Caṅkam literature and its public
p. 25-56
Texte intégral
1The Caṅkam at the origins of Tamil literature around the beginning of this era is, in a word, a collection of anthologies (eight, and a series of ten “songs”), 2,381 poems, by 473 poets and 102 anonymous, a grammar that may be contemporaneous with these texts, 18 works, probably later and considered minor, and some miscellaneous fragments.1 To this is attached, by an historically contestable and contested extension, some long poems which are more narrative than properly epic and of later date, and various works of imitation or transition. The corpus is extended by a series of medieval commentaries and theoretical works that gloss a posteriori the poetic principles which essentially are a structured collection of literary conventions defining an “interior landscape” (Tamil, akam) and an “exterior” (Tamil, puṟam) universe of civil or warlike lyricism: Callimachus and Pindar in other words.
2Why should we broach the question of the Caṅkam by way of its public? As all literary texts, it has obviously two publics, the first contemporaneous with its creation and the second, spread throughout history and consisting of those who seek to draw from their reading historical or sociological evidence or aesthetic pleasure. Fairly common too is the fact that the Tamil Caṅkam has no precise historical or social context beyond what it itself says. We have then a text which is its own context and its own first, if not only, witness, in quest of an initial public which its future public will not be able to comprehend or imagine except through the information it itself gives. There is some similarity here with the history of the Homeric poems, though the archaeological context is not nearly as rich, neither the structure of the works identical, or with that of Vedic literature, though with a much more modest retinue of exegeses; there is further analogy with the Veda in any case, since the Caṅkam has its medieval commentators just as the Veda has Sāyana, both disparaged but necessary, having the same irreplaceable function in the tradition and the same lacunae and philological weakness of interpretation.
3Before concluding, however, that we are dealing with nothing more than a regional variation on the theme of vanished cultures preserved in our memories by the evidence of their, fortunately lasting, literature, let us remember that the Caṅkam, unlike the Veda, is a literature claimed to be “retrieved”, or we might say, twice rediscovered, first during the middle ages and then in the current epoch, by a population which is today attaching itself to it with existential fervour as though it were a creation myth or a myth of the golden age. This somewhat modifies the historical perspective if not the methodology.
4What in fact was the extreme south of the peninsula at the beginning of the present era apart from the Caṅkam? D. D. Kosambi replies brutally to this question: “a state of late stone-age Savagery, with a trifling amount of metal in use at a few undated spots”.2 This is not very flattering nor completely accurate. Elsewhere, on the sole testimony of the Caṅkam, he himself evokes a possible extension of the conquest of the armies of Chandragupta Maurya as far as Madurai3 but, more generally, his resolutely integrated northern perspective overlooks the Caṅkam when speaking of ancient literature; this is quite surprising in view of the place he gives to the Anthology of Hāla, the only trace left of the output of the Satavahana. The waves of his analysis of Aryanisation break against the Vindhya mountains even though Agastya crossed them, though that of course is in the realm of myth. Thence comes a rather offhand conclusion which leaves to “professional” historians the profusion of the Southern dynasties, which “make a nice but generally meaningless list”, to retain a single, but fundamental, principle, that of “mutual acculturation”, defined thus: “The higher Brahmin culture was imposed upon or adopted by tribesmen while primitive elements were reciprocally absorbed by the Brahmins”.4 That says it all, though perhaps hastily, since against the myth of the Aryan pioneers is set another: the irreducibility of the South, a myth that is very much alive, and supports this statement by Wendy Doniger O’Flaherty: “One of Stephen Potter’s best ploys was the one in which the Lifesman, finding himself involved in a conversation on a subject of which he was totally ignorant, while Opponent was an acknowledged expert, would simply remark from time to time, “Not in the South”. I fear that any Lifesman could go far with that phrase, were he forced to discuss this book. South Indian Tamil texts are a world unto themselves…”(italics ours).5
5If this is what American anthropologists believe, or affect to believe, it is understandable that Tamil nationalist elites have, since the second half of the 19th c., been at pains to rediscover through the Caṅkam texts, forgotten though they were for three centuries, the historical reality of an indigenous culture, while adhering to the “historical evidence” above all, even more perhaps than to the pleasure of the reading. The story of this long progression which pits those biased in favour of the Aryans and those in favour of the Dravidians against one another, and in which the rather separatist non-Brahmin movement has forged its weapons, is fairly well known. It is evoked, for example, in the works of Eugene Irschick,6 and conveniently taken up in an article by Suvira Jaiswal, “Studies in the Social Structure of the Early Tamils”7 in which the author rather too facilely denounces the equivocators. Writing under generally innocuous titles: Tamils eighteen hundred years ago (V. Kanakasabhai; written between 1895 and 1901), History of the Tamils from the earliest times to 600 AD (P. T. Srinivas Iyengar, 1929), Social Life of the Tamils: the classical period (Singarevelu, 1966), A social History of the Tamils (K. K. Pillai, 1969), The Poems of Ancient Tamil (George L. Hart III, 1976) and, in Tamil, the very classical Tamiḻar cālpu, (caṅka kālam) by S. Vithiananthan (drawn as of 1954 from his...) much plundered Ph. D. thesis from London University), are all in pursuit of a primitive Dravidian culture, or at least the earliest: as pure as possible in either case. All of them make claims for an historical progression as does N. Subrahmanian who, for “historical” purposes, groups, by subjects in his Caṅkam Polity (1966), and in alphabetical order in his Pre-Pallavan Tamil Index (1966) fragments of “literary” quotations whose value as historical evidence is precisely what is in question. All this effort collapses before the definition of the corpus itself which is strictly literary and has no other chronology than that invented by literary critics.
6Thus, the Muttoḷḷāyiram is excluded because the literary tradition has not retained it even though its subject and language belong to the epoch defined; N. Subrahmanian evokes the same tradition to include Takaṭūr Yāttirai, but drops it to draw profusely, in defiance of the evidence of the commentary, from Iṟaiyaṉār akapporuḷ. He has perfectly understood that the reference in Nālaṭiyār to the family of Muttaraiyar, known only from the 7th c., is an argument that cuts both ways: it may be a terminus a quo for the text, but, inversely, the text may bear witness to the antiquity of the dynasty.
7The scholastic discussion on the designation by the medieval commentators, Nacciṉārkkiṉiyar and Pērāciriyar, of the authors of the eighteen “minor” works rests upon a subtle distinction: are they piṟ cāṉṟor “posterior authors” or piṟa cāṉṟōṟ “distinct authors”? the latter reading is a little less grammatical but the palaeography does not allow for a conclusion. This is as much as to say that the literary fact itself is uncertain, as is the case with all these “heroic” or “bardic” literatures, stuffed with problems, larded with interpolations, covered with scholia, and embedded in commentaries. With such texts as both the source and sole verification of the hypothesis, it is obvious that there is bound to be doubt. But it is also necessary to remember that these texts have been responsible, since the beginning of the 20th century, for sufficient credibility being given to the idea of a Tamil culture flourishing at the beginning of our era: an idea up to a point accepted today even if, in India and abroad, it is periodically reconsidered, by Vaiyapuri Pillai or B. G. L. Swamy, for example.8
8We are thus left with a mass of convergent information despite various excesses of zeal or of subtlety, but we must also guard against another type of confusion: the more or less “Dravidian” quality of the civilization of which these texts are the echo. In other words, do they give evidence of a “Dravidian” public as some believed? A number of learned Tamils and various Tamil scholars see in South India, if not the original dwelling place, at least the most finished expression of the Dravidian family, when it was still uncontaminated by Aryan invasion. If, however, we still know very little about Dravidian origins we do know enough, at least in the negative, to consign this illusion to the domain of racial myth.
9The supposed, and doubtful, linguistic affinities between the Dravidian and the groups of the “agglutinating” type of the Mediterranean (Elamite) and the Finno-Ugric world tended to support ideas both of an extra-Indian origin and of a probable intrusion by the Northwest to which the Brahui could bear witness (and perhaps too the “agglutinative” character tentatively identified in the language of the Indus, unless we accept the assumption of S. R. Rao9 that the Indus script, being already Indo-Aryan, the Aryan-Dravidian fusion must have been pre-Vedic). But all these are only fascinating speculations, condemned to remain so.
10Considering only the most reliable works, those of experts such as Emeneau, Burrow, and Kuiper, and of a younger generation, Southworth, Colin Masica, or even McAlpin, we know that the Dravidian influence is perceptible from the most ancient strata of the Veda, not only in phonology and vocabulary but also in the structure of phrases, and that it may go back to the Indo-Iranian. If we don’t, therefore, know much about what a Dravidian is on the racial or linguistic level, we are at least sure, (and this is, in a way, the recognition of an old alliance), that we may no longer, at least since the Indo-Iranian, speak of the Aryans in the original sense and with a racial connotation. Not only can the Sanskrit heritage no longer be vindicated as of exclusively Indo-European origin, but Iran appears as a curious melting pot,10 the Hindukush as a precious conservatory,11 and Indo-Aryan culture as composite and permeable to foreign influences. The degree of speculation in these considerations does not render less serious the need to be on guard against the notion of a “Dravidian” culture in which the first public for the Caṅkam would have been immersed.
11In speaking in rather general terms of comparative culture or linguistics, however, we have changed the module; in this chronological zoom effect the Caṅkam and the modern Indo-Aryan make up one stage. Let us turn to the more strictly historical view, for example in the results of an enquiry by Colin Masica into the agricultural vocabulary of North India.12 On the one hand, there are fewer Dravidian radicals in the earliest strata of the Ṛg Veda than there are in recent ones and, on the other, “many a Dravidian word current in Sanskrit has left no living descendant in Hindi”; the far from negligible importance of the Dravidian element appears to diminish between Sanskrit and Hindi. The period of optimum Aryan-Dravidian contact thus seems to have been in a “Vedic” epoch.
12For our part, when looking for Sanskrit and Brahminical elements in the Caṅkam, we note numerous traces of Vedic ritual and especially essential terms: yūpam (yūpa), avi (havis), āvuti (āhuti), tavam (tapas), and “Dravidian translations” such as, tūṇ for sthūṇa or muttī for tretāgni.13 This is all quite coherent. These are terms found in the earliest collections, as we naturally do not take into account a work such as Ācārakkōvai, without doubt the latest Caṅkam opus; treating of rules of conduct and quotidian observances, it has deliberately taken from Sanskrit Brahminical texts such as Dharmaśāstra and smṛti. This borrowing is clearly acknowledged in its preface which the, unfortunately too widely distributed, edition of the Caivacittāntak kaḻakam preferred to ignore.
