Left) and Gangaikondacholapuram (Right) (courtesy Adam Hardy).


so meticulously at Tanjavur. This rather episodic pattern of donation presents the historian with a serious problem if we are to assume that the city was founded as a grand imperial gesture, and tends to reinforce the suspicions of Pichard and others that the decampment to Gangaikondacholapuram, though politically motivated, may not have been economically and institutionally sustainable, leading to a sort of ‘fragmented urbanism’.

The preponderance of later Chola inscriptions at the temple, beginning from the reign of Vīrarājendra, is notable. It tends to suggest that both the temple and the city, whatever the depth of their institutional support, continued to be important centres of Chola activity throughout the later eleventh and early twelfth centuries. Inscriptional fragments at the site which mention the presence of female servants (peṇṭāṭṭi) from Chola palace establishments (veḷam) and royal ‘intimates’ (aṇukkiyār) – are all consistent with the presence of a Chola palace at Gangaikondacholapuram, known from contemporary inscriptions,38 from the middle of the eleventh century. These include two fragmentary inscriptions, built into different parts of the temple, mentioning a single palace servant (peṇṭāṭṭi) of the royal bathing establishment (tirumañcaṉattār veḷam) which may be attributed to the short-lived reign of Vīrarājendra’s son Adhirājendra,39 an undated fragment built into the doorframe of the entrance to the Durgā shrine, which registers gifts of gold to the temple by unnamed intimates and peṇṭāṭṭis of the royal bathing establishment,40 and a fragment built into the gopuram, ascribed to the reign of Kulōttuṅka I, mentioning the establishment of a palakaippaṭai at the temple by a royal intimate (aṇukkiyār) by the name of Sembikulamāṇikkattār.41 As is well known, Adhirājendra’s mysteriously short reign was overshadowed by the controversial accession of the Vengi Chalukya-Chola king Rājendra II, also known as Kulōttuṅka I, in 1070. Kulōttuṅka I has four inscriptions at the site, at least one of which seems to pertain to another temple. Engraved on the western side of the central shrine, this damaged inscription records a devadāna endowment to a Śiva temple at Pottālimaṭam. Its relevance to Gaṅgaikoṇḍacōḻīśvara is currently inexplicable. Among the remaining three, we have already mentioned the gopuram fragment above naming Sembikulamāṇikkattār. The other two are of interest because neither seems to record any fi nancial transaction – but are purely eulogistic in nature and suggest royal interest in the site for its symbolic potential.

The fi rst is a fragment fi xed into the eastern side of the outer wall supporting steps leading up to the entrance platform of the large mahāmaṇḍapa which contains two mutilated Sanskrit verses in Grantha script in the sragdharā metre which, though damaged, may be reconstructed through the help of an identical inscription found on the outside of the eastern wall of the innermost prakāra of the great Śiva temple at Chidambaram:42

Daud Ali 11


[sidda]ṁ

[pāṇḍyān=da]ṇḍena jitvā prac[ura-śara-mucā pañca] pañcānana-śrīr=d[agdhvā Koṭṭāra-]durggan=tṛṇam=iva sa [yathā Khāṇḍa]vam Pāṇḍu-sūnuḥ |

[piṣṭvā tat Kera]ḷānām balam=atibahi[lam śrī-Kulo]

tuṅgacolaś=cakre ś[akra-pratāpas=tri]bhuvana-vijaya-stambha[m=ambhodhi-tīre] ||

puṇye sahyādri-śṛṅge [tribhuva]na-vijaya-stambha[m=ambhodhi-pāre] svacchandam pārasī[nān=taruṇa-yuv]atibhir=ggīyate yasya kīrttiḥ |

[śrīmā]n=asta-śatruḥ prabala[-bala-bharaiḥ pa]ñca Pāṇḍyān=vijitya k[ṣubhyat-kṣmāpāla]-cakraṁ savidhikam=akaroch=Chrī-Kulotuṅgacolaḥ || 43

Following Hultzsch’s lead, the verses may be translated as follows: ‘Success. Having subdued the fi ve Pāṇḍyas by an army which discharged numerous arrows, having burnt like straw the fort of Koṭṭāram, just as the son of Pāṇḍu (Arjuna) (had burnt) the Khāṇḍava (forest), (and), having crushed that extremely dense army of the Keraḷas–that glorious Kulōttuṅka Chola, who resembles a lion in majesty and Śakra in valour, placed on the shore of the ocean a pillar (commemorating his) conquest of the three worlds. Having subdued the fi ve Pāṇḍyas by masses of powerful armies, that glorious Kulōttuṅka Chola, who has scattered (his) enemies (and) whose fame is spontaneously sung on the further shore of the ocean by the young women of the Persians, duly placed on the holy peak of the Sahyādri (mountain) a pillar (commemorating his) conquest of the three worlds, before which the crowd of kings is trembling’.44 The chief purpose of these verses was to praise Kulōttuṅka I’s conquest of Pāṇḍya and Keraḷa kings, and to note his far-fl ung fame and his erection of victory pillars (jayastambha). The inscription is probably early, as it seems to describe campaigns that were completed by his eleventh year (1081).45 Though we cannot be sure of the inscription’s original epigraphic context and location at the site, its lack of any transactional aspect, coupled with its perhaps contemporary duplication at Chidambaram, together suggest that its purpose was primarily a statement of royal authority.

Such a statement, issued at both Gangaikondacholapuram and Chidambaram, is particularly signifi cant given the circumstances of Kulōttuṅka I’s rise to power in the Chola heartland. As is well-known, the roots of Kulōttuṅka’s career may be traced back to Chola imperial ambitions in the Vengi region at the beginning of the eleventh century, which was



held by an ‘Eastern’ branch of the Chalukya family but coveted by both the Cholas as well as the ‘Western’ Chalukyas based in Kalyāṇi. Rājarāja I married his daughter Kundavai to the Eastern Chalukya king Vimalāditya while her brother, Rājarāja’s son and successor, Rājendra I, builder of Gangaikondacholapuram, strengthened the relationship some years later by bestowing his own daughter, Ammaṅgadevi, to Rājarāja Narendra, his sister’s son. The son of this marriage, the Eastern Chalukya prince Rājendra II, in turn married a Chola princess (daughter of the Chola king named Rājendra II) by the name of Madhurāntakī, sealing three generations of marriage between the families. The eastern Chalukya prince Rājendra II (later Kulōttuṅka I), after the death of his father Rājarāja Narendra in 1061, seems to have been forced into exile by his uncle Vijayāditya VII. Rājendra II’s career at this point is obscure. Violently disinherited from his patrimony by his uncle, he may have carved out a small sphere of infl uence for himself in the region of modern Bastar, where inscriptions of his second year mention that he had levied tribute from the Nāgavaṁsi king Dhārāvarṣa, near Cakrakūṭa.46 Meanwhile, in the Chola heartland Vīrarājendra appointed an heir, likely known by the name of Adhirājendra, whose successorship seems from the outset to have been contested. So contested, in fact, that the Western Chalukya king Vikramāditya VI (also related to the Cholas through marriage) apparently had to secure the throne for him at Vīrarājendra’s death.47 But shortly thereafter, Adhirājendra disappears from the epigraphic record and the eastern Chalukya prince Rājendra II suddenly appears in the Chola lands, taking the title ‘Kulōttuṅka’, or ‘one from lofty family(ies)’. Kulōttuṅka I’s role in the demise of Adhirājendra (if he played any) is unclear, but his sudden and almost immediate assumption of power at the death of Vīrarājendra suggests that there were factions loyal to him at the heart of the Chola political establishment.48 As has been pointed out, the poem called the Kaliṅkattupparaṇi, composed at Kulōttuṅka I’s court, probably in the last ten years of his reign, represents a seamless succession with no mention of Adhirājendra.49 Some sort of struggle may be alluded to by the text in its description of the strife and disorder ensuing in the Chola kingdom at Vīrarājendra’s death, and Kulōttuṅka I’s fortuitous arrival from northern regions to assume the throne. 50

Given this somewhat mysterious sequence of events, both the fragmentary inscriptions of royal servants dated to the reign of Adhirājendra and the rather early eulogistic verse of Kulōttuṅka I may be signifi cant, at the very least suggesting the presence of factions loyal to both kings within the city. While we have no evidence of a direct struggle to date, it is clear that Kulōttuṅka I’s accession, which constituted a clear break in the Chola line, precipitated a crisis in ‘legitimacy’. A considerable amount of evidence from Kulōttuṅka I’s reign

