Version classiqueVersion mobile

Deep rivers

Kannan M.
Jennifer Clare

Classical Tamil

Kāraikkālammaiyār: between her legend and her works

To the memory of Julien Vinson (1843-1926)

Texte intégral

  • 1 «Les études dravidiennes et particulièrement les études tamoules de 1900 à 1906», Année linguistiq (...)
  • 2 Tampi-purosottamap-pillai, Kāraikkāl Ammaiyār purāṇam, Madras, 1906, pp. III, 8, 165, 1. The copy (...)
  • 3 Murdoch, John, Classified Catalogue of Tamil Printed Books (first ed. 1865) reprint Madras 1968, p (...)
  • 4 «Sur les noms de Pondichéry et de Karikal» Revue historique de l’Inde française, vol. 4, 2nd part, (...)
  • 5 Lambezat, Bertrand, Karaikal Ammear, Patronne de la Ville de Karikal, Karikal, 1952, 26p.

1More than a century ago, Julien Vinson, the first French master of Tamil studies, published in the Revue orientale et américaine (new series, vol. 4, 1880, pp. 117-136) his translation, from the Periya purāṇam, of the complete version of the life of Kāraikkāl Ammaiyār (here after KA), of which Karavelane only wrote a summary for the first, 1956, edition of his book. Far from seeing her as a local saint, treasured in the region where he had spent his childhood, Vinson knew how to rank this text as “a very important contribution to the study of Saivism which is one of the Indian religions least studied in Europe. However it should receive more attention because it is the most important religious cult in the south of India and also because it has incorporated, through a very obvious artifice, local traditions, legends and deities.”1 That is why twenty five years later, Vinson published, a revised version of his translation in the Revue de linguistique (vol. XXXIX, 1906, pp. 258-275) after he had received a new edition, printed in Madras, of that same chapter of the Periya purāṇam by the director of “indigenous” studies at Calvé College in Pondicherry, under the title Kāraikkālammaiyārpurāṇam.2 Finally in 1920 he came out with some interesting new bibliographical information. Let us compare here what the J. Murdoch Catalogue was obliged to acknowledge in 1865: “Her poems are said to have been printed; but the compiler failed in his attempts to obtain sight of a copy”3 with Vinson’s description: “That first one [separate edition of that chapter of the Periya purāṇam] which was published in Madras in the year 4966 of the Kaliyuga (1865) is a square size in-8° of 20-39 pages which contains the text, an explanatory commentary, and religious poems attributed to Puṉitavati. The small book has three lithographic plates comprising seven pictures which depict the principal episodes of the saint’s story; these are drawings, the spontaneity of which compensates for their primitive craftsmanship.”4 Vinson adds one more important piece of information: “For about thirty years [he writes this in 1920], that is since the movement of renewal of Saivism, the festival of Puṉitavati has been celebrated with great pomp and ceremony, attracting a large number of devotees.” The remark links the famous “Mango fruit festival” in Kāraikkāl, Māṅkaṉit tiruviḻā, with the modern Saiva revivalism that haunts the beginning of the twentieth century and finds its expression in numerous publications. This information does not formally contradict the musing of M. B. Lambezat5 on the analogies with agrarian rituals of the summer solstice, as he puts back their origins to before KA herself, but it is a serious warning against any historical extrapolation from this festival, of which, on the other hand, he gives an excellent description that includes all the episodes of the puranic story staged over four full days, carefully distinguished by him from the paṅkuṉi ceremonies (March-April) evoking the beatification of the saint solely by the singing of her antāti, without any popular festivities. As we can see, the developments of the legend have a life distinct from the fate of the work.

History and legend

2The body of work itself supplies only very sparse information about its author. At most her name appears three times in four different works, each time in the final stanza, and under the unique form Kāraikkāl pēy (the demoness of Karaikkal). In both the two patikam it is in the 11th stanza, a sort of envoi or colophon added to a poem supposed to contain only ten stanzas: as in the case of Tēvāram, where this occurs quite often; critical minds will denounce and exorcise the devil of interpolation there too. In the antāti, the concatenation of rhymes would, however, render the hypothesis untenable. In fact, nobody has ever seriously questioned either the authenticity or the authorship of these four poems. They open the 11th Tirumuṟai (supposed to be arranged in chronological order), following a poem attributed to Siva himself and followed by a poem of 24 veṇpā attributed to Aiyaṭikaḷ kāṭavar kōṉ, identified by Auvai C. Doraisami Pillai, T. P. Minakshisundaram and K. Zvelebil with Paramēsvaravarmaṉ 1, a Pallava king of Kanchipuram whose rule may be dated approximately 670-685 CE. But, since the publication of the Mysore Annual Archaeological Report, 1925, Aiyaṭikaḷ kāṭavar kōṉ has been consistently assimilated by all epigraphists to one Cimmavarmaṉ of the end of the 6th c. (550-575 CE). The poems which come next in the collection are by Cēramāṉ perumāḷ, considered to be a contemporary of Cuntarar.

  • 6 The Religion and Philosophy of Tēvāram, Madras Univ., Book II (vols. III and IV) p. 971. No refere (...)

3The latter is the junior in the Tēvāram trio: Appar, Campantar, Cuntarar, who are situated, in that order, between the 7th and the 9th c. without its being possible to refine the chronology further. According to the Periya purāṇam, the eldest, Appar, derives his title (“Father”) from his first encounter with Campantar. There is, therefore, nothing to indicate that it should be seen as a parallel with “Ammaiy(ār)” (“Mother”) the title given, according to the same source, to KA by Siva himself when she reached Kailasa. This source also establishes that KA is earlier than Campantar, as in his story it is recounted that he refused to walk on the ground of Tiruvalaṅkāṭu where KA had walked before him, and that he entered the site walking on his head in order that his feet not desecrate a place already so powerfully consecrated to Siva. All these elements are of course drawn from the hagiographic legend, and cannot be confirmed from the internal evidence of the corpus itself. We indeed hesitate to see, as suggested by Dorai Rangaswamy,6 a reference to KA in a line of Appar “pēyt toḻil attiyaip-peṟṟuṭaiyīr”(Tev. V, 95, 4) “you gave birth to a woman who ruled over the activities of the pēy”. This probably designates simply a terrible form of the Goddess, such as that of Vīri “created in the past with a vēl in her hand” (Tev. VI, 53, 6), cited by Appar too. Cuntarar in the Tiruttoṇṭattokai, (Collection of the 62 Saiva saints who are the Servants of Siva, Cuntarar being the sixty-third) designates KA by one single word pēyar, “demoness”:

Perumiḻalaik kurumparkkum, pēyarkkum aṭiyēṉ (Tev. VII, 39, 3)
(“I am the slave of the Kurumpar of Perumiḻalai (near Kumpakonam) and of pēyar”)

  • 7 We owe to J. Filliozat, in his “Introduction” to the Chants dévotionnels tamouls de Kāraikkālammai (...)

4The term he uses is pēy7 and not, therefore, “Mother”, a word that, though seemingly attached to the legends, is without echo in the work itself.

  • 8 South Indian Inscriptions III, No 205, pp. 383-439 (see p. 384 and the Tamil portion of the inscri (...)

5The only other proper name quoted by KA, and then in only one patikam as a refrain, is Tiruvalaṅkāṭu. The fuller form, ‘Paḻaiyaṉūr Tiruvalaṅkāṭu’ is confirmed by epigraphs, namely the 31 copper plates of the 6th year of Rajendra Cola I, dealing with the donation of the Paḻaiyūr village (in Sanskrit Purāṇa grama), alias Tiruvalaṅkāṭu in the Paḻaiyaṉūr nāṭu, to the Siva temple at Tiruvalaṅkāṭu.8 Campantar devotes a full patikam (Tev. I, 45) to the site, and the tradition reports that he sang it at Paḻaiyaṉūr, next to Tiruvalaṅkāṭu, and not at the site itself, in memory of the respect shown to the place by KA. However, the text of the poem alludes to the terrifying goddess Nīli, who induced 70 vēḷāḷar to commit suicide, and not to KA at all. Appar is equally silent on two occasions, in the hymn VI 78 of which the enumerative pattern offers opportunities for all possible allusions, and in IV 68 devoted to Paḻaṉai- Tiruvalaṅkāṭu. Lastly, Cuntarar overlooks KA in one hymn (VII-52) which, however, recalls Siva dancing in the burial ground along with the pēy and the refrain of which repeats “...Ālaṅkāṭa! uṉ aṭiyārkkaṭiyēṉ āvēṉē” (O you of Ālaṅkāṭu, I enslave myself to those who are your slaves.”). But no further remembrance resounds in the line.

