1The present edition of the Paripāṭal does not claim to replace the excellent edition of U. Ve. Swaminatha Aiyyar. It is, on the contrary, nothing but a complement, necessary for those who are not very familiar with Tamil poetic language approaching the text itself without a guide, and indispensable to those who are unfamiliar with Tamil but who wish, nevertheless, to get to know its literary richness and to draw something from this important source of ancient Hinduism through the medium of a translation. This is the first complete one in an occidental language and is published in a collection not meant for the general public; it thus deliberately sacrifices something of elegance to rigour. As far as the opposing structures of two languages permit, it endeavours to conserve the order of appearance of ideas and images, verse by verse if possible, which clarifies the use, if not the abuse, of the inversion and of the parentheses which indicate additions deemed indispensable. Where Tamil indefinitely juxtaposes words to produce a riot of images and sounds, the translation, paying tribute to French “clarity”, has to present the grammatical tools of the logical connections that the poets prefer to leave indeterminate, and to choose a single meaning where they cultivate multiplicity or ambiguity. Allusions, images and obscure expressions are not, however, elucidated except in the notes; clarity is not the first concern of the original text all of whose implications can be made plain to the unprepared reader only by a long paraphrase rather than a translation. The unique ancient commentary, the work of Parimēlaḻakar, often consists only of that paraphrase, and we have judged it fruitless to give a complete translation of it, choosing rather to avoid any superfluous counterpoint to the translation of the text itself. On the other hand, we always indicate in a note the complementary information it provides and the occasions where we have thought it necessary to adopt an interpretation other than the one in the commentary.

Introduction

2Literary traditions, like conformist societies, have difficulty in appreciating independent individuals. Suspicion and sometimes ostracism are the prices paid for superficial or profound divergence from a norm which is all the more strict for being the more arbitrarily defined. There can, however, be no doubt about their membership in a group that seems to reject them. Such today is the literary fate of Paripāṭal. It is accepted into the collection of Anthologies which constitutes what is called the Caṅkam, since its presence is necessary to the conserving of the traditional physiognomy of the totality, but as a poor relation, mutilated and misunderstood or ignored. It is quoted certainly, but most often according to the résumés by Swaminatha Aiyyar,1 and not in terms of the text, which leads to it being credited with the thought of its commentator, Parimēlaḻakar, who is posterior to it by six centuries or more. It is usually avoided since it has the reputation of presenting its readers with more pitfalls than riches. Those who want, arbitrarily, to see in ancient Tamil literature a pure manifestation of the Dravidian genius are ready to take for truth the conceit of the learned Sanskritist of Madras, V. Raghavan: “The authors of Paripāṭal are the fifth column of Caṅkam literature!” and to accuse these texts of being a later compilation of puranic stories, garnished with Sanskrit vocabulary. This idea is reinforced by the number of parallel Sanskrit texts offered, almost to satiety, by the rare modern Vaishnava commentators on the hymns to Tirumāl,2 perhaps somewhat more often cited than the others, either because they are religious or simply because they open the collection, thus giving credit to the idea of Paripāṭal as, first and foremost, a devotional text even though almost half of what remains of it is profane. That proportion was most probably the same for the complete text, for only a little more than a third of the content of Paripāṭal has been rescued from oblivion and the remainder seems to be irrevocably lost. Moreover, it was meant to be sung, but to melodies about which we know almost nothing except the names, and not always those, so that it is somewhat of a libretto without the score.

3A meagre assessment indeed! How then is it to be explained that this work has some enthusiastic admirers,3 and that we ourselves are convinced of its grandeur and importance? It is unnecessary to defend it. The simple reading of this small group of poems: twenty-four and some fragments, will be sufficient in itself. They constitute what are most probably the very earliest religious hymns in Tamil literature, thus prefiguring the Aḻvār in singing the glory of Tirumāl (notwithstanding the fact that they are conspicuously ignored by all the medieval commentators of the Vaishnava canon) and they introduce us, on a par with Tirumurukāṟṟuppaṭai, to the popularity of the cult of Murukaṉ, the Tamil equivalent of Skanda. They bring to the rich literature celebrating the city of Madurai some of its most beautiful verses and, to the ludic, erotic lyricism widely represented in the Caṅkam, a particular strain of playfulness and of joyful games on the banks of the Vaiyai. In it, finally, we get almost the only example of a metre and of a literary form, half way to theatre, which is not without analogy to the Greek Bucolics or the Mimes of Herondas, for example. To facilitate the reading, we shall cast an eye on its historical and literary context before presenting a brief inventory of its principal themes.

PART I. Background

4The term Paripāṭal designates one of the “eight anthologies” (eṭṭuttokai) which, with another collection of ten songs (pattuppāṭṭu) and a treatise on grammar and rhetoric, Tolkāppiyam, constitute the essential of the earliest Tamil literature, called collectively Caṅkam literature. The word Paripāṭal refers to each poem as much as to the complete collection since it defines a particular sort of poem, and since the collection, made up – in the image of the other seven – according to a strict criterion of form, contains nothing but Paripāṭal; for example, Kalittokai, another anthology, is the collection of poems composed in kali metre. We know of no other Paripāṭal than those we are presenting, apart from a single later imitation, in praise of Tirumāl, in Pappāviṉam, a sylistic exercise by a 17^th c. man of literature.

5If pāṭal means “song”, then the verbal root pari- is unclear. The ancient Tamil commentators of Tolkāppiyam explain everything by the verb parital, with its multiple meanings, of which they here select “to accept”, “to participate in”, and “to take upon oneself”.4 This may amount simply to wordplay for a Paripāṭal is defined as a metre composed of all other known meters. If we accept this version for want of a better, we may attempt to translate Paripāṭal by “mixed song”, a sort of pot-pourri analogous to the well-known satura of early classical Latin literature. The hypothesis remaining moot, we shall turn to a definition of the form and content of this type of poem and to place the collection in relation to the totality of Caṅkam texts. First of all, the Caṅkam itself must, however, be briefly discussed without, at the same time, entering into the detail of a complicated controversy, in itself the subject of many theses, and whose often polemical bibliography is virtually endless.5

I - Caṅkam: an academism and an elusive academy

6Tamil literature begins brilliantly with a set of anthologies arranged according to criteria of form as well as of content and consisting of more than a hundred thousand verses. These texts, covering an indeterminate period, came to be called Caṅkam from the name of the assembly of poets who were the authors. Legend has it that three Caṅkams succeeded one another over a fabulously long period. History still has only a very tentative grasp of these assemblies: the internal chronology and the external synchronisms6 have so far been established only as probabilities. Not to speak of the authors who have taken legendary givens for authentic and take us back ten millennia before the beginning of this era: dates go from 6^th c. BCE to 10^th c. CE. Serious discussion tends, even so, to converge upon the hypothesis of the original compositions being spread between 2^nd c. and 4^th c. CE,7 and the compiling of poems into anthologies, probably with the addition of some later pieces, as belonging to the period immediately following. Some falling off seems therefore to be hinted at from the 5^th c., if not earlier. Histories of Tamil literature ordinarily encompass the period from 4^th c. to 8^th c. with the epics and the collections of moral works, whilst religious poetry was coming into being with the development of sects. The problem is, in fact, extremely complex and depends upon the problematical examination of doubtful data. These texts in their entirety, which are not essentially historical documents, are displayed at will to give a semblance of consistency to erudite ignorance, to which we are often reduced as regards the first six centuries of our era. In other words, the well known religious literature of the Bhakta of Vishnu and of Siva was not in existence before the middle of the 6^th c. even though the kingdoms of the South are attested to from the 3^rd c. BCE, as the Pāṇṭiya kings are, that is, from the time of Megasthenes. We therefore read the Caṅkam texts as the literary expression of generations which border upon the beginning of our era, and find ourselves faced with the double necessity of covering a large expanse of centuries and of respecting the homogenous coherence of that literature. The period is sometimes lengthened by the distinguishing of an early group anterior to the 3^rd c., and a later group covering the 4^th c. to the 6^th c. with Kalittokai, Paripāṭal, eighteen minor poems and the epics.8 Behind all this is primarily the preoccupation with presenting early Tamil literature in terms of a coherent development, without any missing links. This may seem intellectually satisfying but is not to be relied upon, and could easily be a modern version of the legend of the three Caṅkams.

7This legend is presented in full in its ancient form for the first time by the Commentary of the 1^st sūtra of Iṟaiyaṉār akapporuḷ, attributed to Nakkīrar and dated between 750 and 1000, in the following terms:9

“The Pāṇṭiyas maintained three Caṅkams, the first, the middle, and the last. It is said that to the first Caṅkam belonged Akattiyaṉar, the god with huge tresses who burnt Tiripuram (Siva), Murukavēḷ who destroyed the mountain (Krauñca), Muṟiñciyūr Muṭinākarāyar, the master of wealth (a name of Kubera, perhaps here a poet), and in all five hundred and forty-nine. It is said that, these included, four thousand, four hundred and forty-nine composed the poems. Their compositions comprise innumerable Paripāṭal, Mutunārai, Mutukuruku and Kaḷariyāvirai. It is said that they held this Caṅkam for 4440 years. Those who maintained them in the Caṅkam were eighty-nine, starting with Kāyciṉa Vaḻuti to Kaṭuṅkōṉ. It is said that amongst them seven Pāṇṭiya were poets-laureates. It is said that it was at Madurai, swallowed up by the sea, that they held their Caṅkam and studied Tamil. It is said that their (grammatical) treatise was Akattiyam.
Then it is said that belonging to the Middle Caṅkam were Akattiyaṉār, Tolkāppiyaṉār, Mōci and Karuṅkōḻi of Irantaiyūr, Kāppiyaṉ of Veḷḷūr, Ciṟupāṇṭaraṅkaṉ, Tiraiyaṉ Māṟaṉ, Tuvaraikkōmāṉ and Kirantai, fifty-nine in all. It is said that, these included, three thousand, seven hundred composed the poems. It is said that their compositions comprised Kali, Kuruku, Veṇṭāḷi and Viyāḻamālai akaval. For these, it is said, the treatises were Akattiyam, Tolkāppiyam, Mā-purāṇam, Icai-nuṇukkam and Putta-purāṇam. It is said that they held their Caṅkam for 3700 years. It is said that those who maintained them in the Caṅkam were fifty-nine, beginning with Veṇṭērcceḻiyaṉ to Muṭattirumāṟaṉ. Amongst them, it is said, five Pāṇṭiya were poets-laureates. It is said that it was at Kapāṭapuram that they held their Caṅkam and studied Tamil. It is said that it may have been during this epoch that the Pāṇṭiyaṉ land was submerged by the sea.
Then it is said that studying Tamil in the last Caṅkam were Ciṟumētāviyār, Cēntampūtaṉār, Aṟivuṭaiyaṉār, Peruṅkuṉṟūrkkiḻār, Iḷantirumāṟaṉ, Maturai āciriyar Nallantuvaṉār, Maturai Marutaṉiḷanākaṉār and Nakkīraṉār, son of Kaṇakkāyaṉar, forty-nine in all. It is said that, these included, four hundred and forty-nine composed the poems. It is said that their compositions were Neṭuntokai nāṉūru, Kuṟuntokai nāṉūṟu, Naṟṟiṇai nāṉuṟu, Puṟanāṉūṟu, Aiṅkuṟunūṟu, Patiṟṟuppattu, a hundred and five kali, seventy Paripāṭal, kūttu, Vari, Ciṟṟicai and pēricai. It is said that their treatises were Akattiyam and Tolkāppiyam. It is said that they held their Caṅkam and studied Tamil for 1850 years. It is said that those who maintained them in the Caṅkam numbered fortynine, beginning with Muṭatirumāṟaṉ, present when the sea came in,10 to Ukkirapperuvaḻuti. It is said that amongst these three Pāṇṭiyas were poetslaureates. It is said that it was in the Uttaramaturai that they held their Caṅkam and studied Tamil.”

8This essential text, in which evidently mythic numbers and mythic elements may perhaps hint at historical memories, would thereafter continue to be referred to. Prior to it, some fleeting allusions had been interpreted by critics11 as possible references to these Caṅkams, or at least to the last of the three. None of them is convincing. It is in fact particularly surprising to find in this history no explicit reference to the Caṅkam in the so called Caṅkam texts themselves: the very earliest (7^th c. CE) and the most probable, in fact, goes back to the Saiva saint Tirunāvukkaracar, and again in a hymn to Tirupputtūr and not to Madurai.12 In 10^th c. is cited the great inscription on copper of Siṉṉamanūr (S. I. I. vol. III, 4, pp. 454-460) but it is already posterior to the text cited above.13 In 12^th c. and then 14^th c., several of the legends of the Tiruviḷaiyāṭaṟpurāṇam clearly refer to it, but we are in the age of Commentators for whom our text already has the value of dogma. It is still today the principal and, in fact unique, document, and it is interesting to note that the “hypercritical” Vaiyapuri Pillai himself refers to the given figures for the number of poets in the Last Caṅkam to support the argument for excluding Paripāṭal, which would otherwise be a rather singular opinion, given that the same text explicitly cites the seventy Paripāṭal amongst the products of that academy.14

9It is slightly surprising also to find these Tamil academies bearing a Sanskrit name15 and we wonder if they may not perhaps have taken their name from that of Buddhist, or more likely Jain, communities (sangha), the latter having certainly had more literary influence.16 We know for a fact that a Jain Dravida-Sangha was founded at Madurai in 470, and that the Jains played a considerable role in Tamil studies from the 5^th century. But this tells us fundamentally very little, since the Caṅkam is for us essentially a literary corpus with a great homogeneity of vocabulary and themes and faithful to a set of conventions to which Tolkāppiyam and some other later treatises give us the keys. It is this homogeneity and the constant presence of this conventional framework, which stimulates or limits the originality of the poets, which ultimately constitute the strongest support of the legend. Is it possible for such an academism to have existed without something resembling an academy? Is it possible for all these poets to have had such an acute awareness of the rules of their art and of their own precise place in society and in the courts of kings or chieftains, without their community having some kind of consistency?

10It may not be idle to remember that in the Caṅkam, as in Sanskrit and pretty generally in Indian literature, formalism is essential to poetic exercise. Such an affirmation may sometimes be misleading. It aims less at excluding the aspect of personal artistic creation as at emphasising that it is expressed only in a recognised order and by variations on approved formulae. Poetic invention is not engaged with the raw richness of the common language but with a vocabulary and a system of conventions which constitutes a kind of “metalanguage” in which literary expression ties and unties itself and outside of which no communication is possible at the level of art. In this perspective, we may understand how the Tamil literary tradition had the authority, against historical probability perhaps, to proclaim the anteriority of the treatise on grammar and rhetoric, Tolkāppiyam, to all known anthologies: every poem in each of the collections in fact presupposes the necessary existence of an entire system; no situation makes sense except with reference to the totality; each image and symbol speaks only from within a unique thematic. The Poruḷ, material or meaning, is hardly more than a collection of constitutive themes of the poetic substance of the Caṅkam, an inventory, not of an enumerative type but well organized according to a system of inter-reference. It will certainly not do to rhapsodise over the astonishing modernity of that structural(ist) vision of the poetic universe, especially at a time when a perfectly ordinary compliment risks losing all significance. It will rather be preferable to draw out the consequences for the subject of our study.

11On the literary level we first of all have, rather than the right, the absolute duty, never, in understanding a poem or episode, to lose sight of the system of references which explains it. Certainly, the situations, especially here the amorous ones in the poems to the Vaiyai, are never quite clear except in the prose epilogue which accompanies the poem and which is probably the ulterior draft. At first sight this interpretation appears to be imaginary and to be the work of the commentator and one might be tempted to substitute a different reading for it. It is in fact, of all the interpretations, the least arbitrary and the least imaginary, not only because all it does is to explain the situations, the sentiments, and the images according to the strictest code there is, the knowledge of which is essential to the understanding of the text, but also because the vocabulary of the poet, the images to which he resorts and the situations he sets up in cryptic words are as much implicit references for an informed reading. The author himself is, moreover, sometimes explicit to the point of defining his themes,17 and pedantic to the point of giving a lesson in rhetoric, for example on the two kinds of amorous life and their essential characteristics.18 This pedantry is elsewhere redeemed by the healthy self-esteem of a man of letters wherein the Tamil sense of pride most probably finds the first expression of its claims. With a fairly clear idea of the particular literary character of these songs and in the face of a quite direct appeal to the grammar they require, how may we proceed if not by according to traditional rhetoric the place to which it has every right?

