Version classiqueVersion mobile

La Bhakti d'une reine

Charlotte Schmid

Annexe I. Résumé Anglais

Texte intégral

The Bhakti of a Queen, Śiva at Tirucceṉṉampūṇṭi

1The Bhakti of a queen draws on the iconography and epigraphy of South India to investigate the establishment in the Tamil land of the deities of Bhakti, that is, Viṣṇu, Śiva, as well as the female deities sometimes gathered for convenience under the name of “the Goddess”. From North India to the tip of the peninsula, between texts and archaeological material, the work conducted in situ and the interlinking of several types of documents (texts and carved tradition for the most part) has guided this study.

2The book has been structured by the corpus itself. After the introduction and a chapter (1) presenting the material, the reader is taken on a tour of the temple to scrutinize the four deities represented there: Viṣṇu (chapter 2), Brahmā (chapter 3), and Śiva considered together with the many forms of female deities (chapter 4). In chapter 5, the study of the queen and donor Māṟampāvai allows for these gods to be united in the devotion towards a specific place, that is an essential part of “the Bhakti of a queen”.

3The introduction explains how research on the ground carried out in a local site like the one here presented fits into the broader perspectives of the construction of Hinduism, relations between the royal and the local and the kernel of regional identities.

4The temple to the god of Tirukkaṭaimuṭi is the base on which the present study is built. This temple consecrated to a local form of Śiva is provided with many inscriptions and sculptures still in place. Between the 7th and the 9th c., one of the hymns of the Śaiva Tamil anthology was composed to honour the deity of Kaṭaimuṭi. The poem, presented in Annex III, is the first known testimony of devotion to a deity named the Lord of Kaṭaimuṭi. Through his name, the deity is attached to a precise place called today Tirucceṉṉampūṇṭi, situated in the Kāvēri delta, 12 km northeast of the contemporary Trichy. The most ancient vestiges of the site are of the 9th c. and initially we seem to be dealing with one of the first temples built in the kingdom of the Cōḻas. The glory of these kings in the Tamil country overshadows the achievements of other dynasties, including that of the Pallavas who reigned over the north of the country well before the establishment of the Cōḻa empire. The latter may be said still to rule today through a Tamil identity very much present in Tamil Nadu, where the amount of Tamil inscriptions dated to a Cōḻa regnal year may account for the importance the Cōḻas themselves are given. In any case, this dynasty is part of an identity phenomenon where language and history merge. In the last part of the 20th c. several people killed themselves to support the Tamil language, recognized since 2006 as one of the classical languages of India, that is, on a level with Sanskrit. The Cōḻas, with “their” art and “their” inscriptions can be considered as the first massive archaeological evidence of something that looks like a Tamil culture. One of the issues here is to question the precise role played by the Cōḻas in the establishment of such a cultural background as well as to cast light on the part taken by other agencies.

5I have established for my study of the small temple known today as the Caṭaiyar of Tirucceṉṉampūṇṭi two corpus, both given in the annexes. One is the epigraphical corpus of the site (Annex IV). The other is iconographic (Annex V). Those two corpus are the main base on which the study is founded. A special emphasis is put on the Tamil inscriptions as none has yet been translated in its entirety. While their data has not been accessible, their study has proved to be of the utmost importance. It enhances the part played by the first donatrix of the temple, the “honorable Aṭikal Kaṇṭaṉ Māṟampāvai”, “mahādevī (chief-queen) of Nantippottaraiyar, from the famous family of the Pallavas”. Her devotion is studied to cast light not only on what Bhakti may have meant in the 9th c. southern part of India but also on the nature of this phenomenon in early Medieval South India.

6The intellectual element, more or less important in the devotion and asserted first in Sanskrit in the Bhagavadgītā, composed perhaps as early as the 2d c. BC, is often contrasted with more emotional forms of devotion attested in the south of the Indian peninsula, initially in Tamil texts. The poems of Tamil Bhakti, that is, the Tēvāram (the exact translation of which is still debatable), a collection of hymns in honor of Śiva, and the Nālāyira Tiviyapirapantam, the “Four Thousand Sacred Stanzas”, composed in honor of Viṣṇu, echo the voices of devotees that can be heard in the inscriptions engraved in the Tamil territory. I assume that the devotees shape their personal deity in a double movement of definition of themselves (poet or donor) and their god to which I give value as characteristic of Bhakti. The Tamil hymns are, in my opinion, rather close to the tradition of carvings whose originality is not always recognized as such. Their inventiveness is made somewhat clearer by the normative discourse found first in the Āgamic school on which secondary literature heavily draws; it cannot be used for the times dealt with in this book since the Āgamas were not in existence then, at least not under the form we know today. One hymn of the Tēvāram was originally attached to the temple of Tirucceṉṉampūṇṭi but this is not acknowledged in the present editions of the anthology. The loss of the link between the text and the site demonstrates how little is known of the history of the texts themselves and how much is to be gained in this domain by a precise investigation of inscriptions.