  • 9 Ed. Venugopala Pillai, GOMS, Madras, pp. 286-289, ceyyuḷiyaḷ, at the end. Naccinārkkiṉiyar attribu (...)
  • 10 K. Zvelebil, Tamil Literature, in Handbuch der Orientalistik, E. J. Brill, 1975, p. 155, n. 164. S (...)

6In the 11th c., the commentary on the Yāpparuṅkalam by Amitacākarar,9 on the subject of the genre called aritac ceyyuḷ (rsi’s work), cites a stanza attributed to both Pūtattār and Kāraikkāṟpēyār, after citing three veṇpā attributed to Poykaiyār; this is the first encounter between KA and the first three Āḻvār; only Pēy Āḻvār is missing from the Trio, but we take our support here from the persistence of the early name of the saint rather than looking for chronological indice, no matter what may have tempted K. Zvelebil to consider those authors as contemporaneous, thus forcing him to put the date of KA around 700 CE, evidently embarrassed at having to disagree with the arguments of Vaiyapuri Pillai in favour of a later date for Pūtattār, who makes reference to the site of Mahabalipuram.10

7During the 10th c. Nampi Āṇṭār Nampi compiled the collection of the first Tirumuṟai and wrote several poems all in some way connected with history, such as the Tiruttoṇṭar antāti which, in 89 stanzas narrates the achievements of the 63 Saiva saints, and the author’s own life. Stanza 28 deals with KA; for the first time there is no question of a pēy but only of her ascent to Kailasa on her head, and the appellation “my mother”(eṉ ammai) given to her by Siva, in response to a smile from Umā. For the first time too, the stanza ends by suggesting what a treasure KA might be to her family, revealing her in her female dimensions at last. Here is an essential element of her legend which, though quite foreign to her writings, Cēkkiḻār would develop in the Periya purāṇam, in the 12th c.

  • 11 Many segments of the narration probably appear in the Jataka, the Kathāsaritasāgara and in the inn (...)
  • 12 Aṃsa 7 (uttarārdha), adhyāya 36, śloka 7-11, ed. Sri Cāmrajendragranthamāla, No 32, with a Kannada (...)

8Nobody knows from where, or from which local or oral tradition,11 Cēkkiḻār drew the lavish details, along with all the proper names, which constitute the content of his narrative, and which fill up the first 50 out of the total of 66 stanzas; the so-called Sanskrit sources are either apocryphal, as the references to Brahmāṇḍa purāṇa which cannot be traced in Sanskrit, or are posterior to the Tamil text, as probably is the Śivabhaktavilāsa which, in its 46th chapter (pp. 106-110 in the Madras edition dated 1931), retells the story of Puṉitavati as narrated by Cēkkiḻār. The Śivarahasya, a text of no precise date, said to belong, as do so many others, to the Skanda purāṇa, mentions12 a kṣetra named Karakāleśvara, east of Trinaleśvara (Tirunallar), where Siva bears the name of Nityakalyāṇa and is worshipped by Vishnu and by the sage Jahnu; KA is, however, nowhere to be found.

  • 13 See st. 28 of the 1977 Tanjore edition, but st. 24-26 in the older popular editions. One Tiruvālaṅ (...)

9Even if it is true that a large number of Jaina texts might have provided the author of the Periya purāṇam with a hagiographic framework and series of 63 characters, in the later history of modern Saivism it is Cēkkiḻār’s text which has been imitated and adapted in other languages in the Deccan. With respect to Tamil, the 28th stanza of the Tiruttoṇṭarpurāṇacāram by Umāpati Civācāryā (first quarter of the 14th c.) is a perfect example of the state of affairs: proper names (Puṉitavati, Taṉatattaṉ “a merchant from Karikal”...), episodes mentioned (marriage, mangoes, and the husband running away before his wife is transformed into a pēy, and so on), all follow the text of Cēkkiḻār exactly, that text being the unquestioned and unquestionable source of all later popular developments, such as the pious staging of the saint’s festival as described by Lambezat around 1950, subsequent to the period during which Vinson observed an earlier development of the festivities, as well as being the source of the gamut of later literary writings, as represented by, for instance, the Ātticūṭi purāṇam.13

  • 14 Those wood engravings have been reproduced from the Periya purāṇam edition by Sri Arumukar tampira (...)

10The irresistible development of the legend of KA, independent of her work, is even better displayed by her iconography. Cēkkiḻār’s first narrative emphasis is on the miracle of the mango. That episode, which we are not able to link with any source prior to the 12th c., is responsible for a great number of the most recent and crudest stone or bronze images. When the question arises of the selection of wood engravings to be included in the printed versions of the Periya purāṇam, and only one image is to be selected out of a set similar to the ancient one we have reproduced in our new edition,14 it is invariably the miracle of the mango which is retained, for instance in the countless popular chap books which, for more than a century, have offered abridged prose renderings of the Saiva Golden Legend.

11On the other hand, in the 12th c., in Darasuram or Melakkatampur, it is KA reaching Kailasa walking on her head who is represented in the bas-reliefs of the temples, a feature which had already attracted the attention of Nampi Āṇṭār Nampi. Whenever we deal, so to say, with a major form of art, it is the image of the musician, preta, alone, or at the feet of dancing Siva which prevails from the earliest representations: the one, of course, that most accords with the work, and also with the finale of the legend in which Cēkkiḻār inscribed the testimony of the work itself.

12It is after about fifty stanzas inspired by other sources that Cēkkiḻār’s heroine finds herself, in fact, in the situation evoked in st. 13 of Iraṭṭai maṇi mālai: having assessed what family means (“wife and children” she says) the sole refuge remaining is God, our Father. The words of the P. P. (st. 48-50), at the moment Puṉitavati becomes “the Karikal demoness”, are the same as those through which KA expresses the deep knowledge of the inner path (mēl neṟi uṇarvukūra) which characterises the register of her Antāti. Cēkkiḻār selects exactly the words (cintai/oṉṟiya nōkkil mikka uṇarvu koṇṭu, uraiceykiṉṟār, st. 48) which express this dialectic of vision and reflection: the obsession with retaining the vision in order to keep the mental realisation, so often found in the Antāti (“seeing him to the filling up of my eyes... thinking of him through my thinking to the filling of my thinking” (st. 85. See st. 18-19, 45-46, 61-62, and 72-73). It is quite naturally, within this state of deep meditation that Cēkkiḻār sets the composition of the main work of KA, the Aṟputat tiruvantāti, the words of a seer but one whose vision is not guaranteed to last.

13That is why the hagiographer immediately goes on to the song of Iraṭṭai mālai antāti, a similar but shorter work, in which the metric concatenation of the stanzas creates a breathless rhythm, rendering the quest, and the request, of the saint more insistent. If, after the longer poem of the first antāti, the second one “makes all her acquired wisdom shine out” (P. P. st. 43) it also appears as an incantation against the risk of losing that wisdom, as a conjuration against rebirths, the dangers of evil, and the oceans of acts (i. 9, i. 12), and against the obsession with drowning, which is quoted three times (i. 4, 6, 9), from which we can be protected only by the One whose hair an entire river, the Ganga, failed to make wet (i. 13, 18; also i. 1, 5, 7, 14: here the Ganga is the main leitmotiv). The poem’s impact is due to the importance given to the help and refuge (tañcam) Siva represents, he who is always depicted as ‘strong’, ‘triumphant’, and ‘luminous’ (st. 3, 4, 13), since, though salvation may come through love and the control of the senses (i. 16, 17) it is ultimately god’s grace which is called for (i. 19).

14So it is inevitably Siva’s grace KA requests when she faces Him after ascending Kailasa, and as an excellent exegete of her work, Cēkkiḻār sums up in a single stanza (P. P. 60) the prayer offered by KA: apart from the desire for an ever fulfilled love (iṟavāta iṉpaṉpu vēṇṭippiṉ) she formulates three wishes: not to be reborn, if reborn not to forget him, and, lastly, to remain at his feet, hymning him happily whenever he dances. The words Cēkkiḻār uses as rhymes, piṟavāmai, and maṟavāmai (non-birth, non-forgetfulness) pick up exactly those used as rhymes by KA in the second stanza of Iraṭṭai maṇi mālai: piṟavātu (var. piṟavāmai) and maṟavātu; as for the three wishes, they sum up the message of the whole work: to be liberated, but, more than that, never to be separated from Siva, and best of all, to enjoy the ecstatic vision of his dance. The Lord has then only to grant his devotee a place on the spot where that dance takes place, Tiruvalaṅkāṭu, to enable Cēkkiḻār to bring together, in the last stanzas of his tale, the eulogy of the southern land with the two patikam devoted respectively to the site and to the master of the dance, authenticating thereby the legend of KA, thanks to two final and precise references to her own text.