Daud Ali 13


suggests a conscious attempt on the part of the king and his allies to secure this legitimacy–including extensive literary patronage51 and temple building. Notable in the latter case was the temple at Chidambaram, the location of the twin inscription of the verses found at Gangaikondacholapuram, mentioned above. Chidambaram was perhaps the most sacred site for the Tamil Śaiva tradition in the Chola period. The temple there saw extensive building activity during the reigns of Kulōttuṅka I and his immediate successors.52 Hermann Kulke has further shown that the sthalapurāṇa associated with the site, the Cidambaramahātmya, most likely composed during the reign of Kulōttuṅka I, by temple priests he settled in the region, memorialised his accession in the text under the name of Hiraṇyavarman.53 It is within this larger context of textual production and monumental construction, in light of a potentially contested accession, that we should see these two early praise-verses, one at a key religious centre and the other at a key royal centre of the kingdom.54

Perhaps the most enigmatic inscription at Gangaikondacholapuram belongs to the end of Kulōttuṅka I’s reign. In the Annual Report submitted summarising the tours undertaken in 1907 and 1908, the Assistant Archaeological Superintendent for Epigraphy, Southern Circle, Rai Bahadur V. Venkayya, noted a peculiar inscription copied during his visit to Gangaikondacholapuram in February of 1907. On the west wall of the central shrine of the temple at Gangaikondacholapuram is an inscription, damaged in parts, which begins with the meykkīrtti of the Chola king Kulōttuṅka I commencing with the words pukaḻ cūḻnta puṇariyakaḻcūḻnta dated to the 4[1]st year of the king’s reign (1110–1111 AD), and is followed by four verses in Sanskrit written in Grantha (south Indian) characters.

[a]kuṇṭhotkaṇṭha-vai[k]uṇṭha-kaṇṭha-pīṭha-l[uṭha]t-karaḥ |saṁrambhas=surat-ārambhe sa śriyaḥ [śreyase=stu vaḥ] ||

[āsīd=aśītadyuti-vaṁśa-jāta-kṣmāpā]la-mālāsu divaṁ gatāsu |sākṣād vivasvān=iva bhūri-dhāmnā nāmnā Yaśovigraha ity=udāraḥ ||

[tat-s]uto=bhūn=Mahīcandraś=candra-dhāma-n[i]bhan=nijam |yen=āpā[ram=akūpāra-pāre]-vyāpār[i]taṁ yaśaḥ ||

tasyā=bhūt=tanayo nayaika-rasika[ḥ] k[r]ānta-dviṣan-maṇḍalo |vidh[va]stoddhata-dhīra-yodha-timira[ḥ] [śri]-Candradevo nṛpa[ḥ] ||55

Venkayya pointed out that these verses were a verbatim copy of a portion of the praśasti found in the copper plates of the Gāhaḍvāḷa kings of north India, suggesting some relationship between the two families. The inscription has been commented on by numerous historians but to date has defi ed any



wholly convincing explanation.56 The fi rst extant inscription of the Gāhaḍvāla dynasty was issued in 1090 AD, during the reign of Candradeva, though the standard genealogy includes the existence of two earlier kings, Yaśovigraha and Mahīcandra, unattested by their own inscriptions. The role the Gāhaḍvāḷas played in north India during the earlier quarters of the eleventh century when the Candellas of Jejabhukti, Kalacuris of Cedi, and Paramāras of Dhāra, vied for control of the Ganga–Yamuna Doab–and when Rājendra dispatched his expedition to the Ganges–is unclear.57 What is certain is that under the able military leadership of Candradeva (1089–1106 AD) and his son Madanapāla (1104–1113 AD), during whose reign the Gangaikondacholapuram inscription was apparently written, the Gāhaḍvāḷas consolidated control of what seems to have been their homeland around Varanasi and conquered the doab region as far as Kanauj, perhaps defeating kings further westward. Though all historians have recognised that ‘some sort of relationship existed between the Chola king Kulōttuṅka I and the Gāhaḍvālas of Kanauj’, the exact nature of this relationship has remained obscure. Venkayya surmised that ‘either Madanapāla or Govindachandra or some other member of their family went to the Chola capital Gangaikondacholapuram on a friendly visit to the Chola king Kulōttuṅka I and wished to make a grant to the temple. Accordingly the genealogy of the donor was put in. For some reason or other either the grant itself was not made or it was not engraved in full on the stone.58 Venkayya further hypothesised that the original friendship between the families may have originated in Rājendra I’s time, perhaps in connection with the famous trip to the Ganges. Beyond the obvious lack of any corroboratory evidence, it is unclear what status the Gāhaḍvālas had in the second decade of the eleventh century, when their fi rst public record cannot be dated until at least sixty years later. It may be more fruitful to look for contact between the families early in Kulōttuṅka I’s career, which, if it was indeed spent in the Bastar region, may have resulted in the establishment of friendly relations with the Gāhaḍvālas in common cause against the Kalachuris (who invaded Andhra during the early part of Kulōttuṅka I’s reign in 1072–1073 AD, according to their own inscriptions) or other spatially intervening kingdoms. One might, in this case, suspect evidence of a marital alliance – but among the many queens of the Chola kings from the time of Rājendra I to Kulōttuṅka I and his successors, none can be traced to the Gāhaḍvālas. So this line of interpretation remains highly speculative. In lieu of evidence supporting a direct dynastic connection, however, commentators have pointed to evidence of religious and cultic exchange. 59 It has been noted that sun-worship, prominently patronised in the Gāhaḍvāla kingdom, may have been introduced into Chola lands under Kulōttuṅka I, as the only surviving Sun temple of the period, called

Daud Ali 15


in its inscriptions the Kulōttuṅka-cōḻa-mārṭṭaṇḍa-ālayam, was established in his reign.60

The Kulōttuṅka I/Gāhaḍvāla praśasti remains thoroughly enigmatic. As Venkayya recognised, the inclusion of a date suggests that the inscription was an order or donation, but the absence of any operative portion, seemingly by design (ie, it is not apparently damaged or worn away), is peculiar. Moreover, in inscriptions containing two praśasti, one is often ‘nested’ or ‘enclosed’ within the other, making it clear which monarch is the issuing authority.61

In this case we have two juxtaposed inscriptions, almost certainly incised by the same hand.62 The date notation would suggest that in this case the issuing authority was Kulōttuṅka, save for the fact that the Gāhaḍvāla praśasti is placed after the date-citation where we would normally expect to fi nd the operative portion of the inscription.63 The stone face is not damaged after the conclusion of the Gāhaḍvāḷa praśasti verses, suggesting that the inscription was either not completed or did not contain any grant portion (see Figure 3).

After the reign of Kulōttuṅka I there may have been a lull in activity at Gangaikondacholapuram. Though recent discoveries found in the temple precincts have revealed three inscriptions dated to the reign of Kulōttuṅka’s son Vikramacōḻa (1018–1035 AD), and one in the reign of Kulōttuṅka II, none seem to suggest an imperial presence in the city.64 Thyagarajan has reported that no inscriptions dated during the reigns of Kulōttuṅka I’s immediate successors (Vikramacōḻa, Kulōttuṅka II, and Rājarāja II) refer to royal orders being made from the palace at Gangaikondacholapuram.65 This may likely be the result of royal activity at other palatial residences like Chidambaram,66 Muṭikoṇḍacōḻapuram,67 ‘Vikramacōḻapuram’68 and ‘Rājarājapuram’. Later


Figure 3: End of the Gāhaḍvāla praśasti


Chola inscriptions at the site include a relatively large number of records dated to the early reign of Kulōttuṅka III, who Nilakanta Sastri called ‘the last of the great Cōḷa monarchs’.69 Though mostly fragmentary, the inscriptions testify to the presence of royal offi cials or titled nobility at the site at the end of the twelfth century.70 There are no further Chola inscriptions at the temple after 1192 AD, and we learn little about the history of Gangaikondacholapuram in the thirteenth century from Chola sources–a fact perhaps pointing to the decline of the city or even royal decampment. Pichard has argued that the omission of Gangaikondacholapuram from the list of towns conquered in the Chola country (while Tanjavur, Uṟaiyūr and Āyirattaḷi are included) by the Pandya king Māṟavarmaṉ Sundara Pāṇḍya I (1216-1238 AD) in his elaborate meykkīrtti indicates the declining signifi cance of the city.71 But this might be too hasty a conclusion, as it is notable that the central shrine and maṇḍapa can boast of no fewer than seven Pandya inscriptions dating from the latter half of the thirteenth century–more, notably, than Rājarājeśvara at Tanjavur. These inscriptions, which record land gifts to the temple and special arrangements for services to be performed for specifi c Pandya kings, testify to the continued importance of the city and temple even as the Chola empire teetered on the brink of collapse.