7The vestiges of the temple dedicated to the god of Tirukkaṭaimuṭi belong, moreover, to a crucial intermediary period, the 9th c. and the beginning of the 10th c., between the few royal foundations of the Pallava and the numerous temples due to local initiatives of the Cōḻa period—and before the first royal foundations of the Cōḻas. New programs, epigraphical and iconographic, were put in place at that time and were to dominate Medieval India. The site of Tirucceṉṉampūṇṭi is both an exemplary sample of the period and a site so specific that it demonstrates how much the area where a temple is built, including its physical characteristics, can shape iconography and epigraphy. The relations between the royal and the local are here precisely documented. The material provides new answers to questions about, for instance, the role kings, queens, royal and local courts, merchant communities, Brāhmin assemblies and many others have played. Does this material available in Tirucceṉṉampūṇṭi give sufficient substance to order an artistic production along with a directional, supposedly royal, pattern as K. A. Nilakanta Sastri thought? Is it the product of an incorporative model of kingship such as the one conceived by Burton Stein? Or is it a sample of the Kāvēri-style convincingly put forward by Padma Kaimal?

8The question of the link between royal power and local temple is embodied at Tirucceṉṉampūṇṭi in the donatrix after whom the book is named, spouse of a Pallava king, and princess of a local dynasty she patronizes with gifts sometimes recorded in Cōḻa regnal years. Her endowments are made in two Śiva temples and in one of a female dynastic deity. Thus “Māṟampāvai” makes it possible to figure out the contours of relationship between rulers and regions during a period crucial for the formation of so-called Cōḻa art, at the crossroads of dynasties—Pallava, Cōḻa, Muttaraiyar, Irukkuvēḷ, Paḻuvēṭṭaraiyar—, territories (local, regional, dynastic and royal), and cults.

9Chapter 1 (“The Caṭaiyar of Tirucceṉṉampūṇṭi”) is devoted to a presentation of the material on which the book is constructed. The Caṭaiyar of Tirucceṉṉampūṇṭi is located in the heart of the Cōḻa territory to the northeast of the oldest capital of the Cōḻa dynasty, Uṟaiyūr (located in the suburbs of Trichy). Now abandoned, the temple is organised around a Śaiva shrine partially reassembled in the 1970s by the Archaeological Survey of India, in fields lying between the Kāvērī and one of its tributary canal, the Koḷḷiṭam. This is the territory of a god of the Iṭaiyāṟṟunāṭu, the “country between the rivers”, as it is described in the Cōḻa-period inscriptions from the site itself.

10The temple opens to the east. The three main niches of the sanctuary are occupied as follows: a standing Śiva holding a vīṇā to the south, a standing Ardhanārīśvaramūrti to the west and a standing Brahmā to the north. This brief description shows that the iconographic scheme of the temple of Tirucceṉṉampūṇṭi does not correspond to the programs of the earlier royal foundations of the Pallava, or to the most common models of the Cōḻa period. Can we uncover the logic that presided over the organization of the figures? Is it a form of transition between two periods, a scheme that points toward a regional culture? The exploration of the way the temple was built is completed with a review; the concept of Cōḻa art itself has been developed to contrast the theoretical inputs with the concrete data. To give as precise as possible a date to the temple has been the aim of previous authors. The goal of this book is different. I do not think it is possible to give a precise date to local foundations which were constantly being built, renovated, enlarged, etc. as is proved by the study of the archaeological clues. In the same way the royal model is not only often distant but always many-sided as is demonstrated by the epigraphical corpus.

11This corpus has thirty-two inscriptions, of which only one was previously partially translated into a Western language. They are here edited and translated into French (annex IV). This corpus is peculiar for its temporal brevity. The inscriptions of Tirucceṉṉampūṇṭi were engraved for only about a century, which is rare in the Tamil country. Like many others, this corpus has been used in the secondary literature on the art of the Cōḻas to locate the temple in time and constitute it as a product of dynastic art. It is shown here that the epigraphs were considered in too specific a manner in these perspectives. The corpus is indeed mutilated as the inscriptions dated in Pallava regnal years are either misinterpreted, or not recognized as such. Engraved mostly on pillars today separated from the temple, these inscriptions belonged to a vanished sanctuary, presumably built largely in perishable materials. What may be called the second part of the corpus is composed of inscriptions dated after the regnal years of Cōḻa kings and these, engraved as they are on the base and the walls of the temple, are still visible today.

