Version classiqueVersion mobile

Deep rivers

Kannan M.
Jennifer Clare

Classical Tamil

Towards reading the Tēvāram

Texte intégral

  • 1 This presentation is neither a substitute for nor a synopsis of the Tamil introduction. Both imply (...)

1The Tēvāram is a canonical collection of hymns to Siva, a kind of psalter and is essentially the perfected lyrical expression in the Saivite domain of the most powerful movement of religious fervour, or Bhakti, to emanate from the Tamil world between the 6th and 9th centuries.1 From a more technical point of view, the Tēvāram constitutes, in the body of twelve Tirumuṟai (sacred work), the first seven divided between three authors: Ñāṉacampantar (I-III), Appar (III-VI) and Cuntarar (VII) who form the Saivite "trio" (Mūvār), whence the appellation Mūvār Tēvāram frequently given to their work in its totality. They are part of a group of saints called Nāyaṉmār who numbered sixty-three and whose legendary history is set out in the Periya purāṇam by Cēkkiḻār, the twelfth and last book of the Tirumuṟai. According to another grouping the Trio becomes a Quartet (Nālvar) by the addition of the author of the VIIIth Tirumuṟai, Māṇikkavācakar, thus forming the assembly of the camayācāriyar, masters of Saivite "religion" as opposed to the four cantānakuravar, masters of the Saivite "line": Meykaṇṭar, Aruṇanti, Maṟaiñāṉacampantar, Umāpati, three of whom are the principal authors of another, later (12-14th c), philosophical-religious corpus called Meykaṇṭa śāstra, more technical in content and abundantly provided with commentaries.

2We have to assume that this meticulous arrangement was a posteriori and came in to replace the mysterious chaos of history with a formal classification, and scattered manuscripts with a definitive version, proof against variations and alterations. Leaving aside the chronology and factual history which exist only in vestigial form, there remains a coherent literary grouping within the boundaries of its premises, the cultural environment constituted by the apogee of the Tamil middle ages.

3In reading the Tēvāram we must therefore bear in mind three essential correlations. The first is the body of this Saivite literature, lyric, narrative and philosophical-theological, within which the Tēvāram fulfils the specific function of praise, more ceremonial than doctrinaire but indispensable nevertheless to the spread of faith, and whose literary and historical impact must not be underestimated.

  • 2 For example, L. Siegel, Sacred and profane dimensions of Love in Indian tradition as exemplified i (...)
  • 3 Kāraikkālammaiyār - Chants devotionnels tamouls PIFI, 1982 (new ed.) see postface p. 102, see the (...)

4The second is the parallel corpus of the Vaishnavite religion, the work of the Aḻvār, the Nālāyirappirapantam and its retinue of commentaries and later sectarian works. Vaishnavite Bhakti literature occupie the forefront of the scene: in Sanskrit from the Gitā to the Bhāgavata purāṇa and, in Tamil, mostly with Āṇṭāl and Nammāḻvār and, today, in occidental indology when the stress is on literature and philosophy rather than on history.2 In particular the exegesis of Bhakti lyricism, as well as its emotional content, reaching to the point of depth-psychology, and its linguistic expression, essentially rests on an analysis of Vaishnavite literature whilst the Saivite texts are almost always victims of an analogical reading whose Vaishnavite presuppositions, which we have referred to elsewhere as the “lyric temptation" are to be discounted.3 We recalled, as well, the most apposite example, the chapter, built entirely upon Vaishnavite quotations, entitled "Bridal Mysticism" in Dorai Rangaswamy's thesis devoted to the work of Cuntarar, The Religion and Philosophy of Tēvāram. This correlation, artificially emphasized by the parallelism of compilations, came into being naturally because of the community of sources (Veda, Upaniṣad, Purāṇa), and of literary forms; nevertheless, we should not lose sight of the specificity of Saivite lyricism in which love is service and respect rather than the mystic union of being, nor of the audience, nor of the influence of texts stimulating religious practice and the literary sensibility of the Tamil people.

  • 4 Burton Stein, Arjun Appadurai, R. Kennedy etc. And lastly, C. J. Fuller, Servants of the Goddess, (...)

5The third essential correlation is with the cultural environment and the historical scope. The times which produced the songs of Tēvāram also raised the towers of the temples in the Tamil countryside. With the growth of the Cōḻa empire and the flourishing of royal foundations from the 9th c. is confirmed that synergy of royal power with the authority of the temples, the study of which so fascinates the anthropologists of our day.4 Whether cause or result, driving force or simply symptomatic epi-phenomenon, the Tēvāram, probably since the birth of hymns but more certainly from their official introduction into the daily ritual in the temple, plays a part in the theological-political triumph of medieval Saivism which is restored to us by the literature and the epigraphy and whose workings historians are endeavouring to disentangle.

6In this respect the chronological elements are especially deficient. Even the term Tēvāram is only a much later appellation for the texts it applies to. Limited in a great number of manuscripts to the designation of the Tirumuṟai of Appar, the word tēvāram in its current usage applies to the seven Tirumuṟai from the 16th c. onwards (Tiruvaruṇaikkalampakam of Caiva Ellappanāvalar). In the Ēkāmparanātar ulā, in the middle of the 14th c. the term tēvāram signified the "worship" or “cult”, and the hymns of the Trio are still called tiruppāṭṭu.