13If we class, but even more speculatively, amongst the words of probable Dravidian origin, terms of quite significant cultural importance which also appear in the Veda, such as kuṇḍa “(sacrificial) pit”, nagara and paṭṭaṇa, two terms for the city and port-town, as well as pura, “town” or “fortress”, (which may come from the Dravidian pura- “to guard, protect, defend”?), or if we play on more abstract connections such as pēy/piśāca; māya/māyam; tānū/tān,14 we are willingly falling in with the hypothesis of contacts between “Dravidians” and “Indo-Aryans” between 2000 and 1500 BCE, either in or to the west of, the Indus valley, but without doubt through some form of (post-)Harappan civilization whose official language is unknown and the exact extent of which is still under archaeological investigation.
14Abandoning the spinning of hypotheses before it becomes gratuitous, we note that it is reasonable, in any case, to reverse the terms: if the concept of the cultural and linguistic unity of the Indian area has gained ground little by little, and its evolution been postulated in terms of fusion, it is appropriate to study it in terms of differentiation too, which means being attentive, in the case of each phase of a regional culture, to the effort towards the specificity of that culture’s own expression if not of its patrimony. From such a perspective, the Caṅkam appears to be inserted into an historical and cultural network which attaches it much more closely to the overall movement of the civilization of the Indian peninsula than can be said of it if its particularities are too exclusively insisted upon. At the same time, it would seem to be strongly characterised by a literary milieu with very elaborate conventions, as we have written elsewhere “l’académisme d’une académie introuvable” (the academism of an elusive academy), inseparable from a grammar, Tolkāppiyam, which is in a sense the linguistic and poetic consciousness of the group.
15The Caṅkam texts give the impression of being isolated if approached solely by way of Tamil, but all around them, in Sanskrit, Pali and Prakrit, the literature of the Maurya period, the historical evidence of Jain and Buddhist texts, the inscriptions of Ashoka, etc. frame the Tamil land within a network of correlations which we are bound to evoke: and this whilst that proto-historical civilization the pre-historians have been investigating was still in existence, and the first local inscriptions in Brahmi script were appearing.
16Certainly, the earliest texts in the Buddhist tradition describing the Dakṣiṇapatha are not very precise south of the Godavari. Their source is hearsay just as the Chinese pilgrim, Fa Hien’s, was as late as the beginning of the 5th c. CE. The Tamil realms were known however, Uraiyur and Madurai being mentioned for example, as well as the port of Kavirippattinam, but it is evident that Buddhism had not yet really crossed the Vindhya,15 even though it was establishing itself progressively by way of Nagarjunakonda and Kanchi. The maritime route was very probably more frequented, and it is by way of Ceylon and its chronicles that we learn the most. Without a doubt, the country that Indian historians today still readily refer to as the “Far South”, is more naturally cited for its wealth than for details of its population and its culture; in this regard, Kautilya, who praises the southern commercial route, speaks of Tamiraparani pearls and Madurai cotton as if he himself were a foreigner like Megasthenes. However, the overland route leading to the ramparts and golden gates of the Pāṇṭiya capital through forests, mountains, rivers and caves is, as the Kiṣkindhākāṇḍam describes it (section XLI) in the Rāmāyaṇa, more romantic than frightening; it is difficult of access all the same, as is appropriate for a an expedition of the monkeys’ army. Agastya, it is not to be forgotten, was already there, apparently master of the place, and that is what matters. It is still the case however, that the “Far South” has been left out of the great descriptions of Buddhist society drawn from the canonical texts and the Jātaka, as well as of the descriptions, classical thenceforth, of the interior and external commercial development, based on Neolithic agriculture reinforced by the appearance of iron introduced from the North, concomitant with the development of the city culminating in the ideal of the nāgaraka. The schema of D. D. Kosambi are easily recognisable here. The best exercise on the application of these theses to the culture of Andhra was written by K. Satyanarayana,16 with a chapter on the Satavahana age which takes numerous illustrations from t7he Anthology of Hāla.
17The question thus arises of how to know what amongst all this is also pertinent to Tamil Caṅkam culture. In other words, does the social evolution reflected by the Caṅkam texts take up on its own account the movement going from Hāla to Vatsyayana, from village to city, and from “son of the village chief” to nāgaraka? This is only a rough schema which it is up to history to refine or revise. If the administrative structures and the dimensions of the kingdom are not the same at the level of beliefs or daily practices, conjunctions are going to be abundant nevertheless, though the illumination may be different. Thus, the Caṅkam texts on the poets or their mixed community of wandering bards and minstrels are very ethnocentric, whilst the occasional references from the Jātaka look at them from a distance. It would nonetheless be the same itinerant troupe of wandering minstrels, baton in hand, even did the texts of the North not confirm for us the speculations of G. Hart on the “sacred powers” connected with the baton of the rhapsode. Whether it is lovers in flight in the Caṅkam, or merchant caravans in the Jātaka, the dangers they face are the same: robbers, wild animals, thirst, demons, famine…17 These may be details but they are details of the weave of the texts and of the daily life of their public. More precision is possible when it comes to technical parallels as is shown by the enquiry of Thomas R. Trautmann into Dravidian Kinship.18 On the literary plane, we may discern comparisons between the erotic themes of the Caṅkam (tuṟai) and the picture of feminine wiles in the Jātaka. The parallel with Hāla has been pushed even further, right up to the translation of the anthology into Tamil and its rearrangement according to the canons of the Caṅkam.19
18With the epigraphy new coordinates start to become clear. The Lalitāvistāra fixes the number of scripts existing in the epoch of the Buddha at sixty-four, with amongst them a dravidi or damili. Then, all Dravidian scripts are derived from the Brāhmī script of Ashoka in use in South India and Ceylon until the beginning of our era with complementary signs in the Tamil to mark the Dravidian latent phonemes. Tamil inscriptions on rock faces and caves, and on fragments of pottery at Korkai, Uraiyur, Karur, Arikkamēṭu, Kanchipuram etc., are spread over a period approximately from 3rd c. BCE to 2nd c CE, which presumption favours the possibility of written communication between Ashoka’s administration and the independent kingdoms listed in edicts on rocks II and XIII: Cōḻa (Coromandel), Pāṇṭiya (around Madurai), Kerala (Malabar), Satiyaputra and Tambapanmi (Ceylon). It has been adequately demonstrated, or so it seems to us today, that the Brāhmī Tamil script is derived directly from that of Ashoka, in conformity with its conventions as followed and adapted by the earliest Tamil grammar, Tolkāppiyam.20 There’s a sense here of the cultural influence of the North and, moreover, the pressure on the “limitrophes” from a ruler, already cakravartin, who does not appear to recognise any other cultural existence than that of Indo-Aryan expression.21 We recall, however, the testimony of Megasthenes, who caught in the North an echo of the Pāṇṭiyaṉ kingdom Tamils of today are infatuated with, even though we hear in it rather indirect homage to Chandragupta with his troops, roads and police, that is, briefly speaking, homage to a pressure from the North that the conquest of Kaliṅka by Ashoka only confirmed. In this context, it is easier to appreciate the interest of a Tamil inscription discovered at the beginning of October 1981 at Jambai in a South Arcot village 15 kms from Tirukkoyilur,22 which reads “satiya puto atiyan neṭumāṉ añci iṭṭa pali” (pali the shelter of a, probably Jain, ascetic given by Atiyaṉ neṭumāṉ añci, Satiyaputra). The donor’s Tamil name is that of a prince celebrated by the greatest Caṅkam poets, Paraṇar, and more especially the poetess, Auvaiyār. His capital Takaṭūr, identified with Dharmapuri, about 80 kms. from Tirukkoyilur, was the site of a warlike exploit by the prince, narrated by Paraṇar; the testimony is corroborated by Auvaiyār. According to the palaeography, this inscription belongs to the first century CE, and should therefore be contemporaneous with another at Pukalur in which the name of Ātaṉ Cēral Irumpoṟai figures, who could be the conqueror of Atiyaṉ in the final battle fought at his town, Takaṭūr, when it was besieged by a Cēra king, Peruñcēral Irumpoṟai, who was ultimately the victor. As for the name Satiyaputo, which figures in this spelling in edict II of Ashoka, it has been variously interpreted, the most plausible suggestion being that of K. G. Sesha Aiyar,23 supported by Burrow and taken up by K. A. N. Sastri, that the Sanskrit form be seen here in a local compound *Catiya-> Satiya-, whose second terms puto (Skt. putra) means “son”, just as in “Kelala-puto”. The parallel continues in the doublets: as we have Cēramāṉ (-man <makaṉ “son”) opposite to Kerala-putra, so we find Atiyamāṉ next to Satiyaputra. And as regards the territory, it is certainly that which we might expect if geographical meaning is to be given to the order of principalities listed by the Ashokan edicts: that between the Pāṇṭiyas and Kerala. We note that the order is immutable in the variants of edict II. Some minor details in the written form of proper names and toponyms may certainly still give pause, such as the meaning of a title surviving three centuries after Ashoka. This is not, however, without parallel; we think of the bilingual coins, in Prakrit and Tamil, of Vaśiṣṭhiputra Satakarṇi, a century later (with the alternates – puta- / -makaṉ –there too).24 Lastly, an inscription of the 11th c. (Epigraphia Indica, vol. VI no 34 A, pp. 331-332,) recalls an ancient allegiance of the Atikaimāṉ Eḻiṉi to the Cēra monarchy, with a very well defined Jain connotation, which also accords with our account.
19The epigraphy must take us further however. Already factual history defines the area of the Caṅkam with landmarks hard to overlook, just at the time when commercial contacts with Rome and the Mediterranean world were at their peak, leaving coins and pottery (Aretine at Arikkamēṭu, red glaze, amphora) along the coast and into the interior, an interesting cross reference with the palaeographic dating of the inscribed fragments (majority 1st and 2nd c CE). We may stop there, and archaeology may at the very most offer an exploration of a megalithic civilisation flourishing in the South from 4th c. BCE to 4th c. CE. over an area which corresponds well enough to that of the supposed Dravidian speakers for the connection to be made, the more so as, for instance, the sketchy descriptions in the literary texts perfectly match almost all the various types of burials. That the citations suggested by K. R. Srinivasan25 are relevant is one thing: apropos the “hero-stones”, or funerary stele of warriors, belonging to the centuries immediately following, there has been a flood of similar literary references, but none of this proves anything, given that the phenomenon is for the most part pan-Indian. The interpretation of the Neolithic culture provides a convenient analogous schema of agricultural towns in valleys, as different from one another as are the Danube, Tigris, Euphrates and Indus: the same urbanisation, clay structures, seals and undeciphered pictograms or scripts; no relationship has been demonstrated however. The same goes for the megalithic culture of Andhra and the Tamil country.