15Once more, the mastery of Cēkkiḻār is evident here in the last quarter of his narrative, where into the woof of the work as it exists he weaves the warp of a spiritual itinerary which though invented is still plausible since it allies literary verisimilitude with religious orthodoxy: that this texture could not fail to be well received is a given. It credits by its genuine and truthful coloration a mass of legendary fabrics, which, though looser, and less clear as to origin, share from then on, with the revealed and consecrated canonical corpus a privileged place in the memory both of devotees and of the lettered. We have not reached as far back as the source of KA’s history, but at least we have been able to catch the instant when the ‘author’ enters into literature, immortalised along with her work: that work we are now dealing with.

The work

16Excepting a few stray references in Caṅkam literature and a few lines in Cilappatikāram, the work of KA is the first Saiva monument of Tamil literature. In order to appreciate it fully, one must remember that it is contemporaneous with the hymns of Paripāṭal (or according to some, slightly later or earlier) which celebrate Murukaṉ and Tirumāl, and is more or less contemporaneous with the first Vaishnava trio of Āḻvār, each of them having written one antāti, as KA did and, finally, that in both religions one large corpus of religious lyricism was soon to blossom, the Saiva Tēvāram and the Vaishnava Nālāyirappirapantam.

17It must not be forgotten that it is the latter which gives the pitch to all the Bhakti exegetes, nor that, even if modern commentators on Saiva hymns all have a leaning towards predating the Saiva medieval doctrinaire texts to before the Vaishnava medieval interpreters, it is only in the light of the latter and taking for granted the phraseology they created, that the Saiva authors in their turn gloss the emotional Bhakti, which they do only to the extent of an “analogical” reading that would be quite different without the existing literary and philosophical Vaishnava tradition. (See The Religion and Philosophy of Tēvāram by Dorai Rangaswamy in which the chapter dealing with “Bridal mysticism” quotes only Vaishnava references). But while we must certainly be aware of Vaishnava presuppositions, which could be seen as a lyrical temptation of some kind, we must also, with regard to the Saiva tradition, be on guard against two other attitudes. One of these would be to show more enthusiasm for the images of the saint than for what she actually said, and the second would be to reduce her work to a meditation on the cosmic dance of Siva; two temptations: the gruesome and the cosmic sublime. Truth is elsewhere, more diversified and deeply pondered.

18Before going into the details of the work, let us remember evidence that is too often forgotten: within the Saiva universe, however little of an intellectual KA may be imagined to have been, she still belongs, within the group of the 63 Saiva saints, to that small minority of those who became famous through a work, both sung and written, rather than for a particular kind of behaviour. The Tiruttoṇṭattokai calls this little group “paramaṉaiyē pāṭuvār” (those who sing the Supreme), the intellectual ones, making up about one seventh of the entire set, and differing both from the yogin with their mental concentration and from those practising religious activities or service, which latter occupy the greater part of the hagiography. Saivism, as a matter of fact, found its expression in a way of life and, more specifically, in particular behaviour within a civilisation of temples and before exemplary devotees; the legend of KA, though fitting perfectly within such a perspective, remains exterior to her work, which is in no way indebted to the legend, even if it has partly inspired it.

  • 15 Kandiah, Arumugam. A critical study of early Tamil Saiva Bhakti literature with special reference (...)

19The various poems are differentiated both in terms of content and metrical scheme. The two shortest, or Patikam, are the most cited, because they appear together at the beginning, not only of the manuscripts but even of most of the printed versions of the Tēvāram, representing, there some kind of introductory invocation. The great majority of the Tēvāram hymns are, in fact, composed in the same mould. But is it enough to confer the role of model on the two poems of KA? Some people have believed that it is, and editors have extended to both the patikam the epithet of mūtta ‘ancient’ which the manuscript tradition, and Cēkkiḻār too, reserve for the first one alone. Such a title is trivial and quite haphazardly bestowed: it is also conferred on a poem of Māṇikkavācakar, the Kōyil mūtta tiruppatikam dealing with Cidamparam, probably because that too is followed by a second Kōyiṟṟiruppatikam (and many more patikam). The fact that Caṅkam literature also offers early examples of the grouping of stanzas by tens, in Aiṅkuṟunūṟu and Patiṟṟuppattu makes it all the more difficult to draw any argument just from such an epithet. KA may, however be, at the least, the first example of the tradition of accompanying each decade with an eleventh stanza in which are recalled, along with the author’s name, the benefits derived from the recitation and from the worship of Siva. The first Āḻvār differ from KA in this regard but the convention is a common feature of the Tēvāram, mostly Campantar’s, and M. A. Kandiah has devoted a chapter of his Ph. D. thesis to the study of those final stanzas and their phraseology;15 KA has thus already provided us with an exhaustive example of the tirukkaṭaikkāppu, where the mention of the pure Tamil (centamiḻ) has not been forgotten either.

20Notwithstanding their many analogies, the two patikam are free of overlap and admirably complement each other. In the first one, the vision concentrates and sharpens itself on one site, the grove of the burial ground of Tiruvalaṅkāṭu, with the refrain repeated ten times “It’s Tiruvalaṅkāṭu, the place where our Father dances.” In the second one, the vision attaches itself to the Dancer and enlarges the scenery to the dimensions of the universe. The site is not otherwise mentioned by name, except in the probably apocryphal title of the poem which is belied by the last stanza, “the demoness of Karikal with many groves”, and even by the text which does not give a place-name to the “haunted groves” (aṇaṅku kaṭṭil, pp. 12-14). This double movement which carries us back, first to the meditation on a site, and then, through the mediation of a privileged vision, to the universal contemplation of the cosmic Dancer, admirably closes the work of KA, taking us from fascination with a place to the rapture of the Actor who transcends that place. We must also underline the unity of style and the perfect continuity between the two forever indissoluble texts.

  • 16 Five images of the Buddha were found buried near the famous tower near their College. See Ph. Ed. (...)

21What may be striking at first sight is the gruesome and even ghoulish, atmosphere. It mixes the horrible with a kind of humorous, rather jarring, tone, found in Tamil literature from Kaḷavaḻi nāṟpatu, to Kaliṅkattupparaṇi and Pācavataipparaṇi; but with regard to Siva as an ascetic haunting the burial grounds, it goes back to the Mahabharata itself (see XIII, 128) and to Sanskrit puranic texts. So much has been said of that “devilish” vision, that other aspects have been obscured. We should, however, insist, as does J. Filliozat, on the place of the Buddhist tradition in this theme of meditation on ghastly images; the memory of Buddhism was certainly alive at the time of KA, in the coastal area, near Kavirippattinam and Nagappattinam, where the French Jesuits found vestiges of it as recently as 1856.16 The conventional scenery of arid forest (kāṭu) or desert (pālai): cactus, thorny bushes, euphorbia, as well as the fauna of carnivores and predators, the crowd of the pēy, famished, swarming, grotesque, the mass grave and the burial ground, with corpses in all possible states of decomposition, are all the conventional ingredients, but they take on here a more forceful power of evocation by addressing all the senses (noises, odours, colours, movements, repulsive physical contacts, strange tastes) in raising the backcloth against which stands the Dancer, the Beautiful and our Father, all in One.

  • 17 P. Z. Pattabiramin, Trouvailles de Nedoungadou. Tandavas de Siva, Pondichéry, 1956.
  • 18 Āṭalvallāṉ, published by Tiruvāvaṭutuṟai ātīṉam, 1967, 354 pages with many illustrations; Wendy Do (...)

22Speculation on the dance of Siva flourishes. When the first edition of this book was issued (1956) an illustrated document on Siva’s tāṇṭava was also published, including a contribution by V. Raghavan and many borrowings from the Tamil book of Mayilai S. Venkataswamy;17 all these elements were taken up again two years later by Dorai Rangaswamy who critically underlined the unavoidable confusion in labelling and the contradictions in the descriptions, many of which are only theoretical systematisation or, as in the case of some Tamil sthalapurāṇa, simply modern classifications and of little value. More recent erudition18 has enriched our documentation to some extent but at best only prohibits us from being dogmatic.

  • 19 Dorai Rangaswamy, work quoted, I, pp. 485-486.