12As regards the historical interpretation, we must never forget that what we are considering is a literary work, very conscious of being one and, as such, it is the concern of literary criticism and only accidentally bears sociological witness. All the efforts of the Caṅkam poets to elaborate their poetry outside the common language run counter to any attempt to treat their work as a simple document of material or social life or of popular beliefs and concepts. Between the Anthologies and quotidian reality in Tamil Nadu in the first centuries of our era, there is as much distance as between the poetry of the troubadours and the civilisation of the middle ages as described by historians and economists. Rhetoric, learning, and a refined culture: we are far from the popular, and later, writings familiar to us in other Indian languages; we are closer to Kālidāsa, who may, after all, be almost contemporaneous with these poems. Then again, who would think of treating the Raghuvaṃśa as a sociological document without first referring to the rules of kāvya?

13Let us then, as regards the Caṅkam, before any interpretation of Paripāṭal, at least remember that of all the historical realities to which it does bear witness, artistic and intellectual life is the most indubitable, and the hypothesis of a tradition of poets grouped around the princes and rich citizens of Maturai, the least risky. We may certainly find the word “Caṅkam” unsatisfactory since, objectively speaking, it refers to a Collection19 of texts rather than to an assembly of poets but, given that a modern tradition has established it, we maintain it as symbolic homage to the intense literary activity at Maturai during the 1^st centuries CE, under the Pāṇṭiyaṉ kings.

II - The Paripaṭal

14Notwithstanding a continuous tradition of lettered readers and commentators, the memory of this magnificent literature faded over the centuries and we owe its resurrection to the activity of various scholars and researchers of the 19^th c. and beginning of 20^th c. The exhumation of Paripāṭal was, alas, only partial. The text was edited for the first time in 1918 by Swaminatha Ayyar who had discovered and identified the poems from some incomplete manuscripts. Since then no new discovery has been possible. The original manuscripts are the following:

• a ms. belonging to the Tiruvāvaṭuturai Ātīṉam, missing its first and last leaves, contains the end of the commentary on the first poem and goes up to the commentary on v. 38 of poem XIX;
• two mss. acquired by SA from Te. Lakshmanakavirayar of Alvartirunakari, one a copy of the other and both in a fairly bad state, constitute the basis of the edition. They contain the text and commentary of poems II to XXII (incomplete). They are in the Swaminathayar Library at Adyar today;
• a ms. made of two leaves, belonging to Tarumapura Ātīṉam, contains poem V and gives at the end the name of Parimēlaḻ akar as the author of the commentary;
• a ms. of two leaves (text only) has been given to SA. by Ra. Irakavayankar, a poet from Ramnad.

15The identification of the text is formal. Poem V is in fact summarised and quoted by the early commentator of Tirumurukāṟṟuppaṭai (v. 58 & 255), Naccinārkkiṉiyar, who also noted the fine sanskritism of XXI, 3, carumam “skin” (cf. sk. carma) in his Com. of Tolkāppiyam (ecc. su. 6), and the first poem was known to the commentators of Tolkāppiyam. We may notice moreover that these commentators never challenge the membership of the Paripāṭal20 in the Caṅkam and that they don’t, when speaking of it, differentiate it especially. It is in their works too that most of the fragments that complete our edition are collected, but the attribution of these fragments to Paripāṭal is never accompanied by any indication that would allow of their being allotted a number in the order of the collection.

16A further confirmation of the authenticity of the Paripāṭal that we know of is found in the glosses that accompany the only other Paripāṭal which exist in Tamil literature: a collection entitled Pāppāviṉam21 composed according to all known Tamil metres which ends with five Paripāṭal, oddly given as if in an appendix, for the word “END” appears twice in this text, before and after these five poems. The author must have been, if not the actual author, at least the commentator of the Māṟaṉalaṅkāram22 a very interesting treatise on poetry from the middle of the 16^th c. (a work by the same author is precisely dated: 1548), who also cites our poems. Several critical notes by the author of Pāppāviṉam concern Paripāṭal as we know it. These were of use to SA in its identification and have helped us to define the rules. They date, at the latest, from the beginning of the 17^th century.

17The identity of the commentator of Paripāṭal is no less certain and was known before the text itself. The editor in chief of the review Centamiḻ23 in fact revealed in 1901 that a commentary by Parimēlaḻakar to Kuṟaḷ has on the first page the text glorifying its annotator which appears today at the head of the edition of Paripāṭal by SA. This text is contained in a ms. belonging to Tevarpiran Kavirayar and Tayavalantirtta Kavirayar of the family of Tirumeni Irratina Kavirayar of Alvartirunakari; it is followed by the words, “Know by these presents that he who wrote a Commentary for Paripāṭal was named Parimēlaḻakaraiyaṉ”. Here is a good example of what an old family tradition means to the preservation of ancient Tamil literature: the kaviraja kecari Tirumeni Irattina kavirayar had composed already in his Tirukkuṟaḷ nuṇporuṇmālai,24 in order to elucidate the subtleties of the commentary of P. on Kuṟaḷ, several times cited passages of P’s commentary on Paripāṭal and confirmed that these two commentaries were by the same author. The abundance of comparisons proves this beyond any doubt.

18Parimēlaḻakar lived towards the end of 13^th c.25 He possessed a solid culture in Tamil and Sanskrit which he often displays, especially in commenting on hymns to Tirumāl or in referring to Yoga, Saṃkhya or Mīmāṃsa systems or Pāñcarātra and Vyūha. Much as he is believed to have been Vaishnava, he exhibits no sectarian tendency and never even quotes from the Nālāyira TivyapPirapantam in support of his comments on those hymns. SA believes that he belonged to a family of chief priests of the temple of Ulakaḷanta Perumāḷ at Kanchi and we concur, but some other texts have him as native to Okkur and he may also have stayed temporarily in Maturai. He would have been a slightly younger contemporary of Cēṉāvaraiyar, and earlier than Nacciṉārkkiṉiyar (who criticizes his Com, in IX 57 under Muruku. 106; cf. also note SA, under X 126-131). [See the article on Commentators of Tirukkuṟaḷ].

19A final exercise of doubt may not, however, be dispensed with. Our evil genius here is a woman, Kantiyār, who is held responsible for interpolations in Paripāṭal which are supposed to be in rather doubtful taste. The poem to the glory of the commentator forewarns us that the text we know has been carefully expurgated by P. But it is a common convention, shared with Sanskrit, to praise the commentators for having amended the text; we are told, moreover, that this same Kantiyār, perhaps herself a poet from Perai, even suspected of being a Jaina ascetic named Kavunti, might have contributed as many as 400 additions to Cīvakacintāmaṇi that Naccinārkkiṉiyar has either not detected or not preferred to other readings. Though this is unconfirmed, an anonymous author reveals that an old ms. from Arrur, called Palakavittiraṭṭu and given by A. M. Malaiyappillai, also attributes to that same Kantiyār the ancient and anonymous veṇpā which we are going to quote in order to set down the subjects of the 70 original Paripāṭal;26 we seem to revolve within a very small world indeed.

20But this throws a faint suspicion on the document, which does not make it any the less important, it being the only one to instruct us on what we have lost and on the value of the sample which remains. It also confirms the figure of 70 given by the commentator of Iṟaiyaṉārakapporuḷ mentioned above, and by Pērāciriyar under Tolk. Poruḷ. 461. The content of the poems is, in fact, given in the form of the following mnemonic quatrain:

For Tirumāl, eight, for Cevvēḷ, thirty-one,
One for the Guardian of the forest (var.: for the beginning of the season of kār)
For sweet Vaiyai where union takes place, twenty-six, four for great Maturai,
That is, so it is said, the nature of the perfect Paripāṭal

21The reading kāṭukaṭkoṉṟu for kārkoḷukkoṉṟu is found in two ms., from Alvartirunakari and Arumukamankalam with a clarifying note taken from the commentary to Muruku. by N.: “at that time kāṭukāḷ was the name of the one called Kāṭu kiḻāḷ today”, that is, Durga, goddess of the forest,27 the mother, according to a Tamil tradition, of Murukaṉ, under the name Koṟṟavai.

22We nowadays possess almost all the poems to Tirumāl: seven out of eight, with a quarter of those to Cevvēḷ: eight out of thirty-one, and a third of those to Vaiyai: nine out of twenty-six; we have only a few fragments on Maturai. On the basis of 70 poems, Tirumāl gets above 11%, Cevvēḷ more than 44%, the Vaiyai more than 37%, Maturai less than 6% (but Vaiyai plus Maturai, 43%) Kāṭu kiḻāḷ 1.5%.

23This detailed accounting is also of interest in that it helps us to define Paripāṭal by its content. Tolk. (Poruḷ 433) gives it love (kāmam) for its principal theme, treated in an ideal manner or according to the usages of the world (ibid 53; 56 according to the com. by Ilamp.). The commentators, concerned with adjusting a rather different and more complex reality to that dictat, are embarrassed. Pēraciriyar, however, maintains that whether in the case of an invocation to God, of games in the mountains or of games in the water, etc., love must be the main reference (com. to 433). Iḷampūraṇar more prudently accommodates the descriptions of mountains, rivers and cities (com. to Ceyyuḷ. 117) and Nacciṉārkkiṉiyar accepts the invocations to God as well as the poems treating of love (under Akat. 53). This in fact gives us one of the reasons for keeping Paripāṭal apart: the great division in Caṅkam literature between the themes of akam (love) and puṟam (the exterior and everything else) is not respected in it, and, even though Tolk. assigns akam to it for subject, it freely goes beyond that and mixes genres too.

24This formalism is a shade puerile since a mixture of genres exists, in detail, in plenty of other Caṅkam texts. The comparisons in particular, often prolonged or developed as allegories, are the frequent occasions of the borrowing of akam themes for a puṟam poem, and the reverse.28 Furthermore, this mixture of genres is often emphasized, starting with Nacciṉārkkiṉiyar himself, in the longest Caṅkam poems, those of Pattuppāṭṭu. We may even suppose that it was the theoretical necessity of arranging the scattered elements of each theme that obliged the commentator to reconstruct the framework of such poems, sometimes abusing the procedure called māṭṭu which, in practice, consists of linking together verses or fragments of verses “according to meaning”, even if this entails disrupting the order of the text with a carelessness we find scandalous, but which is perfectly justified in that tradition. In other words, the fundamental distinction between akam and puṟam has to be understood as an ideal exigency, necessary to the Tamil genius but to which the individual temperament of the poets bows only when it suits them.

25On the other hand, in admitting that we are bound always to accept the arbitration of Tolk. and of its scholars, we note that the formal characteristic of Paripāṭal is diversity and liberty. Why then does this not extend to the content itself? The fact is that the interrogation of the texts has been neglected. The answer lies in the declaration of XVII 46, in which religious festivals and profane rejoicing are declared “to alternate and be mistaken for one another according to usage” and, the author adds, “it is good thus”. The words used by the poet express much more than a simple co-existence: the word taṭumāṟṟam implies deliberate ambiguity; it is connected with the stylistic figure called taṭumāṟutti, that is an inversion of cause and result (P. refers explicitly to it) and to the one called taṭumāṟuvamam, that is either an inverted comparison, or an equivocation between two terms so similar that it is easy to mistake one for the other: to speak about the one as if taking it for the other: is this not precisely the rule followed by these poems? It therefore seems to us that it is the deliberate originality of this poetic genre to unite the flights of the heart and those of the soul in a single song, and that there sometimes is in it, apart from a very natural human movement, an original pre-figuration of the Bhakti attitudes which were to come into being soon after. We may consider, for example, that Tirukkōvaiyār of Māṇikkavācakar is composed entirely of akam themes but often interpreted, as a better reading (uṇmai viḷakkam) in a purely religious sense, and that Āṇṭāl too always plays on this double register. Less subtle in this regard are the authors of Paripāṭal who sing the praises of their god unambiguously or finish their erotic games with prayer; they are always perfectly aware that everything of theirs that is good must be pleasing to God.

26The same confusion reigns over the formal definition of a Paripāṭal as over the content. It is the name of a type of poetry that “takes it upon itself” to assume all the others, that is to say that it consists of all possible metres. It is shared by all and cannot be entered in any one of the great divisions of pa (metric feet). In the detail, even its composition is confused; it may have four parts: koccakam, arākam, suritakam and eruttu, but it sometimes has five parts, far less, or a few more. The commentators define them more or less precisely; P. is mute, and the editors have never analysed the poems as they do not find the necessary indications in the colophons; the only two poems which contain such comments by Iḷampūraṇar (I), and by Nacciṉārkkiṉiyar and Pērāciriyar (Fr. I) do not make the basis for a rule. The remarks of the author of Pāppāviṉam are useful in that they demonstrate the necessity for these annotations if the reader is to follow the subtleties of the composition, and also in that they contradict Tolk. on several details.29 We understand that a poem must begin with the taravu which “gives” the person addressed or spoken of (it is also the first part of the kalippā). There are poems, however, which have no taravu and there are taravu that appear in the course of a poem. The eruttu, or “nape” would normally come after the taravu, the “face”. This is the principle of Pāppāviṉam, but Tolk. does not clearly distinguish the two, sometimes considered as synonymous. The suritakam contains the ideas already expressed, and is also called aṭakiyal (content) and is therefore found at the end of the poem, but the metre is not certain: veṇpā or āciriyappā. The arākam is a fast movement where short syllables dominate. It is distinguished, if roughly, from the muṭukiyal which is a variety characterised by a great many short syllables and a finale in veṇpā. Neither of these two is featured at the beginning of a poem which, anyway, may leave out, or add one or the other. Finally, the koccakam is the first to be named by Tolk. because it almost always appears. This is the arrangement in strophes of several kuṟaḷpā, forming, as the “folds” of a garment, the meaning of its name. Moreover, the last line may have a foot or two less and, in that case, an acai takes the place of a foot. Even though this is the most characteristic formula, it is sometimes absent in a poem. Lastly, the isolated words, or hyper-metres, called coṟcīr, and stanzas of two verses of four feet, called pēreṇ (“great number”), are also intermingled in these large categories. There is overall an appearance of license, if not of confusion, but in practice we have a difficult composition, which would explain its subsequent total abandonment, with the sole exception of the fancy of an erudite epigone. Most readers have failed to distinguish the originality of this poetic formula in relation to the kalippā, and Pērāciriyar is the only one to have fought for their acceptance as two clearly demarcated genres30 even though both Iḷampūraṇar and Nacciṟṉārkkiṉiyar had the tendency to incorporate Paripāṭal into kali, which doubtless contributed even more to the extinction of the first genre in later Tamil poetry, although, as we have seen, the commentary of Iṟaiyaṉār had already attributed the composition of Paripāṭal to the first Caṅkam, which are lost forever.

27Our own perception of the Caṅkam poems in the context of their recitation also leads us to give to the “performance” perspective all its due. So, we shall, lastly, add two remarks on points not always sufficiently insisted upon. The first is that under Tolk. Ceyyuḷ 242, Pērāciriyar lets it be understood that Paripāṭal correspond to a musical language, that is, they must seek the simple expression suitable for song or drama, loaded with sweet and agreeable syllables. The second is that the recitation may be accompanied by actual mime and gesture (avinayam) which develops the meaning. This explains why some poems are complete scenes with several characters and tells us that, notwithstanding the frustration of western philologists and “poetologists” looking for pure internal evidence, we must have recourse to the commentary which alone is responsible for providing some kind of scenic indication and for distributing the elements of dialogue between the concerned partners. It also justifies the appearance of various terms from the spoken language and for so-called grammatical “neologisms”. We consider these indications as of more importance than sūtra 474 of Tolk. Poruḷ., which defines the length of Paripāṭal as between 25 and 400 verses. What remains to us varies between 32 and 140 verses.