12The constitution of what has been called the “iconographic corpus” of the temple, which corresponds to the presentation and the identification of small carved panels forming two superimposed friezes one above the other all around the base of the temple, is then introduced. Annex V presents these panels, together with references to several of the texts which allow for their identification. They represent exploits of Viṣṇu, Śiva and the goddess, as well as scenes of music and dance, and decorative foliage. Forming a whole narrative in the case of Rāmāyaṇa, these friezes are inspired by Epic and Purāṇic literature, but also, I think, by more openly devotional texts of which the Tamil Tiviyapirapantam is probably the closest representative. Giving access to a very large repertoire of divine figures, they have thus an importance that their size and their location do not at first glance suggest.

13Chapter 2 (“Love and battle: Viṣṇu on the banks of the Kāvēri river”) considers the set of representations of Viṣṇu in the temple. This deity is not mentioned in the epigraphical corpus and his representations are met with only on the base of the temple. It is however a major figure: in the friezes of panels, forms of Viṣṇu are much more numerous than those of any of the other deities carved here. Moreover, some are so specific that they constitute the temple to the great god of Tirukkaṭaimuṭi as fundamental evidence of the composition of the Vaiṣṇavism of the Tamil land. To find the equivalent in the texts, one must take into account not only versions of this or that episode of the legend of Viṣṇu in Sanskrit Epics and Purāṇas but Caṅkam literature as well, that is, the most ancient known Tamil texts, and the hymns of Tamil Bhakti.

14From ground level up, the first frieze encountered on the base of the temple consists of panels devoted to the first three books of the story of Rāma (there is only one panel which might belong to the fourth book). This is the first known representation in the Tamil country of a whole narrative cycle of representations consecrated to Rāma. The versions of the Rāmāyaṇa of Vālmīki, a Sanskrit text, seem to be the most relevant textual correspondents. Still, in the scenes that define a country of rivers, corresponding closely to the nomenclature in the inscriptions of the administrative territory where this shrine stands as that of “the land in between the rivers”, those that depict a territory bounded by a sea filled with demonesses or where one may marry women, can be read clearly the phenomenon of adaptation to the local found in many other parts of the temple.

15The representations of the other major avatāra of Viṣṇu, Kr̥ṣṇa, proved to be particularly original in Tirucceṉṉampūṇṭi. This young god is mainly a slayer of monsters in Sanskrit literature. Here he sometimes endorses costumes of Tamil origin and, among well-known exploits of Kr̥ṣṇa, several panels point to some of the most distinctive figures of southern Vaiṣṇavism. On the one hand, the first known figure of Kr̥ṣṇa playing the flute has been carved here. This divine musician is inherited from the profane flautist of the most ancient Tamil literary tradition, that of the Caṅkam. He is destined to be subsequently represented in the whole of India, where it is assumed in this book that he was disseminated mainly through the Bhāgavatapurāṇa, a major Purāṇa written in Sanskrit but inspired by the Tamil Tivyapirapantam. On the other hand, Kr̥ṣṇa dances with pots, according to an iconography that appears only in the Tamil country and which could be a sign of the specific importance given there to dance, and more precisely to a type of folk-dance. Finally, the identification of one of the panels proves to be difficult. It can correspond to a wellknown episode from the ancient Sanskrit texts, where Kr̥ṣṇa uproots trees by pulling a mortar between them. But it may also correspond to an enigmatic episode known only from southern texts, where Kr̥ṣṇa breaks what the Tamil texts refer to as a kuruntu, which may be a shrub or a branch. It has been demonstrated by F. Hardy that the motif of the theft of clothes by Kr̥ṣṇa sprang from the kuruntu episode. The small panel of Tirucceṉṉampūṇṭi might thus represent one of the steps leading to the constitution of the theft of clothes first attested in Tamil literature: a north-Indian legendary element so greatly transformed that it came to constitute a new mythical theme in the South of India—what I call a shadow-motif.

16The whole formed by the representations of Kr̥ṣṇa in the Śaiva temple of Tirucceṉṉampūṇṭi thus reflects the fruitfulness of the contact between the most ancient Tamil literature and legends related in Sanskrit texts to lead to a new iconography. In particular, the appearance and evolution of the carved representations plead for a reassessment of themes of northern origin in Tamil literature, first in the Cilappatikāram, “The Tale of the Anklet”, one of the Tamil epic (5–7th c.) and, more broadly, of the beginnings of Kr̥ṣṇa devoted cults in the Tamil land. I believe that in many cases the literary tradition has played an important role.