20It would be more judicious to consider the extent to which this culture had already been penetrated and how penetrable it was, or even to wax rapturous over its remarkable refinements: irrigation, rich and diversified funerary material, decorated and painted pottery of high quality (which then declined), international contacts (Greece, Rome, China), and, finally, writing, as well as, in brief, its elevated degree of symbolism and of sophistication of ritual. It was in this very rich and cosmopolitan context that the first poems of the Caṅkam were probably composed for sophisticated farmers and cattle breeders, warriors, horsemen, and metal workers, and not within some narrow and primitive Dravidian culture. The connection between the material culture apprehended by archaeologists and the literary texts is not, however, easy to establish, and the less so as the archaeological exploration, especially of coastal or river harbour sites, is still in a sketchy state. A systematic effort at a synthesis has been made by Clarence Maloney26 but, due to the uncertainties of the information, it has not attracted all the attention it deserves. “Classical” archaeologists27 remain very careful in their descriptions and look for incontestable historical evidence before undertaking the speculative ventures, in which the entire generation of R. Ramachandra Dikshitar28 and T. Balakrishnan Nair29 to some extent lost themselves on the pathways of origins.
21The epigraphy then, has a surprise in store for us once again. A still unpublished30 Tamil inscription on a gigantic rock (15 x 2 m), discovered at Pulāṅkurucci in Tirucirappalli, may be paleographically attributed to 3rd c CE. It is dated 192 in an unknown era presumed to be the Saka era (thus 270 CE) and would be the earliest dated inscription of that era. It concerns a land transaction but what is far more striking here than the toponyms, all of which have not been identified, is, 1) the variety of territorial divisions mentioned: kūṟṟam, kōṭṭam, nāṭu, ūr; 2) the reference to gifts of land to Brahmins under the term kuṭumpu, an administrative division of the Cōḻa administration known of at a later date and, lastly, 3) the references to various governmental officials who would seem to have their equivalents in the Maurya regime. Briefly then, the technical vocabulary used is already that of the first Pallava and Pāṇṭiya charters. What are we to think of this inscription, localised as it is in a territory, that of Ollaiyūr, well known to the Caṅkam as the conquest of a Pāṇṭiya prince called Pūtappāṇṭiyaṉ who was the author of several poems? Without waiting for its publication in full, we may already measure its range: at its date the “Far South” is believed to have been still only minimally impregnated by the social structures of the North and thus to have been virgin soil for all the restitutions born of readings of the Caṅkam poems and the wishes of their interpreters. But we discover that the principal institutions of medieval India, of the brahminical culture and probably of that temple civilisation that was thought to have developed only with Bhakti literature, were in fact already in place and constituted the framework of the life of those who were the Caṅkam’s first public.
22Parallel to Buddhism and to Brahminism, Jainism, present everywhere in inscriptions of the first centuries CE and in the imagery of the caves, was evidently also introduced in the Tamil language thus participating in the literary flourishing of the late Caṅkam. The connivance of the Jains with the rulers before the age of Bhakti is often referred to. Might it not be better to remember that they very soon integrated themselves into a Tamil landscape all of whose fundamentals they accepted? Their role in the early literature cannot otherwise be explained, nor the fact that Peruṅkatai is a Jain work, but at the same time the last brilliant bouquet of Caṅkam traditions, artificially preserved at the height of the middle ages, probably 9th c or 10th c: a rare example of perfect integration. The non-brahminical literatures of the Buddhists and Jains are likely to have been popular for their narrative and humanistic aspects with the Caṅkam public whose tastes they spontaneously shared. The success of Cilappatikāram and of Maṇimēkalai also shows that the hypothesis is not gratuitous. It is therefore quite natural that the story treated in vernacular Paisaci by the Bṛhatkathā of Gunadhya (usually considered as belonging to the 1st c. CE and attached to the Satavahana dynasty) became the theme of another big kāvya which does not, on this occasion, have as principal subject a narrative familiar in Tamil lands but is an adaptation of an entirely northern plot that needed to be clothed in Tamil garb.
23It is interesting to look at the literary procedures by which this was done; Dr. R. Vijayalakshmy31 defines them thus: 1) the conformity of the description of the landscapes to the Tamil theory of five regions (aintiṇai) devoted to specific chapters; 2) the whole episode of the meeting between Patumāvati and Utayaṇaṉ, treated according to the conventions of love of kaḷaviyal with each step developed in a chapter as well; 3) the precise reference to “the Tamil custom” in the episode of the marriage of Viricikai to Utayaṇaṉ; 4) the description of the city of Rajagiri with its lengthy elaboration of the image of the city compared to a lotus, borrowed from the Paripāṭal32 but found elsewhere in the literary tradition of the Caṅkam too. This is less a question of an esoteric pastiche than of the restitution to the Tamil public and to us as well of a poetic language, revealed in its most characteristic aspects. We believe that the Caṅkam is a literary phenomenon and an original poetic language first of all, and Peruṅkatai demonstrates this in masterly fashion by recreating, centuries after its beginnings, this world of particular poetic conventions in the appropriate vocabulary. The enquiry pursued in the cultural milieu, at a distance from the texts, brings us back again to the texts whilst the meeting with an experienced public plunges us into the literary problem once more.
24It is this acute awareness of their poetic art which is striking in the tradition of the Caṅkam poets, the anecdote which is the best example in this regard being that of Nakkīrar refusing to submit to Siva in the course of a controversy over a mistake in rhetoric in a poem (the second of Kuṟuntokai), written by the god and offered as a gift to the starveling poet, Tarumi, to allow him to compete successfully for a literary prize: the impossible paradox of a supreme God who teaches the grammar and unexpectedly finds himself condemned in its name by his best poets.
25That anecdote is known to the earliest of the three authors of the Saivite Tēvāram, Appar (Tev. VI, 76, 3), who was one of the very first Tamil authors, dated with some certainty at 7th c., probably in the first half of that century. Thus he comes a generation before Arikēcari Parāṅkuca Māṟavarmaṉ (circa 650-700), who is today generally considered to be the hero of Pāṇṭikkōvai,33 that is, the stanzas which illustrate the well known prose commentary on Iṟaiyaṉār Akapporuḷ, the first text to give in full the legendary history, which appears in the preface, of the three Caṅkams or successive academies. The first two of these academies were swallowed up by the waters34 so that the texts we know today would represent nine of the fourteen titles cited as works of the third Caṅkam only. Without embarking upon a discussion of the historicity of the legend, where myth and events identifiable as something called “historical reality” interact with each other at will, a subject that merits a whole book,35 we shall simply highlight the point at which “history” and literature coincide here to such an effect that no historical deduction is possible other than on the presumption of literary likelihood. This is the core of the problem where we find the inter-dependence of the Caṅkam and its public exclusively in the evidence constituted by the words of the poets.
26In terms of the literati, what then is the message here? The existence of three academies succeeding one another over the course of long centuries is affirmed. Now to discern a demonstrable reality behind the unlikely legend! In the first place, without such an institution the keys of the poetry would be lost, and, in fact, the third academy at Madurai died out due to a terrible famine that raged for twelve years in the Pāṇṭiya kingdom. When the poets and grammarians who had fled the court returned, the phonetics, the morphology and the versification were retrieved but the poetics had been lost. This is a disaster which only the intervention of Siva can remedy. He engraves on three copper plates, which he causes to be discovered by chance in his temple, the sixty sūtra which summarise the love poetry of the Caṅkam, known by the name of Iṟaiyaṉār Akapporuḷ. It remains for the king to institute a competition, judged by a mute five year old child, an incarnation of the god Kumara, or Murukaṉ, to select a commentary having authority, the indispensable complement to the text, so essential that the intervention of the god Murukaṉ is no less necessary here than was that of Siva to establish the text. It is the commentary of Nakkīraṉār, son of Kaṇakkāyaṉār (“accountant” or “writer”) of Madurai which triumphs over the one by Marutaṉiḷanākaṉār of Madurai; the preface actually ends with the enumeration of those through whom this commentary was transmitted, apparently from generation to generation for ten generations, that is, approximately, three centuries.
27Though without illusions as to the “historical” value of that preface we must treat it according, first of all, to what the literary Tamil public has made of it in the course of the ages. First of all, the idea of a series of stanzas devoted to a single hero and illustrating the successive aphorisms of a treatise on rhetoric seems to be familiar, for this is the arrangement of a celebrated treatise of Nāṟkavirāca Nampi, Akapporuḷ viḷakkam thus brought into relation with Tañcai Vāṇaṉ Kōvai, in the 13th c. or, a little later, of Māṟaṉ Akapporuḷ linked with Tiruppatikkōvai. Also obvious is the disassociation of the style of these stanzas from the language of the Caṅkam poet Nakkīrar; obvious too is the conclusion that they are most probably contemporaneous with the last stage of the commentary in which the written text is established. Moreover the story itself, confused as we imagine it, introduces a time lapse between the Pāṇṭiya king, a contemporary of Nakkīrar and his commentary, and another Pāṇṭiya who is the, at least presumed, hero of the exploits of Pāṇṭikkōvai.
28If these last go back to the end of the 7th c. we shall still be in the realm of likelihood in inferring that Nakkīrar, and the close of the third Caṅkam are situated at the end of the 4th c. or beginning of 5th c. Whatever be the date however, if we are to accept the evidence of the colophons (it being understood that they don’t always have any evident connection with the texts they gloss and are sometimes barely coherent; but whoever accepts a commentary and the preface to that commentary had certainly better consider the scholia), the epoch in which our preface places Nakkīrar is one of many other coincidences which invite us to reconstruct further, not the history but a history, which makes some sense in spite of itself.
29The five year old mute who incarnates the god Kumara has a name, Uruttiracaṉmaṉ (Rudra Sarma), son of Uppurikuṭikiḻār, which is also the name the colophon of Akanāṉūṟu attributes to the compiler of that anthology made at the behest of that same Pāṇṭiya king (cited in Puṟam 21 and 367, author of Akam 26) Ukkirapperuvaḻuti. The fact that Akanāṉūṟu is already cited in the commentary of Iṟaiyaṉār Akapporuḷ as a constituted anthology allows for a degree of irony regarding the precocity of the miraculous child. It is more positive however to entertain the idea that the same epoch saw the completion of a corpus of amorous lyricism and the drafting of the treatise which would define its norms.
30Furthermore, Nakkiraṉār, son of Kaṇakkāyaṉār of Madurai, is known under this full name (with one or two variations in ending) as the author of several poems of the Caṅkam anthologies (Kuṟuntokai 143, Akam 93, Puṟam 56 and 189), and of a good many more if all the texts signed with only the shortened surname, Nakkīrar/n, are taken into account.