28A final problem is that of the music. The colophon of several poems indicates, after the name of the author, the name of the musician and of the corresponding melody: an indication that would be very valuable if we knew better what the terms concerned corresponded to. The classification principle in the fragments we possess seems, in fact, to be only that of the music: we are given successively poems composed according to the modes pālaiyāḻ, nōtiram, and kāntāram. The second poem may be composed according to kuṟiñciyāḻ, as a variant suggests. Most interestingly, a musician, P. Cuntarecan, (1914-1981) offering modern equivalencies, reinstates the music which he sings in a convincing way. We refer to the exposition he gave on “the modes of the Paripāṭal”, an article in Dr. R. P. Sethu Pillai Silver jubilee Commemoration volume, Madras, 1961, pp. 253-258. His reconstruction, it must be admitted, is not unreservedly accepted by all musicologists because it assumes a continuity of the musical tradition at the historical level which is rather surprising to theoreticians. It is, however, coherent and conforms to the teachings of Vipulananta31 an uncontested authority. We are personally indebted to P. Cuntarecan for several clarifications of the musical allusions in the poems, including, for example, the identity of musical instruments (XIX 40-41) or XI 126-130 where Nallantuvaṉār is technically challenging when describing bees humming the arumpālai paṇ or the viḷarip paṇ. It must be conceded that the Cuntarecan’s theories are about the original paṇ and scales of Paripāṭal perfectly adequate to clarify the musical details as put into words by the poet so many centuries ago. Further, his sung performance of the four arākam of the first Paripāṭal, lines 14 to 25, considered as a “locus desperatus” by all modern editors, starting with SA, enabled V. M. Subrahmanya Aiyyar and myself to attempt and propose a more readable version of the text in keeping with the musical rhythm, which was clear enough to suggest a meaning. The first and fourth arākam deal respectively with the dazzling ornaments Māl wears along with the Kaustubha, and with the bow and arrows he uses to destroy the asuras (tāṉavar), while the second and third arākam are devoted to warlike aspects of the god, his broad beautiful chest, bright ornaments and flashing weapons. In contrast to the commonplace, found in the verses to the glory of P., denouncing mistakes and introduction of spoken language by the musician-interpreters of the text, and to the poor opinion professed by regular pandits about the philological capacities of the ōtuvār, it was a privilege to hear a great musician boldly opening to a renowned Tamil vidvan a novel access to a retrieved fragment of Caṅkam poetry. The result of this encounter appeared in the notes of our Paripāṭal edition.

III - Paripāṭal and the eight anthologies: A chronological dilemma

29Thus defined, Paripāṭal traditionally belongs to the principal “Eight Anthologies” which constitute one of the collections of the Caṅkam which are all, after the same fashion, anthologies arranged by theme and according to even more formal criteria, such as the length of the pieces that make them up. This traditional ordering is, however, challenged by authors who, wanting to see in Paripāṭal a later accretion, have thus sought to eject it from a primitive corpus, which has it must however, be said, no definition other than the one they have chosen. Kalittokai and Tirumurukāṟṟuppaṭai share this disgrace. We must now attempt to examine the reasons for such discrimination.

30The first is the linguistic argumentation. Vaiyapuri Pillai32 gives twenty or so Sanskrit words which attest to the later origin of, for example, Paripāṭal. But the same author also compiled a much longer list of Sanskrit terms which he attributes respectively to Brahmins, Buddhists and Jains, thus attesting only to a very significant Aryan presence at the heart of the Caṅkam and destroying at a blow the substance of this argument against Paripāṭal. The only obvious thing is that the importance of Sanskrit terms in Paripāṭal is partly the function of the subjects treated: several of the terms mentioned appear in a religious context. Moreover, we have not, up till now, had any reliable statistics, the handling of which would be quite delicate anyway, precisely because of the absolute necessity of taking into account the differences in genre between one anthology and another.33 Furthermore, to establish a chronology relating to the Caṅkam texts according to the proportion of Sanskrit vocabulary in them is, without a doubt, to depend on shaky reasoning based upon the postulate of a progressive “Aryanization” of Tamil and the summary idea that that process can be measured by a single formal criterion: Sanskrit vocabulary. We are aware of the more complex reality. The cultural fusion between Dravidian and Indo-Aryan elements has had, first of all, to operate preference through the intermediary of the prakrits and what’s more, those of an early date; then again, borrowings by the Caṅkam from Sanskrit and Middle-Indian have not been sufficiently catalogued, even though, in fact, they do give proof of antiquity.34 Tamil, furthermore, sometimes employs terms that are indigenous or considered to be so to designate factors of a culture which was pan-Indian, and sometimes even specifically Vedic,35 a likely indication that a double register of terms was available to the authors. The subject of a text ultimately determines its vocabulary and it is evident that puṟam texts offer a priori more Sanskrit terms than do the akam anthologies, unless, like Kalittokai, they shine with their eclecticism. We are therefore inclined to speak of style rather than of chronology, and even to conclude with the paradox that there are, proportionally, less Sanskrit words in Paripāṭal than in some of the earliest Tamil Brāhmī inscriptions.

31Another linguistic argument instances lexical and, more often, grammatical forms that are later or more common.36 This too, weightier though it may be is not deciding, because, as we have already seen, the Paripāṭal, to a greater extent than other texts, is by definition related to the spoken language. We are sorry to say that the grammatical remarks put forward towards a chronological argumentation demonstrate, once more, nothing but a difference in register.

32For more reason still, the presence of puranic themes in the religious portions of the texts does not reveal these texts to be later but only that they are treating puranic subjects. We shall later show that the religious beliefs of Paripāṭal are far from being unique within the Caṅkam corpus and, on the contrary, often have their echoes elsewhere in it. The Vedic elements in the Caṅkam and the Aryan influences in these texts have been listed too often for there to be any point in repeating them. K. A. N. Sastri speaks of a pronounced cultural fusion and he demonstrates it in two pages37 containing thirty quotations, of which there are none from Paripāṭal, and only one from Muruku.38 The refutation of reasoning on thematic dating was also taken up by M. Rajamanikkam;39 interested readers may refer to him.

33It would be gratifying to be able to construct an argument from the names of the authors of the poems but we unfortunately know so little about any of them; there are moreover possible homonyms and numerous orthographical variants. Here too, Vaiyapuri Pillai comes out against the suggested comparisons. A matter of temperament: where others see an opportunity for connections and seize it, he sees reason to doubt and takes hold of that. Ka. Kovintan recognises in his extensive Tamil compilation on History of the Caṅkam Poets40 that he knew nothing of Nalvaḻutiyār beyond the fact that he sings the Vaiyai in Paripāṭal (XII). The same avowal would have to be made regarding the other authors; their “biographies” have been written all the same, on the lines of a genre which resorts to eulogy more than to history and demands a certain abundance. Here this is provided, in default of documents and by way of risky speculations on the names of authors and of the exegesis of their works; the result is sure to be precarious, but how are we to go beyond this?

34K. A. N. Sastri41 places Nalvaḻuti in the Pāṇṭiya family and accepts the identification of Iḷamperuvaḻuti (poem XV) with the author of the same name, but preceded by the epithet “who died in the sea”, and who is the author of Naṟṟiṇai 55 and 66, and of Puṟam 182? SA. too accepts that identity, but Vaiyapuri Pillai naturally refutes it on the pretext that the form Intirar employed in Puṟam 182 proves that the author is Jain and consequently cannot have written a Vaishnava bhakta poem.42 On the other hand, it may be somewhat arbitrary to attach Nalleḻuṉiyār (XIII) to the family of Atiyamān Neṭumāṉañci on the sole pretext that the son of the latter was called Eḻiṉi. Kīrantaiyār (II) is mentioned among the members of the second Caṅkam which has no Paripāṭal among its assets, and the same name appears in the Cilap. (23, 42); these mentions may in fact apply to three different figures. Because Kēcavaṉār also wrote the music for his poem (XIV) it has been deduced that musicians too may have been poets and that there is a relation, as well, between Kaṇṇākaṉār (music for IX) and the author of Naṟṟiṇai 79 and Puṟam 218 who has the same name, and Naṉṉākaṉār (music for V and XII) and the author of Puṟam 381. Nothing either affirms or invalidates these connections.

35There remains the case of Nallantuvaṉār. According to some colophons he is the sole author of four poems (VI, VIII, XI, XX) which are amongst the finest in the collection. Also attributed to him is the invocation poem of Kali., the last part of this collection (118 to 150: neytal), as well as the responsibility for having compiled the anthology. Some authors, such as Sivaraja Pillai,43 even feel that Kali. has such unity of expression that it may be the work of a single author, who would thus be our poet.44 Vaiyapuri Pillai refuses to identify him with Maturai Āciriyar Nallantuvaṉār of Akam 43, or with Antuvaṉ cited in Akam 59 as having sung of the mountain of Tirupparankunram. But there is another Nallantuvaṉār, author or Naṟṟiṇai 88, and it is, in fact, possible that this author appears in four of the Eight Anthologies. Vaiyapuri Pillai judged it safer to see different authors there.

36As for poem 59 of Akam which cites “the high mountain, full of sandal, sung of by Antuvaṉ, the fresh Paraṅkuṉṟu of Muruku of great wrath”, its author is Maturai Marutaṉ Iḷanākaṉ, a name found in several other anthologies; to him is also attributed the third part (marutam) of Kali. Any reader will understand those two lines as an obvious homage to the author of Paripāṭal VIII, the beautiful poem to Cevvēḷ, with lavish descriptions of His mountain, Paraṅkuṉṟu. But we also see how these similarities of name are bifurcated: an argument for better integrating Paripāṭal and Kali. to the rest of the anthologies, they can as well point at Nallantuvaṉār as a talented forger, or at the very least arouse suspicions of poem 59 of Akam as being a later interpolation. There is not, in fact, any decisive argument on either side.

37[However, we must notice: 1) that I. Mahadevan infers from the inscription 55 on Tirupparaṅkuṉṟam of his corpus the existence, as early as that Brāhmī inscription, of a learned family of Antuvaṉs attached to that hill, 2) that both Marutaṉ Iḷanākaṉār and Nallantuvaṉār, are considered as belonging to Maturai, 3) that they are cited side by side amongst the poets of the Last Caṅkam listed in Iṟaiyaṉār Akapporuḷ commentary, 4) that the first plays a prominent role there as the next best interpreter of the recovered Akapporuḷ, and, 5) that the second is repeatedly given the title of Āciriyar (“master” or perhaps “professor”, as David Buck would have it) which makes sense when we deal, as below, with precise technical astronomical data in a literary milieu quite familiar with the subject, as several poems, such as Puṟam 229, attest. See also the article on Caṅkam in this vol.]

38All these vagaries make it even more regrettable that this same author, Nallantuvaṉār, is considered to be the author of poem XI, which starts with the horoscope that now claims our attention. We may imagine, in the face of uncertain discussions of chronology, the value of an exterior argument, absolute and irrefutable, such as an astronomical dating. This one was believed to have been found in the first lines of poem XI which describe an eclipse of the moon with a complete horoscope. The prudent in general avoid drawing conclusions from it,45 but it is presented often enough to make it necessary for us to summarise the dispute once more here.

39There was a fashion, in Madras, between 1920 and 1930, for posing this type of problem. Thus, in its first volume, the Journal of Oriental Research suggested two dates for Āṇṭāl in two different articles: 731 and 850 CE; the controversy was continued without any final result in the next volume. As regards Paripāṭal it was Centamiḻ that served as the stage for a long and passionate debate which is to be found spread out between vols. XVI and XXII.

40Subrahmanya Sastri of Tancavur, following the general considerations of the musicologist Abraham Paṇṭitar on the astronomical knowledge of the early Tamils, was concerned with demonstrating the falsehood of the horoscope. Naraswamy Ayyar concluded similarly but, on the insistence of Narayana Ayyankar and R. Ganapati Ayyar, he reconsidered the problem to conclude, on purely hypothetical grounds: “372 CE, 15^th day of the month of Āṭi”. This did not satisfy Somasundara Desigar, who protested and arrived at the date of 161 BCE, a date recorded as definitive by an editor of Paripāṭal, who subsequently presented it with a manifestly unrealisable schema.46 This entire part of the controversy seemed to be nothing but the apologetic desire to award an ancient date. The date of 161 BCE almost coincides with only a fifth of the information given in the text, which proves how irrelevant it is.47 Naraswamy Ayyar’s position was more solid and there is an historian of Tamil literature who continues to maintain it as a chronological argument even now.48

41The real scientific debate is found, in fact, in the work of Swamikkannu Pillai. After hesitating between 634 and 17, he concludes in favour of 17^th June 634, and that date, which matches the quasi-totality of the information given in the text, has never been seriously questioned, for we may not consider as refutation the pure and simple rejection this date usually provokes amongst literary historians, with the notable exception of Vaiyapuri Pillai. Only K. G. Sankar, who had first suggested 254, reverted to 27^th July 17, the other probable but finally excluded date, because it more closely coincides with the heliacal rising of Agastya; this though accurate, forced him to leave out of account the position of Venus and Mercury and to suppose, quite gratuitously, that Hindus didn’t know how to calculate these planets correctly.49 Further, the discussion on this point appears futile if we consider that the mention of Agastya is not relevant for dating the horoscope itself, while the positions of the planets are.

42Must we then accept the conclusion of Sw. P in An Indian Ephemeris (I, 1, § 243-271). We have long hesitated to do so, for several reasons of varying importance.

43The first is that the dating of a horoscope has to take into account all the elements of that horoscope. This elementary idea has too often been lost sight of in the course of the controversy, too many authors having been satisfied with partial solutions which must, in the light of scientific rigour, be rejected as insufficient: and, as Sw. P. himself writes, “… unless one was certain of all the elements in an horoscope having been correctly recorded, the time inference drawn there from may turn out to be widely discrepant from the truth.” Then, Sw. P. himself leaves room for discussion, rather than for serious doubt, about three details:

1. The expression of v. 7: aṅki uyar niṟpa (lit.: “the fire standing high”) has been interpreted by P. as meaning that the Pleiades (aṅki, the fire) have reached the meridian and he has deduced from this that the Sun is in Leo. At first sight the deduction is irrefutable: if the Pleiades are at the Zenith at daybreak, the Sun is naturally in the constellation that is on the horizon at sunrise, with a difference of 90^o, and is thus one of the constellations that make up the sign Leo. But according to Sw. P., who obviously had till then done his best to follow P.’s hypothesis at all costs, on the 17^th June 634, the Sun must have been in Cancer. P’s deduction must therefore be rejected: uyar niṟpa does not mean “at the zenith” but only “standing high up”;50 for that matter, the Pleiades were then, in the course of the eclipse, only half way to the Zenith, but fairly “high up” when they probably became visible thanks to the darkness of the eclipse, that is, in the very terms of the poem: “at the first dawn” (pular viṭiyil). Sw. P’s thesis here might seem a bit stretched but he argues it well even though this detail is given no great significance in his reading.51
2. The exact position of Saturn seems at first sight to be debatable. On the date chosen by Sw. P. it is still in Sagittarius. The expression in v. 9 allows for two readings. According to the first, which is P’s, Saturn is in Capricorn, the sign after Sagittarius. According to the second, Sw. P’s version: “…Saturn was at the end of Dhanus (Sagittarius) and going to Makara (Capricorn)”, its own house. The date of 634 fits or not according to which version one accepts: the question can continue to be raised at any time since it depends less upon astronomy than on the interpretation of the text. However, here, Sw. P.’s does more justice to the subtlety of a poet’s text and follows grammar more strictly.52
3. In general v. 10 and 11 are understood as allusion to the heliacal rise of Agastya, the star Canopus which was certainly not visible from Maturai at the time of the eclipse of 634. Much as there are good reasons for thinking that the mention here of Agastya had no relevant place in the actual horoscope itself53, we might however concede to G. K. Sankar (see above) that this information, as given by the text, could perhaps also be taken into consideration, recognising that Sw. P. has come to grief on this point.

44The second fundamental objection to the learned endeavour of Sw. P. is that the horoscope might very well be fictive. P., who proves himself perfectly qualified as a competent astrologer, was the first to notice that every planet was in its own house in the Zodiac. Mars, Venus, Mercury and Jupiter are respectively in Aries, Taurus, Gemini and Pisces. And, to complete this beautiful schema and make it perfect, P. very astutely also places Saturn in Capricorn and the Sun in Leo. That is going too far, and if his reading were right, we should indeed, necessarily, be in the realm of fiction. In his perspective, the eclipse of the moon, the imaginary or actual, is not dateable by insertion into the absolute chronology, which makes no sense to him, but only by insertion into a liturgical cycle: the full moon of the day of āvaṉiyaviṭṭam, the date for the renewing of the Brahmin thread and, as well, that of a festival celebrated at Maturai, which had already dropped into obscurity at the time of the commentary of Iṟaiyaṉārakapporuḷ in which its memory is conserved.54

45The problem cannot unfortunately be resolved simply by recourse to its negative. It would have been quite easy in fact, for a poet as familiar with astronomy as ours was, to set up a much more satisfactory fictive horoscope than the one imagined by P., in more or less the same terms, but with Mercury in Virgo and Venus in Libra which, keeping the Sun in Leo, does not present us with the obstacle of theoretical impossibility which the preceding version does because of the maximum elongation of the two inferior planets, Venus and Mercury. Why would the poet, or even P. himself, not have thought of that? And who, besides P. who absolutely requires it for his liturgical purposes, says that the Sun was in Leo?

46Coming back to purely astronomical considerations, with the Sun in Cancer as Sw. P. correctly has it, the very extraordinary data of the actual horoscope of 634, which offers us an eclipse of the moon while Mars, Venus, Mercury and Jupiter are effectively in their own houses, would surely justify the poet, first in having fixed the memory in his verses, and then also in having determined the position of Saturn in relation to its own house which it is “longing to join” (to give to mēva, its full poetic charge) when it is still on its way, at the end of the neighbouring house; the fifth planet is thus given a more ideal, but still rigorously exact, analogy with the four others.