The earliest chronologically fixed Pandya inscription at Gaṅgaikoṇḍacōḻīśvara is dated to the second year of the reign of Jaṭāvarmaṉ Sundara Pāṇḍya II (1251–1270 AD) and records the remission of all taxes and dues on a number of lands belonging to the temple in order to conduct an evening service in Sundara Pāṇḍya’s name.72 As mentioned above, by this time the Chola kingdom had already been overrun by the armies of Māṟavarmaṉ Sundara Pāṇḍya I, which did not, apparently, visit Gangaikondacholapuram. This was earlier in the reign of Rājarāja III (1216–1256 AD), whose fortunes were abysmal. After defeat by the Pandyas, he was captured by the Kāḍava king and was restored to the throne only by the intervention of the Hoysala Narasiṁha.73 Toward the end of his reign, his inscriptions decrease in number and are largely confi ned to the North Arcot and Nellore regions, while in the central Chola lands we hear of another Chola monarch, Rājendra III (1246–1279 AD) who seems to have made some effort to restore Chola power, judging from his claims to have defeated Pandya kings. He seems also to have been active at Gangaikondacholapuram, as he calls himself in a Sanskrit praśasti, ‘the chosen overlord of the city named for the Cōḻa who seized the Gaṅgā’ (gaṅgāgrāhakacōḷanāmapuravarādhīśvara).74 This may explain Jaṭāvarmaṉ’s appearance at Gangaikondacholapuram in 1253, presumably during or shortly after his conquests of the regions held by the Cholas and Hoysalas around Trichy. The conquests of Jaṭāvarman Sundara Paṇḍya II seem

Daud Ali 17


to have inaugurated a time in the 1270s and 1280s when the Chola kingdom became the playground of various Pandya princes, in a period marked by what Marco Polo called the rule of ‘fi ve royal brothers’.75 In this period, two inscriptions found on the southern side of the main shrine, which seem to record largely local transactions, are dated during the reign of Māṟavarmaṉ Kulaśekhara, suggesting his presence and power in the region.76 Another in the same location, dated sixteen years later in the reign of the Pandya king Māṟavarmaṉ Vikrama Pāṇḍya, known in his inscriptions for having defeated the Chola king on the banks of the Veḷḷāṟu river, makes arrangements for a special evening service to be performed in the king’s name, Rājakkaṇāyan.77 Beyond these dated Pandya inscriptions, three others, of seemingly ambiguous ascription, all in the name of one Tribhuvanacakravartin (a common Pandya and later Chola title), and which remain unpublished, came to notice in 1908 elsewhere on the main temple. Two of these three inscriptions are located on the east wall of the maṇḍapa in front of the main shrine,78 and the third apparently on the south wall of the central shrine itself.79 An apparent feature of the Pandya epigraphical legacy on the main temple is that they are predominantly on the southern side of the monument. While a fuller study of the Pandya inscriptions here (as elsewhere in the Chola country) has yet to be undertaken, it is possible to contend with some surety that Gaṅgaikoṇḍacōḻīśvara was still active in the latter half of the thirteenth century, and was important enough to attract the patronage of the Pandya rulers.

After the thirteenth century, inscriptions at Gangaikondacholapuram dramatically decline. A late fourteenth-century record found on the north wall of the mahāmaṇḍapa near the central shrine details an agreement between the nāṭṭār of a village and the temple regarding some lands, and three fragmentary inscriptions, dated toward the end of the fi fteenth century (on the west wall and north walls of the central shrine), record what appear to be the presence of Vijayanagara nobles or rulers, one of which mentions a street by the name of Rājendratiruvīthi.80 It seems that whatever the fate of the town, the temple of Gaṅgaikoṇḍacōḻīśvara remained an important enough site to attract the patronage of key regional rulers and lords from as far away as Hampi.81 If Pandya inscriptions seem to be on the southern façade of the temple, Vijayanagara inscriptions are often found inscribed on the padma of the main shrine, beneath Chola inscriptions on the kumuda, suggesting perhaps some need on the part of the local agents of Vijayanagar to inscribe a presence on the central shrine near well-known Chola inscriptions – though this remains speculative, and we have very little knowledge as to what extent the memory of the Cholas circulated in Vijayanagara times.82 Vijayanagara records also need to be placed in the context of post-Chola construction at



the site, which includes a number of repairs and new buildings, including, most notably, the Durgā shrine, which has built into its structure a number of Chola-period fragments. Overall, it would seem that despite the urban contraction of Gangaikondacholapuram into apparent insignifi cance from the

Daud Ali 19

93 2

11 10

1

1

1

1

19

19

19

19

15

21

20

2216/17?

2518

14

8?

12?

6

7

4

5

central shrinevimāna

mahāmaṇḍapa

mukhamaṇḍapa

Figure 4: Identifi ed inscriptions on main shrine and adjoining mandapas (plan after Pichard)


thirteenth century, the temple continued to draw some patronage from regional elites. The location of a considerable number of these later inscriptions on the central shrine stands in stark contrast to the temple’s fi rst Chola patron, whose presence at the site, to the extent that it remains foundational, is shadowy and poorly preserved. Further on-site research, particularly in comparison with other monuments like Tanjavur and Chidambaram, and with the details of Kulōttuṅka I’s reign, may help answer some of the unresolved issues pointed out in this essay.



Vīr

arāj

endr

adev

aye

ar 5

(106

8 A

D)

Kul

ōttu

ṅka

Iye

ar 4

1(1

111

AD

)

Kul

ōttu

ṅka

I ye

ar 4

9(1

119

AD

)

Kōṉ

ēriṉ

mai

-koṇ

ṭāṉ

Sund

ara-

Pāṇḍ

yade

vaye

ar 2

(125

3 A

D)

Kō

Māṟ

avar

maṉ

K

ulaś

ekha

rade

vaye

ar 5

(127

3 A

D)

Cho

la

Cho

la

Cho

la

Pand

ya

Pand

ya

Mut

ilate

d; R

ecor

ds th

e ki

ng’s

ord

er is

sued

whe

n he

was

in h

is p

alac

e at

Kan

chip

uram

rel

atin

g to

a n

umbe

r of

vill

ages

ass

igne

d as

teva

tāṉa

to

the

tem

ple

of G

aṅga

ikoṇ

ḍach

oḷīś

vara

m. N

eces

sary

ent

ries

wer

e m

ade

in ta

x re

gist

ers

rega

rdin

g th

ese

villa

ges

by a

num

ber

of th

e ki

ng’s

of

fi ce

rs. S

ome

of th

e vi

llage

s ha

d be

en a

ssig

ned

to th

e te

mpl

e in

the

reig

ns o

f R

ājen

dra

I an

d R

ājād

hirā

ja I

. Min

ute

deta

ils o

f th

e as

sess

men

t of

all

the

assi

gned

vill

ages

are

giv

en in

the

insc

ript

ion.

Dam

aged

and

unfi

nis

hed

Tam

il an

d G

rant

ha in

scri

ptio

n, b

egin

ning

w

ith th

e m

eykk

īrtti

of

king

beg

inni

ng w

ith th

e w

ords

pūk

āḻ c

ūḻnt

a an

d fo

llow

ed, a

fter

the

date

, by

two

vers

es o

f a

Gāh

aḍvā

la p

raśa

sti..

Beg

inni

ng o

f ea

ch li

ne lo

st; C

omm

ence

s w

ith th

e po

etic

intr

oduc

tion

begi

nnin

g p

ukaḻ

māt

u vi

ḷaṅk

a. R

ecor

ds g

ift b

y th

e ki

ng o

f la

nd a

s de

vatā

ṉa-iṟ

aiyi

li in

Cuṅ

kata

virn

ta-c

oḻan

allū

r to

a te

mpl

e of

R

ājen

drac

hola

Īśv

aram

uṭai

yār

at P

ōttā

limaṭ

am. T

he g

ift w

as m

ade

by

the

king

whi

le h

e w

as in

his

pal

ace

at G

anga

ikon

dahc

olap

uram

on

a pe

titio

n m

ade

by a

n as

cetic

(ta

pasi

) of

the

tem

ple

that

the

serv

cies

wer

e at

a s

tand

still

for

wan

t of

endo

wm

ents

.