  • 5 To quotations commonly given add S.I.I. XXIV 262 (Srirangam), of the 8th year of Vira Somesvara, w (...)
  • 6 See A.R.S.I.E. 1936-37 pp. 60-61; Ep. Ind. XXIX pp. 199-201, ARIE for 1972-73 (published in 1983), (...)

7The meaning of the word tēvāram is itself obscure. Modern explanations resort to popular etymology. For example, we break (v) "divinity" + āram "garland" in referring to the hundred hymns of Cuntarar or Campantar but to only one of Appar, where the envoi verse uses the fairly common metaphorical formula Tamil-mālai "Tamil garland" apropos the songs. Or, from vāram "song" or "dance" in honour of the divinity, we gloss "tevā pāṇi" "song or melody in the service of God". The medieval inscriptions however bring us back to the conception of worship or, more directly, of a private cult and of the body of a specific building, as also connected to a temple and to the practice of a cult,5 which leads us to the Sanskrit semantic equivalence of the deva-āgāra, deva-ārha type, which designates a cult place and may be the origin of the Tamil term. There are as well several instances of the word tēvāram used in a Jain context, whether monumental or iconographic, constructions, sculpture or development of a natural site, for example rock sculpture,6 with dehāra as equivalence (Bellary Dt.; S.I.I. IX, No 115), probably derived from the Sanskrit devagṛha.

  • 7 Ch. II., especially st. 20 ff., even if it refers to an altar rather than to a lectern or tabernac (...)

8If we retain the "private" element emphasized by certain compounds of the tēvāratēvar type, " an image intended for the private cult of the king" or tēvāra-p-peṭṭi, "chest containing what is needed for the king's pujā" (a possibly dialectal usage(?), we perceive a semantic evolution which will progressively reduce the cult place of a building to a chest and the chest to its contents, especially when we recall that the manuscripts of the Tēvāram were often kept with the divine images and worshipped along with them. We note, for example, the care with which the Civatarumōttaram of Maraiñāṉacampantar (16th c.) describes the receptacle for the manuscripts of sacred agamic texts7 But this evolution which could be said to be somewhat the reverse of what is meant, in French, by the term bureau, where the entire office is built up from a small piece of green baize (in French “bure”) covering the office table, remains an hypothesis which is unlikely to find favour with those who cleave to a " Tamil" etymology for the word tēvāram.

  • 8 The simplest formula has cymbals (tāḷam) to beat the measure and the drum in the form of an hourgl (...)

9Of little importance here is the diversity of designations of different Tirumuṟai, whether found in the manuscripts (especially the colophons) or in the works of the authors; we will note only that many refer to the prosody (tirunericai, tiruviruttam, tirukkuṟuntokai, neṭuntāṇṭakam etc.) or are variations (sometimes interchangeable) on the names of verses or of poems: pā, paṉuval, patikam, pāṭṭu, pāṭal, since a single term is more commonly used in inscriptions to signify the chanted recitation of the Tēvāram hymns in the temples: the word tiruppatiyam, doubtless relating to Tamil patikam and Skt. padya and designating a poem i. e. usually a sequence of ten verses and very often eleven, counting the envoi, where in practice only the first verse and one or two others would be sung. The largely repetitive vocabulary of the inscriptions consecrates them with formulae such as tiruppatiyam pāṭuvatāka (in order to sing the T.) or tiruppatiyam pāṭuvār nālvarkku (for the four singers of T.). The number of singers is ordinarily from one (S.I.I. XIX, 69, 8) to two (S.I.I. XIII, 74, 7 or 141, 3) or four (ibid. 14, 4; 50, 17; 51, 4) and the accompanying orchestra may vary considerably.8 In principle the endowments are permanent and go with the three daily services (mūṉru cantiyum). This classic schema would seem to have come into being in the 9th c. and to have spread first and most importantly to Tancavur and, by the 13th c., to have encompassed all of Tamil Nadu. As of the present day, the oldest inscription rendering account of these recitations seems to be the one in the temple of Tiruvallam in Chingleput district (S.I.I. III. 43). It dates from the 17th year of Nantivarmaṉ: probably Nantivarmaṉ III (846-869) rather than Nantivarman II (731-796); having been re-engraved it remains questionable.

10Alas, for N. Sethuraman (Early Cholas, Mathematics Reconstructs The Chronology, 1980, p. 32) would also seem to confirm that the inscription of Rajakecari AR 129 of 1914 and seven others collected that year at Tiruverumpur with Cempiyaṉ Vaṭivēlan as donor, are dated from Gandarāditya (950-958); the precise date would thus be 953, the third year of the reign. The same goes for a Pālūr inscription (AR 349 of 1918) from the sixth year of this Rajakecari. If it were Āditya I (871-907) who was responsible for this, that would have given us more evidence from the 9th c. but we are compelled to recognize that there is in fact very little and, at the moment, nothing from before c. 850.

  • 9 See the epigraphic documentation volume III of this edition and the article by B. G. L. Swamy, "Th (...)
  • 10 Article quoted n. 9 above and, in the last instance, in the Bulletin of the Institute of Tradition (...)