47There is even a fairly strong motivation here for us to dispense with searching for another raison d’être for the horoscope: truth is more wonderful than fiction.

48We would add a complementary argument to that, one to which no attention has ever been given. The second part of the poem (v. 75-77) also opens with astronomical data. To mark the start of the Tai baths the author evokes the full moon in Ardra when the Sun is in Sagittarius. This was effectively the position of these planets on Saturday 10^th December 634!55 The phenomenon is periodic so this argument adds but little weight to the calculation, which it nevertheless supports positively. On the other hand, and from the point of view of literary beauty, if we accept something of which there is little doubt, that is: that the poet, familiar with the astronomy of his time, searched the skies for a harmony that would accord with his song, we shall note that, in the first part, at the time of the floods, by the middle of June 634, the Sun is in the beginning of Cancer and the Moon at the end of Sagittarius, that is to say that their positions in relation to the second part, for the start of the Tai baths, in mid-December 634, are almost completely reversed. Is this striking symmetry between the construction of the poem and the chiasmic movement of the stars nothing but an imaginative illusion of ours, or is it the key to a poem written out of his enthusiastic vision by a great astronomer?

49All attempts to supply another explanation have run aground. The horoscope, in fact, is an aesthetic part of a record, which dates nothing else of any significance in relation to the eclipse itself, except perhaps the breaking of a monsoon over the mountain or the occurrence of more or less anticipated floods in June.56 It does not bring together any of the aspects considered as auspicious or as favourable to rain; those have been supposed, but without any justification being given.57

50What remains is therefore a unique poetic achievement inspired by the extraordinary conjunction of planets observed in the sky over Maturai in 634 CE.

51So, the best literary explanation, as well as the most careful reading of the text available to us, concords very well with the dating of 17^th June 634. Why then should we still hesitate before drawing “historical” consequences? Because, whatever may be our intimate conviction about the historical relevance of those seven lines only, to propose, a dating so crucial to what is at stake here, based on that evidence, would amount to an enormous extrapolation from literature to history.

52We should then face the consequences: there is no solution left other than to place Nallantuvaṉār between the middle and the end of 7^th c., and, once this is accepted, we must either definitively separate Paripāṭal and even the two anthologies, Kali. and Paripāṭal, from the other six and reconsider as futile all attempts to find a coherence and a unity within the entire Caṅkam, or make it all later than it is supposed to be by several centuries and consider the spreading of the Caṅkam as posterior to 5^th c., at the earliest.

53Thus, in the uncompromising terms of Sw. P. himself, the emblematic book of Kanagasabhai Pillai, Tamils 1800 years ago,: “if reprinted, as it ought to be, would have to be renamed “Tamils 1200 years ago”, for the whole work is concerned with events, persons and poems centred in the last Caṅkam which, as we have seen, flourished not more than 1200 years ago… these somewhat overdrawn pictures of the state of civilization in Southern India 1800 years ago will have to be revised in the light of our present day knowledge of epigraphy and chronology, and the scenes of the Maturai Caṅkam will have to be transferred from the first century A.D. to the late 7^th and early part of the 8^th century A.D., the period which witnessed, along with the decay of Buddhism, the rise of the Saivite and Vaishnavite teachers, Tiruñāṉacampantar, Sankaracharya, Nammāḻvār, etc.”58 (Italics are author's).

54There has always been a tendency to draw back from a solution so extreme, and this is why the Paripāṭal horoscope has been left out of most discussions of the chronology or, in contrast, tentatively and wrongly interpreted as fitting a date more in conformity with traditional sentiment.

55In solidarity with the entire discussion which opens these pages, the due consideration of this horoscope should, nevertheless, also continue to be understood within the limits of the discussion of the whole.59 On the one hand, new archaeological and epigraphical data, unknown in the days of Sw. P., will be most probably having more and more weight in the argumentation, even though no significant discovery has enriched the textual contribution to the extent of transforming “proto”-history into actual history; such considerations, however, would invite us to see the seventh century as some kind of a terminus ad quem for a Caṅkam literature which would not be extended much later. On the other hand, the very recent resurgence of various hypotheses shifting the Caṅkam literature to as late as the second millennium CE, unexpectedly and paradoxically turns the very same horoscope into an embarrassing terminus a quo, thus standing as a serious obstacle to that chronology. Quite obviously, it will be necessary to account for this. The long standing debate is therefore not yet over.

PART II. Towards an inventory

I - Maturai, everlasting lotus

56The great city, present in every text because these poems are the song of its people, is only described in a selection of fragments. There is nothing comparable with the scale of the 13 chapters of Maturaikkāṇṭam in the Cilappatikāram, nor even in the Maturaikkāñci which breathlessly describes it in a single phrase of 782 verses, and nothing like the bouquet of legends collected in Tiruviḷaiyāṭaṟpurāṇam, Cuntara Pāṇṭiyam or the Kaṭampavaṉapurāṇam concerning Siva, nor of the somewhat isolated elaboration of their Vaishnava counterpart, Kūṭaṟpurāṇam.60 If we must be content with fragments explicitly dedicated to the town itself, we may sum up what they are saying in two words: lotus and everlasting.

57Neither of these themes is completely original: the image of the city-lotuscentre of the world appears with reference to Kanchi in the Perumpṇ. (v. 402-404), but here61 it is extended with more amplitude and more detail and it is possible that these six verses have inspired the lengthily developed image in the Peruṅkatai for the description of Rajagiri. The glory of Maturai, sung by the poets who found her worthy to balance the entire world by herself in a weighing ceremony that Parañcōti attributes to Brahma himself,62 is the subject of several of the anthology’s short pieces (Fr. IX to XII), constructed on the same theme, which attest to the loyalty of the poets to their king and his capital city, their river, their mountain and their faith.

58More interesting is the information on the city which we glean from poems to Vaiyai or to Tirumāl, though there is nothing there either that has not been described elsewhere, especially in Maturaikkāñci and Cilap., which confirmed that Maturai with its great buildings and high ramparts is a prosperous commercial town, a temple town, a town of virtue and vice, of peasants and craftsmen, under the guardianship of a powerful king with invincible armies, where there is law and order and where the protection of irrigation, river-banks and sluices is assured.

59If the general description of the city in Fragment I, on Tirumāl, corresponds well enough with the terms of Maturaikkāñci this does not mean that overmuch realism is to be sought here since many of the characteristics are interchangeable with the vision given of Pukār or Kavirippumpattinam in Cilap. or Paṭṭiṉappālai. The tendency of these songs is towards eulogy rather than exactitude. Further, archaeological research is particularly difficult at Maturai which has never ceased to be a living agglomeration, a crowded business centre; the very site of the ancient city, remains loosely defined.63 The dominant impression is always the same in the texts,64 that of a civilisation at its apogee where political equilibrium and economic prosperity favour the spread of arts: briefly speaking, that which for classical Rome represented the urbanitas, there being nothing surprising in such ideal convergence. If the history, as well as the puṟam texts, leads us to suspect numerous conflicts between the monarchies of the South, only victories are echoed here and the everlasting presence of Maturai is never in doubt. The poets of Paripāṭal are in accord with the beautiful image of the Kali. in which the Pāṇṭiya king is “the master of the town encircled by water, whose ramparts have never known the siege of any adversary other than the hostility widespread around these walls, in flood times, by the Vaiyai, charmingly flowered, like a garland around the vast earth”.

60One or two features are worthy of more attention: the two names of the town first of all, Maturai and Kūṭal. The first term is explained in three ways, none of which is conclusive, nor explain the only form of the name found in four early Tamil-Brāhmī inscriptions: maturai. 1) In the first place, it is treated as a reference to Mathura in the North, the city of Krishna. No feature of Paripāṭal suggests this rapprochement, nor is there any Vaishnava allusion to a localisation in the north and even less to a migration of a religious, mercantile or ethnic character, as assumed by some historians who consider that the other name, Kūṭal (junction) is a reference to traders from the north, accompanied by Jains or Bhagavathas. 2) No epithet for the city in Paripāṭal better supports a Tamil explanation than the word maturam (sweetness), as taken from a legend of Tiruviḷaiyāṭaṟpurāṇam: the sweetness conferred by Siva, grace of a few drops of ambrosia, on the venom of the snake which was going to spread throughout the city.65 3) Paripāṭal, on the other hand, frequently celebrates the river bank with its marutam trees apropos the Vaiyai.66 It is possible that this refers to a locality, park or esplanade beside the river, suitable for the frolickings of town-dwellers. We would, today, like this to be the origin of the name of the town, called after a tree characteristic of the countryside around.67 On the linguistic level, the metathesis is possible: it is done nowadays in the reversed sense, in the colloquial pronunciation which calls the town [Marudai] and the horse [kurudai] instead of kutirai. The whole puranic tradition confirms, however, that Maturai venerated the tree kaṭampu68 for preference, the sthala-vrkṣa of its Saiva temple; however, the marutam, if it grows as well on either river bank, still seems quite fit to be the most emblematic here. The explanation is perhaps inconclusive and is favoured, or rejected, on the grounds that it rules out any suspicion of a northern origin.

61Kūṭal is, on the other hand, a name charged with legendary significance: it is the junction, the confluence of the four clouds of Siva which protected the city against the wrath of Varuṇa, and which became four edifices (Nāṉmāṭam) or four temples. The Tiruviḷaiyāṭal of Parañcōti has preserved the legend, but that of Perumpaṟṟapuliyūr Nampi gives us the names of four protecting deities of the town and devotes a quatrain to each one: Kaṉṉi, Kariyamāl, Kāḷi, and Ālavāy (the Virgin, Vishnu, Durga, Siva).69 Whether this concerns those four temples already, or only its large tall buildings,70 the town is often mentioned under that name. Vaiyapuri Pillai claims that it is not earlier than 3^rd c. to 4^th c., yet it is found in the Maturaik. 429, in a poem by Kapilar (Puṟam 58, 13), etc., not to speak of Kali. 92 v. 65 where Naccinārkiṉiyar also cites the four temples in his commentary: Tiruvālavāy, Tirunaḷḷāṟu, Tirumuṭaṅkai and Tirunātavūr.

62Another problem is grafted onto the identification of those four temples, because of Fr. I, on Tirumāl, which, moreover, cites two little known places for which our text is an essential reference: Iṟantaiyūr and perhaps Kūlavāy. The first of these is the name of a temple of Tirumāl beside the Vaiyai. Gopinatha Rao and Mu. Irakavaiyankar71 identify it with “the holy dwelling of Kariyōn” spoken of in the Kallāṭam (61, 26; but considered as a much later text) in an enumeration given there of the ten temples of the town, and with the temple of Neṭumāl also beside the Vaiyai of which the Cilap. speaks (XVIII, 4), called Iruntavalamuṭaiyār by the “old commentary”, and ‘Antaravaṉat temperumāḷ’ by Aṭiyārkkunallār. It could be the temple of Kūṭal Aḻakar, the great Vaishnava temple of Maturai, to the south-west of the town on the bank of a branch of the Vaiyai called Kṛtamālā, which has a seated Vishnu in the lower part of the shrine, with a standing Vishnu above it and above that a reclining Vishnu. It is possible that the Vaiyai had, at an early date, more or less surrounded the town from whence comes the name Kṛtamālā which the purāṇā also give it. Fr. I of Paripāṭal would thus be the earliest literary attestation to this temple, celebrated by Periyāḻvār and Tirumaṅkaiyāḻvār,72 the former even speaking of the snake with five heads which serves as Tirumāl’s couch. These later references cannot be made to bear witness against the antiquity of Paripāṭal poem, for it seems that the site of the temple is itself very old. As for Kūlavāy (the bank of the pond),73 it is associated with a Māṭakkuḷam to the west of Maturai which even constituted a taluk in itself. The first Tiruviḷaiyāṭal, 12, 13, also speaks of Maturai to the east of Māṭakkuḷam and SA quotes the proverb: “If Māṭakkuḷam overflows, Maturai will be annihilated”.74 These fragments thus reveal to us something of the topography and of the ancient history of Maturai, and these poems surely make its people real to us, divided (cf. Frg XIII) between its river Vaiyai and its mountain Tiruppaṟaṅkuṉṟam, while the mountain of Aḻakar (poem XV) introduces a tripartite tension; thus is affirmed as well the unity of the collection, a geographical unity around the town, its two guardian deities and its river in flood. It is, in fact, in their jubilation and their prayers that we shall follow the people of Maturai, in the poems which celebrate its three gods, its three sources of grace and happiness: the Vaiyai, Murukaṉ and Tirumāl. The contrast is quite striking with the four hymns of Campantar on Tiruvālavāy, which are full of the tensions between the three religions: Jain, Buddhist and Saiva, which divided the city: a contrast which any speculation on relative chronology should obviously take seriously into consideration.

II - The Vaiyai

63Geographical reality: the Vaiyai is deceptive: it flows for 52 kms to die out in the district of Ramnad; its uncertain climatic conditions have been ameliorated only by modern works. Before this, “the Vaiyai had one or two weeks of heavy floods which quickly drained off and was down to a miserable trickle of water the rest of the year”.75 The Ramnad Manual76 considered it to be capricious and uncertain: “its periodic floods are so irregular that they can never be predicted with any certainty nor counted on with any confidence. Due to the inclination of its course, the current of this river is very violent. This, and the fact that the sand on its bed moves at a depths of several feet makes crossing it, even in normal water, an operation of considerable difficulty, sometimes with fatal consequences”. This is interesting in that it situates the frolics our poems tell of in their proper perspective, showing, for example, how the lover may be delayed when crossing the river in flood (poem VI). We must imagine the river otherwise than transformed by the modern works of the Periyar dam, and imagine the more wooded expanses of countryside most likely made the rainfalls heavier than those of today. From this same manual we draw the following information from a table of floods from 1813 to 1889: floods were abundant for 15 out of the 75 years observed; their usual period was from October to January; there might, however, be flash floods of exceptional strength (in Oct. 1884 the Ramnad reservoir filled up in 8 hours), or out of season (in May 1882, for example; we recall that the dates proposed for the horoscope in poem XI vary from June to August). Without affirming that these observations might have been valid fifteen or twenty centuries ago, we may all the same notice that the rendezvous of the river with the people of Maturai evinced the uncertainties and risks of the clandestine love-making dear to the Caṅkam poets and that invocations to the flood are thus not empty words, nor the association of untimely floods with an eclipse of the moon.

64If today the Vaiyai ends its course in the big Ramnad reservoir this has not always been so, for it is thought that the many tanks in this district were, in fact, installed in the ancient delta of the Vaiyai which the river must afterwards have abandoned.77 Paripāṭal seems to consider that the river throws herself into the sea when it is comparing her with the heroines of clandestine love: for she must rejoin the ocean her spouse and end her course according to the marital law of regular marriage,78 in spite of the efforts of the inhabitants of Maturai to put obstacles in the way, and those of the poets to suggest to her that she retain the more enviable fate of a free woman of the mountains. We don’t know exactly when the Vaiyai refused to bend to conjugal law. There are two words in Takkayākapparaṇi, along with the text of its early commentary, which affirm most clearly, and probably for the first time, that the Vaiyai “is not enamoured of the ocean” and doesn’t throw itself in the sea.79 The text is from 12^th c., and the commentary may be slightly later. Parañcōti in his turn confirms that the waters of the Vaiyai do not join the sea80 and so does the author of Tiruvātavūrār purāṇam81 but that was in 18^th c.

65More than geographical reality, it is the river magnified by the poets that we find here. There is little doubt of its celestial origin though no poem refers to the heavenly Ganga other than by comparison. Only in both Tiruviḷaiyāṭaṟ purāṇam does Siva make her descend from his hair to quench the thirst of Kundodara. She is then called Vegavati (Tam. Vēkavati) for “With impetuosity, in abundance, comes the river, the Impetuous”. Is this the Sanskrit origin of the Tamil name Vaiyai? The Tamil Lexicon supposes so, but the Kūṭaṟpurāṇam suggests another etymology:

For her swiftness, Vēkavati,
Because she is in space, Vaiyai,
Because she becomes a garland, Kirutamālai,
This is the river to which the nakar gave three names.

66The second verse reminds us of another Sanskrit etymon, Vaihayasi, “the aerial”, the name of a river which figures in a geographic enumeration precisely where the Vaiyai would be expected to have a place (Bh. Pur. V, 19, 18). The name Vegavati applied elsewhere to other rivers remains, with that of Kṛtamālā, the one most commonly associated with the Vaiyai, as witnessed by the stanza by Parañcōti, who added Saiva considerations:

Because she descends from the adorned koṉṟai tresses of the Pure, Civakaṅkai;
Because, to those who meditate upon the Pure, she gives knowledge, Civañāṉa tīrttam;
because she comes swift as the wind, Vēkavati tīrttam;
Kirutamālā; these it is said are the names of the Vaiyai.