31The texts which offer a “complete signature” have already filled us in: Puṟam 56 compares the Pāṇṭiya king, Māṟaṉ, to Siva, Balarama, Krishna and Skanda successively and further on (v. 18 ff.) speaks of libations of alcohol and vases of gold with reference to the Yavana: classic brahminical mythology, and contacts with the Roman world; Akam 93 is arranged to feature the three Tamil kingdoms and their capitals in twenty-three verses, a brilliant rearguard exercise which does agree with the moralistic accents of the Kuṟuntokai 143 in which critics see reflections of the Kuṟaḷ. It would be risky on the basis solely of the fairly widespread name of Nakkīrar to credit the same man with two long poems in Pattuppāṭṭu: Neṭunalvāṭai and Tirumurukāṟṟruppaṭai,36 for the list would have to be enlarged even further and the irreconcilable presumed dates for these two texts engaged with, the former supposed to be quite early37 and the second late (according to literary criteria of content and style).
32And the same harvest is reaped by turning to the unfortunate rival of Nakkīrar, Madurai Marutaṉ Iḷanākaṉār. Strictly speaking, there are two poems of Puṟam, 55 and 349, attributed to him under that name; the first begins with five verses describing Siva, the destroyer of Tripura with a single arrow, with the snake for string and the Mēru for bow, with his black throat, the crescent moon and the frontal eye. A little prod at the variants of the colophons could take us deeper into the known texts of Puṟam and of other anthologies, and even towards a glorious paternity for our author. Zvelebil38 and John Marr39 have felt obliged to accede to these assimilations which do not seem to us to be indispensable. For, again, under his “complete” name appear twenty or so Akam poems (thus bringing the anthology, by its material, into close connection with the Iṟaiyaṉār Akapporuḷ) which are often very fine, and contain a precious series of realia of which some examples, taken at random, follow.
33The body of the heroine, sheltered from the eyes of the indiscreet, is compared to a “tall (sacrificial) stake (tūṇ = yūpa), strictly guarded, beautiful to behold with a cord tied at its middle, for that sacrifice accomplished with enormous difficulty in other times by the great one whose war axe destroyed the race of princes” (220 v. 5-8-a very explicit allusion to Paraśurāma). Poem 77 speaks of a village abandoned by its inhabitants because of drought and persistent heat, and also of bulletins written on palm leaves, placed in an earthenware urn, tied with strings and sealed, which seems to evoke the electoral system described subsequently for village assemblies.40 Elsewhere there is maritime commerce with the port, the lighthouse standing close to the village and huge vessels like the very earth on the march (255).
34Finally, we select several allusions to funerary stele of slain warriors. These show not only that these stele are decorated with garlands of peacock feathers or stand in complete rows (387) but that the names and exploits of the warriors are engraved upon them (131). More than that, they are already immemorial, forgotten, illegible and deserted (343); this shows that dependence is already upon the poetic vision of an earlier tradition, to which the poems of Puṟam were to bear plentiful witness, often quoted and collected together (compare with ex. Puṟam 264: lustration, garland of peacock feathers, inscription),41 the codification, or “ritual”, of which is presented in the texts on rhetoric parallel to the Iṟaiyaṉār Akapporuḷ, but devoted to civil and martial lyricism. We are dealing here with a pan-Indian phenomenon, let us remember, and the site of Nagarjunakonda has provided some of the earliest specimens, but the mention of the inscriptions is interesting in this regard since it may perhaps coincide in time with the 4th c.-5th c. date of the earliest, recently discovered stele, engraved in Tamil or vaṭṭeḷuttu script.42
35What we want definitively to retain of these philological explorations is not, however, a desperate attempt by the historian to reconstruct an exact chronology of literary data or to bring together in a coherent fashion the sparse and rarely concordant indications of the colophons of which the authority will always remain doubtful; it is a more general and consequently perhaps more generally exact impression, disengaged from these samples and sketchy synchronisms. The sum total of the notations taken from these, already late, texts offers us a modest assessment which does not override the texts nor go beyond likelihood. The vision is of a royal court, with its king who has around him the more or less organised college of his poet-grammarians, often in competition, and who also has his temple, let us not forget, the texts being sufficient reminder that the brahminical religion was already established; and, in this context, a very important literary phenomenon: the compilation of anthologies and reflections presented in the form of debates on poetry.
36Further, is it after all surprising if, in that anthology intended to bring together several generations, that of the compilers, which is the last, is represented therein by two of its most brilliant and possibly rival authors, each in the proportion of five per cent of the texts? Is it further surprising that the texts which are the longest and most sophisticated in their composition (those collected in Pattuppāṭṭu) seem to be linked to authors belonging also to the more recent strata? This is to some extent what we have glimpsed.
37The most important phenomenon is, however, most probably elsewhere. We noted that the famine lasted twelve years, a conventional length for this type of calamity43 and one which should not suffice to empty memories; in any case the capacity remains, when the catastrophe is over, to retrieve the art of versifying but not that of composing fine poems: prosody is there but not rhetoric with its chapter on themes and subjects. This is prodigious, as is the care taken by the medieval commentator Pērāciriyar to maintain for the theory of love poetry the preponderance of this treatise of Iṟaiyaṉār, which is the mutaṉūl, against the chronology itself, which made him place Tolkāppiyam before Iṟaiyaṉār. We should emphasise that he did not object to the fact the kaḷaviyal is treated as mutaṉūl in spite of its being a later text; he states it clearly: “piṟkālattup perumāṉaṭikaḷ kaḷaviyal ceytāṅkuc ceyyiṉum piṟkālattāṉum mutaṉūlāvateṉpatu aṟivittārkkum” (Per. Com to Tolk. Poruḷ. Marapiyal 94) this, of course, after the obligatory reference to a treatise of Agastya which would have to be the first, were it not lost. It would doubtless be tempting to anchor here a theory according to which the last part of Tolkāppiyam, dealing with poetics, would be later,44 recent, and by a different author. Speculation by philologists who conveniently forget that the poetic consciousness represented by Tolkāppiyam has to be contemporaneous with the writing of the poems is, however, of little importance. Let us above all hold to the indication given here that nothing is possible and nothing makes sense without a treatise on subjects in terms of poetics. This is the whole teaching of the account that we have followed. It is also what is emphatically declared by the author of a song of Paripāṭal which stigmatises the blindness and lack of sensibility of those “who have not studied the pure Tamil in its treatise on subjects” (IX 11-26). It is what demonstrates the concern with having, in parallel with the chapters of this treatise on love lyricism, a treatise on “exterior” lyricism which will be Puṟapporuḷ veṇpā mālai of the 10th c., a later work and a “secondary” one, derived from a Paṉṉirupaṭalam of the 9th c. which is lost but for a few fragments. In the amalgam of later centuries Nakkirar is not great simply by virtue of his works or for having rebelled in the name of poetic conventions against Siva himself, but because he alone knew how to recapture the essential: the poetics which takes account of, and gives meaning to, all the versified babbling.
38Obviously, the Caṅkam texts attest here to all the characteristics of a classical tradition, but rather than being satisfied with their own internal evidence, the idea came into vogue in the late 1950’s of serving the literature of the Caṅkam by treating it as the poems of Homer are often treated, that is according to the way we reconstruct the conventions of both the heroic age and the formulaic style proper to oral poetry. This was the thesis of K. Kailasapathy, on Tamil Heroic Poetry45 which, perhaps in a spirit of homage to Sir Cecil Bowra, was well received in England. The author is reverently indebted to Milman Parry and George Thomson for the Greek epic parallels. But the same year he was accepted by John Marr in the brotherhood of the “Et ego in Arcadia vixi”, Prof. Italo Siciliano, in a very well written and diabolically intelligent book,46 ferociously denounced the clichés that have accumulated around all oral literatures and their proceedings in a vain attempt to reduce the written masterpieces to such clichés in all their precariousness. With this book, a pamphlet really, but dazzling in its culture, the work of K. Kailaspathy was dated and qualified before ever it was published.
39To be properly understood, the debate on formulaic style, first connected with the epics and then extended to other genres, must be considered in its chronological perspective. After the first step initiated by Antoine Meillet and then by Milman Parry came the blossoming of “orality”, from Lord, to “the Chadwicks” (H. M. and N. K.) many others such as William Whallon and Ruth Finnegan.47 Finally came the witty attacks of I. Siciliano against the excess of formulaic and oral in the genesis of the Chansons de geste, despite which formulaic studies have survived. “The poetry of the formula and of the oral composition continues to rage and to proliferate, the more so as it allows for the effortless compilation of fine articles full of the formulae which fill the chansons de geste”.48
40It was in 1923 that a French linguist specialist of Comparative Grammar decided in a peremptory manner: “Homeric epic is entirely composed of formulae handed down from poet to poet. An examination of any passage will quickly reveal that it is made up of lines and fragments of lines which are reproduced word for word in one or several other passages. And even lines, parts of which are not found in another passage, have the character of formulae, and it is doubtless pure chance that they are not attested elsewhere.” (A. Meillet, Les origines indo-européennes des mètres grecs, Paris, 1923, p. 61)
41Milman Parry’s two theses published in Paris, in French in 1928 on L’Epithète traditionnelle dans Homère, Essai sur un problème de style homérique (The Traditional Epithet in Homer, Essay on a Problem of Homeric Style) and Les formules et la métrique d’Homère (Homeric Formulae and Homeric Metre), though more or less written under the impulse and guidance of Meillet introduce in fact many more nuances. But it was a landmark. It was, in fact, the academic teaching of M. Parry and the large echoes of field works conducted by him with A. B. Lord in Kosovo (Parry & Lord Serbocroatian heroic songs, 1954) that popularised the themes of the formulaic style and its links with epic or narrative literature, oral or written. Parry’s thesis, however, has been accessible in English only since the publication by his son, Adam Perry, in 1971 (Oxford Clarendon Press) of his collected papers under the title The making of Homeric Verse.
42So, it is only after a generation that an exact and comprehensive overview of the positions of M. Parry can be assessed. It is highly commendable to re-read his work because it has a much wider horizon than the technicality of his Hellenic or Slavic studies and, for our purpose, it remains completely relevant and current. It is in fact a model of historical criticism, as Ernest Renan defined it in a few words which Parry has repeatedly quoted, commented on or illustrated: “How can we seize the physiognomy and the originality of early literatures if we do not enter into the moral and intimate life of a people, if we do not place ourselves at the very point in humanity which it occupied, in order to see and to feel with it, if we do not watch it live, or rather if we do not live for a while with it?”49 This is a permanent invitation always to practise a strict relativism in our readings by taking into account the relationship between a text and its different publics, perennially renewed throughout history.