17Chapter 3 is devoted to Brahmā (“In the northern niche of the shrine, Brahmā”). This god is the counterpart of Viṣṇu in many Śaiva images, starting with those of the Somāskandamūrti and the Liṅgodbhavamūrti, religious manifestations par excellence of Śiva in the Tamil land. Such a link with Viṣṇu is clearly visible at Tirucceṉṉampūṇṭi in the representation of the Rāmāyaṇa which begins with a figure of Brahmā. But earlier associations and those documented elsewhere are not represented in Tirucceṉṉampūṇṭi where some choices were made. Carved in very high relief, the statue of Brahmā which guards the north side of the shrine is a body of evidence quite other than the low-relief of the small panels on the base of the temple. The analysis highlights the complexity of relations between Brahmā and Śiva. In the area and during the period where the temple of Tirucceṉṉampūṇṭi was built those two deities present the specificity of a sometimes strongly marked rivalry.

18Brahmās, such as the one standing in the northern niche of the sanctuary of Tirucceṉṉampūṇṭi are found in many of the northern faces of the Śaiva temples of the Cōḻa age in the Kāvēri delta, as well as in subsequent periods and in a larger area. These Brahmās are, however, one of the notable innovations of the iconographic programs of the Cōḻa period. If Brahmā appears in the royal foundations of the Pallavas, he is much more subordinated to Śiva there and is a far more minor figure than he appears to be in Tirucceṉṉampūṇṭi.

19During the Pallava age, one can constitute a dominant iconographic model of Brahmā. The deity of the Cōḻa period, three-headed, beardless, with four arms, carrying a rosary in his upper lefthand shares characteristics with the dominant Pallava model. But the latter period presents numerous iconographic variations that express the complexity of the iconographic history of a god whose figures are very ancient. There were Buddhist Brahmās, included on the sites of Andhra Pradesh, land of origin of the Pallavas. On the other hand, in the Tamil country, the relations of Brahmā with Skanda-Muruku, the son of Śiva in the Hindu context, are close.

20The analysis focuses then on the figure of the ascetic teacher, be it a form of Brahmā, Skanda or Śiva himself. It reassess, inter alia, the identification of Skanda in what is called the Trimūrti cave in Mahābalipuram, to enhance the importance of the Hindu tripartite schema to which Brahmā belongs. It traces the emergence of Brahmā in the north niche of the Śaiva shrines, firstly, with the iconographic programs based on the Trimūrti principle, from the 9th c., secondly, with the disappearance of a form of Śiva as both fighter and ascetic placed in the north of the first royal Pallava temples, and, thirdly, with the importance given to the ablution rituals in the Śaiva worship, clearly shown in the epigraphy of Tirucceṉṉampūṇṭi.

21Eventually, the importance of Brahmā in the surroundings of Tirucceṉṉampūṇṭi is attested through peculiar statues of a seated Śiva carved in the round and provided with four heads, mimicking the iconography of Brahmā closely enough to be mistaken for him. In my opinion, those carvings express a kind of competitiveness of Brahmā with Śiva, echoed in the site of Kaṇṭiyūr, some 15 km from Tirucceṉṉampūṇṭi, where a Śaiva Tamil tradition, already effective in the Tēvāram, locates the beheading of Brahmā by Śiva, of which the only known representation is encountered in the Pallava Kailāsanātha of Kāñcīpuram.

22Chapter 4 (“The Mahādeva of Tirukkaṭaimuṭi”) examines the evidence that pertains to the god Śiva at Tirucceṉṉampūṇṭi, where the Lord is often paired with a female figure who may be considered as his wife or his complement, but also appears as an autonomous goddess. Referring to those women who gave to the temple, chapter 5 will allow these mythological figures to come into resonance with characters that can be described as historical. Such resonance does not appear here as a bias introduced by the analysis and the links between Śiva and the goddess have much to tell us about the importance of the shrine when it was active. They illustrate close correlations between devotion and policy, devotion and regional identity, definition of a particular deity and assertion of the very self of the devotee in his Bhakti.