Reg

iste

rs r

emis

sion

of

all t

axes

and

due

s on

num

ber

of la

nds

belo

ngin

g to

tem

ple

to c

ondu

ct s

ervi

ce in

tem

ple

calle

d Su

ndar

a Pā

ṇḍya

San

dhi.

Lan

ds w

ere

free

fro

m ta

xes

such

as

mīṇ

paṭṭa

m, t

ari-i

ṟai,

taṭṭā

pāṭṭa

m,

kaṭa

i-āya

m, e

tc…

.

Inco

mpl

ete.

Rec

ords

the

sale

of

land

s be

long

ing

to n

umbe

r of

in

divi

dual

s to

tem

ples

of

Gaṅ

gaik

oṇḍa

choḻ

īśva

ram

.

On

the

kum

uda,

pad

ma

and

jaga

ti of

the

nort

h an

d w

est w

alls

of

cent

ral s

hrin

e.

loca

ted.

On

the

wes

t wal

l of

the

cent

ral s

hrin

e of

the

mai

n te

mpl

e. lo

cate

d.

On

the

kum

uda

of th

e w

est

wal

l of

the

cent

ral s

hrin

e.lo

cate

d.

On

the

kum

uda

of th

e so

uth

wal

l of

the

mai

n sh

rine

.lo

cate

d.

On

the

kum

uda

of th

e so

uth

wal

l of

the

mai

n sh

rine

loca

ted.

SII 4

.529

ARE

82 o

f 18

92

Unp

ublis

hed

ARE

of 2

9 of

1908

SII 4

.527

ARE

80 o

f 18

92

SII 4

.524

ARE

77 o

f 18

92

SII 4

.525

ARE

78 o

f 18

92

1 2 3 4 5

Ap

pen

dix

: L

ist

of

Insc

rip

tion

s at

Gan

gaik

on

dach

ola

pu

ram

K

ING

/DA

TE

D

YN

AST

Y

DE

TAIL

S

LO

CA

TIO

N (

confi

rm

atio

n)

PUB

LIC

AT

ION

Insc

rip

tion

s on

th

e C

entr

al

Sh

rin

e

Daud Ali 21


Kō

Mā[

ṟava

rmaṉ

] K

ulaś

ekha

rade

va y

ear 4

(127

2 A

D)

Kōṉ

ēriṉ

mai

-koṇ

ṭāṉ

Vik

ram

a-Pā

ṇḍya

deva

year

6

(128

9 A

D)

Tri

bhuv

anac

akra

vart

in

no k

ing

men

tione

dsu

bhan

u ye

ar

---

Saka

138

5 ex

pire

d,

subh

anu

(c.

146

3 A

D)

Śaka

138

4 E

xp.

(146

2 A

D)

Prau

ḍha

Vir

upak

ṣarā

yaŚa

ka 1

405

exp

Śobh

akṛt

(148

3 A

D)

Adh

irāj

endr

a

Pand

ya

Pand

ya

Pand

ya

Vija

yana

gara

? Vija

yana

gara

Vija

yana

gara

Vija

yana

gara

Cho

la

Rec

ords

the

purc

hase

of

land

s in

Dev

anip

uttū

r b

elon

ging

to

cert

ain

Kum

āram

aṅka

lam

Uṭa

iyāṉ

, Civ

ataṇ

ḍaṉ.

Reg

iste

rs g

ift o

f 20

vēl

i of

land

by

the

king

for

a s

peci

al s

ervi

ce

calle

d R

ājak

kaṇā

yan-

Śand

i ins

titut

ed a

fter

the

nam

e of

the

king

as

tax

free

land

to th

e te

mpl

e au

thor

ities

of

the

tem

ple

of

Gan

gaik

oṇṭa

coḷīś

vara

m-u

ṭaiy

ār a

t Gan

gaik

oṇṭa

cōḻa

pura

m in

M

aṇṇa

ikoṇ

ṭacō

ḻa-v

aḷan

āṭu

in V

ikra

mac

ōḻav

aḷan

āṭu

Prob

ably

bel

ongs

to r

eign

of

Sund

ara

Pāṇḍ

ya.

Dam

aged

. Rec

ords

gif

t of

the

chie

f T

iruv

eṅka

ṭam

uṭai

yāṉ,

m

entio

ned

in n

o. 1

0 be

low

. May

hav

e be

en s

ubor

dina

te o

f V

ijaya

naga

ra k

ing.

Dam

aged

. Men

tions

one

Tir

uveṅ

kaṭa

muṭ

aiyā

ṉ E

kām

ranā

tha

Gaṅ

geya

n of

Vai

ppur

, a c

hief

who

see

ms

to h

ave

been

in

cont

rol o

f th

is a

rea

at th

e en

d of

fi ft

eent

h ce

ntur

y.

Muc

h da

mag

ed; m

entio

ns D

evar

āya

once

and

M

allik

ārju

narā

ya tw

ice.

Son

of M

allik

ārju

na.

Frag

men

tary

. Men

tions

Taḷ

ikku

ḷam

(ol

d w

ord

for

Tanj

avur

).

Seem

s to

rec

ord

prov

isio

ns m

ade

for

a pe

rpet

ual l

amp

prob

ably

by

one

Mar

utaṉ

, a p

eṇṭā

ṭṭi o

f th

e T

irum

añca

ṉattā

r ve

ḷam

. Thi

s w

oman

is a

lso

men

tione

d in

no.

29

belo

w (

326

of 1

965)

.

On

the

kum

uda

of th

e ea

ster

n si

de (

sout

hern

fac

e) o

f th

e m

ain

shri

ne. l

ocat

ed.

On

the

kum

uda

of th

e so

uth

wal

l of

mai

n sh

rine

. loc

ated

.

On

the

sout

h w

all o

f the

cen

tral

shrin

e of

the

mai

n te

mpl

e. n

ot

loca

ted.

cou

ld b

e un

iden

tifi e

d in

scrip

tion

on th

e ma

hāma

ṇḍap

a.

On

the

wes

t wal

l of

the

cent

ral

shri

ne. l

ocat

ed

On

the

padm

a of

the

wes

t wal

l of

the

mai

n sh

rine

. loc

ated

.

On

the

wes

t wal

l of

the

Cen

tral

shr

ine

of th

e m

ain

tem

ple.

loca

ted.

On

the

righ

t of

the

entr

ance

in

the

nort

h w

all o

f th

e sa

me

prak

āra.

not

loca

ted.

Bui

lt on

to th

e so

uthe

rn s

teps

of

the

plat

form

abu

tting

the

cent

ral s

hrin

e of

the

mai

n te

mpl

e

SII 4

.522

ARE

75 o

f 18

92

SII 4

.523

ARE

76 o

f 18

92

Unp

ublis

hed

ARE

31 o

f 19

08

SII 4

.528

ARE

81 o

f 18

92

SII 4

.526

ARE

79 o

f 18

92

Unp

ublis

hed

ARE

30 o

f 19

08

SII 4

.530

ARE

83 o

f 18

92

Unp

ublis

hed

ARE

323

of 1

965

6 7 8 9 10 11 12 13

Insc

rip

tion

s on

th

e M

ah

ām

aṇ

ḍap

a



---

[Kul

ōttu

ṅka

II]

[Kul

ōttu

ṅka

III]

year

[13

](1

191

AD

)

Kul

ōttu

ṅka

III

[Kul

ōttu

ṅka

III]

Tri

bhuv

ana

[cak

rava

rtin

]

Śaka

131

9 (1

397

AD

)

Kul

ōttu

ṅka

I

Cho

la

Cho

la

Cho

la

[Cho

la]

char

acte

rs o

f tw

elft

h ce

ntur

y

---(

Pand

ya?)

Vija

yana

gara

Cho

la

Frag

men

tary

. Con

tain

s po

rtio

ns o

f th

e m

eykk

īrtti

beg

inni

ng

pūm

ēviv

aḷar

… S

eem

s to

rec

ord

gran

t of

land

for

the

proc

essi

on o

f th

e de

ity d

urin

g a

fest

ival

. Men

tions

the

nam

es o

f a

few

mer

chan

ts s

uch

as

Till

aiyā

[ḷi]

Ātiv

iṭaṅk

aṉ, e

tc. a

nd a

nd r

efer

s to

Kar

ikāl

aśoḻ

acat

urve

dim

aṅga

lam

and

a d

evat

āṉa

iṟai

yili

Dam

aged

, with

beg

inni

ng li

nes

lost

. Com

men

ces

with

the

eulo

gist

ic

intr

oduc

tion

puya

l vāy

ttu, i

ndic

atin

g it

is p

roba

bly

date

d to

the

reig

n of

K

ulōt

tuṅk

a II

I. R

efer

s to

a r

oyal

wri

t and

an

inst

alla

tion

in th

e te

mpl

e.