11What is at stake here is of some importance. From the 11th c. onwards, the epigraphical evidence multiplies and diversifies: donations for recitations of the Tēvāram (under the name of Tiruppatiyam), donations for statues erected in the names of the three Nāyaṉmār in a group or separately, the most diverse religious and secular foundations bearing their names, innumerable individuals also with their names, and, in supreme consecration of the work, explicit allusions to episodes or passages recalled in a literal manner and taken from the texts themselves.9 This points up the misapprehension of Tamil chronology, under which the British authors of the 19th century laboured, such as G. U. Pope, the Tamil specialist or Nelson, the editor of gazetteers who placed the Tēvāram in the 13th or even 14th century. But are we justified in asking if, when P. Sundaram Pillai, in 1895, in a revolutionary article entitled, "Some Milestones in the History of Tamil Literature or the Age of Tiruñāṉa campanta", which was widely distributed and printed, dared to place this poet in the 7th c., he was succumbing to the opposite type of excess? Eighty years had to pass before the question was again posed, in iconoclastic terms, in a series of articles and talks by B. G. L. Swamy which tended to show that the Tēvāram could not have existed prior to the 10th c.10

  • 11 Early Chola Art, Part One, Bombay, 1966 and Early Chōḻā Temples, New Delhi, 1971. For another view (...)
  • 12 Work quoted above n. 2., p. 107.

12At first sight, the wealth of references, the precise dates, and the dialectical game used to denounce irreconcilable contradictions in the literary tradition and to combat accepted ideas, is certainly all very attractive. That apparent scientific offensive will be seen to have definitively generated more light on, or at any rate more questions about, the sectarian history of medieval Saivism than on its beginnings in the land of the Tamils. By giving greater prominence to the movements of sects, that is to say to the often temporary conquests of those who patronized them, than to their ideological content and, equally, by giving prominence to the chronology of monuments (defined solely on the basis of the studies of S. R. Balasubrahmanyam11) which is to say to evidence of an institutionalisation of religious practice rather than to the geographical reality of the sites and, even more significantly, with only a surface view of Tamil literature, B. G. L. Swamy lost sight of the essential: the Tēvāram's entrenchment in Tamil and its congenital adherence to regional history. Evoking, with a rare felicity of expression, the cultural explosion under the Pallava of Kanchi and the Pāṇṭiyās of Madurai, A. K. Ramanujan writes, "Meanwhile, saints' pilgrimages and celebratory songs about hundreds of holy places mapped the country much as the king's institutions did; they literally sang places into existence" (our italics).12 Thus is defined the fundamental intuition of the Periya purāṇam and, by association, its presence in the history, since it is the one thing which gives shape to that history. If the historian wanted to crown the legend, the meeting of Appar and the most cultivated of sovereigns, Makēntravarmaṉ I, would be imperative as a transcendental and ineluctable extrapolation.

  • 13 Placed in the reign of the Cōḻa king Abhaya Kulacēkara, the recovery of the Tēvāram hymns in a sto (...)

13But how are we to remain satisfied with that kind of side-stepping in the face of the exigencies of historical methodology? Certainly, where the negative insistence of B. G. L. Swamy upon historical methodology is concerned it should be remembered that the cultural and religious history of the Tamil lands in the course of the first centuries A.D. seems to have had more weight than is generally supposed and that the tide of Bhakti did not have to wait for the 10th c. before finding doctrinal bases and cultural supports. Moreover, the Saivism of the hymns is primarily that of popular piety and of the lyricism of the authors. The object of eulogists and of the faithful is the celebration of their god to invoke his presence in the place of their act of faith, and not the deepening and affirming of the doctrine; such an objective would be that of sects anxious to defend their conquests and dependent upon the secular arm to safeguard the fervour of popular devotion. The Periya purāṇam may thus appear as the felicitous integration of the Bhakti current into the agamic cult sanctioned in the temples. It should be remembered that it dates from the 12th c. but that, since Rajaraja I, statues of the authors of Tēvāram had been erected in the great temple of Tancavur and that only a century earlier (or, according to some, a century later13) Nampi Āṇṭār Nampi had compiled the collection of hymns we know today.

  • 14 We have analysed the creative steps of Cēkkiḻār and his manner of integrating the hagiographic tra (...)

14Those who saw a contemporary of Rajaraja in Nampi were not fundamentally mistaken: it is this same process which determines the characteristics of the portraits of the Nāyaṉmār of Tancavur and also sets their hymns in a canonical collection. The Periya purāṇam, written under royal patronage and integrated into the canon, crowns this double enthroning of the cult. It is probably anachronistic to assume, as does K. Zvelebil, a plunge into the royal archives by Cēkkiḻār for that would have been rather unrewarding, but it should be recognized that this work is an exceptional memorial. The erudition of its multiple facets reaches the highest standard of Indian scholarship: an awesome recollection of texts, of puranic stories and of anecdotes, a feeling for narrative, characters and situations and all this in the service of a noble cause. Cēkkiḻār's strength lies in his use of the hymns as the warp whilst he remains faithful to them, literally and in terms of the prosody. Thus the shuttle may run between the various threads, loaded in turn with history, legend, mythology, doctrine, authentic traditions or pious tall stories. At the end the most diverse woof is inextricably interwoven with higher truth and with the authority of the hymns themselves.14 Cēkkiḻār is entirely incorporated into the Tēvāram and nobody after him has been able to read it except through him. B. G. L. Swamy may revile this puranic story as lacking in historical value but here it is the genius of Cēkkiḻār which has brought the history into being and which has conferred a new reality upon the Tēvāram hymns and given them a perceptible presence on every acre of Tamil soil. The historian can only note that they are rooted there and is hardly able to unravel the true from the conceivable.