23KA uses technical terms only when she deals with the instruments of the orchestra. She does not give a name to the “great evening dance” (anti P. 10-4; cf. Skt. sandhya) which is also the “midnight dance” (naṭu nāḷai, P. 3.1) but, according to V. Raghavan, all dances are evening dances; the attributes of the dancer are not enumerated, neither are his postures. He “dances the fire” (aṉalāṭum) which can be interpreted both as above and inside the funeral pyre, unless it refers to a dance in which he “holds the fire in his hand” (aṉal kai ēnti, P. 8, P. 13, P. 18); parallel quotations are evenly distributed.19

  • 20 To be precise he is “the lord who practices dance” (P. 14.4) but never “the Lord of the dance”.

24The posture with the legs in vertical splits (P. 6 and 7), one foot pointing to the sky (P. 4), evokes the ūrdhvatāṇḍava; here, in conformity with the puranic tradition; the ‘competition’ with Kali is, in fact, clearly alluded to (P. 4), and a little further on she indeed looks at Siva ‘astonished’ (P. 19). But we notice that, in KA’s text, none of the plastic evocations of dancing Siva20 is ever accompanied by a speculation on the mystic values of that dance, and that it is nowhere said to be a dance of destruction, [even though the elements of the universe may be ‘whirling’ (P. 8) or being ‘shaken’ (A. 77)], any more than a dance of blessing. Siva, in any case, is never a ‘destroyer’; he is “our Father” in the first patikam and, in the second one, he is Young, Beautiful, especially Beautiful, Lord, Supreme, Magnanimous, Enchanting, Immaculate; sin alone is annihilated, not the world (P. 22.4). Briefly, the scenery surrounding him is more frightful than he is, and KA is evidently not to be duped by his accoutrements (A. 29, P. 15) nor by his whims. Yogic mastery, the result of a meditation on the horrifying, or a simple state of clairvoyance reached through asceticism and good conduct lets KA see in Siva predominantly life, salvation, liberation, just as recounted with some lyricism by J. Filliozat’s introduction.

  • 21 Iraṭṭaimaṇimālai means “a garland of two kinds of gems” rather than “a collar made of a double ran (...)

25However, if these two poems, which are the most popular, contribute to the fixing in our heart of the vision of, and the desire for, the Dancer of the burial ground, such images do not exhaust KA’s work. Obviously we find them again and again. The theme of the dancer is repeated in the Antāti (A. 2, 25, 30, 51, 70, 77-78, 85, 98-99) and Iraṭṭaimaṇimālai (i. 14, 15), approximately ten times with the same connotations: fire, burial grove, pēy; and about ten times also, separately from the patikam, Siva appears adorned with bones (A. 2, 26, 29, 30, 31, 37, 95) or skulls (A. 8, 52, 56, 74) or both (i. 17), and covered with ashes (A. 7, 59, 65, 95) or dust (A. 39, 56). The frequency of these visions is certainly important, as it is equal to that of the evocation of the frontal eye of Siva, and is just a little inferior to the number of mentions of the Ganga in his locks; all major themes in Siva’s most standard imagery, but all superseded by the image of his throat darkened by the poison he swallowed and neutralised for the world’s sake (approximately twenty allusions). There are, fortunately, other aspects of the vision of KA which are her own, and of an originality which does not consist in the nature of the major myths and iconic stereotypes, but in the selection of colours and attributes and in the frequency of repetitions. In completing our inventory we shall bring together the Antāti and the Iraṭṭai antāti, as being close enough in that respect to be joined here.21

  • 22 Cf. Glenn E. Yocum and his book quoted above n. 6.

26Siva is the Master of the constellation of Ārdrā (i. 8, 10), of the Veda (ibid.), the murderer of Yama (i. 12), of Kāla (A. 80, 81) and of Kāma (A. 89). He appears before Arjuna as a hunter (A. 62), dresses in an elephant skin (A. 7, 60), and rides a bull (i. 17, 18, 19; A. 94, 100). He is the bowman who destroyed Tripura (i. 5, 9, 15; A. 27, 32, 34, 37, 81, 84), a fairly wide-spread representation. When he is a mendicant ascetic (A. 25, 43, 57, 74) mad (p. 15) or slightly evil (A. 29), we are reminded of his previous frightful or sordid forms, with the same accessories. In the Tēvāram, the two most often quoted myths are the killing of Rāvaṇa and the story of Liṅgodbhavamūrti; about nine out of ten of the patikam of Campantar evoke both of these in that order in strophes 8 and 9. Curiously, it is the same sequence that KA uses twice in the Antāti (18-19 and 80). In the Iraṭṭaimaṇimālai they are also present but in the reverse order (8 and 14). Though the Saivism of KA is not aggressively sectarian, Siva is twice described as wearing round his neck the tusk of the boar, an avatara of Vishnu defeated by Siva. However the image of Harihara, a fusion of the two gods, appears five times (A. 41, 52, 54, 59, 83), less to belittle Vishnu than to show how elusive Siva can be; this image is joined with his Ardhanāri image (A. 41 or 59), which stands alone in several representations (A. 39, 47, 50, 51, 58, 71, 94; i. 5, 12, 19). The latter theme, which is very often quoted in the Tiruvācakam of Māṇikkavācakar, (more than 70 occurrences) is used here mostly to express an intellectual enigma, not the myth of the androgyne, but rather half woman and half Vishnu: this Siva, who is he really? It also evokes, but rather coldly, the literary game of jealousy between spouses: Siva has Uma at his side and Gaṅgā in his locks (i. 5; A. 71). The elsewhere extremely strong symbolism of the union between Siva and Pārvati is never used here for mystic lyricism, nor for any outburst of feminine sensitivity. Even if some enquiries reveal her preoccupation with conjugal love or feelings (mainly A. 94-95), KA does not entertain any erotic emotion, as opposite to Māṇikkavācakar, and never attempts, as he does, any feminisation of Siva.22

  • 23 F. E. Hardy, Emotional Krishna Bhakti, Ph. D. London 1976 p. 378 in the mss.

27Shall we conclude with F. E. Hardy that “her Bhakti lacks any eroticism”?23 We have yet to examine the representation of Siva which KA proposes most insistently: about thirty times Siva is shown wearing the crescent of the moon, about thirty times adorned with snakes, and about twelve stanzas associate together these two antagonistic attributes whose symbolic significance is, however, complementary. Both evoke desire and suggest eroticism, but the crescent moon, luminous and cool, is associated, without restriction, to the attraction of the Youth of the beautiful Dancer, while the snakes, whose embrace traditionally evokes, in Sanskrit and in Tamil literature as early as the Caṅkam period, the enjoyments of physical union, have a much murkier symbolism and repel as much as they attract. Further, KA is quite sensitive to the contrast between these two symbols, and even emphasizes their antagonism: on the body of Siva, the snakes, a projection of desire as well as a prohibition, threaten to devour the crescent, the weak child of the moon, which is alarmed by such cohabitation (A. 36). The snakes hiss angrily against the moon (i. 7, cf. A. 90), and want to catch it (A. 22); their relationship is like that of “the candid stag and the aggressive tiger” (A. 67). Fascinated and attracted by the crescent, fascinated and repelled by the snakes, doubly fascinated, but torn apart between attraction and reverence, KA confesses the violence of her desire only through these veiled but nevertheless transparent expressions; a skeleton consumed by an inner fire, she allows herself only this psychological symbolism to avow her passion. Capable of investing as much energy in her transports as the poets in love with Krishna do, she refuses to surrender to such transports or make them the modalities of her quest for god; still less is she willing, in the manner of Vaishnava authors, for them to be the main theme of her lyric. Far from being cold and inhibited, she finds outlets for her exaltation elsewhere than in mystic eroticism.

28She speaks of love as much as others do, and Hardy himself notes the terms: aṉpu, kātal, virumpu, inpu, acknowledging that “emotional undercurrents are by no means absent”. We rather quote a few beautiful expressions coined by her to name Siva: “inhabitant of my heart” (A. 6), “sweet treasure of the heart” (A. 10). But even if KA speaks of love from the first to the last stanza, an ambiguity remains: the heart is the seat of the mind and if it is the heart that enlightens the mind, (A. 17, 47) what is involved is some kind of wisdom, an understanding through intelligence, as much as a current of the heart. Words meaning ‘to know’ or ‘to understand’ are numerous (eṇ, cintai, niṉai, etc.) culminating, with uṇarvu, in the deepening of knowledge and experience. While we may uncover a multiplicity of intellectually oriented statements: “He is also the great Intelligence knowable to others” (A. 30), “thinking of him over and over, through the thought up to the fulfilment of the thought” (A. 85), “Thou, residing in clear thought” (A. 89), we must immediately quote as counterpart, the admirable stanza 96 of the Antāti: “covering him with a cover of love and by a glorious privilege, locking him up in our unique heart, we keep him hidden in mystery”, we must remember too the importance of the pleas for grace (A. 9-10, 44, 59, 88...) necessary to understand Siva and unite with him (A. 7 and 8). The eternal conflict between mind and heart has its solution in fact only in the mystical experience; thus the work of KA conveys to us the agony of the great visionaries: the vision is not everlasting; so how is that captivating vision to be recaptured “which was only a flash of lightning” (A. 24)? It is agreed that stanza 16 is an experience of the absolute, corroborated by stanza 86 (“we who have known him to some extent...). How may such an experience be made eternal, with the help of Siva’s grace? In three words: 1) by meditating on him transcendent; 2) by seeing him fascinating; 3) by possessing him mysterious.