Men

tions

Vir

udar

ājab

haya

ṅkar

a va

ḷanā

ṭu a

nd M

ēṟkā

l – [

nāṭu

].

Frag

men

t con

tain

ing

a po

rtio

n of

the

hist

oric

al in

trod

uctio

n of

the

king

.

Frag

men

tary

. See

ms

to r

ecor

d a

gran

t for

two

sand

hi la

mps

.

End

of

each

line

mut

ilate

d

Ver

y da

mag

ed. R

ecor

ds a

n ag

reem

ent b

etw

een

the

nāṭṭā

r an

d th

e [t

andi

rim

ar]

of P

arai

yūr-

nāṭu

and

the

sthā

ṉattā

r in

the

tem

ple

of

Gaṅ

gaik

oṇḍa

chol

īśva

ram

-uṭa

iya-

naya

ṉār

rega

rdin

g so

me

tem

ple

land

s

Con

tain

s tw

o Sa

nskr

it ve

rses

in th

e w

here

in th

e co

nque

st o

f a

fort

, vi

ctor

y ov

er th

e Pa

ṇḍya

and

set

ting

up o

f a

pilla

r of

vic

tory

are

m

etio

ned.

Fra

gmen

ts b

uilt

int

o th

e pa

dma

form

ing

the

base

of

the

sam

e pl

atfo

rm o

f th

e m

ahām

aṇḍa

pa o

f th

e m

ain

shri

ne;

On

the

mah

āmaṇ

ḍapa

, eas

t w

all,

out

er t

ier;

on

kum

uda

of n

orth

east

cor

ner

of

mah

āmaṇ

ḍapa

;

On

the

nort

h w

all

of t

he

maṇ

ḍapa

in

fron

t of

the

ce

ntra

l sh

rine

.

Sto

nes

buil

t in

to t

he

sout

hern

ste

ps o

f th

e pl

atfo

rm a

butt

ing

the

cent

ral

shri

ne o

f th

e m

ain

tem

ple

On

the

east

wal

l of

the

m

ahām

aṇḍa

pa.

Nor

th w

all

of

mah

āmaṇ

ḍapa

, clo

se t

o ce

ntra

l sh

rine

; lo

cate

d;

mis

desc

ribe

d in

AR

E r

epor

t.

Sto

ne fi

xed

int

o th

e w

all

of

the

plat

form

at

the

entr

ance

in

to th

e m

ukha

maṇ

ḍapa

;

Unp

ublis

hed

ARE

315

of 1

965

Unp

ublis

hed

ARE

319

of 1

965

Unp

ublis

hed

ARE

32 o

f 19

08

Unp

ublis

hed

ARE

324

of 1

965

Unp

ublis

hed

ARE

34 o

f 19

08

Unp

ublis

hed

ARE

320

of 1

965

Unp

ublis

hed

ARE

709

of 1

963

14 15 16 17 18 19 20

Insc

rip

tio

ns

on

th

e M

uk

ha

ma

ṇḍ

ap

a

Daud Ali 23


Kul

ōttu

ṅka

III

year

11

(118

9 A

D)

Kul

ōttu

ṅka

III

year

9(1

187

AD

)

---

---

Tri

bhuv

anac

akra

vart

inK

ōṉēr

iṉm

ai-k

oṇṭā

ṉ

---

Śaka

---

Cho

la

Cho

la

char

acte

rs o

f el

even

th c

entu

ry

[Cho

la]

char

acte

rs o

f th

irte

enth

ce

ntur

y [C

hola

]

---

(Pan

dya?

)

---

---

Rec

ords

the

dona

tion

of tw

o fl

ywhi

sks

to th

e de

ity m

ade

with

an

am

ount

of

poṉ

of 8

māṟ

i wei

ghed

by

Koy

ilaṅk

āḍik

-kal

by

Am

maṅ

kaiy

āḻvā

r, da

ught

er o

f C

oḷat

evar

and

one

fl yw

hisk

m

ade

with

20

poṉ

of 8

māṟ

i wei

ghed

by

the

sam

e st

one

by

Puka

ḷiyāḻ

vār,

daug

hter

of

Vāṉ

avaṉ

Māt

evīy

āḻvā

r.

Beg

inni

ng li

nes

lost

and

inco

mpl

ete.

Com

men

ces

with

puy

al

vāyt

tu. S

eem

s to

rec

ord

gift

of

482

kaḻa

ñcu

by T

uṇṭa

nāṭ-

uṭai

yaṉ

Eka

vācc

akaṉ

Kul

ōttu

ṅkar

, son

of

Can

dram

auliy

āḻvā

r fo

r so

me

purp

ose

in th

e na

me

of th

e la

tter,

who

app

ears

to

bea

r so

me

rela

tions

hip

to S

uṅga

ntav

irtta

ruḷiṉ

a K

ulōt

tuṅk

acōḻ

a.

Frag

men

tary

. One

of

the

frag

men

ts r

efer

s to

an

unde

rtak

ing

give

n by

an

indi

vidu

al a

fter

rec

eivi

ng fi

ve k

ācu

and

anot

her

men

tions

Maṇ

itaḷa

kuḷa

m u

ṭaiy

aṉ a

nd T

iruc

ciṟṟ

ampa

lam

- uṭ

aiya

ṉ an

d a

thir

d to

land

not

fi t f

or c

ultiv

atio

n of

pad

dy.

Frag

men

tary

. Men

tions

one

Tiṭṭ

ai P

iṭāra

ṉ

Reg

iste

rs o

rder

of

king

rel

atin

g to

em

olum

ents

of

a te

mpl

e se

rvan

t.

Frag

men

tary

. Ref

ers

to a

Gaṅ

gavi

ḷand

arap

urav

arād

hīśv

araṉ

, in

late

cha

ract

ers

Frag

men

tary

. See

ms

to r

ecor

d gr

ant f

or a

san

dhi l

amp

Tie

rs o

n th

e ea

st w

all

on th

e en

tran

ce to

the

muk

ham

aṇḍa

pa;

Tie

rs o

n th

e ea

st w

all

on th

e en

tran

ce to

the

muk

ham

aṇḍa

pa

Frag

men

ts b

uilt

into

the

padm

a fo

rmin

g th

e ba

se o

f th

e sa

me

plat

form

of

the

mah

āmaṇ

ḍapa

Fr

agm

ent b

uilt

into

the

padm

a fo

rmin

g th

e ba

se o

f th

e sa

me

plat

form

of

the

mah

āmaṇ

ḍapa

On

the

sout

h w

all o

f th

e m

ukha

maṇ

ḍapa

; pho

tos

0012

-001

4 lo

cate

d Fr

agm

ent b

uilt

into

the

base

of

the

left

pla

tfor

m o

f th

e ma

hāma

ṇḍap

a of

the

mai

n sh

rine.

Frag

men

t bui

lt in

to th

e pl

atfo

rm o

f th

e N

andi

.

Unp

ublis

hed

ARE

321

of 1

965

Unp

ublis

hed

ARE

322

of 1

965

Unp

ublis

hed

ARE

316

of 1

965

Unp

ublis

hed

ARE

317

of 1

965

Unp

ublis

hed

ARE

33 o

f 19

08

Unp

ublis

hed

ARE

314

of 1

965

Unp

ublis

hed

ARE

318

of 1

965

21 22 23 24 25 26 27

Insc

rip

tion

on

th

e N

an

di

Sta

tue



---

---

---

Kul

ōttu

ṅka

III

year

14

(119

2 A

D)

[Kul

ōttu

ṅka

I]

---

char

acte

rs o

f th

e el

even

th c

entu

ry[C

hola

] [C

hola

]

char

acte

rs o

f th

e tw

elft

h ce

ntur

y[C

hola

]

Cho

la

char

acte

rs o

f th

e tw

elft

h-th

irte

enth

ce

ntur

y C

hola

[Cho

la]

Frag

met

nary

. See

ms

to b

e pa

rts

of a

rec

ord

regi

ster

ing

gift

s of

gol

d by

per

sons

incl

udin

g fe

mal

e se

rvan

ts (

aṇuk

kiyā

r an

d pe

ṇṭāṭ

ṭi) o

f th

e T

irum

añca

nattā

r ve

ḷam

. Als

o re

fers

to T

ribh

uvaṉ

amāt

ēvip

-per

uṅkā

ṭi.