23All the epigraphs of the site record grants made to the same god, bearing a name that varies little, tirukkaṭaimuṭi mahādeva. It turns out that the temple erected in honour of the god of Tirukkaṭaimuṭi is what is called a temple of the Tēvāram, that is, a temple dedicated to a god celebrated in one or several of the hymns of the Śaiva anthology. This is a particular case where the link between a hymn composed by Campantar (1.111) in honour of the god of Kaṭaimuṭi and the temple of Tirucceṉṉampūṇṭi has been lost. The hymn is today attached to another temple but the god of the hymn and the Tirucceṉṉampūṇṭi epigraphs use the same name, Kaṭaimuṭi, rare and difficult to analyse, of which no trace is found on the site to which the 1.111 poem is now attached. The original name of the site might have designated a god of the “confines”, as Tirucceṉṉampūṇṭi was situated on the southern border of the Pallava kingdom. In the Periyapurāṇam composed by Cēkkiḻār in the 12th c., the link between the site and the hymn is still productive. But called there “Caṭaimuṭi” (a name of the deity which appears in some of the latest inscriptions of Tirucceṉṉampūṇṭi), the place-name Kaṭaimuṭi was subsequently reinterpreted as one of the names of Śiva himself, the one with the matted hair (caṭaimuṭi).

24Such a study underlines the possible input of the epigraphical data in the assessment of the history of the corpus of Tamil Bhakti. As regards Tirucceṉṉampūṇṭi more specifically, the characteristics of the poem dedicated to Kaṭaimuṭi in the Tēvāram are very fitting to a displacement of the place, as the name of the abode of the deity seems to be the single specific element of the hymn.

25In this perspective, the inscriptions of the site of Tirucceṉṉampūṇṭi and the Tamil Śaiva Bhakti corpus seem to attest the same phenomenon: the emphasis being gradually put on Śiva for whom a specific place provides a possible embodiment. This need not necessarily be seen as the “Śaivaisation” of a territory inhabited by local deities gradually assimilated to Śiva, but may perhaps be considered as a more sectarian presentation of Śiva. In all cases, if the place is not presented specifically in the Tamil hymns, these poems allow the places they name to be part of a broad religious network, which was a Śaiva one.

26Offerings to the god recorded in the epigraphy give the same reading of the god and his abode; they are given for the bathing and feeding of a lacto-vegetarian deity. Such offerings are very common in Tamil epigraphy but appear distinctive nevertheless in a Tamil country where Caṅkam literature may refer to mountains and rituals as implying the shedding of blood. By contrast, the donations to the deity of Tirukkaṭaimuṭi associate their god with an extensive network, spread not only in the South of India, but also in the whole of India and even in Southeast Asia.

27It appears also that donors, whether they give a date with a Pallava regnal year or with a Cōḻa regnal year, hand over gold rather than land to the deity. This is not a god ruling a vast territory. The abandonment of the temple is consistent to a degree with the absence of land, even if the deity somehow consumes land through the numerous food offerings made to him.

28The analysis of the stone-figures of Śiva focuses on three forms.

29Firstly, Tirucceṉṉampūṇṭi hosts one of the most ancient known representations of a Śiva dancing the “Bliss-dance”, with which the Cōḻa dynasty entertained a specific link from, at least, the second half of the 10th c. Two questions are raised here. One pertains to the creation of the image, the second to its links with the site of Citamparam, today considered as the place of origin of the cult developed around this figure. It is proposed here that the link with the royalty of the dancing figures of Śiva is more ancient than the association of this figure with any worship in Citamparam. This hypothesis is based on the analysis of the formation of the iconography and the symbolism associated with this form. The dancing figures of Śiva are certainly prominent in the Pallava royal foundations, where they were imbued by a royal symbolism which I assume to have been taken on by the Cōḻa dynasty, while some amongst them inspired the concrete form itself.

30Secondly, the south niche of the sanctuary shelters a walking Śiva holding a vīṇā. Such a musician of the Cōḻa age appears as one of the variants of the mendicant form of Śiva, who is sculpted under two different iconographies (the one practised during the Pallava period and one of those which appeared during the Cōḻa period) below the musician on the base of the temple. To me, the most appropriate perspective is to consider this musician as inspired by the Dakṣiṇāmūrti, the teaching, seated form of the god in the South, and a Bhikṣāṭanamūrti, the mendicant form, walking, which are both depicted in a number of Pallava foundations. These two figures are also associated here as Dakṣiṇāmūrti is represented on the crowning above the musician, topped by the figure of the dancing Śiva. Already associated in Pallava times and linked by a peculiar narrative in the Citamparam myth, these three forms of Śiva stand between the two distinct types of foundations that are the royal Pallava temples and the very particular complex of Citamparam.