15In their bareness the hymns are offered to us in a state of utter chronological disorder and the allusions, often veritable litotes, to facts or to actual people, cannot be understood in isolation; hieratic and eternal, their emotional content waning, the manuscripts stand in danger of becoming food for termites. But, let Nampi dust them off and Cēkkiḻār enter the picture and the universe of the Nāyaṉmār is updated; the reading given by Cēkkiḻār during the apogee of the Cōḻa empire is so contemporaneous that it gives everyone the lie for it can only be read as the entire Cōḻa period. It is from the mirage of Cēkkiḻār that the historian tries vainly to escape, since the past he represents can no longer be dissociated from our magician of the 12th c. nor from his imperative suggestions.

  • 15 K. R. Venkataraman, Devi Kāmākshi in Kānchi (A short historical Study) 2nd ed., Tiruchirapalli, 19 (...)
  • 16 Verse 9: “I am the servant of the servants of Kaḻaṟciṅkaṉ, Kāḍava Prince, the Eminent who watches (...)

16Even when it comes to a past closer to Cēkkiḻār, would we, without him, be so sure of the contemporaneousness of Cuntarar and Cēramāṉ Perumāḷ? Whereas the Cēra king of Mahodaya (Makōtai, Tev. VII, 4, 1) risks dragging Cuntarar into the 9th c. if the identification of Cēramāṉ with Rājacēkaravarmaṉ (c. 820-844), suggested by A. Sreedhara Menon and sustained by K. R. Venkataraman15 is correct: the statues of the two saints are honoured even today in the temple of Tiruvancikulam in celebration of their mutual ascension to Kailāca, a scene which, amongst others, is represented in the famous paintings of the temple at Tancavur. It is more difficult for Cuntarar to put back the time of his royal companion but he has to do it all the same if they are both to be dated in terms of another synchronization with the Pallava king Kaḻaṟciṅkaṉ, mentioned in the present tense in the Tiruttoṇṭattokai16 and most often identified as the builder of the Kanchi temple, Narasimhavarman II (700-728?), whilst C. Minakshi thought him more likely to have been Nantivarman III (846-869?) which, to be exactly compatible with the thesis of S. Menon, seems somewhat late. If the contemporaneousness of Cuntarar with the first recitations of the tiruppatiyam in the temples is likely to prove an embarrassment, we may remember that before referring to the whole Trio, the term may designate the works of Ñāṉacampantar alone. Thus the debate concerning Cuntarar's dates, which K. Zvelebil, following in the steps of Vaiyapuri Pillai, says is "fairly certain", concludes on a non liquet with a strong emphasis in favour of the 9th c.

  • 17 For example, K. R. Venkataraman see work quoted, n. 15. For an opposite view, see R. Nagaswamy, se (...)
  • 18 Abundant literature and controversy between R. Nagaswamy, partisan of the synchronization and of t (...)

17All the more reason not to confirm the synchronization of Tiruñāṉacampantar with Ciṟuttoṇṭar or, more precisely, the identification of the latter with Parañcōti, the Pallava general who was the victor at Vātāpi (Badami) in 642 and responsible for the sacking of the Calukya capital. Nothing that the hymnist tells us about Ciṟuttoṇṭar leads us to see him in any way as a warrior, except by going through the complementary evidence of Cēkkiḻār. Even so, the confrontations between Pallava and Calukya are long-lasting and repetitive enough for us to conceive of other, later, identifications17 for which there is no shortage of suggestions. This constrains us to make another attempt at double synchronization: Appar is an older contemporary of Campantar; this may be accepted on the strength of the works-the contemporaneousness that is, not the ages-without the intervention of Cēkkiḻār. But what is there to say about this edifying story of the conversion of Mahendravarman I to Saivism by Appar, except that it rests on the authority of the Periya purāṇam alone? The Sanskrit inscription of Tirucirappalli, so much discussed and often reinterpreted, is extremely interesting in that it reveals something of Saivism to us along with the cult of the liṅga and the attachment of the king to that cult: it suggests, besides, a high degree of intellectual culture and refinement on the part of the monarch and his scribes. It is unfortunately not possible to ascribe, on the basis of a single and disputed interpretative expression (vipakṣa-vṛtteḥ parāvṛttam), the responsibility for the conversion of Mahendravarman from Jainism to Saivism to the influence of Appar.18 Yet, if we do not find confirmation in these three words of a story from the Periya purāṇam nor consequently an historical synchronization between Appar and Mahendravarman I, this does not prevent us from envisaging the possible interest of the monarch in the hymns of Appar, if the latter had indeed been living in his reign.

  • 19 See note 17.