Frag

men

tary

. See

ms

to r

ecor

d a

gift

by

Mar

utaṉ

Civ

akko

ḻunt

u (a

peṇ

ṭāṭṭi

) Als

o re

fers

to tw

o m

erch

ants

of

the

pēra

ṅkāṭ

i, K

ūtta

ṉ Ā

divi

ṭaṅk

aṉ w

ho w

as c

harg

ed to

con

trib

ute

inte

rest

out

of

the

gift

and

T

utuv

aṉ…

who

sto

od a

s su

rety

for

the

form

er.

Rec

ords

that

the

pilla

r w

as d

onat

ed b

y Ā

ṇṭap

iḷḷai

-put

tanā

yaka

ṉ th

e ki

ḻāṉ

of P

aḻai

yāṟu

. In

char

acte

rs o

f th

e 12

th c

entu

ry.

Frag

men

tary

. See

ms

to r

ecor

d gi

ft o

f 12

0 aṉ

ṟāṭu

-naṟ

kācu

to th

e te

mpl

e of

Civ

aññā

ṉīśv

aram

uṭai

yār

by T

iruv

aiyā

ru-u

ṭaiy

aṉ.

Frag

met

nary

. Rec

ords

that

the

pala

kaip

paṭa

i pro

babl

y fo

r K

ūttā

ṭunt

ēvar

(Ś

iva

as d

ance

r) in

the

tem

ple

was

cau

sed

to b

e m

ade

by S

embi

kula

-m

āṇik

kattā

r, an

aṇu

kkiy

ār (

clos

e fe

mal

e at

tend

ant)

of

the

king

.

Frag

men

tary

. Lis

ts v

ario

us g

ifts

like

a la

mp,

two

cow

s fo

r its

m

aint

enan

ce a

nd g

old

for

som

e or

nam

ents

.

Fra

gmen

ts b

uilt

int

o th

e ri

ght

side

doo

rjam

b of

th

e en

tran

ce t

o th

e D

urgā

sh

rine

.

Frag

men

t bui

lt in

o th

e ba

se

of th

e w

all o

f th

e of

fi ce

of

the

Exe

cutiv

e of

fi ce

r of

the

tem

ple.

On

a st

ray

pilla

r ne

ar th

e E

xecu

tive

offi

ce o

f th

e te

mpl

e

Frag

men

t bui

lt in

to th

e w

all

of b

etw

een

the

hous

es o

f th

e E

xecu

tive

offi

cer

and

tem

ple

prie

st.

Frag

men

t bui

lt in

to th

e si

de

wal

l of

the

entr

ance

of

the

gopu

ram

of

the

tem

ple.

Ano

ther

fra

gmen

t in

the

side

wal

l of

the

entr

ance

of

the

gopu

ram

of

the

tem

ple

Unp

ublis

hed

ARE

325

of 1

965

Unp

ublis

hed

ARE

326

of 1

965

Unp

ublis

hed

ARE

327

of 1

965

Unp

ublis

hed

ARE

330

of 1

965

Unp

ublis

hed

ARE

328

of 1

965

Unp

ublis

hed

ARE

329

of 1

965

28 29 30 31 32 33

Insc

rip

tion

on

th

e D

urg

ā S

hri

ne

Exec

uti

ve

Offi

ce

Insc

rip

tion

s on

th

e G

op

ura

m

Daud Ali 25


Vik

ram

a-C

hola

,ye

ar 3

(112

1 A

D)

Vik

ram

a-C

hola

,ye

ar 6

(112

4 A

D)

Vik

ram

a-C

hola

year

9(1

127

AD

)

[Rāj

ādhi

rāja

II]

year

27

---

---

Cho

la

Cho

la

Cho

la

Cho

la

---

---

Frag

men

t. D

amag

ed a

nd in

com

plet

e. R

efer

s to

the

slui

ce o

f th

e la

ke in

Gaṅ

gaik

kond

acho

lapu

ram

of

Kiḍ

āraṅ

goṇḍ

acho

lava

ḷanā

ḍu

Frag

men

t. D

amag

ed a

nd in

com

plet

e. R

efer

s to

the

gift

of

mon

ey

(nār

kācu

) by

a la

dy, o

ne T

ribh

uvan

amuḍ

aiyā

ḷ with

the

inte

rest

am

ount

of

whi

ch a

san

dhi l

amp

was

to b

e m

aint

aine

d by

tem

ple

auth

oriti

es.

Frag

men

t. D

amag

ed a

nd in

com

plet

e. R

efer

s to

the

acce

ssio

n of

land

and

men

tions

pūp

paḍa

i of

Gaṅ

gaik

oṇḍa

chol

apur

am o

f K

iḍār

aṅgo

ṇḍac

hōḻa

vaḷa

nāḍu

Frag

men

t. D

amag

ed a

nd in

com

plet

e. C

onta

ins

port

ions

of

the

praś

asti

of th

e ki

ng a

nd r

ecor

ds th

e gi

ft o

f tw

o pe

rpet

ual l

amps

. [r

egna

l yea

r is

too

high

]

Frag

men

t. D

amag

ed a

nd in

com

plet

e. R

ecor

ds d

onat

ion

of a

do

orfr

ame

(tir

unila

ikkā

l) b

y on

e M

ādha

va-c

hēṭṭi

and

his

wif

e (a

kam

paṭiy

āḷ).

In

char

acte

rs o

f th

e 12

th c

entu

ry.

Frag

men

t. B

adly

dam

aged

. Pur

port

not

cle

ar.

On

loos

e st

one

piec

es k

ept

in th

e te

mpl

e pr

ecin

cts.

On

loos

e st

one

piec

es k

ept

in th

e te

mpl

e pr

ecin

cts.

On

loos

e st

one

piec

es k

ept

in th

e te

mpl

e pr

ecin

cts.

On

loos

e st

one

piec

es k

ept

in th

e te

mpl

e pr

ecin

cts.

On

loos

e st

one

piec

es k

ept

in th

e te

mpl

e pr

ecin

cts.

On

loos

e st

one

piec

es k

ept

in th

e te

mpl

e pr

ecin

cts.

Unp

ublis

hed

ARE

524

of 1

994

Unp

ublis

hed

ARE

525

of 1

994

Unp

ublis

hed

ARE

526

of 1

994

Unp

ublis

hed

ARE

527

of 1

994

Unp

ublis

hed

ARE

528

of 1

994

Unp

ublis

hed

ARE

529

of 1

994

34 35 36 37 38 39

Oth

er F

ragm

ents



NOTES

† Tamil and Sanskrit diacritical marks will not be used for modern place names or historical dynasties common in English language scholarship. Abbreviations include ARE (Annual Reports on Indian Epigraphy), SII (South Indian Inscriptions), EI (Epigraphia Indica) and IA (Indian Antiquary). In the case of fragmentary inscriptions, square brackets […] will be used to indicate letters supplied from context. I would like to thank Leslie Orr for her comments on a fi nal draft of this paper.

1 The fi rst academic study of the site was R. Nagaswamy, Gangaikondacholapuram (Chennai: State Department of Archaeology, 1970). This has been supplemented by the exhaustive work of the French Institute at Pondicherry, published in Pierre Pichard et. al., Gangaikondacholapuram: Vingt ans après Tanjavur, Mémoires Archéologiques 20, 2 vols., (Paris: L’École Français d’Extrême Orient, 1994) and the related tome by Pierre Pichard, Tanjavur Bṛhadīśvara: An Architectural Study (Delhi and Pondicherry: Indira Gandhi National Centre for the Arts, L’École Français d’Extrême Orient, 1995).

2 Leslie Orr, ‘Preface’, in G. Vijayavenugopal, ed., Pondicherry Inscriptions Part I (Pondichéry: École Française d’Extrême Orient, 2006), pp. iv–v. Though in my own previous work claims were made about moving away from the positivist framework of seeing inscriptions as self-evident fact-bearing objects, the approach was largely ‘textualist’, and still largely ignored their material and physical ‘life’. Daud Ali, ‘Royal Eulogy as World History’, in Ronald Inden, Daud Ali, and Jonathan Walters, Querying the Medieval: The History of Practice in South Asia (New York: Oxford University Press, 2000), pp. 165–229.

3 The critical survey offered here remains necessarily incomplete, as I was able to consult the estampages and transcripts of only some of the unpublished inscriptions in Mysore relating to the site, and have thus had to rely in many cases on the summaries included in the Annual Reports. I was also unable to consult transcripts copied from the site by surveyors under the direction of Colin Mackenzie at the Government Oriental Manuscripts Library, Madras.