67All its Sanskrit names notwithstanding, it is in essentially Tamil terms that the poets wanted to celebrate it in song.

The erotic theme

68The specific aspect of the amorous element in poems to the Vaiyai is only intelligible from inside the symbolic conventions of akam literature,82 even when seeming to be at variance with them. Erotic situations, as other characteristics of life, are shared out between the Caṅkam’s five ideal geographical regions: the mountain (kuriñci), forests and pastoral land (mullai), cultivated plains (marutam), the littoral (neytal), and lastly, desert (pālai). Corresponding to each region is a strictly defined set of customs and practices; each has its god, its chiefs and inhabitants, its fauna and flora, music and instruments, hours and seasons, its work and chosen pastimes etc., and, in the erotic realm, its particular situations. The specific theme (uripporuḷ) of each region bestows on the mountains the union of lovers and their clandestine midnight assignations; on forests, the brief wait, serene and accepted; on the shore, the long wait, sad and unbearable; on deserts, the anguish of separation; and lastly, on the rich cultivable plains where the action of our poems takes place, the separation due to courtesans, with the growing anger of the heroine, and the sometimes bitter quarrels between spouses and between lovers on such occasions. The midnight hour and the coolest and most rainy months are compatible with the mountains, autumn rains and evening with forests, twilight at all seasons with the sea shore, and noon and the torrid months with the desert; the hours before sunrise, time of the return of the unfaithful husband and the first light of dawn which reveals on his person the traces of his exploits,83 is compatible with the quarrels of the plains at any time. Some modern commentators wax rapturous over this natural psychology, seeing in it not even the memory of a phylogenesis of the Tamil people, from nomadism to hunting in the mountains and to the pastoral life of the jungle, before the discovery of agriculture on the plains and commerce on the shore. We shall rather conclude from this very brief schema, which excludes details and nuances, that the vision of the poets is necessarily inscribed in the solidly structured forest of symbols: the ideal system of situations according to which every poem interprets the usages of the world.

69Propitious to the joys of lovers on the mountain, the Vaiyai, on the Maturai plain, must be celebrated for the savour of their spats and reconciliations. The other anthologies with amorous themes encompass poems of all the regions: two of them, Aiṅkuṟunūṟu and Kalittokai, are even constructed in five parts with a single criterion each; in Paripāṭal, on the other hand, all we have is a very partial view of the amorous thematic of the Caṅkam. It seems, however, that the poets of Paripāṭal, with the exception of Nallantuvaṉār, hesitated to devote the essential of their poems to an obligatory theme, but that they nevertheless generally respected the rules of the genre. The geographical situation of Maturai, in fact, obliges them to keep love off the scene except on occasions when it has been tarnished. This gives us some fine lively dialogues (VI and XX for example). But when they failed to make brilliant variations on this theme, there remained that of gambols in the water, which they preferred to describe since such scenes gave occasion to develop lovely images of a crowd at leisure. The commentary that accompanies these accounts often considers them as some kind of indirect message from the heroine to the hero, to close the door on him84 or, on the contrary, to let him know that his return is awaited after a long absence;85 here is all the ambiguity of lovers’ quarrels: whether bitter or sweet, is in accordance with what is in her mind is the deception or the expected reconciliation.

70The characters attending the couple on the scene are those sanctioned by the conventional rules: musicians and singers of both sexes, the attendant or companion (tōḻi), are the messengers accepted by Tolk. The intervention of the chorus of old women accompanying the heroine also conforms to usage. Above all, the scenery of the poems to Vaiyai favours all kinds of promiscuity, namely between “chaste wives, courtesans and their retinues”, a situation likely to generate misunderstandings as well as genuine explosions of jealousy. As for the courtesan, she has a status and a level as well defined as in Sanskrit literature, as listed, not surprisingly, in Vīracōḻiyam.86 This treatise cites three occasions when the hero may be separated from his beloved by courtesans: festivals, groves and aquatic games (here we are) and four kinds of courtesans (parattai) who are, as described by the modern editor: 1) illiṭaip-parattai, who have a residence of their own where they welcome people attracted by their reputation; they have social status and independence, [one detail is not clear, their status in front of the hero: are they the co-wives whom he keeps and visits in separate residences, called even to-day “the small house” (ciṉṉa vītu)?]; 2) iyaliṭaip-parattai, who outdoors, on the roadside, seduce and attract passers-by, they are of no status. 3) kalviyoṭu paḻakiya kātaṟparattai, who attract to their residence by their capacities for performing various fine arts. The modern editor stops at this geisha-like definition, however, and overlooks its kātal (love) component; 4) nalliyal nilaimai nayappup-parattai, who attract by their reputation in various games and songs, by their beauty, and by their happy mood and good nature. Commenting on Paripāṭal, P. is satisfied with two types, the concubine (iṟparattai) who shares a residence with the hero and whose status seems to be that of a “co-wife”, and the most cherished mistress (kātaṟparattai) to be met in secret. In fact, the dialogue between hero and mistress sometimes waxes quite acrimonious, often to the dismay of the beloved herself who, aware of the quarrel, may be a witness or may even become a participant; but we are indebted only to the commentator for the distribution of the replicas, while it is not really clear whether the purely descriptive, and often lengthy, passages devoted to the scenery and to the crowd enjoying itself in the water-games are sung by the hero, the heroine or a third party. Other partners are often rather elusive or even mute witnesses. Further, in the colophon to poem VI, it seems that P. himself simply followed the general scenario suggested by that colophon when he decided about the casting. Finally it is always a question of the lover who cannot keep away from there, or of the heroine who doesn’t know how to defend herself against this. Paripāṭal is therefore following overall the rules of akam, even though the only poet to take upon himself completely the erotic situation imposed by the marutam region was Nallantuvaṉār, the most brilliant of the authors,87 who is traditionally credited with equal mastery of every aspect of the subject in Kalittokai. [Once again, when we try to pierce the anonymity of the authors, parallels and coincidences are too many to be always fortuitous, so one wonders which attitude is the “anti-historical” one: the one which brings the Caṅkam poets closer within a small world with the help of the colophons, or the one which points at philological differences or analogies to treat them as definitive chronological markers.].

71Amorous concepts, other than the frustration created, spread out and dissipated, also appear, but the poets could not, in poems to Vaiyai, treat the actual themes of clandestine amours which are the privilege of the mountains88 and may sometimes be taken up in poems to Cevvēḷ. It is evident, however, that the beauty and force of passion that is part of this love remains the preferred theme for which the poets are nostalgic. It is in fact a poem to Cevvēḷ (IX) that gives the clearest definition of kaḷavu as secret and spontaneous love, the communion of two bodies and two souls in perfect accord, in contrast to kaṟpu, chaste conjugal life, with the idea unambiguously expressed that, between those two fundamental concepts of the erotic in the Caṅkam, the first is superior to the second,89 whilst the power of passionate love can be reawakened in conjugal life only in amorous quarrelling.90

72It is for this reason that celebrating amorous frustration, bringing onto the scene married couples separated by a courtesan, giving full range to the anger of the wife and insisting on the confusion of the husband is still to celebrate love and have it triumph. This is the context of the wise counsels of moderation given by the old women who know that the husband will return, repenting of his escapades, and that the wife will be as unable to keep him at first, as to drive him away later.91 They know too that Vaiyai, whose floods and caprices are very close to the movements of passion, is capable of all union, that Murukaṉ, hero of the union with Vaḷḷi on the mountain, also suffers the reproaches and the jealousy of his other wife (IX), and that Tirumāl himself shares his embraces between two wives.

The ludic theme

73The verses not devoted to amorous conflicts or to the description of the river in flood are filled up by the poets of the Vaiyai with frolics in the water. Here too is a theme familiar from other anthologies: in Aiṅkuṟunūṟu for example, the puṉalāṭṭuppāṭṭu, or “ten (songs) about games in the water” constitute an entire chapter of the first part (marutam),92 and elsewhere we often catch echoes of the joys of these baths, of the lovers’ meetings they promote, and of the display of luxury, elegance and sportive vigour of which they are the occasion. The often allusive references are numerous enough to allow us to consider that these games are spread throughout the whole of the Caṅkam; they refer to all the places, from mountain torrents to sea beaches and to all ages, from young girls playing on the sands to adolescent ball games, to less innocent adult games on in the river or in nearby groves, up to the regrets of the aged,93 or the attentive regard of the duenna and to all classes as well, from the most humble artisans’ quarters to royal festivities. The jalakrīḍa are otherwise known in Sanskrit kāvya and Kālidāsa, (Raghu. XVI 70) describes them here and there; treatises on rhetoric numbers them amongst the conventional entertainments. Even śṛṅgakrīḍā are mentioned, ritual aspersions with coloured water through syringes, as represented in Rajput paintings.94

74In this tradition, the originality of the Paripāṭal brings to the poems to the Vaiyai the most complete and systematic description that is, at the same time, one of the liveliest: preparations for the games are made and one leaves in procession at dawn to return in the evening, in festive garments, and with the instruments for the bath, toilet utensils and dress for the bath: there are the battles of flowers and of perfumed water dyed red, not forgetting the drink consumed, sometimes to excess.

75This is the business of the entire city, people pushing and jostling each other, to the sound of conversation, of songs and of music, and the literary merit of Paripāṭal lies, in fact, in its restitution of that festive atmosphere when the whole population joins in the celebration. The inevitable promiscuity is obviously an opportunity for the young who meet lovers and courtesans there, where one may become enamoured of a woman met by chance, or where there may at least be, in contrast, the chance encounter of the wife with the courtesan adorned by the husband with jewels belonging to her, or a misinterpreted wink leading to embarrassing results. We thus come back to the erotic theme of amorous quarrels and fallings out, and it would be superficial to add that, in each poem, love is grafted indissolubly onto the games. If the Peruṅkatai takes up the ludic aspect in several very rich descriptions, the Kūṭaṟpurāṇam in its turn devotes a large part of its amparitapaṭalam (1-69) to describing the games of men and women in gardens, on the riverbank and in the water. Gathering flowers, plunging into the water, fighting playfully, drinking to satiety and fondling pretty girls, who are often consenting but held back by the presence of their companions: these few pages of great sensuality, in which lovers and beloveds celebrate Anaṅkaṉ (Kāma) seem to us deliberately to echo all the harmonics of love and joyous games of the poems to the Vaiyai. Lastly, we must remember that only one third of the poems on the Vaiyai have been preserved, while they constituted two-fifth of the original anthology, in which the Vaiyai is the most treated subject.

The religious theme: controversy

76The preceding analysis represents an attempt to bring into evidence two specific aspects of the poems to the Vaiyai, both relating to the particular literary character of these songs: an erotic aspect, itself developed within the Tamil conventions of akam, and a ludic aspect more liberally developed solely for the poetic pleasure, as is found in examples of Sanskrit kāvya. This formal explanation is not the only possible one, but it may be less superficial than a hasty anthropological extrapolation resorting to some vague religious background, as for example the theme of fertility. We have sufficiently insisted upon the part of rhetoric and of pure literature as fundamental to be able to give our attention now to certain underlying elements of the whole. It is essential to emphasise that not every aspect of religion is excluded from these aquatic festivities. It is not only their ornaments and toilet requisites that the citizens of Maturai carry in procession, but also the objects of the cult. They offer oblation to the river and pour out ex voto-es there.95 This detail indisputably indicates a religious ritual, no doubt meant to maintain the fertility of the waters. Fecund and beneficial, the Vaiyai is here implicitly a gift from heaven. It is striking that she is not divine at any particular moment and that she is “celestial” only in response to the needs of the metaphor.96 Markedly secular, she belongs only to the king and his people, the frolics of the latter having the gratifying effect of distancing the brahmins, horrified by the pollutions of the waters.97

77Were these then princely games? We may emphasise the “warrior” character of some games which were certainly simulated combats in fact, as well as the analogy with the Sanskrit jalakrīḍā from Kālidāsa to the later Renukāmāhātmya. In the Cilap. again, the Cōḻa king Karikālaṉ is represented as having inaugurated the games in honour of the first floods of the Kāviri.98

78Such a conjecture must take into account the extremely composite character of the participants: “Literate, illiterate, people of no importance”, the text also says.99 We may perhaps be closer to the truth in thinking of patiṉeṭṭāmperukku (of the 18^th day flood) which still today in Tamil Nadu welcomes the arrival of fresh spates at the peak of the floods, on the 18^th day of the month of āṭi. V. Raghavan evokes this thus: “Towards the evening families repair to their village brook or river with all sorts of freshly prepared dishes. Elderly ladies sit on the bank dispensing the delicacies and boys in hip deep water eat, disporting themselves”. He rightly associates it with, amongst the rain festivals (varṣā ṛtu), the navodaka-abhyudgama, the festival of welcoming the floods.100 The beginning of poem XI and the numerous passages relating to the appearance of the fresh spates after the rains invite us to think that the songs consecrated to the Vaiyai have something to do with such celebrations. The aspersions with red water, on the other hand, are too universal for a detailed relationship to be made with the Holi festival, except according to the vaguest possible sense of the theme of fertility which will not take us very far; the poems to the Vaiyai have nothing to tell us of the gods associated with these festivals and the erotic aspect of these poems is far indeed from the popular obscene songs sung at Holi.101 The Paripāṭal is, moreover, not a canticle either, even though some of its melodies reappear later in the religious songs of the Tēvāram. Aside from the considerable amount of hypothesis involved in the suggested identifications, we often have, elsewhere, the example of profane harmonies being applied to religious hymns over time.

79In contrast, the particular case of the baths of Tai to which many Caṅkam poems make allusion102 and which are always cited in a female context,103 there, the religious element is dominant; all the same, it was only after the Caṅkam that it became exclusive. One will speak then of the festivals of Mārkaḻi, accompanied by specific observances, and especially of the dawn baths; here one recites, according to persuasion, the Vaishnava Tiruppāvai by Āṇṭāl or the Saiva Tiruvempāvai by Māṇikkavācakar. Much has been written on the identity of the Caṅkam tainīrāṭal with the later mārkaḻi nōṉpu. It appears that the former, as in the most complete description of it we have in poem XI (v. 74-92) of Paripāṭal, is, however, more profane than the latter, or at the very least more worldly in the content of its prayers: the young girls ask for the rains and a husband; Āṇṭāl does penance to obtain Krishna. We note that in Paripāṭal there is no question of the sand statuettes of Kātyāyanī that Mu Irakavayankar deems it necessary to discover in the pāvai of Tiruppāvai, by reference to Bhāgavatam,104 even though sand dolls are often mentioned elsewhere in a profane context. The interpretation of ampāvāṭal (to bathe (with) the mother, ampā) as relating to the mother-goddess is nothing but conjecture, and the presence of Siva, master of the constellation Ardra, is certain only in the commentary of P. and thus does not support the presence of Pārvati. Counter to the thesis of the identity of Tai of the Caṅkam with the Mārkaḻi of the bhaktas, we remember as well that Āṇṭāl herself sang of the two months consecutively and that, after celebrating the mārkaḻi baths in the Tiruppāvai, she sings of the Tai baths in the first poem of Nācciyār Tirumoḻi “Taiyorutiṅkaḷ”, “during the whole month of Tai”, and the 2^nd stanza declares:

Having decorated the street with white sand,
Having bathed from the Ghat before sunrise,
Having burned dry sticks without thorns (cf. XI 85-86)
O Kamadeva, I do penance with effort…

80These austerities will continue during the month of Paṅkuṉi; those of Tai are addressed to Kāma to obtain from him the love of Krishna. Then, in Kerala, the festival of the day of Tiruvāthira, essentially female, also celebrates Kāmadeva, in the form of Anaṅga burned up by Siva, and comprises notably a bath in cold water at dawn and the game of the swing.105 We must be careful with this connection: it may be as well to stop drawing on the probable, but unverifiable, hypothesis of a reform of the calendar to claim the ancient festivities of Tai as direct ancestors to current Markaḻi festivals.106 It would be better to consider that the Caṅkam often mentions water games, especially at flood season, and that those of Tai are, more than any of the others, linked with young virgin girls and with prayers for material benefits: rain and a husband.