18Once more, under the hypnotic influence of Cēkkiḻār, this beautiful legend would ask to be taken for truth by history. In parallel, but without more certitude, Campantar, a contemporary of Appar, would also have converted a Pānṭiya king from Jainism to Saivism. According to details provided by Cēkkiḻār, we end by identifying that king with Arikēcari Parānkuca Māṟavarmaṉ (650-700), the hero of Pāṇṭikkōvai. This is a fascinating synchronization, admittedly as lightly founded as the preceding one yet perfectly compatible with the identification of Ciṟuttoṇṭar with the victor of Vātāpi. R. Nagaswamy, who takes these hypothetical convergences for granted, therefore places Appar between 580 and 660 and Campantar between 640 and 656. He strongly attacks, for their absence of internal coherence, the attempts made by K. R. Venkataraman to place the age of Appar and Campantar between 650 and 760.19

19But more than of the divergences of these authors, we are aware of the similitude of their systems, notwithstanding the critical spirit of K. R. Srinivasan who inspired the thesis of his brother K. R. V., but who takes as given too much material exterior to the text and obviously drawn from Cēkkiḻār's poem which is treated as a document of verified history. We must be attentive as well to the relative convergence of conclusions. If a century separates the two datations they are still not later than the middle of the 8th c. which, in the face of the radicalism of B. G. L. Swamy, continues to be the essential point to retain and perhaps the only certitude in this confused debate.

20When it comes to the point of a critic refuting in detail every single generally held proposition, there would seem to be nothing left of the tradition; however, instead of returning to the postulates of B. G. L. Swamy, we have recourse to the classical pattern, accepting in general the chronology whose detail we have rejected, and that simply because it constitutes the least unlikely periodization of Tamil literature. The 7th century is going to find itself swollen with apparently mutually exclusive riches, since it must make room for the end of the Caṅkam and the beginning of Bhakti, but it is in this way that things are linked and even the Kalabhras will find their true place when it is possible to de-emphasize their role and to cease to blame them for all the ignorance and lacunae of history. Our bias, if it is one, will appear at first to be the refusal of all absolute datation. It is in fact coloured with a certain amount of this very pragmatism which seems to have been a characteristic of the periodization of Vedic and brahminical studies at a particular stage in their history, and not such a distant one. We work somewhat like a landscape painter who, in the boundless desert tries to distribute, along a single track, the obscure landmarks that he has. This off-hand vision brings us back, advantageously, to the essential, which is an “in perspective” view of the literary and religious developments. The better to read the Tēvāram, we are going to make a brief attempt to come to grips with it through what successively precedes, accompanies and follows it.

21Before the Tēvāram, comes a Tamil literature, very conscious of itself as a language out of the common, essentially profane in content but incidentally religious, especially in its recent developments. There is a Sanskrit literature already imported, some texts of which are extremely technical and religious. There is, lastly, evidence of a religious, and notably Saivite, practice already linked to some degree with temple ritual. Surely the fact that all this existed before the 7th c. should suffice to render natural and plausible the blooming of Tamil Bhakti literature as of that date.

  • 20 F. Gros, Le Paripātal, Texte tamoul, introduction, traduction et notes, 1967; Jean Filliozat, Un t (...)
  • 21 See work quoted above, note 2.

22Tamil literature of the Caṅkam anchors its beginnings in epigraphical references in Tamil Brāhmī and in the Greco-Roman relations of the first centuries A.D., to close with the commentary of the Iṟaiyaṉār akapporuḷ, as illustrated by the verses of the Pāṇṭikkōvai, which is now put at the end of the 7th c. This literature being essentially profane, the Saivite references are scattered in the Anthologies and find their place by predilection in the invocatory poems, which may reasonably be considered as being of a slightly later date, and of which we would consider that of Kalittokai to be the most remarkable, or in the works which in the uncertain chronology of the epoch are considered as being later, notably the Paripāṭal and the Tirumurukāṟṟuppaṭai, already published in this collection.20 The exceptional evidence in the Paripāṭal on the development of Bhakti literature has been given its due weight in the study of Vaishnavite devotion by Friedhelm Hardy who adheres to the strictest possible procedures.21 The demonstration is striking where Vaishnavite literature is concerned but the literary inspiration of the Caṅkam was not disowned by the poets of the Tēvāram.

23With the Paripāṭal appear the first specifically devotional poems in Tamil literature, their subjects being Tirumāl or Vishnu (8 poems in principle of which 6 have been preserved), Durgā, under the name of Kāṭukiḻāḷ, "the Ancient of the forest" (lost poem). Cevvēḷ alias Murukaṉ or Skanda (31 poems of which 8 have been preserved) and, lastly, more profane themes, the Vaikai river and the city of Maturai. The colophons transmit to us, besides the name of the author, that of the paṇ (melody) and that of the composer of the music but the equivalences between the paṇ of the Paripāṭal and those of the Tēvāram are still the subject of interminable controversies. We have a predilection for the thesis of the musician Pa. Cuntarecan, who endeavoured to reinstate the music of the Paripāṭal, a music specifically distinct from that of the Tēvāram and which links the words very closely with the melody according to a musical model which would not have survived. Two poets are musicians as well, one for his own poetry, Kēcavaṉār who, notwithstanding his name, sings Cevvēḷ, the other in the service of his confreres. The alliance of poetry with music is, therefore, a Tamil literary tradition older than the Tēvāram; this is an important point and the text of the Cilappatikāram brilliantly affirms that alliance, noted as well by grammarians and commentators. The literary compositions of the Paripāṭal have a more elaborate prosody than do the hymns of the Tēvāram and do not divide up into identical verses. The form closest to this can be seen in the invocatory poem of the Kalittokai with its regular verses and its suggestion of a refrain. Lastly, let us say that the anthology as a whole is not sectarian: one poet, Kaṭuvaṉiḷaveyiṉaṉār eulogised Cevvēḷ as well as Tirumāl and others, such as Nallantuvaṉar, are more celebrated for their profane songs. With the Tēvāram, we enter into a more sectarian phase of Tamil literature where it is easier to agree that the eulogists of Siva only extolled him, an extreme attitude which is also the subjective argument invoked for refusing to credit Mānikkavācakar, the author of the Tiruvācakam, with the paternity of Tirukkōvaiyār, a poem of Saivite obedience considered to be very profane in tone, a sort of devout eroticism not well integrated into the prevailing pattern, even though, interestingly, several modern editions offer a commentary called an “uṇmai viḷakkam” a true clarification of its religious significance.