29The way the soul of the devotee is possessed by its god is not described here in sensuous and carnal language. While the Vaishnava poetess Āṇṭāḷ wishes to marry Krishna, KA neither identifies herself with Pārvati nor substitutes herself for her. It is not Siva’s nature to inspire such dreams, and his erotic ‘games’ play a very minor part in the Tēvāram, negligible as compared to the līlā of Krishna; the erotic desire he arouses, as Bhikṣāṭaṇa for example, is collective, so to speak, and never particularly individualised; for him, there is no marriage except in heaven, and he is not surrounded by gopi to play with, but by pēy who serve him. If in the work of Māṇikkavācakar there are instances when the devotee feminises himself, KA sticks more strictly to Saiva tradition by shifting the concept of possession to the different register of service, where Siva is both Master and Father, and where she lies at his feet as an attentive slave (cf. the frequency of words as aṭi, āḷ, paṇi etc.) The theme of the service of god is ubiquitous in all Bhakti literature, not to mention the initial etymological connotation of the word itself. That the development of religious services and temple culture generally was taking up more and more space, is shown by a redundant vocabulary of praise and eulogy, adoration and prostration. KA often reiterates her eagerness to be the servant or the slave of Siva, to the extent of making such service “the pride of our mind” (A. 79), and the more so because it is, as well, the best safeguard against other forms of slavery: such possession is liberation. Master and Lord, Siva is at the same time Father; affirming once more a relationship rather of dependence than of kinship: Cēkkiḻār makes Siva, when answering her, call her “Mother”, in sharp contrast with the vocabulary of address of the Āḻvār, that of lovers or the married couple. However, as a price for her enslavement, KA brings forth from her innermost depths the claim of possessing a treasury, cēmam, a word connoting also safety, warranty of protection, or vaippu, a deposit that grants her joy, strength and salvation (A. 10, 14, 31).

30We have referred to Siva’s possession as mysterious (A. 96) as he is, in fact, transcendent and beyond the reach of knowledge. KA’s poetry is not primarily speculative, but she several times happens to express herself in abstract vocabulary that renders the intellectual memory of upaniṣadic culture visible. She does not speak of Siva’s avataras then, but of Siva as the object of truth (meypporuḷ, A. 9, 20), the principle of the universe (in the same words as Tirukkuṟaḷ); she quotes his eight manifestations (A. 21) and sees in him primarily the sapience (ñāṉam) which resides in thought (A. 45); she never despairs of understanding him, even though she meditates on him as “by essence not to be known” (A. 92); because he also possesses her, he is the grace (ibid.), “he knows and creates knowledge” (A. 20). Only a very few passages in her work linger on the meditation on the Absolute without speedily, and more concretely, turning to the avatara. Can we really see KA as a cerebral person, sensitive to the most intellectual forms of Bhakti, when she scornfully abandons “those whose knowledge is purely bookish” to their ramblings (A. 33)? There is, in fact, no meditation on transcendence in KA which is not ongoing and does not open up into vision (ibid.).

31All the dialectic movement of her work is between vision and thought, summarised in two neighbouring citations by no means to be considered separately if their intention is to be properly understood: “I would not be able today to understand the nature of our Father” (A. 18) and “Today we easily obtain the gift of his vision” (A. 19). The obsession of KA, whose prayer is her poetry, is the always effulgent form of Siva (A. 59-62, 93), the form whose multiple splendours she never ceases to string into litanies without ever being satisfied, and which is the only object of her thoughts because in the form resides the reality of the essence (A. 59, 33). The primacy of this vision, both justified and acclaimed, (see word frequency of kaṇ, kāṇ, nōkku...) ends our encounter with KA’s work, the true work of a seer (ariṭacceyuḷ) in which the transcendence of Siva and the ardent longing to serve him are united in a love as passionate as that of the first Āḻvār for Krishna, but in which feminine sensitivity has chosen, instead of asserting itself, to be consumed and ultimately to become, in this emaciated body, nothing but a glaring gaze in quest of samādhi:

Vos yeux sont deux flambeaux mystiques où mon âme
Vient s’éclairer au pur amour… (Paul Verlaine, Poèmes divers, Juillet 1890)

(Your eyes are the two mystic torches where my soul
Comes searching the light of pure love...)

The images

32After the publication of the first edition of the Chants dévotionnels of KA in 1956, the identification of the figure of KA on some Khmer monuments amply justified an additional note by J. Filliozat for the new 1982 edition. The interpretation of that discovery raises, however, a serious methodological problem.

  • 24 J. Filliozat’s references are: Mireille Benisti, «Notes d’iconographie khmère IV. Au sujet du lint (...)

33It is not at all the identification itself that is in question, but its real impact and the clues it may offer. Summarised objectively and in context, the phenomenon is as follows: within a Khmer iconography which is quite resistant (as confirmed by Bernard Groslier in person to the present author) to any gruesome representation (different in that regard from Campā or Java images, for example) there suddenly appear (we might say “accidentally”), four or five perfectly stereotyped images of KA, and this within a limited period of time (hardly more than a century) and in an area with a particular cultural imprint: the Angkorian territory between Banṭāy Srĕi (founded in 967 CE), Văt Basĕt (first or second half of the 11th c., according to various authors) and Pimāy (last quarter of the 11th c.). To this is added the Phnom Chisor and a presumption in favour of Vat Ek (here again within the limits of the 11th c.).24

34In the absence of any contextual Khmer reference, it is tempting to look for some kind of a signature, admitting, or rather proclaiming its origin. Specialists in Khmer studies are all inclined to designate Yajñavarāha, the learned brahmin who looked after the young Jayavarman V, who knew everything about every Indian language, imported the āgama and their doctrine, and served Saivism with all his culture and astonishing personality. In addition to all that, he may well have imported KA too; but when a borrowing is not immediately fertilising or fertilised, it remains ephemeral. In contrast to the KA imagery in the Tamil land, the Khmer KA are without any posterity or variation. In short, by its very rigidified character, the Cambodian representation of KA proclaims its ignorance of its own history and its incapacity to found a lineage: the image was imported, and perhaps a text as well that we do not know of. It is therefore overreaching, when speculating on the many other indications of South Indian Saiva presence in Cambodia, to speak of “the diffusion abroad of the Saiva hagiography of that Tamil saint” when such a diffusion had already ceased even before the legend itself was fixed in writing in its country of origin.

  • 25 See J. Dumarçay et F. L’Hernault, Darasuram in Mémoires Archéologiques de l’EFEO 1983.

35More precisely, we have already carefully distinguished the late popular iconography from the much earlier imagery of a little skeletal musician-attendant at the feet of dancing Siva. So, when did this musician start to turn into KA, in its female version, that is? The question is crucial, because KA became what she now is only through the diversification of her iconography and the staging of episodes of her legend, mainly after the 11th c., and starting with Darasuram and Melakatampur.25

  • 26 work quoted, pp. 365-367.
  • 27 in Yazdani, Early History of the Deccan, part I-VI, 1960, Hyderabad, p. 435.

36Even the optimistic presentation by C. Sivaramamurti26 of the cultural diffusion of KA (Angkorian culture extended into Thailand territory, thanks to the modern political localisation of Pimāy today, and into Vietnam) does not allow, as he himself acknowledges, for the presence of KA in South-East Asia, outside Cambodia. In fact, it is of foremost importance to remember that it is Bhṛṅgin whom we first meet at the feet of dancing Siva, in Mī-so’n, as early as 7th or 8th c.: the dancing Siva found in Campā omit KA and it seems that she is commonly overlooked or unknown even in India itself, other than in Tamil lands. The images referred to by Sivaramamurti in Ceylon are Cola bronzes clearly isolated within the chronology, and do not seem to have founded any lineage. Kerala images are not anterior to 17th or 18th c.. We may be a little suspicious when Nilakanta Sastri writes that he finds “KA, the Saiva woman devotee, in an inscription of 1124 from Devanagere, as further evidence of the presence of Tamil elements in the Karnataka of the Chalukyas”27 because, when checking the inscription (Epigraphia Carnatica XI Dg. 155), we find that it refers to one “Karekal-Amme, pure daughter of the Siva Brahmans in Nityavinoda chaturvedimangala” whose son, called Kunta, was born in Tirukālaṭi. To put it briefly, the ‘cultural’ phenomenon here is of little importance. It is the singularity of the Cambodian episode that directs our attention to the original development of the iconography of the Tamil saint.