III - Cevvēḷ

81This is not the place to accord to Murukaṉ the great importance he has in Tamil lands. The literature dedicated to him is immense, especially later, and he is the most popular god in Tamil Nadu. His complex nature requires a wider approach than simply the study of the texts and iconography. The poems dedicated to him in Paripāṭal constitute, along with the Tirumurukāṟṟuppaṭai, the very earliest documents of Tamil literature we have about him. It is therefore remarkable to notice that he is, from the very beginning, fully equipped with his legends: a god of war and of love, as essentially Skanda, born of Siva and Agni, as he is Murukaṉ, god of the vēlaṉ, (his local priest holding his lance, the vēl) and of the hunters and their mountain, husband of Vaḷḷi. All the elements we are about to distinguish are, in fact, coexistent and completely indissoluble, from the age of Paripāṭal onwards, and if some appear to be more local and others more pan-Indian we must accept that fusion as it is and take care not to reconstruct a history of its genesis since it precedes any possible historical approach.

82Poem V gives a version of the birth of Murukaṉ which may legitimately take its place amongst the innumerable narratives of the origin of Skanda in Sanskrit puranic texts.107 It is one of the very oldest known, and it is not the least paradoxical aspect of these verses, which evoke first the typically Tamil frenzied dance of the vēlaṉ, to draw us then irresistibly to the Ramayana and the Mahabharata, much more than do even most of the later Tamil texts. As in the first version of the Bālakāṇḍa (ch 36), and in contrast to the majority of puranic versions, Skanda, here, is not longed for but feared by Indra and the other gods who, faced with the prodigiously long synousia of Siva and Uma over hundred of the gods’ years, are terrified by the idea of the fruit of such a union. In our text, Indra had obtained from Siva the promise to abstain; to keep his word, the latter cuts the gestating embryo with his axe of fire. This gesture seems to us to be unique, given the well established tradition, of his tejas being handed over, according to various texts, to Agni, Indra or Vāyu who, unable to bear its heat any longer throw it into the Ganga or the Saravana.

83Here, the Ṛṣi-s collect up the fragments of the embryo but are too fearful for their wives’ reputations to fecundate them with it. It seems that, by now, they have read the Vaṇaparvan (ch. 223-225) where Sahā, daughter of Dakṣa, approaches Agni six times, disguised as the wives of the Ṛṣi-s, but fails on the seventh try to take the form of Aruntati whose chastity protects her; later, the Ṛṣi-s, disturbed by various rumours, want to repudiate their wives (225,8). Skanda, forced to give in to their request, rehabilitates these wives, by accepting them as his mothers too (229, 1-5). In Paripāṭal, as in the Mbh., Aruntati receives exceptional treatment.

84The Ṛṣi-s of Paripāṭal go on to remember the essential role of Agni, sometimes the first and sole sire, especially in the Mbh. and the second version of the Bālakāṇḍa (37): they decide to throw into the sacrificial fire the fragments abandoned by Siva, with an astonishing, and impressive command: “Let him carry himself”. There is no better way than this to affirm some identity between Murukaṉ and fire,108 and also the invigorating effect of the latter, which naturally, respects the embryo. It is therefore in the form of such a sacrificial offering that the embryo is consumed by the wives of the Ṛṣi-s, –apart from Aruntati–, in the same way as Agni takes the precious semen in his mouth, in Kumāra sambhava,109 Matsya purāṇa110 and Ajitāgama,111 for example. They give birth to six babies in a pond full of reeds in the Himalaya, known to many versions; but although it is usually at this moment that the Pleiades intervene, to nourish the six infants, or the six-headed infant, they are mentioned here only very briefly, and in passing, in only one verse that identifies them with the wives of the Ṛṣi-s.

85As in the Mbh., Indra remains hostile to Skanda. In the epic, however, pressed by the worried devas to kill Skanda he first prudently refuses. In Paripāṭal on the contrary, he less wisely attacks and is beaten by Murukaṉ who, in order to overcome him, turns himself into a single baby with six heads and twelve arms. Indra is here, therefore, the involuntary agent of the union of six babies in one, whilst this is often the role allotted to Pārvati. But in Matsya purāṇa it is certainly Indra who, for the good of the gods, reunites as a single infant Kumāra and Viśākha, issued from Pārvati’s right and left flank;112 later he will give him his daughter Devasenā in marriage. Let us note in passing, moreover, that in the Mbh. Viśākha comes from the right flank of Skanda when Indra, deciding to attack at last, wounds him with his thunderbolt.

86Once Murukaṉ has won the uncontested victory, the gods shower him with gifts. The list of these gifts, as that of the donors, varies perceptibly from text to text; we have mentioned the possible connections in the notes. In this regard too Paripāṭal finds its place in the puranic tradition. Next are numbered the twelve attributes Murukaṉ holds in his twelve hands. All these are known, with a measure of incertitude that subsists despite everything that has been iconographically identified,113 and we have not so far found any statue or image which represents them all at the same time,114 nor any theoretical description of the mūrti of the god that brings them all together.115 The Ārumukaṇ warrior described in poem V seems to be ignorant of the gesture of abhaya which, on most twelve armed mūrti leaves at least one hand empty. The twelve arms, the vēl and the eleven incarnations of the eleven Rudra-s featured in the Tamil Kantapurāṇam116 are also different. Because of this two fold originality, which satisfies neither a purely Tamil vision nor puranic or agamic orthodoxy, poem V of Paripāṭal has often been ignored. It would be wiser to recognise that this unique and ancient version deserves particular mention; Naccinārkkiṉiyar obviously understood that, since he summarised and cited it in his commentary to Tirumurukāṟṟuppaṭai.117

87Apart from this narrative of his birth, many pan-Indian elements of Skanda’s legend are mentioned in our poems: his coming into the world in a reed filled pond in the Himalaya; his marriage to Devasena, daughter of Indra (‘He with a thousand eyes on his body’, cf. poem IX); and the destruction of Mount Krauñca by his arrow, exploits known from the time of the Mbh. and, for the last, from that of Viṣṇu Dharmottara purāṇa; as his double maternity of Uma and the Pleiades and the fact that his father, Siva, carries the Ganga in his locks (IX, 4-8) and has his throat stained with poison. There is nothing original in any of that, not even within the limits of the Caṅkam corpus itself where may be gleaned references to the chastity of Aruntati, to the burning of Tripura (mentioned here immediately before the narrative of the birth in poem V) and to the throat and the locks of Siva.

88Finally, puranic tradition links the advent of Skanda with the conflict between the deva and asura and has him leading the celestial armies. The fight against Taraka is the only battle extensively sung by Kālidāsa,118 and even the Tamil Kanta purāṇam mentions it. But Paripāṭal has nothing to say about Taraka: Murukaṉ’s enemies are the asura in general, often called avuṇar, and his essential adversary is Cūr, that is Sūrapadma, according to the commonly accepted identification, even though this name is never mentioned in the text, and even though Cūr, in Tamil, is explained very well by a radical cūr- meaning “dread, cruelty, affliction”, often used to designate collectively evil spirits of mountain and river, and better attested to at an early date than the Sanskrit radical sūra “brave or hero”; later Cūr would be indifferently called Cūr, Cūra­ or Cūrapatmaṉ. The fight against Cūr and his race occupies the major part of the Tamil Kantapurāṇam and the Sanskrit version which most closely approaches it.119

89In the entire Caṅkam, and later in references as may be gleaned from the Tēvāram regarding Murukaṉ, this is still the most frequent characteristic to appear. Cūr, expert magician, incarnates as a mango tree to escape the rage of Cevvēḷ and takes refuge in the sea where the vēl of the god finally makes an end of him. We may add that two mango trees seem to be distinguished here. In the Kanta pur. (IV, 13, st. 464-491) Cūrapaṉmaṉ himself dives into the sea and takes the shape of a mango tree which covers the entire earth, but in Tirumurukāṟṟuppaṭai (59-61) it is after having killed Cūr that Murukaṉ, to put down the avuṇar, cuts a mango tree of which the clusters of flowers were turned towards the soil; commentators hesitate over whether this was the above mentioned incarnation of Cūr or another old mango tree, a sort of enchanted tree, refuge or emblem of the asura, under the shade of which they did penance to obtain the boons they later misused, and which could have been their tutelary tree. The commentators hesitate to distinguish these (see the various old commentaries on those three lines) but most of the time they correspond. Paripāṭal refers to a single tree only.120 It is, along with the destruction of Krauñca, the most oft quoted exploit, appearing, for example, as common denominator in all the fragments on Murukaṉ collected by Mu. Irakavaiyankar in his Peruntokai (nos. 98-101); amongst those we may quote the anonymous nēricai veṇpā, simply an example of grammar on the multiplicity of actions relating to one and the same subject, but as well a real robot-like portrait of Murukaṉ:

Murukavēḷ cut the roots of the mango tree Cūr; he adorned with garlands
The wavy tresses of Vaḷḷi; having given his protection
He led the great army of the highest celestials;
With the vēl he cleft through the middle the mountain (Krauñca).

90With his Aryan birth and warrior exploits, the husband of Vaḷḷi thus does not seem very different from his puranic image and we shall see later how his Tamil consort would herself be integrated into the one brahminical Indo-Aryan family. The names under which he is addressed in song are, however, most certainly Tamil121 as are some aspects of his cult. Murukaṉ is the most popular denomination: the animated form of muruku, a name of quality which also designates him, the attested meanings of which are “youth, beauty and perfume, divine essence and sacrifice.” [We leave to I. Mahadevan the responsibility of identifying Muruku and of reading it in “Indus script”, but he has persisted in this for the last ten years, and is heard from again in 2008; see his statement in Airāvati, Felicitation volume, 2008, pp. 461-487.] Cevvēḷ is the “Red Desired One” for vēḷ is ‘desire’ (name applied elsewhere for Tirumāl as well) while Cev- is ‘redness’, but with the connotations of beauty and prosperity attached to this desirable colour. People also like to see in him Cēyōṉ, the Red, god of the mountainous region of kuriñci, as opposed to Māyōṉ, the Black, god of forests and pastoral lands of mullai. This opposition, attested to from Tolk. onwards, seems the basis of the most probable etymology,122 but Cēyōṉ remains ambiguous, the same radical also meaning “far off” (the inaccessible god, a theme absent from Paripāṭal but found later in Bhakti texts), and the word Cēy also fairly commonly means “son” Both meanings of Ceyon, “Red” and “Son” are in any case inseperable in the eulogy to Murukaṉ and serve equally well to explain a fusion with the “Aryan” gods (Siva-Rudra, the red; Kumara, the Son, Tamil Kumara­) as to indicate “Dravidian” origin, since the Red of the mountain is also the Son of Koṟṟavai, the redoubtable Lady of Victory, known as well under the names of Kāṭukiḻ aḷ, the Lady of the Forest, and Paḻ aiyōḷ, the Ancient (one), “the beautiful goddess who dances the tuṇaṅkai (dance of battlefields), whose huge womb engendered Cēy” according to the Perumpāṇ. (457-458). It is thus possible that a local pantheon had been renamed Skanda and Durga.123 By the time of Paripāṭal the fusion has taken place and the study of correspondences between Koṟṟavai and Murukaṉ is difficult due to a lack of texts: the poem which is perhaps dedicated to the Lady of the Forest in Paripāṭal has disappeared and the sole mention of Koṟṟavai (XI 100) represents her with the Saiva frontal eye; the series of invocations in the Tirumurukāṟṟuppaṭai124 where this appears is an excellent evidence of an already achieved syncretism.

91More than the vocabulary, the content of certain rituals appears to be characteristic of the cult of Murukaṉ in Tamil lands. The veṟiyāṭṭu, frenzied dance led by the vēlaṉ (the lancer, from vēl, lance or javelin, the favourite weapon of the god) is a common scene in Caṅkam literature. It is often described with a sort of ethnographic curiosity, starting with numerous allusions in the Maturaikkāñci, Naṟṟiṇai, Kuṟuntokai, Aiṅkuṟunūṟu etc. The setting of the dance is known; the veṟikkaḷam or veṟimaṉai, with the vēl planted at its centre by the vēlaṉ; fresh sand spread over the area, and the white poṟi (roasted flour of rice prepared with the superior variety of “red” rice), kaṭampu flowers (Barringtonia racemosa), whose flowers grow in small clusters coiled up around a long axis, so that two branches are enough to make a garland, and other sweet smelling plants, perfume, incense etc.; the kid (maṟi) attached to a sacrificial post daubed with sandal, and then killed in such a way as to mingle its blood with the red millet, (red is the dominant colour, for example, the garlands of kāntaḷ, Gloriosa superba); the various ritual offerings, the songs to the tune of kuṟiñci; the dance of the vēlaṉ certainly, to urge the god to enter into one of his devotees, and perhaps also a variety of kuravai dance analogous to those celebrating Māyōṉ; and finally the oracles, drawn from a count of grains of cereals, and the overall contagion of the dance of possession (veṟiyūṟu nuṭakkam) etc.

92The description given in the Paripāṭal is much more restrained (and even the amplified echo found in the Tirumurukāṟṟuppaṭai is not as complete), but we may understand the importance of this cult in Paripāṭal itself when we see the worshipers of Murukaṉ climbing the hill of Tirupparaṅkuṉṟam holding in their hands everything necessary to celebrate him: sandal, incense, lamps, flowers, drums, bells, axe etc. The text clearly speaks of the sacrifice of a ram and of the songs and dances of the possessed vēlaṉ. It is even characteristic that, amongst all the works of the Caṅkam, Paripāṭal is almost the only one to give these festivals their proper dignity, and to celebrate them seriously, perhaps to the dismay of P. who tends to minimise the importance of the vēlaṉ. In the other anthologies the dance of the vēlaṉ hardly appears, in fact, except as a clandestine love episode and a sort of strategy in the tactics of lovers. Because her daughter seems to be ill, the mother asks the vēlaṉ to come, but the young beauty has no malady but the sickness of love and is concerned with her hero rather than with Murukaṉ; however, “the vēlaṉ who knows nothing of this says that she is possessed”. It is expressed thus in the Aiṅkuṟunūṟu125 in which the ten songs on veṟi (veṟippāṭṭu, “songs of frenzy”, 241-250) recount only how the heroine’s companion tries to postpone the dance of veṟiyāṭṭu and to make the mother understand that, instead of coming there for his dance, the vēlaṉ would have done better to celebrate the renown of the one who is loved and waits somewhere in the mountains. In an analogous manner, in the Kuṉṟakkuravai of the Cilap. the young girl declares that it is laughable to see the vēlaṉ come to cure a malady caused by his lord on the mountain.

This vēlaṉ is a fool, and more fool than he,
If he comes, the Son of the Lord who is seated under the banian tree.

93Nowhere in Paripāṭal, is Murukaṉ treated of as a simpleton or ignoramus, as he is in Naṟṟinai 34. The register is quite different: it is no longer concerned with an amorous (comic) strategy but with the celebration of a god at the heart of his territory, and the vēlaṉ ceremonies are a part of this celebration. Sometimes, today, some authors seem to notice in Paripāṭal and Muruku. traces of survival of the primitive local religion based on terror or on ghastly rites, which was succeeded by bhakta, devotional love, with its more abstract concepts. Rather than being a chronological succession from fear to love126 (in the image of Christianity), we think that the two aspects might have coexisted for a long time, and still may coexist in our texts; relatively numerous accents of Bhakti, and sketches of relatively abstract concepts are in fact intertwined with the expression of more popular local traditions. Both trends are accepted by the poets of Paripāṭal, more liberally perhaps than by their medieval brahmin commentator.

94In Paripāṭal, Murukaṉ is, first of all, the guardian and protecting deity of Tirupparaṅkuṉṟam, his dwelling a few kms. south-west of Maturai,127 a hill on which his temple stands today, celebrated in the story of the god as the place where the purāṇas have him marrying Tēvayāṉai, or Devasena. Two Akam poems at least (59 and 149) attest to the antiquity of this shrine the songs of Paripāṭal describe with images of grandeur and luxuriance, just as processions of pilgrims coming from Maturai carrying their offerings, make an enormous garland with the flowers that decorate their hair. God of the hunters of the mountain, Murukaṉ himself appears on the hill of Paraṅkuṉṟam as hunter and warrior, as he is described in poem XXI, seated on his elephant, a warlike mount, armed with his lance and shod in sandals of cured and stitched leather, and with a garland of kaṭampu on his head, flowers of his favourite tree under which he sits and which elsewhere is used by the vēlaṉ to tie up the sacrificial victim. Other poems corroborate this typical description.