  • 22 Arumugam Kandiah, A Critical study of early Tamil Saiva Bhakti literature with special reference t (...)

24Moreover, even though Vaishnavite Bhakti draws on amorous themes belonging to the Caṅkam to the greatest possible extent, especially when dealing with the condition of separated lovers where the analogy of the soul in search of God is evident, the Tēvāram which is not unaware of this, shows itself to be much more restrained. It has been calculated22 that only twenty-eight poems and about thirty-two verses in all are concerned with akam, themes of interior, passionate lyricism, which is very few, especially in comparison with the Aḻvārs whose main themes these often are. Then too, the treatment of the akam by the Nāyaṉmār differs greatly: the imagery scarcely conforms to the traditional categories, there are more formulae than configurations (tuṟai) and the structural schema is broken up rather than being reconstructed around the heroes of "mullai", the forest, as it is with the Aḻvārs. Although with the latter we sense a deliberate effort to better adjust the pan-Indian phenomenon of Krishna worship to the Tamil phraseology of akam, it is for the Saivites to emphasize the differences in style and to avoid the ambiguity; more hieratic, the service of Siva does not suffer comparison with profane love.

25Nevertheless, the nostalgia for literary themes remains, explaining the flood and resurgence of images but somehow breaking up the framework by which the conventions were defined. Each author has his own particular characteristics and Appar seems to have been the only one to have reintroduced the more elaborate mode to all the traditional roles in the comedy of akam, and the only one too to have accorded a certain place to the feminization of the devotee in relationship with God. Though more confusing and allusive than Appar, Campantar is also able to recreate associations between new images and akam themes; he shows, as well, to exactly what point the poets of the Tēvāram are conscious of writing in a Tamil literary tradition antecedent to themselves. A long debate on the language of Bhakti would normally include the parallel story of Kāraikkālammaiyār and the first Vaishnavite Alvārs, and then a more compelling comparison between the Tēvāram and contemporaneous texts, a study the need for which does not seem to have occurred to B. G. L. Swamy. The roots nonetheless are in the Caṅkam, the love of beautiful language and poetic science are drawn from there and the riches of the erotic and folkloric sublimated to the service of the supreme God. The Saivite tradition seems to centre more around the service of Siva and of his devotees, and the Vaishnavite tradition to be more emotional in its analogies with profane love.

26Like the Paripāṭal, the hymn to Murukaṉ quoted by both Perāciriyar and Naccinārkiṉiyar in their commentary on the sūtra 152 of Ceyyuḷiyal, (which appears as an exergue in our edition of that anthology), and the Tirumurukāṟṟuppaṭai of Nakkīrar bring together the phraseology of the Caṅkam on Murukaṉ and a Bhakti phraseology that is beginning to emerge; before the Tēvāram, Saivite lyricism is practising its scales so to speak. Moreover, the later legend of Nakkīrar contains everything necessary for him to be a Nāyaṉār, a martyr of Murukaṉ, whose work is also used as a recipe for overcoming obstacles: such thaumaturgical merit as some hymns have is to be found in the Tēvāram and, much later, in some of the songs of Aruṇakirinātār. As distinct from the highly literary exercise the Paripāṭal is, the Tirumurukāṟṟuppaṭai like the Tēvāram and Tiruppukaḻ, has been consecrated by popular piety and definitively placed outside the domain of pure literature; its inclusion in the eleventh Tirumuṟai did no more than record this change in audience and in register. All of this draws our attention to what is the fundamental historical fact as far as the chronology as well as the understanding of the text is concerned: the close kinship between the most elevated poetry and the most reverent Bhakti prayers.

  • 23 F. Gros, work quoted, introduction pp. XX-XXIV, referred to by K. Zvelebil, Tamil Literature, p. 4 (...)