37First of all, let us agree that the popularity of the theme of dancing Siva in South India from the apogee of the Cōḻa, which was also the time of Umāpati, and in the area of Chidambaram, along with the commercial links, privileged though circumscribed in time and space, encouraged the flourishing of Saivism in Cambodia. All these are well established facts which should not, however, lead us to neglect one or two other details of equal importance. We have alluded to the dancing Siva on the two tympanums in Mī-s’on, where a skeletal Bhṛṅgin brandishing a drum, figures among the attendants, a scene which recalls analogies in India, at Ellora, for instance, where Bhṛṅgin is clearly visible behind the right leg of Siva, in cave No 14. Another Vākaṭaka sculpture in Elephanta shows Bhṛṅgin with a drummer by his side. And we note that Jean Boisselier has shown analogies between Campā art and the art of Ellora or Badami, and reminded us that Siva is cited as dancing god in a Campā inscription dated 657 CE (inscription of Prakāśudharma). For that date, even if KA would then have been at least a hundred years old, models have to be sought, as regards iconography, outside Tamil Nadu. And they are certainly to be found, from West to East, in Calukya art (Aihole, Alampur...). An iconic tradition does exist before the 9th c., grouping around dancing Siva (and even sometimes around other forms of Siva) gaṇa musicians, the skeletal Bhṛṅgin, sometimes Devī, musicians beating out the rhythm, and even Annapūrṇā. Bhṛṅgin is found again with the Late Calukya in the 12th c., in Hampi, and there are skeletons and emaciated figures, but rarely female ones, in Hoysala art. Kakatiya art would have Andhaka and the Mātṛkā reappearing around dancing Siva, but never KA. It is the same thing as far as Vijayanagara art is concerned, or the Ganga art in the East, at Bhubanesvar, or the Pāla and Sena steles swarming with figures. To sum up: north of Madras a tradition crosses India from west to east, to which even ancient Cēra and Pāṇṭiya examples in the South are no exception.

38We may then formulate the situation in these terms: the dancing Siva’s iconography from its origin includes a surround of attendants, bhūtagaṇa and preta of both sex, often acting as musicians. Among them, a male figure, that of Bhṛṅgin as a skeleton, took on individuality over a very large area extending from Ellora and Elephanta as far as the Campā, and with a frequency in comparison with which KA barely exists. However, over the much smaller geographical area of the Tamil land, and that only from the beginning of the period of Cōḻa art, a female figure is, in her turn, individualised as a skeleton: the figure of KA, exported by deliberate accident to Cambodia almost from its origin but over only one century.

39This formulation seems justified by the actual words of KA herself according to which she evidently did not invent her representation as a pēy, thus conforming and assimilating herself to a pre-existing concept, the troop of pēy attached to Siva (A. 86 and Cēkkiḻār st. 49). It is clear from the literature that even the brisk description which opens the first patikam is not a self-portrait drawn for her modern biographers:

«Seins ridés, veines saillantes, yeux ronds, dents blanches, ventre creux,
Crinière rousse, deux crocs allongés, chevilles hautes, longs mollets,
Une démone hante, haletante la forêt desséchée. » (our translation)

(Withered breasts, jutting veins, round eyes, white teeth, hollow belly, ruddy mane, two protruding fangs, high ankles, long calves, A demoness breathlessly haunts the parched forest)

40But from which period this unidentified, anonymous, description has been taken up as reference for the iconographic identification of KA herself, no one knows. In any case, this text is the oldest and it is the starting point and inspiration today for whatever can be cited in Tamil or Sanskrit on the subject.

  • 28 P. R. Srinivasan, T.A.S.S.I., 1957-58 pp. 47-50. S. R. Balasubrahmanyam Early Chola Art, II pp. 18 (...)
  • 29 Douglas Barret, Early Cola Architecture and Sculpture 866-1014 AD. London, 1974 (see pl. 56, Sempi (...)

41Then comes the question of the period of Tamil iconography in which a transfer was effected to fix KA at the feet of Siva, while anywhere else we find Brahma, Bhṛṅgin, the Goddess, and Hanuman beating the rhythm: the answer depends on the accuracy of iconographic inventories. It seems reasonable to consider the second half of the 10th c. as a date for a few representations which cannot, in any case, be later than the Banṭāy Srĕi image, such as images in Karantai28 or Atuturai, that is, as early as the development of the theme of Nataraja in Cōḻa art;29 it is also the presumed date of Nampi Āṇṭār Nampi who initiated the process of canonisation of early Saiva Tamil scriptures. What is quite remarkable is how quickly this figure was imported (we do not say “spread”) in Cambodia, while everything on Indian soil tends to confirm the regional character of KA. What encounters, collusions, and perhaps confusions, benefited KA for over a period of time within the Angkorian space, we do not know; in India itself she hardly succeeded in travelling beyond Tamil Nadu, to Ceylon, and to Kerala as late as 17th c. Still the seductive and likely hypothesis of the existence of one more piece of evidence of the Saiva eclecticism of Yajñavarāha, remains.

42Serious interrogations arise out of this restricted debate however. For example, what of the other elements of the Tamil Saiva pantheon which were left aside in Cambodia, or, in a wider perspective, how may the impact of the architectural and iconic traditions of South India on Angkorian art of the 10th and 11th c. be more accurately measured? Further, if legends or ideologies, and even iconographic models were able to circulate independently from the texts which are, at least as far as our knowledge goes, their most fundamental expression and support, should we then conclude that texts can no longer be considered as the only sources to look for? And, in this hypothesis, what about the instruments, chronology and modalities of such influences?

  • 30 A striking example. The dancing Siva in a fresco of the Kailasa in Ellora, on the lintel of the en (...)

43Once more, KA’s case study calls for a careful distinction between the written work, the legends and other testimonies, and is a warning against drawing too many conclusions from fragmentary evidence on the exact process of cultural diffusion: all the more so given the still precarious condition of our iconographic inventories.30

44Lastly, while the representations which place KA close to dancing Siva are a typical Cōḻa feature, it seems that after the rule of that dynasty they lost some part of their popularity in the Tamil land itself, perhaps to the advantage of images of the saint alone. From the Vijayanakara period onwards they are rare in the bas-reliefs, and may be looked for in vain in frescoes or in the so called Tancavur glass paintings, or in the wood engravings used from 1850 to print popular images inspired by their style in chap books. The typical entourage of Nataraja is ordinarily composed of Vyāgrapāda and Patañcali, Nārada and Tumburu, some gaṇa, ṛṣi or devotees, but not KA anymore. This cannot be fully accounted for by the important influence of Telugu, omnipresent in printing activity in Madras, when copyists as well as publishers of the Tēvāram text persist in keeping the two patikam of KA as an opening invocation, attesting thereby to a persistent familiarity with the work, quite independent from any visual representation.

45April 1982

This essay was originally published (in French) in the second edition (1982) of the translation of KA’s work into French by Karavelane, which, in 1956, inaugurated the Collection of Publications de l’Institut français d’indologie at Pondicherry. This edition was a revised and updated version of the first one and may be considered as a new book altogether. (verse references are to the French edition). Twenty years on, is there any need to bring up to date again an already updated version?

In 1982 we noticed that a book by Mariasusai Dhavamony, Love of God according to Saiva Siddhanta, OUP, 1971 devoted a short chapter to KA but that the 13 stanzas the author chose to translate were amongst the 20 stanzas J. M. Somasundaram Pillai in his Two Thousand years of Tamil Literature (Madras, 1959) had already reprinted in 1959 from the first original translation dated as far back as 1912; this was the work of N. Nallaswami Pilai, who published it under the title “The Mother of Kāraikkāl” in his journal Siddhānta Dīpikā, (vol. XIII, no 4, pp. 151-160) as no 2 in a series of “South Indian Devotional writers”. Since then, several poems have been translated into English, such as those quoted in the study of Dr. S. Sasivalli, Kāraikkāl Ammaiyār (International Institute of Tamil Studies, Madras, 1984), or T. N. Rama chandran, Hymns of Kāraikkāl Ammaiyār (IISSR Dharmapuram, Mayladuthurai, 1993). Nevertheless, Karen Pechilis Prentiss, in her very interesting essay on The Embodiment of Bhakti (OUP, New-York, 1999) still relies exclusively upon J. Norman Cutler’s Songs of Experience: The Poetics of Tamil Devotion (Bloomington, Indiana University Press, 1987) whenever she refers to KA. All the more reason then to look forward to the publication by Elaine Craddock of her book, Siva’s Demon Devotee: Kāraikkāl Ammaiyār, forthcoming from SUNY Press, with fresh translations made with Dr. R. Vijayalakshmy.