43Kailasapathy was not prepared comprehensively to evaluate the work of Parry, and that included the myths Parry had unwillingly created. Therefore his most positive philological contribution, which Tamil criticism took too long to discover in a Tamil version dated July 2006, could unfortunately not avoid the fact that by entering caṅkam literature into the ready-made category of “formulaic” literature he was driving it into a dead end. The elementary virtue of the formulaic, however, keeps its attraction and magic, and people come back to it today when further refinements are made possible and easier through the automatic concordances and other information technology tools. For example, the approach of “Literary Techniques in Old Tamil Poetry” by Eva Wilden, does not, due to its appearance of pitiless technicality, escape the irony of I. Siciliano which also maintains all its relevance, nor the fundamental criticism by C. M. Bowra who already wanted to know, how, in using formulae, “the poets shift beyond their purely functional use in composition to something that is almost purely poetical.” (Heroic Poetry, London, 1964 ed. p. 240 ff.)
44Greatly to the credit of Kailasapathy still is an analysis of the relations between artists and their patrons, with a subtle attempt to differentiate between authors performers and to classify the latter according to types of performance. But proposed analogies, in fact, mostly remind us of how the situation in that world of “wandering scholars”, (so designated by Helen Waddell in her 1927 book which remains a classic) should be studied in all the diversity which, within each culture, makes for its specificity. More than evidence coming from outside, we appreciate the inventory of themes, motives and metric formulae, as well as the emphasis on the “heroic” part of Sangam literature too often neglected in favour of its erotic lyricism which is better studied by traditional commentators.
45The other fascination of Tamil criticism for the Et ego in Arcadia vixi manifested in the charm and attraction the British Marxist specialist of ancient Greek, at Birmingham University, George Thomson had for the generations of students between 1930 and 1980. In his writings, the origins of the theatre (Aeschylus and Athens, 2nd ed., 1946) the origins of the Greek philosophical systems (The first Philosophers, 1955), the epic of the Greeks with all the ingredients, from tribal cosmology to kingly despotism and democratic revolution are spread out like a marvellous great narrative which could not but attract the young K. Sivathamby; not surprisingly G. Thomson generously greeted his 1979 thesis Drama in Ancient Tamil Society (published in Madras in 1981) with a foreword.
46There again we are less convinced by the Greek parallels quoted or by the proposed periodisation, heroic, feudal and mercantile, than by the author’s study of the sources concerned with a deep understanding of the links between dramatic art and the world of rites and rituals as perceived in all layers of society, not forgetting a worthy attempt to define the artists and the various modalities of their art, meter, dancing, singing and acting. The indologist may perhaps think that there is too much of the Chadwicks and not enough of the Śṛṅgāra prakāśa, but will, nonetheless, strongly recommend perpetual reference to Sivathamby’s book. Nobody is perfect.
47In striking opposition to this intrusion of the modern critics, the early Tamil authors have clearly signified through the traditional myth that something essential at the heart of their society was absolutely not transmitted, and most probably not transmittable by oral tradition. It had, in fact, to be engraved on copper plates: that is the refined but fragile structure of their poetic universe and symbolism, which alone gave to the dust of images and formulae the capacity to crystallise as poems, and to be ordered into collections, the anthologies that have come down to us.
48“In the beginning was caste” is another of Prof. Siciliano’s titles, and, at first sight, we cannot but concur: for its blossoming, the literary phenomenon of the Caṅkam, required a caste of writers to conceive of it, a small or larger royal court to judge and enjoy it and, as we have seen, a temple and a god, that could hardly exist without a priestly caste. Then, the combined attempts (not always in accord in other respects) of our two modern critics to define precisely the role and the social standing in the eyes of their public of each of these varieties of author and interpreter in a now rural, now urban society, that is, in the village square, or at the cross-roads of towns, at ports or at court etc. will surely find their recompense here. Alas, neither the conventional schema of an evolution from the original bard to the writer who reproduced artificially the tics and habits, nor the more or less Marxist schema of an evolution from the heroic to the mercantile phase through feudalism, can alter the fact that “it will always be difficult for us to distinguish clearly the bard from the rhapsode, the thulir from the skalde, the trouvère from the minstrel; and it will always be easy for us, moreover, to confuse the author with the wandering singer, the model with the copy, the necessarily oral style of speech with the improvised composition and we shall always have to admit the self-editing poet or the author-actor”. In conclusion “…let us throw overboard caste and castes which give us the milieu and the framework […] but which can make neither poets nor poems”.50
49This formula could help us to exorcise a last essentialist reduction of the Caṅkam poems to the reconstructed social milieu of their public: the anthropological deciphering of the formal world of their symbols, of which the already forgotten works of Singaravelu51 and Thaninayagam52 have given a rather naïve presentation, and the somehow militant position of which has been expressed later by G. Hart.53 Why, just at the moment when all the “primitives” turned anthropologists were accusing Margaret Mead of having substituted her fantasies for the reality of the societies she visited, should we make candidates of the ancient Tamils, who were the authentic public of the Caṅkam, to succeed the Canuks as the prey of international anthropological speculation? Hart is the possessor of an acute literary sense which first led him to make excellent translations and, thereby, to distinguish quite radically the Caṅkam from a primitive literature; he is therefore also acutely aware of the danger of the poem’s dissolution in the formulae of the oral style. In flirting with comparative literature, however, he has taken another risk, that of a dissociation of the poetic conventions, which are to be taken as a whole, for the benefit of suggestive pointillist themes isolated for the sake of the comparison,54 exact otherwise in the detail but sometimes mistaken overall. Something analogous is produced in the anthropological analysis, the privileged themes of which appear artificially isolated or selected.
50Focusing through the poems, on the beliefs and behaviour of their public, dangerously impoverishes them, introducing an additional dichotomy between elements coming from the North, the abundance of which is recognised, and those from the South, the importance of which is vindicated. This essential contribution of the South, however, makes sense of those “sacred powers” reserved to kings, to sex and to specialists in the sacred, beginning with the very idea of divinity, appearing, as Hart himself says, as no more than a poorly differentiated primitive mana in which no Caṅkam poet would recognise his own works.
51Etymological explanations and historical semantics are two different things.
52Further, the elements isolated as most characteristic of the mythology of the South often have their counterparts in the North. Thus, the famous sacred drum: one is to be found in the puruṣamedha and an apotropaic drum is played on the occasion of the mahāvrata at the winter solstice; also umbrellas, Jain or Buddhist; also the royal generosity so much sung by the poets though not more than by the vedic dānastuti; also the cattle rustling themes of Puṟam, for the gograhaṇa known to the North is attested to very early, especially, it must be acknowledged, in Andhra; also the chastity of women, a theme vital and magical to Tamil sensibility, but defined in Tamil itself from its origins in connection with the star Arundhati; also the magic or “sacred” staff of the bard, for the daṇḍa is the staff of exorcism in the Śatapatha brāhmaṇa, and plays a role, according to the gṛhya sūtra, in the ceremony of initiation (upanayana) and is at least, as everywhere, the helve of a tool, and a sign of power and chastisement.
53The whole difficulty comes, surely, from the fact that we never know exactly which society and which epoch is under discussion since it is on the basis of the texts that the chronological sequence has been mixed up randomly. In maintaining our allegiance to a literary analysis of literary testimonies, we have noticed what we choose to refer to as the “recuperation” of the Caṅkam culture: the story of Iṟaiyaṉār Akapporuḷ is that of the effort made by the Pāṇṭiyas of Madurai to give a purer sense to an already disappearing literary heritage. We may say that, whether it was a twelve year famine or a temporary invasion (historical or mythical or both at once) by the Kalabhra,55 a sort of interregnum separates two irreducibly distinguished periods, one anterior to 4th-5th c. CE and the other beginning later; the problem is precisely that there can be no rigorous periodisation56 without equally rigorous localisation, and that all kinds of overflow are possible here, in terms of space and time.
54Thus, to speak of the Caṅkam is to speak of the three Tamil kingdoms, in perhaps forced and artificial terms, while being at the same time doubtless very motivated by literary concerns to create or recreate that historical atmosphere, as in Akam poem 93, cited above, or in the composition of Cilappatikāram. But it is around the one and only Madurai that all the texts reputed to be later as well as the normative theoretical texts, are centred. Everything happens as though, after the extremely relative “universality” of the megalithic culture and the scattered locations of the earlier poems all over the three kingdoms, the Pāṇṭiya kingdom alone had taken upon itself the heritage of the Caṅkam, consistently substituting for the strict regionalism of earlier times, with its fights and marriage alliances between petty chieftains and kings, a unitary Tamil ideology.57 This does not seem too unrealistic if we consider that during the same period the power of the Pallavas of Kanchipuram was affirmed, and also tamilised (we remember the epigraphic sequence of their charters: Prakrit, Sanskrit, Tamil, from the end of the 3rd c. to the beginning of the 6th c.), and that in terms of literary culture, Bhakti poetry was progressively coming into being. A slightly wider view would allow us to include the Calukya empire in this picture too.
55The difficulty for the historian is that the social structures and the cultural realities prior to the 4th c. are known only from testimonies of two sorts: one is extra-literary (foreign, epigraphical, archaeological) and the others which are properly literary are “recuperated”, that is to say they are supposed to be the substratum of the earliest known literary texts of anthologies whose later interpolations, discreet and pertinent as we suppose them to be, are not to be underestimated. The texts which are probably the latest, far from being artificial or rigid, are among the most brilliant, the most lively and perhaps the most popular in a sense, for their content if not for their style; we think of Kalittokai, of Paripāṭal, or of the great narrative poems, with an epic inspiration previously unknown, from Cilappatikāram to Peruṅkatai. At the same time the dawning of the Bhakti movement created other, not less stimulating, polarisations, and a fascinating phenomenon of cultural integration culminating in the Vaishnavite literature of the Aḻvārs that F. Hardy has analysed with great care in a recent thesis, which is certainly the most complete study so far attempted of the literary expression of that movement.58 However, the subtlety of his minute analysis is sometimes more in the line of fine typological distinctions than of the rigorous periodisation which seems to have haunted him.
56The historians are all in fact discomfited because they are more or less tempted to distinguish an initial phase of the Caṅkam of uncertain chronology ending in the 4th c., having begun, let us say, five or six centuries earlier, the first three or four centuries comprising a period of uncertain elaboration and the two last the framework of the composition of the earliest texts, the whole constituting, so to speak, the “classical” age of the Caṅkam, and succeeded by a phase of “renaissance’ that was much better integrated into the historical period. Yet this framework seems at once fragile and artificial from the moment that it becomes a matter of integrating known works chronologically into it and of placing in it specific ideologies.