27Between the Caṅkam and the Tēvāram there is no profound cultural break nor much of a time interval either. An excellent example of the transition is the invocatory poem of the Kalittokai, a sample of this devotional literature from the Saivite domain, still a little stiff, as is the case with the first Aḻvars, already mingling traditional mythological episodes and, more specifically, literary and technical terms, pertaining here to the dance, with an akam vocabulary (for the invocation of Pārvatī) which is sometimes too precise to be translated literally. In contrast to Saivite invocatory poems from the other Anthologies, mainly attributed to Peruntevanār, author of a Tamil Mahābhārata, and which are a rather bland accumulation of themes descriptive of Siva, as in the introduction to the Akanāṉuṟu, that of the Kalittokai, in spite of differences in metre, is more closely connected with the hymnology of the Tēvāram. What is more, its author is anyway presumed to be Nallantuvanār, the most prestigious of the Kalittokai poets, the one whom the ancient commentator, Nacciṉārkkiṉiyār, most appreciated and who is said to have compiled the whole anthology. We may remember here that there is good reason to set the date by that of the horoscope of poem XI of the Paripāṭal, of which Nallantuvaṉār would also have been the author and which, in all probability, places him in 634.23 It is possible to envisage the possibility of a body of different literary expressions simultaneously in the service of the same religious conviction in the 7th c. The differences in style, vocabulary and metre more than in language point rather to differences in register than in chronology. It is probable that, in Tamil, as of the 6th c. the endeavour has been to extol Siva or to allude to his exploits in a variety of literary forms which were not only to establish themselves but were also to endure through subsequent centuries.

  • 24 Most recently, K. V. Sarma, "Spread of Vedic culture in ancient South India", The Adyar Library Bu (...)
  • 25 I. K. Sarma, The Development of early Saiva Art and Architecture (With special reference to Andhra (...)
  • 26 Work quoted p. 63.

28Everything would tend to suggest the presence, during the same period, of several Sanskrit conveyed cultures. From brief but precise allusions to Vedic worship in Caṅkam texts24 up to the technical vocabulary of logic in the Maṇimēkalai, from the first Tamil Brāhmī inscriptions up to Pallava and Pāṇṭiya copper plates, from the first echoes of the Jataka or of the purāṇas heard in the Caṅkam to the Sanskrit agamic ritual texts, the progressive absorption of Sanskrit culture and the translation of some of its elements into Tamil literary expression constitutes an on-going cultural process which is certainly parallel to the development of brahminical worship in correlation with the establishment of sanctuaries and then of more and more monumental temples. The chief Pallava monuments of known date are contemporaneous however with the growth of the Bhakti movement; the decoration of the Pāṇṭiya caves may look cruder than their splendid and older Cālukya counterparts and the first Cōḻa temples do not predate 850. Consequently, and in spite of recent discoveries by I. K. Sarma in Andhradesa (Andhra)25 relocating in the 2nd c. BCE. the establishment of the liṅga at Gudimallam, and drawing attention to the brick temples which have stretched along the Sātavāhana tīrtha since the beginning of the CE and to the traces of ritual discernible since the 3rd c. amongst the Ikṣvāku of Vijayapuri, the prestige of stone, dressed, sculpted or inscribed remains such as to make us hesitate to repose all our confidence in evidence of another order. Moreover, the same archaeologist specifies, apropos Chidambaram, that "though the temples structurally cannot be pushed back to a date earlier than that of the imperial Cōḻas, the inscriptions take us at least to the 7th c. A.D. and the literature to Patañjali's time"26 and, even though he recognises that the chronology of the Patañjalis and of the Vyāghrapādas is not certain; he envisages the first wave of Pāśupata Saivism in the South occurring in the 2nd c. B.C.

  • 27 Daṇḍin, Avantisundarīkathā, Trivandrum Sanskrit Series no 172, 1954 pp. 12-13, and Avantisundarīka (...)
  • 28 Bāṇa, Kādambari, Nirnaya Sagar ed., Bombay, 1948, pp. 459-464. On the aṣṭapuṣpikā mentioned by Bāṇ (...)
  • 29 A good demonstration of the Tamil perception of the role of the agamas and of their place in the S (...)