A refreshing contribution by Vidya Dehejia, Slaves of the Lord, The Path of the Tamil Saints (Munshiram Manoharlal, New-Delhi, 1988) is praiseworthy for the rare quality of her international iconic documentation. To her too belongs the credit for recording the explanation once suggested by Dr. Robert L. Brown of the Art Dept., University of California at Los Angeles, for the presence of images of KA in Cambodia, a theory which neither Mireille Benisti nor J. Filliozat ever thought of: “When Paramadatta abandoned his wife and sailed off to ‘a distant port’, he may have travelled all the way to a south-east Asian kingdom. Certainly, the hypothesis would explain why KA alone among the Nāyaṉmārs is frequently depicted in the stone reliefs of the Cambodian region.”(o. c. p. 137) Se non è vero, è bene trovato. (if it is not true, it is well imagined).

In another direction, comparative studies between the literary traditions of South Indian cultures should certainly be developed further. See for example, the essay in Tamil by Dr. M. S. Shanta Kaṉṉaṭattil Kāraikkāl Ammaiyār, IITS, Chennai, 1997, on the place of KA and the various puranic traditions around her in Kannada textual tradition, such as the Kāraikkāl Ammayarara Ragale by Harihara (13th c.).

A lavish publication by the Art Gallery Dhyani Asian Arts in the Netherlands, by Peter J. J. de Bruinj, Kāraikkālammaiyār, Part In – An iconographical and textual study; Part 2 Poems for Siva, 168 p., 19 colour plates, 13 B/W plates, offers one more annotated translation of KA and which seems to put particular emphasis on the supposed iconography of the Saint in south-east Asia. The author of this monograph, was unfortunately unaware of our second edition of 1982, published by the French Institute of Pondicherry, and of its postface. The monograph came to our attention only after the present set of articles had already gone to press. It may perhaps enrich and even illustrate the methodological debate we opened on the iconography of KA.


1 «Les études dravidiennes et particulièrement les études tamoules de 1900 à 1906», Année linguistique, tome III, 1905-1907, Paris, C. Klincksieck, pp. 41-108; see pp. 92-93.

2 Tampi-purosottamap-pillai, Kāraikkāl Ammaiyār purāṇam, Madras, 1906, pp. III, 8, 165, 1. The copy dedicated to Vinson is in the library of the INALCO in Paris. Pages 154-165 are presented, with a separate title-page, as being a “translation” of adhyāya 111 and 112 of the sthānapracamcā kāṇṭam of the first part of the Brahmāṇḍa purāṇa which are supposed to narrate the sthalamāhātmya of Kāraikkāl, (kāraikkāl stalamakimai yuraitta… viyākkyāṉam) and here translated into Tamil by Cupparayacastriyar and Ci-rama-mu Iramacuvamipillai, as adhyāya 11 and 12. Differences between this text and Cēkkiḻār are noted on pp. 48-49 of fascicule V B 2, Oct. 1934 of Cittānta Pōta ratṉākaram, which was concurrently publishing an edition of the Periya purāṇam with a commentary by Alala Cuntaram Pillai., As expected, the Sanskrit versions of the Brahmāṇḍa purana, do not contain this version of a local purāṇa.

3 Murdoch, John, Classified Catalogue of Tamil Printed Books (first ed. 1865) reprint Madras 1968, p. LXXXV.

4 «Sur les noms de Pondichéry et de Karikal» Revue historique de l’Inde française, vol. 4, 2nd part, Pondichéry, 1920, pp. 115-123; see pp. 121-122. We have not been able to trace the edition; its page numbering is disturbingly similar to Sri Arumukat tampiran’s edition published in 1890. In a letter dated 30-4-1982 M. A. Gunasekaran, Asst. Librarian, Maṟaimalai Aṭikaḷ Library in Madras has kindly brought to our attention one Kārāikkāl Ammaiyār tivya carittirak kīrttaṉai by I. Irakavamutaliyar of Civakankai, published in Madras in 1869, also described in Tamil nūl vivara aṭṭavaṇai, I, part 2 under the entry 2023 (see also No 2143 and 3663). These are texts derived from the Periya purāṇam; they contain only a summary of it and various popular songs on the same theme. The two patikam of KA had been published already in 1859 in Madras in the edition of the Periya purāṇam by Capapatimutaliyar (vol. 1, pp. 301-306).

5 Lambezat, Bertrand, Karaikal Ammear, Patronne de la Ville de Karikal, Karikal, 1952, 26p.

6 The Religion and Philosophy of Tēvāram, Madras Univ., Book II (vols. III and IV) p. 971. No reference to KA is found in Manikkavacakar, notwithstanding G. U. Pope’s opinion (see his edition of Tiruvācakam, OUP, 1900, under VII, 15); cf. Glenn E. Yocum, Hymns to the dancing Siva, a study of Māṇikkavācakar’s Tiruvācakam, New Delhi, 1982, p. 164, n. 30.

7 We owe to J. Filliozat, in his “Introduction” to the Chants dévotionnels tamouls de Kāraikkālammaiyār, (PIFI, 1956, new ed. 1982) the Prakrit etymology of pēy: pēya (Sanskrit preta) “dead” rather than “devil”, as well as for the link with Pūtattār, Sanskrit Bhūta, considered as a Bhuta of Vishnu in the Guruparamparāprabhāvam. Apart from Pēy Ā ḻvār, Filliozat cites Pēyaṉār, a Caṅkam poet, and Kulacēkaraṉ Ā ḻvār calling himself a pēy in Perumaḷ Tirumoḻi III, 8.

8 South Indian Inscriptions III, No 205, pp. 383-439 (see p. 384 and the Tamil portion of the inscription). Cōḻa references to Tiruvalaṅkāṭu are numerous both in epigraphy and iconography. The complete series of Kāraikkāl inscriptions was published in 2006 by the French Institute of Pondicherry and EFEO. Jouveau-Dubreuil had already noticed the name of Kāraikkāl on copper plates of Kulōttuṅka I about the buddhist monastery located in Nagappattinam and in an inscription of Rājarāja in Tancavur, mentioning a Tēvatāci. Cf. Revue Historique de l’Inde française vol. IX (1955), p. 290.

9 Ed. Venugopala Pillai, GOMS, Madras, pp. 286-289, ceyyuḷiyaḷ, at the end. Naccinārkkiṉiyar attributes the same text to Auvaiyār and KA, as also one of the veṇpā attributed here to Poykaiyār.

10 K. Zvelebil, Tamil Literature, in Handbuch der Orientalistik, E. J. Brill, 1975, p. 155, n. 164. See Vaiyapuri Pillai, History of Tamil Language and Literature, p. 117 n. 1, and Jean Filliozat, work quoted, pp. 12-14.

11 Many segments of the narration probably appear in the Jataka, the Kathāsaritasāgara and in the innumerable edifying tales in Pali or other vernaculars, in all the religions of India: the world of merchants, enriched through their maritime traffic, with its ethics, and the constraints derived from absence, the devotees fulfilling their duty by welcoming ascetics through whom the god himself is honoured; the husband’s unjustified suspicion of his wife; the privileged role of the mango as a miraculous fruit, etc. But what is badly lacking is the origin of the plot which assembles those fragments. In the miracle of the mango fruit, anything at all can be seen, including a fertility symbol; one may even quote the reference of Mary Frere, Old Deccan Days or Hindoo fairy Legends current in Southern India, London, 2nd. ed. 1870, pp. 246-267, on a triple fecundation, thanks to a mango. The thesis of D. D. Shulman who sees in the story of KA a hagiographic retelling of the story of Nīli, the cruel goddess of Tiruvalaṅkāṭu, has the merit of putting forward the analogies (conflict with the husband, emphatic role of the dance of Siva) and of making the presence of KA in Tiruvalaṅkāṭu meaningful, a point which otherwise remains gratuitous in the KA story. But it has the inconvenience of having to minimise irreducible differences which the hagiographic sublimation fails to eliminate, and to attempt an identification of KA’s vision with two contradictory patterns, one leading to Kaṇṇaki, the other to Aṉacuyā or Āṇṭāḷ; this ultimately ends up as an interesting variation on the theme of the eternal feminine for purposes of anthropology. See David Dean Shulman, Tamil Temple Myths, Princeton Univ. Press, 1980, and our article «Tradition tamoule et mythologie hindoue», in Revue d’Histoire des religions, CIC, 1/1982, pp. 74-75.