31Thirdly, the study focuses on the Ardhanārīśvaramūrti which was on the west (now in the museum of Ceṉṉai). This form had a peculiar importance in pallava times, when most examples are Vīṇādharamūrtis carved in the round. In Tirucceṉṉampūṇṭi, this Pallava cult-form seems to have been split into a Vīṇādharamūrti and an Ardhanārīśvaramūrti. The latter exhibits characteristics of the Cōḻa age that attest to new artistic and symbolic trends. First, he is standing, while Pallava Ardhanārīśvaramūrtis are seated and the Vīṇādharamūrti presents a stone incarnation of the deity of the Tamil Bhakti poets in an area replete with Tēvāram temples. Secondly, the Ardhanārīśvaramūrti is provided with only three arms. Such iconography resonates with the story of Kaṇṇaki the powerful heroine of The Tale of the Anklet, who tears her left breast out to set an entire city on fire. Cōḻa-period Ardhanārīśvaramūrtis are known where, contrary to what a wife should do, the female part stands to the right of the deity, allowing for a close identification with Kaṇṇaki.

32The unique arm with which the goddess is provided nourishes thus the ambiguity of empowered women, who stand between goddesses and human females. A similar ambiguity is well illustrated on the base of the temple. In Tirucceṉṉampūṇṭi the pan-Indian undifferentiated character of the iconography of females takes a particular relief. Ogresses and goddesses, humans and she-devils intermingle freely.

33What may be considered to be of typical southern flavour, what could have come from North India, and, once again, the fertility of the meeting between two distinct traditions allow for conclusions on the creativity of what was a buffer area between several kingdoms, Pallava, Paḻuvēṭṭaraiyar, Pāṇṭiya, Irukkuvēḷ, Muttaraiyar and Cōḻa.

34Entitled “The Bhakti of a Māṟaṉ Princess” Chapter 5 ties up the many threads. In the attempt to define the contours of the temple of Tirucceṉṉampūṇṭi the lines of an art that might be called “Cōḻa” emerge gradually, from the royal temples of the Pallavas to a territory using regnal years of Cōḻa kings, where regional, possibly dynastic iconographies appear as significant milestones. In the 9th–10th centuries, before the foundations established by Cōḻa royal figures (the ones of the queen Cempiyaṉ Mahādevī, active in more than thirty sites spread throughout the territory of the Cōḻas, or the Rājarājeśvara of Tanjore established at the beginning of the 11th) local foundations transmitted the royal Pallava heritage in their own creative way.

35In this perspective, the role of female donors proved decisive. Before the establishment of the Rājarājeśvara of Tanjore, the queens, spouses of kings or local dynasts, were indeed responsible for “royal” art. The first donatrix of Tirucceṉṉampūṇṭi is one of them. She is particularly representative of this category and her contribution is examined here.

36At the beginning of the Cōḻa age, the site of Tirucceṉṉampūṇṭi recorded its share of donations from important, dynastic figures—two members of the family of Pallava itself, a woman from the Paḻuvēṭṭaraiyar court, a chief of the Koṇṭanāṭu, one who is perhaps an Irukkuvēḷ and in any case the “daughter of a chief” (vēḷār makaḷār). Here is a place where the dynasties merge. Bearing the name of “Aṭikal Kaṇṭaṉ Māṟampāvai”, the first female donor crystallizes their meeting. This “Lady Māṟaṉ” or “Princess Māṟaṉ” or “Daughter of the Māṟaṉ” bears a name-title associated with all the dynasties of the region Paḻuvēṭṭaraiyar, Irukkuvēḷ, Muttaraiyar and Pāṇṭiya. Even if she is probably a princess of Muttaraiyar origin, as will be shown, in inscriptions written in the first person singular, she claims to be the “great queen of Nantippōttaraiyar, of the illustrious family of the Pallava”. This demonstrates that she considers her identity established as that of the wife of the Pallava Nandivarman III. She never appears in any Sanskrit or Tamil inscriptions directly commissioned by a Pallava king. Māṟampāvai appears implanted instead in the region of Tirucceṉṉampūṇṭi: she made donations in two different temples of the nearby site of Niyamam (6 km from Tirucceṉṉampūṇṭi) which was most probably the capital of the Muttaraiyars. In the 8th c., a Muttaraiyar king founded there a temple to a goddess to whom Muttaraiyars and Pallavas paid tribute as engraved in records that are among the first known to be written in literary Tamil. Māṟampāvai makes donations to this goddess but also to a Śiva of Niyamam. With an inscription on the pillars of the temple dedicated to the goddess of Niyamam, a place usually reserved for men, Māṟampāvai makes it possible to explore the link between the stone of masculine heroism and devotional activity which covered the Kāvēri delta with stone temples. The Cōḻa regnal years sometimes used to date her inscriptions raise questions: do they attest dynastic fluctuations or local freedom?