4 Here I follow the conventional rendition by Nagaswamy and others, despite the irregular vowel sandhi in Sanskrit. In epigraphy the temple is often referred to as Gaṅgaikōṇḍacōḻa Īśvara.

5 T.N. Subramaniam, South Indian Temple Inscriptions, henceforth (SITI), 3 vols. (Madras: Government Oriental Manuscripts Library, 1953), see vol. 1 pp. x–xv for an introduction to these materials.

6 See SITI, vol. 1, p. xiii; vol. 2, p. xi. Those inscriptions from the Mackenzie collections which had been subsequently published by the ASI in South Indian Inscriptions or Epigraphia Indica were in most cases omitted (save where Mackenzie’s surveyors gave superior readings), but in their place Subramaniam prepared summaries and relevant references to these publications, included in appendices to each volume. Apparently, Mackenzie’s assistants also prepared tabular lists similar to the ARE reports, which presumably remain in the Government Oriental Manuscripts Library, SITI, vol. 2, p. xv..

7 See SITI, vol. 2, nos. 671–684, pp. 641–646. No mention is made in ‘Appendix VI’ of any Mackenzie inscriptions from the site which were published in South Indian Inscriptions.

8 Commentators have rarely appreciated their totality. R. Nagaswamy’s fi rst account of the site numbered 12 epigraphs ‘and a few fragments’ at the temple, of which a brief account was given. See R. Nagaswamy, Gangaikondacholapuram (Madras: State Department of Archaeology, 1970), p. 40. T.V. Mahalingam’s list includes only ten inscriptions. See T.V. Mahalingam, A Topographical List of Inscriptions in the Tamilnad and Kerala States, volume 8 (Delhi: Indian Council of Historical Research, 1991), pp. 322–325. In a 1994 study completed by scholars at the École Française Extrême-Orient, both the bibliography and Thyagarajan’s otherwise excellent essay on epigraphy fail to mention any of the later inscriptions copied at the temple, and in the latter case omits fragmentary records there relevant to his discussion of the word veḷam. See L. Thyagarajan, ‘Gangai and its Region: An Epigraphical Perspective’, in Pichard et. al., Gangaikondacholapuram, vol. 1, pp. 184, 203. R. Nagaswamy’s recently published Uṅkaḷ Ūr Kalveṭṭut Tuṇaivaṉ (Pathway to the Antiquity of Your Village), however, includes all inscriptions copied at the site. None of these studies, it should be noted, have included records from SITI.

9 Nagaswamy, Gangaikondacholapuram, p. 19; Pichard, Gangaikondacholapuram, p. 97. I would like to thank Philip Wagoner for interpreting this latter inscription.

10 Pichard, Tanjavur Bṛhadīśvara, p. 16.


11 Nagaswamy, Gangaikondacholapuram, p. 1912 Ibid.; Pichard, Gangaikondacholapuram, p. 97.13 Pichard, Gangaikondacholapuram, p. 55.14 Nagaswamy, Gangaikondacholapuram, pp. 10–11.15 Subramaniam prefaces a number of his texts with short descriptions, no doubt from Mackenzie’s

assistants, which suggest that forest had overgrown much of the complex, and that most of the subsidiary shrines were in a dilapidated state. See SITI, vol. 2, nos. 671, 675, et. al.

16 Cited in Nilakanta Sastri, The Cōḷas (Madras: University of Madras, 1955), p. 234.17 Subramaniam’s presentation of the material is careful but the records themselves and their

locational descriptions as found in the Mackeznie collections are often problematic and do not tally with what we know of current architectural remains. SITI, vol. 2, no. 671, for example, refers to a maṇḍapa in the midst of a forest north of the lion-well (śinga tīrtha) – which according to modern reconstructions could not have been within the temple complex.

18 See Nilakanta Sastri, Cōḷas, p. 238 n.51. His Sanskrit praśastis, as Nilakanta Sastri points out, (ibid., p. 195) though attached to records bearing earlier dates were composed much later and cannot be taken as reliable chronological indicators.

19 Nilakanta Sastri, Cōḷas, p. 227.20 Dated to the nineteenth year of his reign (1031 AD): SII 2.20, pp. 107, 109.21 See Pichard, Tanjavur Bṛhadīśvara p 114.22 Subramaniam includes among his records a fragment found ‘in the vimāna’ apparently dated to

the reign of Rājendra: ‘suba śrī ṟācentira coḻa tevarkku yāṇṭu…. mahesura rakṣa’, SITI, vol. 1, no. 678. There are however, a number of problems in assigning this record to the reign of Rājendra I. First, it contains no other titles like ‘kō parakesarivarmaṉ’ before the proper name of the king, as was common practice in other records dated to the reign of Rājendra. Moreover, the name Rājendra itself does not establish an identity with certainty, as there were a number of other kings who took this name (including Kulottuṅka I himself). The chronology of the Chola dynasty was hardly known with certainty when this ascription was made. Finally, Subramaniam, and perhaps Mackenzie’s surveryors, do not specify where in the vimāna the fragment was found, but it was presumably not built into any permanent structure for it was not noticed by the ASI surveyors who visited the site at the end of the century and has remained untraceable.

23 Pichard, Tanjavur Bṛhadīśvara p 115.24 On the erroneous claim that the Mudikondacholapuram of Chola inscriptions was an earlier name

of Gangaikondacholapuram, see Nilakanta Sastri, Cōḷas, p. 227.25 See Pichard, Tanjavur Bṛhadīśvara, p. 117.26 Nagaswamy, Gangaikondacholapuram, p. 13.27 See for example the remarks of Burton Stein, Peasant State and Society in Medieval South India

(Delhi: Oxford University Press, 1980), p. 334 ff.28 Pichard, Tanjavur Bṛhadīśvara, p. 118.29 Ibid. Pichard cites the case of Burma, but within India one may consider the Sultanate forts in Delhi

(and beyond) as well.30 For the earlier consensus, see K.R. Srinivasan, ‘Gaṅgaikoṇḍacōḻapuram’ p. 241; R. Nagaswamy,

‘Archaeological Finds in South India: Essalam Bronzes and Copper Plates ’, Bulletin de l’École Française D’Extrême-Orient 76 (1987), v. 18, p. 32.

31 ARE 527 of 1994. I have not been able to consult this inscription but the ASI have assumed that it belongs to the reign of Rājādhirāja II (r. 1163–1178) whose rule cannot accommodate such a high regnal dating.

32 SII 4.529.33 See Nagaswamy, appendix. Of the approximately twenty village names which remain in

the inscription, approximately eight (Maruttuvakkuḍi, Karuppūr, Iṅgaiyūr, Paṇamaṅgalam, Kārimaṅgalam, Sāttanpāḍi, Māhāṇikuḍi, Sirusemburai) are also mentioned in the inscriptions at Tanjavur. See, SII 2.4 and 5. Karashima compared details of villages mentioned in the Tanjavur inscriptions with those of some of the more complete Gangaikondacholapuram inscriptions to undermine the notion of the self-suffi cient village in medieval India, see Noboru Karashima,



South Indian History and Society: Studies from Inscriptions (Delhi: Oxford University Press, 1984), pp. 40–55.

34 Pichard, Tanjavur Bṛhadīśvara, p. 118.35 See K.R. Srinivasan, ‘Gaṅgaikoṇḍacōḻapuram, Bṛhadeśvara temple’, in Encyclopedia of Indian

Temple Architecture: South India, Lower Draviḍadeśa (Delhi: American Institute of Indian Studies, 1999), p. 241

36 SII 2.1.37 A few undated inscriptional fragments (nos. 20, 28 and 29 in the appendix) refer to personnel

from Chola palace-establishments (veḷams). Because the great majority of references to such establishments occur in the reigns of Rājarāja I and Rājendra I, these inscriptions are likely dated to the eleventh century. Unfortunately, their fragmentary nature does not allow for any larger conclusions to be drawn.

38 See for example the inscription of Vīrarājendra at Karuvur, which mentions an order made from the palace at Gangaikondacholapuram, SII 3.20.

39 ARE 323 of 1965 and ARE 326 of 1965. Both inscriptions apparently refer to the same person, one Marutaṉ of the tirumañcaṉattār veḷam--the former, dated to the reign of Adhirājendra, records a gift to the lord of Taḷikkuḷam, and mentions the queen [Vanā-?] Tribhūvanamāte[viyār].