95One detail has often shocked the commentators: the leather sandals, which the orthodox do not allow to gods as they don’t touch the ground or, if they did, would wear only wooden sandals as ascetics do. We knew almost nothing about these leather sandals except an analogy, the “sandals covered with cut thongs” of the hunter Kaṇṇappaṉ, one of the 63 Saiva saints128 and it seemed impossible to believe that Murukaṉ, despite his vocation as hunter might also wear them. This was a mistake, for Abbé J.A. Dubois has preserved for us some curious evidence in this regard: “At Palany, in Madura, there is a famous temple consecrated to the god Villeyada129 whose devotees bring offerings of a peculiar kind, namely large sandals, beautifully ornamented and similar in shape to those worn by Hindus on their feet. The god is addicted to hunting, and these shoes are intended for his use when he traverses jungles and deserts in pursuit of his favourite sport. Such shabby gifts, one might think, would go very little way towards filling the coffers of the priests of Villeyada. Nothing of the sort: Brahmins always know how to reap profit from anything. Accordingly the new sandals are rubbed on the ground and rolled a little in the dust, and are then exposed to the eyes of the pilgrims who visit the temple. It is clear enough that the sandals must have been worn on the divine feet of Villeyada; and they become the property of whosoever pays the highest price for such holy relics.”130 If, to these malicious words from a foreign missionary, one prefers Indian witness such is to be found in the Paḻaṉit talapurāṇam, dating from 1628131 in which Murukaṉ also appears thus before Iṭumpaṉ, the companion of Agastya and the first bearer of kāvaṭi:

The god with brilliant leaf shaped vēl, who dwells on this mountain,
Like a king, who wears a black belt around his waist,
Adorning his pure feet with sandals and rings of gallantry
Armed with the lance and the bow with pointed arrows, came to face him.

96Lastly, we still today find leather sandals being brought as offerings by pilgrims to Palani, and the statue of the god, when ritually dressed as a hunter, receives leather shoes of a Muslim type. It is not so much this that is astonishing132 as finding an echo of this popular and localised usage, in something as ancient and literary as the Paripāṭal.

97There is also a fair amount of evidence in Paripāṭal on the mount of Murukaṉ, the elephant Piṉimukam. The name is sometimes mistakenly applied to his other and more usual mount, the peacock. Here it is certainly the elephant that is in question (see V 2 and note, XIX 90-92 and XXI 1-2) and if the iconography of Murukaṉ on the elephant is little known, this may be, first of all, because it is difficult to distinguish it from that of Indra, also mounted on an elephant and oriented towards the east (thus at Mahabalipuram, the central figure, eastern face of Arjuna’s chariot). The elephant and the ram are however found at the foot of the statue of Subrahmanya in the cave at Tirupparankunram, and a Subrahmanya on the elephant at Tiruvalam for example (Gudiyattam taluk) dates from around 10^th c., and another one at Makaral (Kanchipuram taluk) is still later (13^th c.?).133 This mount seems to have disappeared towards the 12^th c.-13^th c., as the commentator of Takkayākapparaṇi only knew the peacock, which is in fact almost always represented.

98In the tradition of the hunter god of the mountain is inscribed the last characteristic we would like to mention: Murukaṉ is the husband of Vaḷḷi, the daughter of the mountain dwellers, and he is a god of love. In Paripāṭal he is, of course, already married to Devasena, but although puranic tradition always places the marriage at Tirupparankunram, on the return from the warlike expedition against Cūr at Tiruccentur, our text also situates at Tirupparankunram the terrestrial marriage of the god, replicating his celestial one, because that hill is paradise on earth, where Vaḷḷi and her entourage triumph, to the great confusion of the celestial wife and her attendants, after the two groups have violently confronted each other in an heroic-comic battle.134 Later texts such as Kantapurāṇam, adopted a more conciliatory version135 in which Devasena and Vaḷḷi are both daughters of Tirumāl and perform the same penance with the aim of marrying Skanda; they obtain their goal and Devasena incarnates as daughter of Indra, (because brought up by Indra’s elephant, Airavata, she is given the name Tēvayāṉai, a Tamil reinterpretation of Devasena, through yāṉai, “elephant” in Tamil) whilst her sister, born of a doe engrossed by the lascivious gaze of the sage Sivamuṉi, is brought up by hunters who find her at the feet of a vaḷḷi plant (yam), hence her name.

99This is the better known version of the story, but it is most probably later. It is true that Paripāṭal on one occasion designates Murukaṉ by the name Māl Murukaṉ, “nephew and son-in-law of Māl”, but the expression may apply to him only through Devasena, apparently the daughter of Indra, but in fact the daughter of Vishnu, whose sister Parvati is also the mother of Murukaṉ. This allusion to the formula of preferential marriage typical of South India, in itself remarkable, does not necessarily mean that Cevvēḷ was playing, as anthropologists would have it, the rule of “polygynie sororale” with the two daughters of his maternal uncle. Vaḷḷi, known as wife of Murukaṉ since Naṟṟiṇai (82, 4) is in fact never mentioned in Paripāṭal and Muruka. other than as daughter of the mountain, without reference to any kind of divine origin, even though we can hardly help noticing the parallel with the mountain dwelling Pārvati, daughter of Himavat. By the same token, she is the favoured one, the triumphant and beloved wife, brave and victorious like her warrior spouse and the courageous hunters of her race. Later Tamil purāṇas have moreover also retained the feeling that the marriage with Vaḷḷi must follow the rules of clandestine love, accompanied by secret meetings and an elopement.136 Vaḷḷi’s triumph will ultimately be attributed to the triumph of love, giving to the true devotee the image of his Bhakti;137 and of the two wives, Vaḷḷi will always be represented on images on the god’s right.138

100The love that triumphs in the Paripāṭal is never separate from the earthly delights which the god has showered upon the inhabitants of his mountain, and the Sanatkumāra of the Chandogya (VII), who teaches the Upanisad, the Kumāra brahmacārin who, in the purāṇas is one of the five mental sons of Brahma, given over to mental concentration and refusing to increase the creation (i.e. to procreate), is met with here as “astonishing Kumaraṉ”, master of the amorous themes of the Caṅkam, in poem IX. Prayed to in the genuine accents of Bhakti at the end of poem V, in VIII he receives the infinitely more worldly prayers of wives and of young girls in love; the prayer in poem XIV is, if we are to believe the commentary, no more than a pretext for an akam episode; some verses in poem XIX (85-96) very likely also relate to magico-religious beliefs which link the cult of the god to the fulfilling of the amorous hopes of women; and in this context, the numerous akam episodes scattered throughout poems to Cevvēḷ are not at all surprising. In the rest of the Caṅkam, Murukaṉ often appears in the treatment of love, while Paripāṭal in its turn considers him as god of love. The girls in the kuṉṟakkuravai episode of Cilap. after mocking the vēlaṉ and his god as we have seen, finish their song by praying at his feet139 for a happy marriage, for after all, the wife of Ārumukaṉ is Vaḷḷi, daughter of the mountain and thus daughter of their race, and it is this sacred couple, the ancient Tamils identified themselves with and that they loved to celebrate and solicit.

101Thus is the outline, in all its complexity, of the portrait of Cevvēḷ in Paripāṭal. These poems to the glory of the god of the mountain of Tirupparaṅkuṉṟam present him to us in most of his aspects: a puranic and pan-Indian god but a local god too, guardian and protector of a mountain people, a warrior and hunter god, married to Vaḷḷi and confronted therefore with the jealousy of Devasena, a god of devotees and, lastly, of lovers. The problems his origin, his nature and his cult raise for historians of religion today, and the mixture of general literary traditions and philosophical speculations with local usages, and even popular superstitions: all this is already reflected in these poems which have allowed us thus to open the first Tamil dossier devoted to Murukaṉ.

IV - Tirumāl

102It is another mountain, to the north and three times farther from Maturai than the preceding one, which honours another god, even though Murukaṉ also has his sanctuary here and is still perhaps the guardian god of the place. This mountain is Tirumāliruñcōlai, the Dark grove of Tirumāl. Its god is Tirumāl, to whom are dedicated the most straightforwardly religious poems in Paripāṭal, the only ones that are integrally religious.

103If we are unable to affirm that, despite their puranic references, the totality of poems to Cevvēḷ are not outstandingly Tamil and do not express the original tones of the devotion of the people of Maturai, the same is not apparently the case as regards the hymns to Tirumāl. Truthfully speaking it is those which have earned for Paripāṭal its reputation as plagiarist of Sanskrit within the Caṅkam. Examination of their content shows, however, that they are more praiseworthy than that illhumoured opinion would let us suspect.

104It is true that Māyōṉ, black god of the forest with the characteristics of Tirumāl, closely resembles Sri Krishna, even in the form of his name: Tiru- is Sri, and māl, like Krishna means “black”. The multiple speculations on the name have, however, caused this sound philological explication to be rejected today by many Vaishnavas. The attempt is to establish a distinction between two radicals, ma- which on its own would mean “black” and māl- which would only mean “big”. In fact, māl-, as also iru-, means either one of these and the philology does not constrain us to decide. It is, moreover, remarkable that this radical māl- is very rich, meaning also ‘attachment” and “enchantment”, as independently attested to by verbal formations. It is consequently quite natural that devotion plays upon these multiple possibilities if we remember as well that Māyōṉ resembles Māyā, giving us yet another analogy. When the name of Tirumāl is pronounced today none of its possible meanings is in fact absent from the mind. If the historian has a preference for the one which is a translation of Sri Krishna, to the extent, certainly, that Krishna is Vishnu in all respects, it is due, above all, to the content of these poems and the relationships they suggest.

105The first verses of the first poem already present the essential: Tirumāl, that is Vishnu, is linked to images of Ādiśeṣa, Śrī (Tiru, Lakṣmi) Balarāma and Garuḍa, and his cult is introduced in the Veda; to celebrate him is to say the unsayable, evoking or conciliating the oppositions, and randomly reciting litanies whose source is in the tradition of the Upaniṣad. This is a desperate but fertile literary undertaking sustained by the intuition of a sentiment which is, in the last analysis, the essence of devotion: the love of God, in all its awkwardness, has its grandeur and the infinitely distant most high is not inaccessible to it for He bestows grace in response to prayer. The poems to Tirumāl are therefore songs of praise, of love and of supplication.

106Praise is expressed mostly in terms of speculation and of the brahminical cult, some elements having direct Vedic reference.140 Tirumāl, for example, measured the earth in three steps to the confusion of the asura (III 54-56); another passage seems to be transposing these three strides of Vāmana from space to time, of which Tirumāl is source and master (XIII 46-47 and n.). Protector of the sacrifice and master of yūpa in the Veda, the post-vedic Vishnu takes on the sacrificial function of Prajāpati, to be the sacrifice. Tirumāl therefore manifests himself in the triple aspect of the word of the performer (the Veda), of the sacrificial post and the sacrificial fire (II 61-68); with, perhaps in the mind of the poet, an expression of the accepted supremacy of Vedic ritual and the brahminical tradition. It is moreover often repeated that it is from the Veda that we may know him and the Veda which prescribes his cult. (I 11-13; III 1-11, 62; IV 65; XV 63-64). What we read in Paripāṭal corresponds to what we learn from the Sanskrit epic, as well as from representations on Gupta coins. Lakṣmi (Mā, or Tiru, that is Śrī) is his wife and rests on his breast, and if she once bore the old Tamil denomination Ceyyōḷ (the Red) she is also, as in Sanskrit, the goddess seated on the lotus. Garuḍa, the enemy of snakes is the mount or standard, celebrated as the equal of his master; we even know the story of his mother Vinatā, enslaved to Kadruva, mother of the snakes and frightful trickster, and the theft of the ambrosia by the bird to free his mother. The attributes the most often named are naturally the disc and the valampuri conch (II 36-51) which, with the bow, the mace and the sword constitute Vishnu’s classic five weapons (pañcāyudha, cited XV 58-61). The avatāras mentioned are those generally regarded as “ancient”: besides Vāmana, Varāha often mentioned in connection with his double marriage to Bhumi and Sri Devi, Narasimha (IV 10-21) lengthily described coming to the rescue of Prahlāda, and, lastly, Krishna, associated with Baladeva. Other familiar images recall the churning of the Ocean of Milk, Mohini, the burning of Tripura, Brahma on the umbilical lotus of Vishnu, his “conscious” sleep on the serpent Ananta, the murder of Kesin, the Aditya, the Rudra, the Asvin: nothing here that would not be immediately recognizable to an indologist quite unfamiliar with Tamil studies, but nothing completely unfamiliar to the Caṅkam either.

107On the other hand, some essential aspects must be emphasised, especially concerning the Krishna aspect of Tirumāl, the principal element of which seems to be his association with Baladeva. The eulogy to the god of Tirumāliruñcōlai (poem XV) never separates them but even claims them to be indissoluble and mingles their attributes. Elsewhere, the attributes of Tirumāl are more easily seen as those of Balarama than those of his brother: a palm tree or an elephant standard, a lethal plough with which he ploughs the bodies of his adversaries (a usage which suggests the underlying gloss, “Saṅkarṣaṇa”, one who ploughs armies). The popularity of Baladeva is common to the entire Caṅkam141 (for example, Puṟam 56 and 58, Kali. 104-105), and we may usefully remember that Tolk. refers to his standard when it gives a special rule for the use of the word koṭi (flag) before paṉai (palm tree). Baladeva subsequently almost disappeared from the Tamil horizon and no trace is to be found today of the shrine that would have been consecrated to him on the Aḻakar hill where the temple of Vishnu is still the centre of a very popular pilgrimage.

108It is again the presence of Baladeva that leads to the reference to Ādiśeṣa, the serpent with five, or a thousand heads on whom Tirumāl rests: his description and that of Baladeva are several times so closely mixed that they are both fused and confused. It is of course common in Sanskrit puranic literature since Harivaṃśa (namely chapter 82) and Viṣṇu purāṇa to mix the description of Ananta with the human form of Baladeva. It is worth retaining here what we may call a parallel characteristic, since we may not speak of source with regard to a puranic theme: Tirumāl is certainly Vāsudeva- Krishna. The whole of Fragment I celebrates him as the nāga-god: Vāsuki and Ādiśeṣa, rope for the churning of the ocean of milk, protection of Meru, bow of Siva for the burning of Tripura, and finally as sustaining the world.142

109The silences are as interesting as the convergences. Krishna does not appear at all in the contexts that are well known in the Bh. Pur., that of the Mathura of the North and that of a “nuptial mysticism”. The erotic aspect, the pastoral life of the cowherd of Gokula, scantily known to the Sanskrit epics though very popular later, is completely ignored by Paripāṭal. In support of A. L. Basham, who sees in the Tamil anthologies the very oldest clear reference to the pastoral Krishna, for the Black One (Māyōṉ) “plays his flute and sports with milkmaids”,143 out of more than 530 verses devoted to Tirumāl, Paripāṭal offers only a single verse from poem III, in which Tirumāl is “Right and left, pot and plough, shepherd and guardian”.

110The dance of the pot refers, according to Cilap., to the defeat of Bāṇāsura and the plough is Balarama’s plough. “Guardian” is a fairly general word but “shepherd”, Kōvalaṉ, very directly evokes Krishna Gopala. The shepherdesses are not far off in fact: the expression “right and left” is most probably an allusion to a dance kuravai which is described in detail in a song of Cilap. called precisely āycciyarkuravai (canto. XVII), dance of shepherdesses, in the course of which Krishna and Balarama are alternatively to the right and to the left of Nappiṉṉai, Krishna’s companion in many South-Indian episodes. But compared with the 150 verses or more of this chapter of Cilap., Paripāṭal’s scant three words give an impression of restraint; this restraint is most probably significant, especially when we think of the wide exploitation of the theme in the later Tamil Vaishnava literature, from the āḻvār to the Bhāgavatam. There is in the Caṅkam itself, in fact, but a single other quite specific reference, to the sport of Krishna with the shepherdesses. In Akam (59, v. 3-6), the gesture of the elephant who bends the branches of the tall trees, yā, to nourish the young females is compared to “Māl who crushed a tree with his feet and bent its branches so that the daughters of the shepherds could adorn their hips with its fresh leaves (when they were bathing) on the wide sandy stretch of the river Toḻuṉai (Yamuna) with abundant waters and which is in the North.” Whilst hesitating to presume too much from this almost total silence, we are inclined to conclude that the Vaishnava Bhakti movement had not yet, in the poems of Paripāṭal, achieved its full development, for the theme which is lacking there would be the essential one in all later literature of this type. It would, moreover, be risky to try to measure according to a simple repertory of themes either the dependence or the originality of the Māyōṉ of the Caṅkam in relation to immense Sanskrit puranic literature. We may just notice that, in the other anthologies, Vishnu, or Māl, is known by his dark colouring, his disc, his conch, his standard with Garuḍa, Lakṣmi on his breast (Gaja Lakṣmi too being known). He is the god of preservation and is celebrated for his strides which measured the worlds and his victories over the asura hordes. Lastly, Balarama is mentioned in contrast to Krishna. This brief inventory144 is mostly a repetition of that in Paripāṭal: as we have stressed in the first part of our introduction, it is impossible to base any chronological discrimination within the Caṅkam itself against Paripāṭal on the presence of such themes. Rather, we could glean in other Caṅkam texts, references to episodes in the Sanskrit epics, Mahabharata and Ramayana, which are not found in Paripāṭal.