37The first of the females to appear in the inscriptions of Tirucceṉṉampūṇṭi, Māṟampāvai belonged to the category of female donors who constitute half of the epigraphical corpus. Present at the very beginning, and until the end, of this corpus, these women insert the Caṭaiyar into a devotional network operative within and beyond the Cōḻa. For her part, Māṟampāvai participates in the Pallava and the Cōḻa circle, and especially in a regional domain, that of the Muttaraiyar. In these three realms she appears as a pioneer in many respects. Several other “Pallava local queens” are known who made donations in the 9th c. Their grants to the Śiva of a specific place are recorded in Tamil. Involved in donations of rituals, at that time of lamps which are so numerous in the Cōḻa period, their gifts are different from those of the previous Pallava royal figures and, like Māṟampāvai, the other local Pallava queens do not appear in Sanskrit epigraphy. They foreshadow the case of Cōḻa queens who do not appear either in Sanskrit epigraphy (royal metal tablets), whereas they are very much present in Tamil stone inscriptions. Māṟampāvai stands as peculiar however: other local Pallava queens are known from a single inscription; The titles of Māṟampāvai have no equivalent in other epigraphs of “Pallava” local queens. She appears thus as a true spouse of Nandivarman III.

38This Pallava king who was the husband of Māṟampāvai appears too in Tirucceṉ-ṉampūṇṭi and Niyamam. He commissioned inscriptions in Sanskrit as well as in Tamil. A royal eulogy in Tamil (the Nantikkalampakkam) was composed in his honour. The Pallava couple of Tirucceṉṉampūṇṭi thus fully participates in the development of a southern crucible, where various influences give rise to an original culture.

39The analysis of the neighbouring Niyamam sanctuaries where this couple made endowments illuminates many aspects of the Caṭaiyar temple. It highlights another phenomenon, which seems to be a process of Śaivaisation of the land or, if one adopts a different perspective, integration of local deities in a religious scheme of greater magnitude. The goddess of Niyamam was a very ancient and important deity of the area. Woman and wife, Māṟampāvai expressed her devotion to the Niyamam goddess whose beauty was related to a synthetic form of goddess where a warrior form meets that of a dispenser of wealth. The pillars of the shrine of this goddess were transported to a Śaiva temple in today’s Centalai, as if this site were fit to absorb the sacredness of the goddess—and the beauty of her temple, as the inscriptions of these pillars are engraved in a calligraphic script which has no known equivalent.

40The cult to the goddess seems, therefore, to have disappeared while the Śiva cult was expanding. The pillars which are the first physical evidence of the temple to Śiva in Tirucceṉṉampūṇṭi have the same shape as those of the temple to the goddess of Niyamam and a direct line can be followed from one temple to the other. The alliance between Nandivarman III and Māṟampāvai takes a mythological relief. Might Śiva, in the manner of a Pallava king marrying a princess of the region, have allied himself with the goddess of Niyamam? The Pallavas did not establish temples to goddesses. Much of the evidence recording the importance of the female deities under the Pallava rule lacks a proper royal horizon. These are steles, representing ritual suicides in honour of goddesses which are sometimes found on royal Pallava temples but the steles are ’local’ by style, form and dissemination (hundreds of them are reported). They disappear in the course of the 10th c., when goddesses of the same iconography take a permanent seat on the north face of Śaiva temples.

41Śiva was not depicted on that particular type of steles. But from the 9th c. temples dedicated to Śiva are built in stone in many places in the Tamil territory. As in Tirucceṉṉampūṇṭi donations made elsewhere record the establishment of a cult, organized according to the detail of the rite and the calendar of the year. These Śivas are called mahādeva, great god (a Sanskrit word) of a specific place, bearing a Tamil name and planted in the Tamil land. As an echo of these ’local’ inscriptions, the Tēvāram makes us ponder. The dates attributed to the Śaiva devotional anthology vary between the 7th and the 9th c., which means they were composed before the first Pallava royal Foundations, or were contemporary with the later ones, and with the first inscriptions of Tirucceṉṉampūṇṭi. One of the difficulties here is the absence of links between the Pallava royal temples and the Tēvāram temples. Tēvāram draws a map of the Śaivism in Tamil country comprising not less than 268 sites. Still, none of them corresponds to a foundation established by a Pallava king. Should we then envisage two rather distinct streams of Bhakti, one “royal”, and one “local”?

42The analysis of the data of Tirucceṉṉampūṇṭi allows for the joining of the two corpus of devotion in the Bhakti of a queen.