40 ARE 325 of 1965 mentions aṇukkiyār and [tirumañcaṉa]ttār veḷattu pe[ṇṭāṭṭi].41 ARE 328 of 1965. The same person appears in an inscription dated to the reign of Kulōttuṅka I (no

regnal year mentioned, but probably late, 47th?) far to the north at the Mukhyācaleśvara temple near Tindivanam. See ARE 553 of 1994.

42 ARE 709 of 1963. For the Chidambaram inscription, see SII 1.155. For an improved reading and translation, see EI 5.13a.

43 This reconstruction is based on a comparison of the transcript at the ASI in Mysore with Hultzsch’s reading in EI 5.13a.

44 See EI 5.13a.45 See Nilakanta Sastri, Cōḷas, p. 336, n.60.46 cakkrakoṭṭattut tārāvaracaṉait tikku nikaḻttiṟai koṇṭaruḷi, see EI 9.1–9.2, and also Nilakanta Sastri,

Cōḷas, pp. 292, 303–304.47 Reported by the Chalukya court poet Bilhaṇa in his Vikramāṅkadevacaritaṁ. For details, see

Nilakanta Sastri, Cōḷas, pp. 294–295.48 For a review of the available accounts, see Nilakanta Sastri, Cōḷas: 285–305.49 Kaliṅkattupparaṇi, ed. A. V. Kannaiya Nāyuṭu (Chennai: South India Saiva Siddhanta Publishing

Works, 1944), see chapter on the royal lineage (irāca pārampariyam), which seems to make no mention of any successor between Vīrarājendra and Kulōttuṅka.

50 Ibid., v. 258 ff.51 The wider literary patronage of Kulōttuṅka I and his immediate successors in Tamil – with almost

all of the important Tamil works of the Chola period composed during these reigns, including not only works like Cayaṅkoṇṭār’s Kaliṅkattupparaṇi and Oṭṭakūttar’s Takkayākapparaṇi and Mūvarulā, but Kampaṉ’s Irāmavatāram and Cekkiḻār’s Periyapurāṇam – suggests a remarkable and unprecedented attempt to raise the royal profi le within the literary circles of the Chola kingdom.

52 On the extensive building activities at the site during the reigns of Kulōttuṅka I and his successors, see Paul Younger, The Home of the Dancing Śivaṉ: The Traditions of the Hindu Temple at Citamparam (New York: Oxford University Press, 1995), pp. 98–104.

53 Hermann Kulke, ‘Functional Interpretation of a South Indian Māhātmya: The Legend of Hiraṇyavarman and the life of the Cōḷa king Kulottuṅga I’ in Hermann Kulke, Kings and Cults: State Formation and Legitimation in India and Southeast Asia (Delhi: Manohar, 2001), pp. 192–207.

54 Kulōttuṅka I is known to have issued orders at a number of palace-cities throughout his realm. See Nilakanta Sastri, Cōḷas, p. 332.

55 ARE 29 of 1908, also, pt 2, pp. 65–66. Compare this with the Bengal Asiatic Society’s plate of Candradeva and Madanapāladeva, IA vol. 18 (1889), p. 11.

Daud Ali 29


56 See Nilakanta Sastri, The Cōḷas (Madras: University of Madras, 1955), p. 325; H.C. Ray, Dynastic History of Northern India I (Delhi: Munshiram Manoharlal, 1973, 2nd ed.), pp. 530–531; Roma Niyogi, A History of the Gāhaḍvāḷa Dynasty (Calcutta: Calcutta Oriental Book Agency, 1959), pp. 81–82.

57 The fi rst kings mentioned in the Gāhaḍvāla genealogies, Yaśovigraha and Mahīcandra, are given very modest titles and on this basis it has been suggested that they may have been subordinate to the Kalachuris of Cedi, asserting independence only after the fall of the powerful Kalachuri king Lakṣmīkarṇa (1041–1070 AD). Niyogi, Gāhaḍvāla Dynasty, pp. 38–42.

58 ARE 29 of 1908, pt. 2, pp. 65–66.59 It has also been noted that a Buddhist monk by the name of Vāgīśvararakṣita hailing from the Chola

country is mentioned (along with another monk from Orissa) as a petitioner for a land grant to the Sangha in the Set-Mahet plates of the Gāhaḍvāla king Govindacandra in AD 1129 (EI 11, no. 3). See Nilakanta Sastri, Cōḷas, p. 325 and Niyogi, Gāhaḍvāla Dynasty, p. 82.

60 See ARE 1927, pt 2, pp. 19–21 and, among others, S. R. Balasubrahmanyam, Later Chola Temple: Kulottunga to Rajendra III (Fardidabad: Mudgala Trust, 1979), pp. 130–133. For inscriptions at the site, see ARE 229–332 of 1927.

61 See the Udayendiram plates of Pṛthivīpati II Hastimalla where a Chola genealogy in Sanskrit is followed by a Western Ganga praśasti, brief details of the order in Sanskrit, and then by a Tamil operative portion dated in the reign of the Chola king, SII 2.76, p. 383.

62 Because the inscription is in Grantha characters, we may assume that the engraver was certainly south Indian and not himself a member of the Gāhaḍvāla court, since Gāhaḍvāla inscriptions are all in devanāgarī script

63 Again, in the case of the Udayendiram plates, the date occurs only in the Tamil portion of the grant just before the details of the land grant are given.

64 ARE 524 and 525 of 1994 and 315 of 1965.65 Thyagarajan, ‘Gangai and its Region’, pp. 181–82.66 Vikramacōḻa is found residing in Chidambaram in his twelfth year SII 7.788.67 Vikramacōḻa is found residing in Muṭikoṇḍacōḻapuram in his fourth year, SII 22.168.68 Nilakanta Sastri (Cōḷas, p. 332) claims that this town was current in the time of Kulōttuṅka I

(presumably founded by him after his title Vikramacōḻa), but the record locating the king there which he attributes to Kulōttuṅka I is actually a record of Kulōttuṅka II (SII 7.461) and should join several others which place the latter king there in the third year of his reign, SII 7.460, ARE 271of 1915 and 533 of 1921. Vikramacōḻa is found residing there as well, see SII 5.458.

69 Nilakanta Sastri, Cōḷas, p. 378.70 See ARE 321 and 322 of 1965. Suffi xes like āḻvār tevar, mateviyār found in the names of the

donors of these inscriptions often refer to members of the royal household or their intimates.71 Pichard, Tanjavur Bṛhadīśvara, pp. 15–16. According to Māravarmaṉ Sundara Pāṇḍya’s

meykkīrtti, the Pandya king performed an ‘anointment of heroes’ (vīrāpiṣekam) in the coronation hall (apiṣekamaṇṭapam) of the Cholas at Āyirattaḷi before moving on to Chidambaram. See S. Krishnaswami Aiyangar, South India and her Muhammadan Invaders (London: Oxford University Press, 1921), Appendix A, p. 208.

72 SII 4.524.73 On the complex events of this period, see Nilakanta Sastri, The Cōḷas, p. 421 ff.74 SII 6.44.75 See Nilakanta Sastri, The Pandyan Kingdom (Madras: Swathi Publications, 1972), p. 158ff. 76 SII 4.522, 525. The former on the eastern side of the shrine facing the stairway leading to the

ardhamaṇḍapa. See Figure 4.77 SII 4.523. Nilakanta Sastri considered the conquests mentioned in Māṟavarmaṉ Vikrama Pāṇḍya’s

inscriptions, which would seem to take place after the last known records of Rājendra III, to be ‘mere poetry’, Nilakanta Sastri, Pandyan Kingdom, p. 168.

78 ARE 33 and 34 of 1908.



79 ARE 31 of 1908. I was unable to locate this inscription on the main shrine, but it may correspond to an identifi ed inscription on the mahāmaṇḍapa. See Figure 4 and Appendix.

80 ARE 320 of 1965 on the north wall of the mahāmaṇḍapa, ARE 30 of 1908 and SII 4.526 on the base of the west wall of the central shrine, and SII 4.530 to the right of the entrance on the north wall of the main shrine (the last of these could not be located).

81 In a remarkable order inscribed in nearly twenty different locales throughout the erstwhile Chola kingdom dated in 1517, the Vijayangara king Kṛṣṇadevarāya mentions a large number of temples in the Chola region which were to receive support – and among them was Gangaikondacholapuram. See EI 25.32 for a copy of the order at Punjai.

82 This applies to three of the fi ve Vijayanagara inscriptions identifi ed on the main temple, with the other two on the northern and southern sides of the mahāmaṇḍapa.

Daud Ali 31


The Epigraphical Legacy at Gangaikondacholapuram: Problems and Possibilities

Documents