12 Aṃsa 7 (uttarārdha), adhyāya 36, śloka 7-11, ed. Sri Cāmrajendragranthamāla, No 32, with a Kannada translation by Gangadhara Sastri, Mysore, 1946.

13 See st. 28 of the 1977 Tanjore edition, but st. 24-26 in the older popular editions. One Tiruvālaṅkāṭṭup purāṇam by Tiyakaraya Kavirayar was published in Madras in 1864. It is silent on KA and centres on the challenge between Siva and Kali; there is no explicit mention of Nīli either. However, one full sarga is devoted to the story of C ōmacāmi, a very religious Saiva brahmana versed in the Veda and conscientious in the performance of his duties, who once fell, when seduced by a young pāṇar girl from the Madurai cēri, to the extant of going to live in the cēri, eating meat, smoking opium and spending a hundred years of his life wandering as a paṇṭāram with his wife and their many children, involved with them in incestuous intercourse. He dies a wreck in Tiruvalaṅkāṭu, but Siva rescues him from Yama’s bhūta. Surprised, Yama asks for an explanation from Siva who answers by extolling the sacredness of the place. At the beginnings of the narration there is a reminder of the plot of Nīli’s story, here transformed into a harmless and edifying tale.

14 Those wood engravings have been reproduced from the Periya purāṇam edition by Sri Arumukar tampiran cuvamikal, first kāṇṭam, second part, Madras, September 1890 (year vikruti, kaṉṉivari; the date given in print as Kali era, 4993 is wrong and should be rectified to 4991). The chapter on KA occupies pp. 220-242, and her works are printed as a 39 pages appendix (numbered separately) with a detailed commentary on Iraṭṭaimaṇimālai and Antāti. The first volume of that edition (Madras, year sarvajittu, miṉavari, Kali 4989) also includes the medieval texts of Nampi āṇṭār nampi and of Umāpati we refer to in this study.

15 Kandiah, Arumugam. A critical study of early Tamil Saiva Bhakti literature with special reference to Tēvāram, Ph. D. Univ. of London, 1973, 2vol.; see Ch. VII, pp. 356-400 of the unpublished manuscript version.

16 Five images of the Buddha were found buried near the famous tower near their College. See Ph. Ed. Foucaux, “Le Bouddha Sakya-Mouni” in Mémoires de l’Athénée Oriental, vol. 1, Paris, 1871, pp. 79-86, especially p. 85-86 and fig. 3 to 6. Consult T. N. Ramachandran, The Nagappattinam and other Buddhist Bronzes in the Madras Museum, Bull. of the Madras Gov. Museum, 1954 (reprint 1965) and T. N. Vasudeva Rao, Buddhism in the Tamil Country, Annamalainagar, 1979, p. 236ff. and the original contribution of Lokesh Chandra, Oddiyana: A new interpretation, New Delhi, n.d.[1978] typed copy, 30 p.

17 P. Z. Pattabiramin, Trouvailles de Nedoungadou. Tandavas de Siva, Pondichéry, 1956.

18 Āṭalvallāṉ, published by Tiruvāvaṭutuṟai ātīṉam, 1967, 354 pages with many illustrations; Wendy Doniger O’Flaherty, Ascetism and Eroticism in the Mythology of Siva, 1973 (OUP in India, 1975); C. Sivaramamurti, Nataraja in Art, Thought and Literature, National Museum, New Delhi, 1974; Stella Kramrisch, Manifestations of Siva, Philadelphia Museum of Art, 1981, and The Presence of Siva, Princeton, 1981.

19 Dorai Rangaswamy, work quoted, I, pp. 485-486.

20 To be precise he is “the lord who practices dance” (P. 14.4) but never “the Lord of the dance”.

21 Iraṭṭaimaṇimālai means “a garland of two kinds of gems” rather than “a collar made of a double range of gems”; it is a poem of 20 stanzas linked together according to the same principle as antāti, but in which two kinds of verses alternate, the veṇpā, here in the even stanzas and, in the odd ones, the kaṭṭaḷaikkalittuṟai. This creates a more variegated and halting rhythm, in short a more “emotional” one than the antāti with the monotony of its hundred stanzas in an identical metric pattern. See for example, Paṉṉiruppāṭṭiyal, (st. 151 and com. ad loc. Kaḻakam edition, Madras, 1963) a treatise written between the reigns of Rājarāja I and Kulōttuṅka III. Note that many variegated poems of the Tēvāram are composed according to that meter: here again KA inaugurates a type of poetry that was to become abundant.

22 Cf. Glenn E. Yocum and his book quoted above n. 6.

23 F. E. Hardy, Emotional Krishna Bhakti, Ph. D. London 1976 p. 378 in the mss.

24 J. Filliozat’s references are: Mireille Benisti, «Notes d’iconographie khmère IV. Au sujet du linteau de Vãt Basĕt», BEFEO LIII, 1967, pp. 513-516, pl. XV, and «Notes d’iconographie khmère, VII. Kāraikkālammaiyār», BEFEO, LV, 1969, pp. 159-161, pl. XVI-XIX; J. Filliozat, «New researches on the relations between India and Cambodia», Indica, 3, 2, sept. 1966, pp. 95-106 and Laghuprabhandah, Leiden, 1974, pp. 454-465; “Sur le civaïsme et le bouddhisme du Cambodge, à propos de deux livres récents», BEFEO, 1981, pp. 59-99.
See also Bruno Dagens, “Étude iconographique de quelques fondations de l’époque de Suryavarman Ier” in Arts Asiatiques XVII (1968) pp. 173-208, with illustrations from Vat Ek, Baset and Pnom Chisor; C. Sivaramamurti, Nataraja in art, thought and literature, New Delhi, 1974, who reviews, through the documents from EFEO or some of its members, all the examples quoted for South-East Asia, p. 352ff.; Mireille Benisti, “Kāraikkālammaiyār in Cambodia”, Proceedings of the Second International Conference Seminar of Tamil Studies, vol. II, pp. 402-404, Madras, 1971.

25 See J. Dumarçay et F. L’Hernault, Darasuram in Mémoires Archéologiques de l’EFEO 1983.

26 work quoted, pp. 365-367.

27 in Yazdani, Early History of the Deccan, part I-VI, 1960, Hyderabad, p. 435.

28 P. R. Srinivasan, T.A.S.S.I., 1957-58 pp. 47-50. S. R. Balasubrahmanyam Early Chola Art, II pp. 183-185.

29 Douglas Barret, Early Cola Architecture and Sculpture 866-1014 AD. London, 1974 (see pl. 56, Sempiyaṉ Mahātēvi temple, Tanjore Dt., c. 981).

30 A striking example. The dancing Siva in a fresco of the Kailasa in Ellora, on the lintel of the entrance of the South portico, revealed by G. Yazdani to the XVIIth International Congress of Orientalists in Oxford in 1928 was published by him in 1930 in appendices to the Annual Report of the Archaeological Department, Hyderabad State, for 1337 F (1927-28 A.C.), Calcutta, 1930, pl. D and p. 22, with the text of his lecture, “The Fresco Painting in Ellora”. Yazdani already emphasised there the affinities of that image with those of the same period from South India, that is the 10th century. But the representation he gives shows no attendant at all beside Siva.
We had to wait for the chemical treatment of the fresco by the Archaeological Survey of India which revealed the details, published at last as pl. XCVI of Indian Archaeology 1976-77 – A review, New Delhi, 1980 (see also p. 138). In the lower portion to the right of the panel, a skeleton-like figure is shown, apparently male, but in the style of the ascetic KA in Nagapattinam, which had remained unknown before that 1980 publication.
Similarly the enigma of the skeletal dvārapāla of the Mahakutesvar temple, raised as early as 1941 by R. S. Panchamukhi (Annual Report on Karnataka Research in Bombay Province for the year 1939-40, p. 17 and pl. II(a) remains unsolved. Are we dealing with Śṛṅgi and Bhṛṅgi, or with devotees of Kālamukha obedience?
A very long list could be made of the books and collections consulted without anything definite resulting. The question of KA does not even arise in the synthesis by K. V. Sundararajan on Cave temples of the Deccan, Archaeological Survey of India, 1981, treating Calukya material in detail; however, he gives, pl. XXVIII, an interesting representation of Kāla and Kali dancing in skeletal form in cave XXI of Ellora in a Saiva iconographic context, and finally confirms the frequent occurrence of Bhṛṅgin.

© Institut Français de Pondichéry, 2009

Licence OpenEdition Books


Volume papier
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search