43The Pallava kings do not seem to be committed to addressing a Śiva of a place but rather created new places of worship for their god. The “foundation” of a temple such as Tirucceṉṉampūṇṭi is of another kind, to which the very structure of the Tēvāram poems corresponds. The anthology bears witness to the Bhakti of a place. The royal temples of the Pallavas played a key role in the emergence of the figure of Śiva in South India: contrary to that pertaining to the location of the shrines, the iconography created in these royal temples has specific resonances in the Tēvāram. The queens and, above all, the emblematic Māṟaṉ princess appear to be the missing link between these regal temples and the local foundations of the Cōḻa age, between royal and local Bhakti. When Māṟampāvai presents the management of her grants to merchants, who appear in other inscriptions as well, she delineates the originality of the model proposed by the queen, who does not rely on Brahmins for her endowments but on another devotional community having its own network.

44The conclusion focuses on “Local queens sowing the seeds of Cōḻa art”.

45Throughout this study, the question of dynastic art appears as linked to that of royal patronage, on the one hand, and to the stylistic identity of a region, on the other. They are both elusive in the current state of documentation and analysis which struggles to delineate the workshops but without definitive results. In assessing a synthesis such as the Caṭaiyar of Tirucceṉṉampūṇṭi, the temples built in the lands of the Pāṇṭiya, the Irukkuvēḷ, the Paḻuvēṭṭaraiyar, and the Muttaraiyar have to be taken into account. Poorly known and little studied, these shrines are still sometimes analysed on the basis of dynastic art.

46It might well be that these categories in the history of the art of the Tamil country are influenced by artistic events that shaped the earlier period, that of the Pallava age. The ascendancy of Pallava art which was almost exclusively royal is all the more important in that it is not often acknowledged. The ambiguity towards the Pallava heritage, marked by Sanskrit and North-Indian inspiration, may have been one element of the regional identity shaped from the 8th c. in the area of Tirucceṉṉampūṇṭi. However this may be, in the exemplary case of the Tirucceṉṉampūṇṭi site, the relationship between dynastic art and local foundations certainly resulted in the definition of a regional culture which participated in the creation of a dynastic one. Between the two capitals of the Cōḻas (Uṟaiyūr and Tanjore), the temple of Tirucceṉṉampūṇṭi was located in the heart of Cōḻa power. Being one of the oldest known in this territory it may be considered as a sample of Early Cōḻa art, not in the sense of a temple sponsored by the Cōḻa dynasty, but as an expression of a sufficiently characteristic cultural complex which could have been a central inspiration for the Cōḻa kings themselves.

47Weaving the links between the kings, from one dynasty to another in space and time, the daughters and wives of the rulers invite the recognition of the local by the royal, first at the end of the Pallava times and then during the Cōḻa period. The dual dynastic affiliation of queens who often indicate a family of origin and a clan into which they marry illustrates the vacuity of applying a dynastic label to their Bhakti. The custom of marriage of cross-cousins in force in the Tamil country strengthens it, as shown by P. Kaimal

48Ultimately, the foundations of the Cōḻa kings do not appear as models but as reflections of the foundations that preceded them. One may easily discern what the royal temples of Tanjore and Gaṅgaikoṇṭacōḻapuram owe to local foundations of the 9th and 10th centuries. There is no foundation inscription in these royal temples which records donations as local foundations do and the parietal epigraphy of the Cōḻa kings is the counterpart of the local, sometimes dynastic, epigraphy but on a much grander scale. On the other hand, if the dimensions of the temples of Cōḻa kings are exceptional and require specific architectural features and construction technique, the basic elements of the buildings, etc. reproduce local foundations on a magnified scale with regard to iconography; the use of the principle of the Trimūrti and the importance given to a specific dancing form of Śiva echo the creations of the local foundations. Eventually, the development of royal epigraphic praises engraved on stone in Tamil and their imposition on a large number of temples signed the reappropriation of the local by the royal at the same time as the myth of the Cōḻa king as builder was being forged on royal tablets engraved in Sanskrit.

49The art of the Cōḻas seems thus to amplify the art of a Cōḻa core territory to which many women, those females of the Dharmamahādevīśvara, the tēvaṉār makaḷ “Daughters of the god” (see Leslie C. Orr works) of Tirucceṉṉampūṇṭi, the dancers of the Rājarājeśvara, but also the queens praised in the chapels of the Kailāsanātha, Dharmamahādevī, the great queen Māṟampāvai, the female donors of the Irukkuvēḷ clan or Cempiyaṉ Mahādevī, daughter of the king of the territory of the Maḻa, wife, mother and aunt of several Cōḻa kings, have contributed while weaving a net of exchanges with their deities—which may be called Bhakti.

© Institut Français de Pondichéry, 2014

Conditions d’utilisation :


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search