57Burton Stein is free to interpret Kalabhra’s interregnum as a break in the ecological equilibrium of the Caṅkam in its diversity in the form of an offensive by non-peasants against the growing movement of the peasant populations which triumphed with the establishment of the Pallava kingdom. The poems of the Caṅkam are mute on such an evolution59 however. Hardy’s careful analyses rightly avoid making class literature out of Caṅkam literature and emphasise that the main thing about it is a style. Once again, just as one is about to pronounce upon a society, one finds oneself grasping nothing but a literary determination. More exactly, the ideology of the Caṅkam is expressed exclusively through a poetic tradition the object of which is not to reflect the fluctuations of society accurately, but to make the vision sublime. This is what is behind the ambiguity of the term cāṉṟōr: these “eminent men” may just as well be the elite of that society whose quasi-secular humanism was made up of stoicism tinged with hedonism, as it also refers often to the authors themselves, the “authors”, holders and interpreters of its poetic tradition and its world of the imagination.
58We will most probably be reproached for favouring the literary approach and minimising the sociological evidence. Perhaps we are too much influenced in the reading of the texts by the theoretical elaboration in which they wrap up the tradition that they have preserved for us. But what do we find in fact as, coming back to Iṟaiyaṉār Akapporuḷ, we leaf through it again? The reiterated affirmation that the world of the poets is not that of everyday life. This is stated in the fourth sūtra in which the meeting of lovers of the kaḷavu type is referred to as existing amongst the poets rather than in life. It is stated in the thirty-first sūtra where it is made clear that the hero knows no fear while the real man is capable, at best, of mastering the exterior physical signs of it. It is signified in the thirty-third sūtra in which the age of the lovers– she eleven years and ten months old and he fifteen years and ten months– is appropriate to gāndharva who are always young, to the extent that their white hair as spoken of is nothing but a fiction, as emphasised by a grammatical particle. It is avowed in the thirty-fifth sūtra which shows the hero always heroic even when he is with the courtesans whom he frequents only for the seduction of the music, the dance, the voices and the charm of it.
59The power of that literature derives from the marvellous alliance between the unreality of its conventions and the realism of its images which reveals to us its public in context and enchants us as well. For almost a millennium this pendulum movement allowed the literary current to remain alive, guarding within a civilisation, which was becoming more and more Hindu in the shade of its palaces and sanctuaries from the beginning of this era, an original humanistic flavour, close to popular indigenous traditions but certainly capable of magnifying them, of mythologising them and certainly of mystifying us too.
Annexe
Appendices
A classic example of the “interior landscape”
It may be necessary first of all to recall the well known division of the earth into five symbolic regions representing five different poetic situations (“situations”, tiṇai rather than “themes”, tuṟai) each with its distinctive attributes, its epoch and hour and lastly its own emotional circumstances, an aspect of the landscape of love.
The mountain, kuṟiñci, has for inhabitants the kuṟavar, as occupation the dances of possession of the god Murukaṉ, the god of the region, agriculture and the guarding of the millet, the gathering of honey and tubers, the hunt and the bath; and for fauna the peacock, parrot, tiger and elephant; for habitat forester’s cabins or villages of huts. The auspicious season is the cold season or that of the early dews, and the hour, midnight. It is the most celebrated region of amorous union.
The “desert”, pālai, consecrated to Durga, frequented by the maṟavar who dedicate themselves to war, theft and banditry has for its fauna the donkey, jackal, predatory birds and ravening wild beasts; it corresponds to the hour of noon, to summer and late dews and expresses the rigour of separation.
The forest, mullai, consecrated to Māl (i. e. Krishna), inhabited by shepherds who are also cultivators and have bullfights for their leisure and the dance kuravai, as well as hunting the small game of the woods (wild cockerels), gazelles and hares. They live in shepherds’ hamlets which recall Vrindavan and have for favourable season the season of rains and for hour the late evening, dedicated to patient and obstinate waiting (as also in Bhakti).
The shore, neytal, where fisherfolk naturally live from the sea: fishing, salt, trafficking, resident at ports with fish for fauna and aquatic birds. All seasons are suitable here, the hour is nightfall and the sentiment that of anguished separation and anxious lamentations due to that separation.
The cultivated plains, marutam, are the region of towns where the farmers abandon their agricultural activities and celebrate in its spate the river whose fish and birds (herons, stalks, swans) with the buffalo and the fauna of ponds constitute life in every season. The favoured hour is morning, at the rising of the sun, the hour of the return of the unfaithful one, since it is the ground of sulks, amorous quarrels and (provisional) ruptures due to courtesans.
Thus presented, the “interior landscape” has something rigid and rather cold and artificial about it. We would therefore emphasize the striking contrast between, on the one hand, the utter vanity of an ethnography or cultural history based on these five regions, given that we are dealing here with a symbolism and with a repertoire of poetic themes and images, and, on the other hand, the extraordinary sense of the actual landscape with its natural features and its inhabitants. It is poetry and not sociology, of the nature of the Bucolics rather than the Georgics. For illustration let us glance at the anthology of Naṟṟiṇai. In the fishing villages of the backwaters the nets are tied to the masts to dry; the salt seller’s cry is heard and, from the bank, the crunching of carts in convoy stuck in the sand. At night the cart wheels have to avoid the crabs; odours of dried fish, cries of herons. At night everything sleeps and cold wind mixed with drizzle blows down the main street with a noise like a whale and penetrates the jointed shelters through every interstice; the dog watches and all that can be heard are the bells of the guards and their cry which marks the time. The prostitute who accuses the hero of neglecting her compares herself to an age-worn boat broken by being left stranded on the riverbank by the sea, which is no longer honoured but which the fisherman attaches to a post again, like the pair of old bulls the labourer keeps in the garden. If the heroine wants to needle the hero, here is how she presents herself: she is nothing but the poor daughter of a fisherman who casts his net into the steely blue sea whose surface it troubles; she chases birds attracted by the fatty slices of shark drying; she smells of fish. Let him stand aloof, that son of the chief who is rich, with fast chariots, who comes from the old city where flags lining the shop filled streets flap in the wind. Another poem describes the arrival of this lord of the maritime villages, a blessing on the prosperity of his beautiful beloved’s family. Elsewhere is depicted the forest and its clearings with their hamlets. Another poem is ecologically oriented and speaks without sympathy of ivory hunters who kill elephants; another compares with the heroine, who wastes away when far from the hero, the gesture of ignorant Kuṟavar who destroy, in stripping off their bark, the small leafed sandal trees which dry up, shrivel and break. Elsewhere again, there is the festival, with dance, music and painted courtesans, and the overworked washerwoman who labours all night because of the village festival.
We are therefore, in a sense, at the extreme opposite of a literature of refined symbolism, and dealing with one in the neighbourhood of realism of the most quotidian with a character close to the “bourgeois” as is inevitably often emphasized with regard to Hāla. This is no longer the chansons de geste or the troubadours’poems of courtly love, it is the Roman de Renart.
It is the adequacy between the most conventional symbol and the natural trait which produces the miracle of freshness of this poetry. To illustrate it, western literary criticism has made well known a short poem from the Kuṟuntokai, for it exists in a number of translations which Zvelebil60 has compared, inciting thereby S. A. Srinivasan of Hamburg to react with what might be called pedantic fury, vain however, since he himself has not come up with a better translation despite five pages of fruitless erudition.61 If we were to join the ranks, we would insist that no translator has fully taken into account, explicitly or otherwise, the symbolism of the interior landscape, even though the anthology as a whole illustrates it perfectly in conformity with the general theory and even though, here, the poem rigorously takes the situation into account.
Everything turns in this short poem on a comparison, a visual image, which is the hinge, in verse 3, at the centre of this five-lines text whose translation has brought up so many problems. The two terms compared are in fact the stems of a variety of millet and the feet of an aquatic bird. Wherein is the analogy? The colour and perhaps the size; this is the whole problem: green or red, short or thin? It is understood that the answer to these basic questions, between whose options translators hesitate, depends upon the learned identification of the animal and the plant. And as, in the words of Srinivasan, “…we yet know nothing about the range of shades covered by the various colour words in the Indian languages..”, we are in a difficult position. In actual fact the feet are qualified by two epithets, ciṟu “small” – translated in English as ‘thin” by most translators whom Srinivasan reproaches on this count – and pacum which he translates as “green” with ciṟu as qualifier, giving us “fresh green”. This orients us in the direction of the heron of the tanks, ardeola grayi, the size of an aigrette and with “legs and feet dullish green”. The comparison is not exact and we may lean towards the bubulcus whose feet are “yellowish green”. But has M. Srinivasan thought about the fact that it is generally considered that the millet mentioned here is “red” and is compared with the feet of the stork as in a Tēvāram poem by Cuntarar on Tiruvarur (Tēvāram, VII, 95, st. 6; PIFI edition 8204) “where are gathered together storks with red feet like stalks of millet”:
Tinai-t-tāḷ aṉṉa ceṅkāl nārai cērum tiruvārūr?
I would put forward the little heron whose feet turn from green to red at mating time, unless another meaning for pacum is discovered, for example in Puṟam 9 in which the king has “given the dancers the good red coloured gold” (cennīr-p-pacum poṉ (the same expression in the Maturaikkāñci 410). Briefly speaking, pacum, when used to designate a colour, accords with that of gold by virtue of being “greenish yellow” with red reflections, gold of superior quality as we are told.
But how, to help M. Srinivasan, are we to translate the joke of Jean Giraudoux in Elpenor (1938) describing how cleverly Ulysses repairs his raft using terms “the explanations (of which) the translators cannot refrain from giving, for the convenience of the reader, through their technical vocabulary: il argua une conasse dans le virempot, puis, la masure ayant soupié, bordina l’astifin: he was saved!” Only the author of Finnegan’s Wake might have been capable of translating this, enriching, as Giraudoux did, the nautical vocabulary in an equally imaginative way.
If we leave aside the ornithologists62 we may make a translation, considering more explicitly than have the known translators the way in which the poet plays with the symbols. Millet is always a symbol of the kuṟiñci region, that is of secret and fortunate loves. The “heron” is more modified since it belongs by right to the marutam region but is mentioned more often in that of the neytal, introducing in both cases a symbolism of absence and of separation, a sentimental malaise which is resolved according to the option chosen, by an anxious wait or by amorous sulking. The image in verse 3 takes us, through the intermediary of a comparison, from the joys of union to the episodes and the anguish of separation, whilst millet stalks are transformed into heron’s feet, in the retrospective vision of the heroine, by the cinema technique of the cross-fade which expresses the changes in the state of her soul through that strikingly cinematic image. As for the tonality of the poem as a whole, it hesitates, caught between anguish and ecstasy according to the refined process, well defined in our texts, of a mixture of genres. But a mixture offers all its savour only to one who knows how to recognise which elements belong to each of its components.
This limpid vision does not pretend to exhaust the emotional content of the poem, but we find much less relevant the arguments of Srinivasan for the heron as dumb witness, whether unique or not, or again as sexual symbol of a seducer in quest of prey, or as a tittle-tattle: all modern western explanations and side-dishes, which do little to enrich the enjoyment of the text. The tradition of the ancient commentators would appear to be decidedly compromised. The following is a possible translation:
No one but he, the rascal,
If he proves untrue, oh! what shall I do?