111It is curious to see Paripāṭal mentioning as a well known characteristic, and following two other puranic references, an episode for which we have no Sanskrit reference:

To dry up the streaming waters of the vast sky
You dry them with your wings, male swan (III, 25-26)

112The hamsa-avatāra is known to be sure, but for revealing the Veda rather than for this singular exploit. This deed is hardly surprising when we think of the solar aspect of Vishnu or of the descriptions the Mbh. gives of Garuḍa and of his terrifying flight which blotted out the sky, but which nevertheless appears to us to be without any parallel version elsewhere. In passing, we note that Puṟam 174, also relates as a comparison that Māl rescued the sun and re-established it in its place when it was stolen by the asura who plunged the earth into darkness by this method: here is another of Vishnu’s exploits that seems not to be mentioned anywhere other than in a Caṅkam text. Finally, Fragment I of Paripāṭal alludes to a battle between Śēṣa and Vāyu to determine which is the stronger: Ātiśēṣa covers the Mēru against which Vāyu dashes himself with so much force it seems the end of the world has come. According to the version by the force of Vāyu alone, or because Śēṣa, at the request of the gods, relaxes his grip for an instant, a fragment of Mēru is detached. In South India this story is repeated in a considerable number of Sthala purāṇa,145 which thus explain the presence of a rock or a hill, for example at Cīkali, at Tiruchengode (mount Nagagiri, in the district. of Salem) and, above all, at Tirupati, where not only is the mountain of Veṅkaṭācala explained in this way, but the seven hills are considered as the seven heads of the Serpent (Srisailam being on its tail and Ahobilam on its body). This narrative is found too in printed versions of the Vēṅkaṭācalamāhātmya (both in Telugu ed. and devanāgarī ed.) but this passage of the māhātmya is not itself found, as it should be, in current printed versions of the Brahma purāṇa (Anandāśrama ed.) and of the Bhaviṣyottara purāṇa (Venkatesvara Press). Thus Paripāṭal, and the Caṅkam, happen to be sometimes a far from negligible source of reference if only for a general study of puranic themes.

113Up to this point we have presented the mythic element which nourishes the praise of Tirumāl, but his praises are equally supported by another theme, more abstract and speculative. Attached to this theme is the idea of a god who is ungraspable although omnipresent and omnipotent, who is identified with the three functions of creation, conservation and re-absorption, with knowledge and order, with all cosmic forces and all elements, a god who, rather than being any object or feeling is the essence of each thing, in everything the best and the perfect; but this is a god too who transcends all these identifications: incarnated in each birth, which he alone makes possible, he knows no birth; origin of time, he transcends and escapes it (we stress here the cosmogonic aspect in which is inscribed the invocation of the varāhāvatāra at the beginning of poem II); nothing manifests except through him, but he himself never manifests as he is; the sole element of reality in all splendour, he too is nothing but illusion.

114There is nothing original in all this in relation to Sanskrit philosophical speculation. Wanting to attach oneself to the theological deepening of these stanzas would moreover be a mistake: it’s a song we have here and not a systematic treatise. What, on the contrary, must be underlined is the borrowing from philosophic, and specifically literary sources, of the most beautiful images and the finest antitheses and the most sophisticated formulae. Let us salute them in themselves, not for their originality of religious thought but for the success of their expression; such formulae as are often cited (III 63-70, IV 25-35) are easily found elsewhere, but in Sanskrit. What is remarkable is that they appear, for the first time in the history of India, in a language other than Sanskrit, and with an admirable formal perfection. Instead of deploring this forceful intrusion by Sanskrit Hinduism into a Tamil literature that some try in vain to describe as immune to such contacts, we should consider ourselves fortunate to have come upon a fusion that was, from the very beginning, so fertile. What is more, the fruits outdo the promise of the flowers since the full blossoming of all this is known as the literature of the Āḻvār and the Tiruvāymoḻi, which the Sanskritist J. Gonda recognises as occupying “one of the first places in the religious poetry of all time and all countries”.146 Before the Nālāyira pirapantam, were Paripāṭal poems to Tirumāl, which actually constitute “the first document of Hinduism of purely religious character written not in Sanskrit, but in Tamil”.147

115An attempt was made once to inscribe the speculative visions of all the poems to Tirumāl in a particular system of considerable importance, the Pāñcarātra. In his brilliant study, Pāñcarātra in Classical Tamil Literature,148 S. Krishnaswami Aiyangar affirms that the five types of cult known to the Pāñcarātra are found explicitly, or almost so, in these poems: adoration of the transcendental form of Paravāsudeva, the Vyūha, the vibhava, or avatāra of Vishnu, his position as Antar-yāmin at the heart of all creation, and his cult forms, or archa. It is true, in fact, that three of these aspects are essential to the poems: the transcendental form, the immanence within all creatures, and the avatāra; poem XV may be interpreted as consecrated to the idol of Tirumāliruñcōlai, even though the poem is much closer to the Tamil genre of āṟṟuppaṭai than to the ritual of the Pāñcarātra; as for vyūha-s, they seem to us to be more present in the commentary of P. than in the text. We are dealing with a more generalised concept of the purāṇā and the Āḻvār: the four colours Māl takes on according to the four ages of the world (cf. III 81-82 and note). Nobody, in fact, needs to draw technically upon the Pāñcarātra to explain how in the Bh. Pur. (X 26, 16)149 four colours correspond to the four yuga, nor to enjoy reading this stanza of the Tiruccanta viruttam (44) of Tirumaḻicaiyāḻvār:

Nature of milk, nature of pure gold, nature like
The green of algae, nature blue as the flower in the lovely pond
The beetles wander on humming: four times thus is time filled up,
The earth hides the nature of Māl from us; of what nature is it?

116This text can be compared instead to the invocatory stanza of the Bālacarita (I, 1) of Bhasa, citing the same four colours. Thus the reference to the Pāñcarātra does not seem evident to us; it is in any case an argument that works both ways: employed by S. K. Aiyangar as a plea in favour of the ancient origins of the Pāñcarātra, it may as well support those who wish to assign a later date to Paripāṭal, as seeing no reason to assign a date before 5^th c. to the Pāñcarātra. What we shall, in any case, retain from that study is that it demonstrates that an almost complete picture of the cult of Tirumāl is attested to in these poems.150

117The counterpart to a relative disaffection with purely philosophic speculation and the unreserved welcome given to a Sanskrit tradition placed under the sign of the Veda, is the personal and profound tone of the religious sentiment. This personal trait is already there in the aesthetic aspect of the song: command of images and a lovely exhilaration of invocations. This flows from the source to express a deep devotion: a call for grace, aruḷ (I 36, 38), for union with God to celebrate and adore him (the quasi-formula concluding each poem); a call for the indulgence of God for the sincerity of clumsy praises (a fine debut in IV, 1-5). The contradiction is painfully perceived between the infinite imperturbable transcendence and the immanence of the God perceived with the heart (IV 53-56), but is apparently resolved in the movement of confidence that ends the poem (IV 70-73), a hope that is taken up by poem XIII where Tirumāl appears as God of deliverance (v. 7-13). But the true desire of the Bhakta is for his perpetual attachment to the feet of God, there to praise him unceasingly (XIII end). Understood by the allusion of IV 2 to the only way (samādhi according to P., but perhaps also Bhakti mārga, path of devotion), Bhakti, in fact, already underlies the religious fervour of these hymns. More than concern with ritual acts or with knowledge, even be it intuitive, it is humble devotion to God which is expressed in some places. It is however to be recognised that Bhakti is less anthropomorphic here and even less emotional than krishnabhakti is.

118It is tempting to see a parallel between Cēyōṉ-Murukaṉ, god of the mountains and Māyōṉ-Krishna god of forests and pastoral lands but that would be rather superficial. Cilap. certainly celebrates each of them by analogous dances (āycciyarkuravai and kuṉṟakkuravai ch. XVII and XXIV), but we are not in the Cilap. here. In Paripāṭal Murukaṉ is closer, Tirumāl farther off, Murukaṉ more concrete and Tirumāl more abstruse. Murukaṉ and Krishna undoubtedly both appear with the characteristic of God as a young man and we are aware of the importance of such a cult of the young hero,151 but in Paripāṭal Murukaṉ is eternally young, while Tirumāl is ageless: the accent is decidedly different. While we don’t know a great deal about either Cēyōṉ or Murukaṉ, it seems to us that however powerful the puranic influence upon him, Murukaṉ remains popular and directly perceptible while Tirumāl, however fervent his celebrations, remains ideally more abstract; in that way he is not, or not yet is, Krishna Gopala, herdsman hero of Mathura, dancing with the Gopis.

119This more abstract, more interior god, however, also knows how to show himself more directly accessible. Thus Fr. I where the praises of all the “good things” in the city precede those of Tirumāl, reminds us of the confusion between the playful festival and the religious procession, already emphasised apropos a poem to Cevvēḷ; the atmosphere of droves of pilgrims must therefore not have been very different no matter which temple they were going to. On the other hand, as an echo of the Tirupparankunram mountain, a poem to Tirumāl, is also devoted to a particular shrine of the god, that of Tirumāliruñcōlai, known today as Aḻakar Kōvil.152 Situated twelve miles to the north of Maturai and standing against the wooded Aḻakar Hills, the site of the present temple is magnificent and is still a popular centre of Vaishnava devotion. The mountain environment is full of historical vestiges and diverse shrines. The vaishnava cult is sometimes associated with the popular beliefs of the Kaḷḷar of the neighbouring villages. There is material here for a separate monograph because, between the Buddhist and Jain references from the past and the temples dedicated today to local deities, stretches a considerable literature, essentially but not exclusively Vaishnava, for a dispute between Saiva and Vaishnava did question the vaishnava hold on the site. All the religions of India, in fact, must have contended for this admirable site, with the hill and its trickle of water that dries up in summer and becomes a torrent during the rains, the Cilampāṟu. There is a detailed but confused description of it in the copious English introduction to the Sanskrit and Tamil edition of its sthalapurāṇa, which includes literary references as well. To poem XV of Paripāṭal which is particularly brilliant and evokes, at the same time as it evokes Tirumāl, the colourful splendours of its mountain, its flora and fauna, its river and the kuṟiñci assignation at the midnight hour, we find hardly any parallel of ancient date other than the last part of the second song of Tiruvāymoḻi, by Nammāḻ vār which emerges as a passionate and lively evocation. These two poems make up a sort of “Tirumāl āṟṟuppaṭai”, an invitation to pilgrims to come to the feet of Tirumāl, which does not at all pale before that of the Tirumurukāṟṟuppaṭai to the six shrines of Murukaṉ. Nowadays, the last of these six high places is ordinarily identified with Aḻakar Kōvil. This would be to admit onto the same mountain the two deities celebrated by Paripāṭal, and we in fact notice that the Paḻaṉit tala-purāṇam, in the kiriccurukkam which evokes the mountains where Skanda dwells, places between a stanza on Tirupparankunram and one on Tiruccentur, a quatrain on Paḻamutircōlai defined thus:

...One who has twice six eyes and was born from the eye,
,..of our Lord hard to see, even for One who has a thousand eyes,
...One who is the colour of clouds that leave not the sky, their dwelling
...Is the mountain of Paḻamutirccōlai which secretes the fresh honey of flowers

120Does the Aḻakarkalampakam bring anything of precision to this pure and simple association? Vāṇagiri, another name for this site, is the dwelling place of “The One who removed the dishonour of the Pañcavar (Pandavas), the One who stays at the side of The One who has twice six eyes”. The conclusion has been drawn that Murukaṉ was there first and was later joined by Vishnu; but, while remarking the intentional ambiguity of the original text, we can do no more than note, once more, their co-existence. Today, the main temple is dedicated to Vishnu, but a modern temple has been built higher up the mountain, though not at the summit; it is dedicated to Subrahmanya and visited by large numbers of pilgrims who come on foot. The way passes by other shrines, to Garuḍa, Hanuman etc. but we look in vain for the temple to Baladeva. It has vanished, at least if his cult has not been absorbed into the principal shrine, thus achieving the fusion sketched out in Paripāṭal. We add, lastly, that another shrine, on the north slope of the Aḻakar hills, ten miles as the crow flies from the principal temple, is called Āti-Aḻakar Kōvil (attached to the village of Mūṅkilpaṭṭi, Melur Taluk) and claims the honour of being the original of the present Vaishnava temple. It is in ruins now and was plundered several years ago. Not all the statues can be found that are mentioned in the English introduction to the Stalapurāṇam of Aḻakar Kōvil, which indicates its existence153 and surrounds it with an aura of mystery. What remains, including the site itself, does not militate in favour of this claim. All the same, the pillars of the mukha-mantapa, only two of which are still standing, are of the Nayak period, and various elements in the reconstructed ardha-mantapa may perhaps be very old (frieze of gaṇa, empty kudu, circa 12^th c.), and a few inscriptions in Tamil and Grantha are also to be studied. In short, the Aḻakar mountain has not yet yielded up its secrets.

121Introducing us thus to the sacred sites of the inhabitants of Maturai, at the very heart of their Vaishnava faith, the poems to Tirumāl constitute therefore an essential document. Despite their style which is more difficult than those of later devotional poems, they are of the nature of genuine prayer and are sometimes very beautiful. Prefiguring what was soon to become the pride of Tamil literature, the songs of the Āḻvār, they are something more than simple precursors. It must be recognised that Paripāṭal, after having given us the first hymns to Murukaṉ, offers us as accompaniment the first Tamil litanies of the Bhakti of Vishnu; it is usually accepted that Bhakti was born in Dravida, most especially beside the Tamiraparani, but it is at Maturai, on the banks of the Vaiyai, that we hear its first song.

“Brittany is the Universe” Saint-Pol Roux

122At the end of a too brief and hasty appraisal, nothing but an invitation to the reading of the texts, we would simply say that the interest of the Paripāṭal is not limited to the erudite satisfaction of the revelation of the details of a difficult text. On the contrary, it shares to the fullest extent that which defines the tone most often attached to early Tamil Caṅkam literature; under formal research its freshness of feeling and its authenticity are never lost. In this mutilated anthology we sense a large number of people living, loving and praying in a small province. But all around Maturai, and its river and two hills, there is Hinduism in its entirety with its myths and its images, in its original form and with the particular conventions of its Tamil expression. And because these poets were great poets, the proud particularism of the subjects of the Pāṇṭiya king offers universally human interest. Whether its muses frolic, celebrate or sing their prayer, the Paripāṭal has the value of a great classic. It is the humble satisfaction of the philologist, who might be imagined busy minutely weaving shrouds for dead gods, to make contact through these songs with the beliefs, the joys and the sorrows of a living people.

123INVOCATORY SONG

124Apart from the Paripāṭal and the Tirumurukāṟṟupaṭai, early texts specifically devoted to Murukaṉ are few: the poem of invocation of the Kuṟuntokai (translated in note under XIX, 100) attributed to Peruntēvaṉār, and a brief chapter of the Cilappatikāram (kuṉṟakkuravai). Later would be added some stanzas of Tirumanitram and the long poem of invocation in the Kallāṭam; but while we may glean from the Tēvāram fifty or so scattered quotations that attest to the fact that Murukaṉ is known, we must wait for the Tiruvicaippā to find a poem addressed to him; we thus understand better the desire to include the Tirumurukāṟṟupaṭai in the twelve Caivat Tirumuṟai. All the same, amongst the short isolated fragments collected for the most part in the Peruntokai of Mu. Irakavayankar, there figures an entire poem in Kali metre, which is very close to the prosody of the Paripāṭal and which two commentators of Tolkāppiyam, Pērāciriyar and Naccinārkkiṉiyar, have preserved for us in their study of sūtra 152 of Ceyyuḷiyal. According to them, it is an ancient text; the subject, the vocabulary and the metre so obviously belong to the poems to Cevvēḷ in the Paripāṭal that we have decided to present it in an appendix to our edition. However, the technical exigencies of the page layout keep us from inserting it there. Felix culpa! The same obstacle that deprives us of the poem of invocation which must have opened the Paripāṭal, as at the beginning of each of the other Anthologies, in a sense impels us to substitute these verses for it. A superb hymn to the glory of Murukaṉ, it ends with the prayer, at once humble and confident, of a Tamil literary figure; we cannot imagine a better opening song.

125This essay originally appeared in French as an introduction to the French translation of Paripāṭal, Le Paripāṭal, Institut Français d’indologie, Pondichéry, 1968. Verse references are to the French edition.