Like millet stalks, slender feet of gold…
In the hide for eels in fresh waters,
The heron was there too, when he embraced me.
The colophon specifies that the heroine is addressing her companion to express her doubts during the wait, which has been going on since the day when the gesture of the hero took on for her the value of a promise, one without a witness. Because of the scene evoked the poem is often associated with the kuṟiñci region by Tamil editors, but Zvelebil has chosen marutam, and we are pleased with the ambiguity. The silence of Tirumaḻicaic Caurip Perumāḷ, for example, in his edition, implies that his choice accords with ours.
To leave the last word to the legend and to Iṟaiyaṉār, or Siva himself, here is the translation of the text referred to above as given to the poet Tarumi by the god, so as to give him the chance to win a contest between the court poets. The poem is, however, rejected, by Nakkīrar, as the elements of the metaphore are not proper: perfume does not belong to the hair but to the flowers it is adorned with: “a mistake is a mistake!” It remains however a fine kuṟiñci poem (cf. bees and peacock, theme of fortunate love) which is intended to express the subject of royal reveries, in the late evening:
Gathering honey is your life,
Bee with pretty wings!
Don’t tell me what I want to hear,
Say what you have seen.
Have you ever known a flower with a finer perfume
Than the hair of my beloved,
With perfect teeth
A peacock gait
Her heart full of me?
This article «La littérature du Sangam et son public» was originally published in French, Puruśārtha 7 No VII, Inde et littératures, Paris 1983.
Notes de bas de page
1 The poet Henri Michaux proposed French readers another reckoning, not factual but more suggestive: “192 important poets, among them 57 agriculturists, 36 women, 29 Brahmins, 17 mountain dwellers, 13 foresters (?), 7 merchants, 13 Pāṇṭiya kings, etc., one potter and one fisherman” Cf. Un barbare en Asie, N.R.F., original edition 1933 (1945 ed., p. 117). For a more scholarly overview, see K. V. Zvelebil, Tamil Literature in Handbuch der Orientalistik, Leiden, 1975, or the first chapters of his The Smile of Murukan. On Tamil Literature of South India, Leiden, 1973. A. K. Ramanujam The interior Landscape, Unesco Collection of representative works, 1967, and George L. Hart III, Poets of the Tamil Anthologies, Princeton University Press, 1979, offer the best English translations and introduction.
2 An Introduction to the Study of Indian History, Bombay, 1956, p. 137. [References to D. D. Kosambi’s vision of Indian civilisation are to a landmark in Indian historical writing obviously in need of being updated. For a balanced appreciation of his role see his posthumous compilation Combined Methods in Indology and Other Writings, compiled, edited and introduced by Brajadulal Chattopadhyaya, O. U. P., 2002.]
3 Ibid., p. 178.
4 The Culture and Civilisation of Ancient India in historical Outline, London, 1965, p. 192. See also p. 90 ff. and 198 ff.
5 The Origin of Evil in Hindu Mythology, Delhi, 1976, p. 12.
6 Politics and Social conflict in South India, University of California, 1969, see ch. viii.
7 Indian Society Historical Probings, In memory of D. D. Kosambi, ed. by R. S. Sharma, New Delhi, 1974, pp. 124-155. [This 1970-1980’s approach is now re-examined, for example, by V. Rajesh. See the paper he presented to the Indian History Congress in December 2007 on “Intellectual Awakening in Late 19th century Colonial South India: A Study of P. Sundaram Pillai (1855-1897)”].
8 And in 2001 by Herman Tieken, Kavya in South India: Old Tamil Caṅkam Poetry.
9 The Decipherment of the Indus Script, Bombay, 1982. We restrict ourselves to this single reference as one of the latest to favour the hypothesis of an Indo-European language, close to Vedic Sanskrit. Other important attempts by Knorozov, A. Parpola or I. Mahadevan, were more in favour of Dravidian. Anyway, no decipherment has yet been conclusive, and the latest contribution of A. Parpola tends to understand the Indus symbols as pictograms rather than a script.
10 Review by Colin P. Masica, Defining a Linguistic Area: South Asia, in Indian Journal of Dravidian Linguistics, X, 1981, 1, pp. 180-187.
11 See G. Fussman, Pour une problématique nouvelle des religions indiennes, in Journal Asiatique CCLXV (1977), 21-70, with bibliography; D. V. Charchan, Glimpses of Rgvedic History in Waziristan in Journal of Indian History LVI (Dec. 1978) 3, 387-398; [and now G. Fussman et al. Aryas, Aryens et Iraniens en Asie Centrale, Paris, 2005, essential in terms of methodology].
12 “Aryan and Non-Aryan Elements in North Indian Agriculture”, in Aryan and Non-Aryan in India, ed. by Madhav M. Deshpande and Peter Edwin Hook, Michigan Papers on South and South East Asia, no 14, 1978.
13 The thesis of John R. Marr (1958) contains in an appendix (pp. 499-511) a “List of loan-words from Indo-Iranian in Puṟanāṉūṟu and Patiṟṟuppattu” of 189 entries, including some Tamil words borrowed from Sanskrit, which earmarks some possible borrowings from Dravidian into Sanskrit on the basis of Burrow’s hypothesis. Interestingly, this appendix was not reproduced in the printed version issued in Madras in 1985 without supervision by the author.
14 See Franklin C. Southworth, “Lexical Evidence for Early Contacts between Indo-Aryan and Dravidian” in, see above n. 12. Note that in the Introduction, A. L. Basham considers that since the Pallava period, if not earlier, in India “arya” qualifies whatever is ‘gentlemanly’ according to brahmanical order, including in its Prakrit Jain derivation ajja which more or less signifies ‘mister’, as does the Tamil aiyar, derived from arya.
15 See Kōsāmbi quoted above (with French bibliography in the translation by Ch. Malamoud and Asoka and the Décline of the Mauryas by Romila Thapar, O.U.P. 1961); regional updating of the pioneering works of B. C. Law on Buddhist geography in K. A. N. Sastri, Age of the Nandas and Mauryas (2nd ed., Delhi 1967) etc.
16 A Study of the History and Culture of the Andhras, Delhi, 1975.
17 R. Fick, The Social Organisation in North East India in Buddha's time, Calcutta, 1920. (English trls.. by S. K. Mitra), and R. N. Mehta, Pre-Buddhist India, Bombay, 1939.
18 Reissued after 14 years, New Delhi, 1995. [Essential for any debate on the theorisation of Aryan and Dravidian views in the making of Indian civilisation, see his recent Aryans and British India, 1997, and Languages and Nations, The Dravidian Proof in Colonial Madras, New Delhi, 2006, both with relevant bibliography].
19 Ira. Mativanan, Āntiranāṭṭu Akanāūṟu, translation of Gāthā saptaśati, Madras, 1978.
20 S. P. Gupta and K. S. Ramachandran (éd.), The origin of Brahmi script, Delhi, 1979, with a commentary by R. Nagaswamy, pp. 72-82.
21 G. Fussman, «Pouvoir central et régions dans l'Inde ancienne; le problème de l'empire maurya» (Central power and regions in Ancient India, the problem of the Mauryan Empire) in Annales Jul-Aug. 1982), pp. 621-647. The introduction which quotes Kim and refers to the British Râj repeats the theme developed twenty years earlier by A. L. Basham: the enduring myth of the unity of India under a strong central power, be it Ashoka, Akbar, the Râj, the Congress... Cf. “Some Reflections on Dravidians and Aryans”, in Bulletin of the Institute of Traditional Culture II, Madras, 1963, pp. 225-234.
22 Express Magazine (Madras, 6 Dec. 1981) and in Tamil Tiṉamaṇi (Madras, 12 oct. 1981).
23 Cëra Kings of the Caṅkam Period, London, 1937, p. 18 ff.; see also Burrow in Bulletin of the School of Oriental and African Studies, XII (1948), 136-137 and K. A. N. Sastri, Journal of the Asiatic Society of Bombay, Sârdha-Satâbdhî Special Volume, XXXI-XXXII, 1956-1957; ed. 1959, pp. 240-243.
24 R. Nagaswamy (ed.), Seminar on Inscriptions, Madras, 1966, p. 200 ff. It includes the Corpus of the Tamil-Brāhmī Inscriptions edited by I. Mahadevan, (pp. 60-73) [now replaced by his monumental Early Tamil Epigraphy, From the earliest times to the Sixth Century A. D., Chennai-Harvard, 2003].
25 “The megalithic monuments of South India-Literature and Tradition”, in Transactions of the Archaeological Society of South India, 1958-1959, Madras, 1959, pp. 1-14; “The Megalithic Burials and Urn-fields of South India in the light of Tamil Literature and Tradition”, in Ancient India no 2, 1946, pp. 9-16.
26 Summary in “Archaeology in South India: Accomplishments and Prospects”, in B. Stein (ed.), Essays on South India, New Delhi, 1976, pp. 1-40.
27 K. S. Ramachandran Archaeology of South India-Tamilnadu, Delhi, 1980, and S. Gurumurthy Ceramic traditions in South India (down to 300 A.D.), University of Madras, 1981. Both are excellent works, with solid information and no bias. [For a glimpse of the progress of archaeological research during the next two decades: K. Damodaran, Editor, Tamilnadu: Archaeological perspective, Dpt. of Archaeology Pub. No 143 [1999].
28 Origin and Spread of the Tamils, Adyar, 1947; Prehistoric South India, University of Madras, 1951.
29 The Problems of Dravidian Origins — A linguistic, anthropological and archaeological Approach, University of Madras, 1977, (posthumous publication). See also P. Joseph, The Dravidian problem in the South India Culture complex, Madras, Orient Longmans, 1972.
30 This is what S. Pollock seems still to believe, in The language of the Gods in the world of Men, Indian ed. 2007, p. 291, n. 12, as if the extrapolation made by Zvelebil in 1992 on the use of a “polish (sic) literary Tamil” makes it difficult, here, for Pollock to integrate Tamil chronology into his global narrative, even though Zvelebil had already followed the dating of N. Sethuraman for that inscription, that is circa. 500 CE. When this communication was first written, for Puruśārtha, we had had a brief look at the “estampages” at the State Department of Archaeology of Tamil Nadu at Madras, in 1980, and we took the first conclusions of R. Nagaswamy for granted. He published them as “An outstanding epigraphical discovery in Tamil Nadu”, in Proceedings of the Fifth International Conference-Seminar of Tamil Studies, vol. I, Madras, Janv. 1981, pp. 268-271. Serious controversies soon arose, and we must refer now, to the article by Y. Subbarayalu and M. R. Raghavavariyar in the first issue of Āvaṇam, 1991, pp. 57-69 which gives their own reading and much more sober interpretation. They confirm that we deal, in fact, with three inscriptions and that the date is ca. 500 CE, on palaeographical grounds. However, the socio-political content of the text maintains all its weight. More than that, this could possibly be “the earliest evidence for the construction of temples and monasteries in brick and mortar…[as it] relates to the administrative arrangements made for three places of worship, two of them Hindu and the other Jaina” (I. Mahadevan, Early Tamil Epigraphy… p. 136) ─ unless these three places are all identified as Jaina. Further, Mahadevan puts the Pallankovil plates on palaeographical ground in 7th c. while the accepted date (c. 550 CE) invites us to connect their content with the Pulāṅkurucci inscription.
31 A Study of the Peruṅkatai, an authentic version of the story of Utayaṇan, Madras, 1981, pp. 147-154. The date of Peruṅkatai is not clearly established. The terminus a quo could be the VII c. but the reference taken by Zvelebil as a terminus ad quem, a quotation of Peruṅkatai, I, 32, 17-18 in the commentary of Iṟaiyaṉār Akapporuḷ, s. 2 (cf. his Tamil literature Handbuch der Orientalistik, Brill, Leiden, 1975, p. 172) is doubtful, as the parallel invoked by the editors of Peruṅkatai is not literal enough to be treated as significant.
32 Cf. edition of Paripāṭal, Pondichéry, Publication de l'Institut Français d'Indologie, no 35, 1968, p. 160 ff. and 307.
33 R. Nagaswamy, “Pāṇṭiya Arikesari and Pāṇṭikkōvai”, in K. A. Nilakanta Sastri Felicitation Volume, Madras, 1971, pp. 106-111, and Studies in Ancient Tamil Law and Society, Madras, 1978, pp. 9-20. For a different opinion, see also K. G. Krishnan, “Madurai Inscription of Pāṇṭiyaṉ Chendan” in Epigraphia Indica vol. XXXVIII (1969), part I, pp. 27-32. There is a separate edition of Pāṇṭik kōvai by V. Turaicami, Madras, 1957.
34 On that aspect of the myth, not considered here by us, see David Schulman, “The Tamil Flood-myths and the Caṅkam Legend”, in Journal of Tamil Studies 14 (Dec. 1978), pp. 10-31.
35 See the introduction to the edition of Paripāṭal quoted above, n. 32, and an article by K. Zvelebil in Indo-Iranian Journal, XV (1973) 2, 109-135, “The earliest account of the Tamil Academies” (with bibliography) but still consult T. G. Aravamuthan’s previous account. See also the review of The Smile of Murukan in Indian Journal of Dravidian Linguistics, IV (1975) i, 162-174. [More recently, K. Paramasivan and David C. Buck have published an integral translation of Iraiyaṉār Akapporuḷ and its commentary under the title The Study of Stolen Love, Atlanta, 1997].
36 Ed. by J. Filliozat, Un texte de la religion kaumāra..., Published by l'Institut Français d'Indologie, no 49, 1971.
37 F. E. Hardy, Viraha-Bhakti, The early history of Krishna dévotion in South India, Delhi, 1982, p. 160 ff.
38 See Handbuch quoted above, p. 270 ff., and, for Nakkirar, pp. 273-275.
39 The eight Tamil anthologies with special reference to Puṟanāṉūru and Patiṟṟuppattu (Ph. D. thesis, University of London, 1958), pp. 156-160. According to Marr, Marutaṉ Iḷanākaṉār might be the son of Māṅkuṭi Kiḻār, author of several poems in praise of the greatest Pāṇṭiya monarch of the period, Neṭuñceḻiyaṉ, and also identified with Māṅkuṭi Marutaṉ, author of the Maturaikkāñci, a long poem celebrating the Pāṇṭiyaṉ capital city. All this is exceedingly hypothetical, but if true it would suggest, given the number of cross-references, that the authors grouped together in the anthologies constituted finally a very small world.
40 F. Gros et R Nagaswamy, Uttaramërûr, légendes, histoire, monuments, Pondichéry, Publ. de l'Institut Français d'Indologie, no 39, p. 101ff.
41 See the article by K. S. Vaidyanathan “Hero Stones” in The Quarterly Journal of the Mythic Society XXXVIII, 3, 1947, pp. 128-138 with many references translated from the poems.
42 Seminar on hero-stones, ed. By R. Nagaswamy, State Department of Archaeology, Madras, 1974. In his preface, mentioning the then recent discoveries in Chengam taluk, K. V. Soundara Rajan points indirectly at the “void” between 3rd-5th c. on one side and 6th-7th c. on the other. Such a void is perceived in the legend we deal with as well as in his difficulties in establishing palaeographical synchronism.
[The bibliography on hero-stones is continuously being enriched. K. Rajan, South Indian Memorial Stones, Thanjavur, 2000, gives a comprehensive recent overview. Tamil compilation in C. Krishnamurtti, Naṭukaṟkaḷ, Manivācakar nūlakam, Chidambaram, 2004].
43 M. Arunachalam, “The Kalabhras in the Pāṇṭiya country and their impact on the life and letters there” (The Sankara-Parvati Endowment Lectures 1976-1977), in Journal of the Madras University, LI, 1, Jan. 1979, p. 68, n. 1-with more examples.
44 This is the thesis of K. Zvelebil (see our review of The Smile of Murukan in IIJDL, Vol. IV Number 1 January 1975.) and of T. P. Meenakshisundaram in his last writings.
45 Oxford University Press, 1968.
46 Les chansons de geste et l'épopée – mythes-histoire-poèmes, Torino, 1968. We are indebted to Prof. Louis Robert for the reference.
47 Her book Oral Poetry, Its Nature, significance and social context, Cambridge University Press, 1977 contains the bibliography available at the time of redaction of the present article, and is relevant enough for our purpose.
48 Work quoted, p. 175, n. The temptation to reformulate the formular style has never disappeared from Dravidian studies, cf. F. Zimmermann, “Les variantes dans un art de tradition. Chants populaires du Sud de l'Inde” in Recherches poïétiques, t. I, Paris, 1975 pp. 115-125, à propos The Toda Songs published by Murray B. Emeneau.
49 M. Parry’s translation in “The Historical Method in Literary Criticism”, an address delivered before the Board of Overseers of Harvard College in 1934, (The Making of Homeric Verse, pp. 408-413).
50 Work quoted, n. 46, p. 260 and 266.
51 Social life of the Tamils. The Classical Period, Kuala Lumpur, 1966.
52 Nature Poetry in Tamil. The Classical Period, Singapore, 1966).
53 The Poems of Ancient Tamil. Their Milieu and their Sanskrit Counterparts, University of California, 1975.
54 See the second part of his book ch. VII, IX and X, and his fascicule The Relation between Tamil and Classical Sanskrit Literature, Wiesbaden, 1976.
55 On the Kalabhra, see opposite views by M. Arunachalam (article quoted n. 43 above) and B. G. L. Swamy The latter may be excessive, but at least he minimises the cataclysmal dimensions too often given to the episode (cf. specially his article, “Kalabhra Interregnum. A Retrospect and a Prospect”, in Bulletin of the Institute of traditional Cultures, Madras, Jan-June 1976, pp. 81-148.
56 For example, if it looks reasonable to date the Ciṟupāṇāṟṟuppaṭai later than the texts devoted to the seven donors and the three kings it quotes in retrospect, the arguments in favour of a late dating for Tirumurukāṟṟuppaṭai are much weaker: the cult of the Vēla is described as vividly there as it was in Aiṅkuṟunūṟu, supposed to be earlier, and in Kallāṭam, a text written after the Caṅkam.
57 This is what is suggested by the conclusion of J. Marr’s thesis quoted above. cf. p. 479 and his reference to Puṟam 357 which perhaps prefigures the situation in the country after the period of the Anthologies. The poem is quoted pp. 44-45 of the printed version of his thesis.
58 Work quoted above n. 37. See the third part, pp. 117-238, and the chronological chart p. 125.
59 Peasant State and Society in Medieval South India, Oxford University Press, 1980, p. 88. [The debate could easily be enlarged to include his controversies with Noboru Karashima, or with references to the Marxist inspired periodisation attempted by Rajan Gurukkal in several publications over more than a decade].
60 The Smile of Murukan pp. 83-84; [more translations are to be added, for example, J. Parthasarathi “The Concept of Faithfulness for the translation of Caṅkam Poetry” in Proceedings of the Fifth International Conference-Seminar of Tamil Studies Madurai-Jan. 1981) ed. by M. Arunachalam, vol. II, Madras, 1981, p. 15, 9-18; M. Shanmugam Pillai and David E. Ludden, Kuṟuntokai, An Anthology of classical Tamil love poetry, Madurai, 1976, p. 83; E. Wilden’s thesis and new Kuṟuntokai translation (in press)].
61 Wiener Zeitschrift (1976), pp. 193-212: “Apropos a recent publication on Tamil Literature (a review-article)”.
62 Not having read in Tamil before a useful compilation which puts the quotations in proper order: P. L. Samy, Caṅka ilakkiyattil puḻḻiṉa viḷakkam, Madras, 1976, p. 5-121, and, in English, J. Murray, The Avifauna of British India and its dependencies, London and Bombay, 1888-1890, 2 vols., as well as, for its colour plates, Salim Ali, The Book of Indian Birds, Bombay, 7th ed., 1964.
Le texte seul est utilisable sous licence Licence OpenEdition Books. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
Le vagabond et son ombre
G. Nagarajan
G. Nagarajan François Gros et Kannan M. (éd.) François Gros et Élisabeth Sethupathy (trad.)
2013
Vâdivâçal
Des taureaux et des hommes en pays tamoul
Cinnamanur Subramaniam Chellappa François Gros (éd.) François Gros (trad.)
2014
The legacy of French rule in India (1674-1954)
An investigation of a process of Creolization
Animesh Rai
2008
Deep rivers
Selected Writings on Tamil literature
François Gros Kannan M. et Jennifer Clare (dir.) Mary Premila Boseman (trad.)
2009
Les attaches de l’homme
Enracinement paysan et logiques migratoires en Inde du Sud
Jean-Luc Racine (dir.)
1994
Calcutta 1981
The city, its crisis, and the debate on urban planning and development
Jean Racine (dir.)
1990
Des Intouchables aux Dalit
Les errements d’un mouvement de libération dans l’Inde contemporaine
Djallal G. Heuzé
2006
Origins of the Urban Development of Pondicherry according to Seventeenth Century Dutch Plans
Jean Deloche
2004
Forest landscapes of the southern western Ghats, India
Biodiversity, Human Ecology and Management Strategies
B.R. Ramesh et Rajan Gurukkal (dir.)
2007