29Due importance as correlations with the cultural background of the Tēvāram should be given, not only to the rather vague agamic references from dedicatory inscriptions of Pallava kings nor to those almost too technical ones of the Tirumantiram by Tirumūlar, the date of which is uncertain, but also to other indirect yet much more valuable witnesses who cannot be suspected of partiality in this debate. As well, the Avanti-sundarīkathā of Daṇḍin (7th c. at Kanchipuram) apropos the quasi-miraculous mending of the broken arm of Vishnu, sleeping on the serpent at Mahamallapuram, by the Sthapati Lalitālaya, in speaking of him, makes mention of "ninety-six kinds of temple" and Daṇḍin himself refers to his visit to this sanctuary. The story is repeated in his Avantisundarīkathāsāra.27 Another sanctuary found in Tamil lands is mentioned in the Āraṇya kāṇḍa (sarga 126, sl. 27, ed. Bombay, 1914, Gujarathi printing press) of Valmiki's Ramayana, that of Tiruveṇkāṭu antakācuracamhāramūrti with reference to the māhātmya of the temple. Late as it may be, the reference is still earlier than the Nāyaṉmār and confirmed in three commentaries. The least disputable and most enjoyable reference comes from the Kādambari28 where Bāṇa takes manifest pleasure in the description of an old priest of Tamil Nadu (jaraddrāviḍa dhārmika), in charge of a Caṇḍikā forest temple, begging Durga for the boon of ruling over all of Dravida territory, wagging his head all day long, as if to avoid mosquitoes but to the rhythm of his mantra, dancing and chanting devotional songs to Bhāgīrathī in his mother tongue (svadeśabhāṣā-nibaddha-bhāgīrathī-Bhakti-stotra-nartakena) which should be seen as a homage to Siva Jaladhārin, bearing Ganga in his tresses, a favourite theme in Tēvāram hymns and known as early as the Paripāṭal; he spends his nights sleeping in different temples in spite of the misadventures this exposes him to and he is naturally outraged when he sees a mistake in the established order of the arrangement of the flowers in the aṣṭapuṣpikāpūjā, a rite which is well defined in agamic texts but which Bāṇa's literary commentators could not understand. This authentic kurukkaḷ, unwilling to relinquish his Saivite pride, (avimukta śaivābhimānena) is not only contemporaneous with Tēvāram hymns but could have been the brother of another hungry dharmin who was awarded a purse of gold before the assembly of poets in the Caṅkam at Maturai (by the grace of Siva), who inspired his text, which is in fact the second poem of the Kuṟuntokai, one of the Caṅkam anthologies. Nakkīrar questioned the perfection of that poem, obliging Siva himself to appear so that he could insist that "a mistake is a mistake" until blasted by the third eye of the irritated god. The incident is alluded to in Campantar's Tēvāram (verse 885) but the same historians who see this as the first literary reference to the Caṅkam often refuse to identify this Nakkīrar, the rebel, with the author of the Tirumuru-kāṟṟuppaṭai which was later integrated into the Saivite canon, whilst Bāṇa himself gives us the best evidence of the perfect internal coherence, whether historical or not, of the legendary tradition; and the story of Nakkīrar, who is a model of the line of the “strong against God”, conforms perfectly to the paradigm of the rebel redeemed, the most frequently repeated leitmotif of Saivite Bhakti and which is incarnate in Rāvaṇa. Much more than does the farce Mattavilāsa, Bāṇa's description attests, in the 7th c. to the context of Saivite places of worship and culture where the hymns of the Tēvāram were able to flourish. It is generally admitted that, from the following century, the agamic texts known to us were largely in use;29 the current swelled under the Cōḻa empire just as there developed around the Nāyaṉmār themselves an abundance of Tamil literature which is more or less sectarian.

  • 30 V. Ramamurthy, art. quoted above n. 16, pp. 334 and 337, connects him with the collateral lineage (...)
  • 31 The flow is not one way. Kashmiri authors like Rāmākaṇṭha or Nārāyaṇakaṇṭha seem well acquainted w (...)

30It is thus evident that Saivite faith of the 7th c. found two already existing currents to help bring it into expression: a powerful Sanskrit movement whose "vedic" and "agamic" components do not necessarily have to be seen as contradictory at the level of popular belief, and a Tamil current which is at once a translation of the Sanskrit one and an aid to its local assimilation. Born with the first Tamil literary allusions to Siva, the linking takes place naturally between the texts which we have called upon and the Tēvāram. If we compare as well the palavakaittiruttāṇṭakam of Appar (VI 93), for example, with the Kṣetrattiruveṇpā the only existing poem by Aiyaṭikaḷ Kāṭavar Kōṉ, a Pallava prince probably contemporaneous with Kāraikkālammaiyār30 we are obviously dealing with a tradition that already has its leitmotifs or clichés. As for the Sanskrit current, it provides the Tēvāram with a great deal of vocabulary and an entire phraseology widely shared with the parallel Vaishnavite Bhakti movement. Certainly the agamic literature is, properly speaking, more specialized and its Kashmiri lines, established in the South as of the Rāṣṭrakūṭa, continue, up to and beyond the 12th c, the diffusion of a Saivism which was to become more and more Tamil, due to its interpreters and for reasons contingent upon convenience more than upon doctrine.31 The Tēvāram hymns are, however, on the margin of sectarian literature and are, as already remarked, texts of eulogy rather than metaphysics.

  • 32 Nanjundan, K. R. Tirumular aur Gorakhnath (in Hindi), Ph. D. thesis, University of Meerut, 1972.
  • 33 T. N. Arunachalam, There can only be one final conclusion in Saiva Siddhantam (according to Tirumu (...)

31We shall better understand their significance through a comparison between Vaishnavite Bhakti texts and the more technical Saivite literature. While the Āḻvār texts are fully integrated with doctrinal thought through the intervention of commentators and the philosophy of Ramanuja, the Tēvāram is the recipient of more respect than it has weight as influence on the doctrine of the Saiva Siddhānta. It remains a source of literary inspiration for subsequent Bhakti authors: Aruṇakirinātar, Tāyumāṉavar, Rāmaliṅka, up to Yoga Swāmikaḷ of Jaffna, author of the Naṟcintaṉai but, as regards the sectarian technical literature, it serves more as occasional illustration and witness and is even used as a grammatical reference which is a rather striking fact. When later Tamil Saiva theologians write in their own language, and not in Sanskrit, the Tēvāram stands as their ideal of literary perfection. The religious authority seems to come from the Tiruvācakam and even more from the very equivocal authority of the Tirumantiram. The latter is actually used as a doctrinaire source though the critic finds in it a more complex meeting of heterogeneous elements, stemming notably from North Indian mysticism.32 Its prestige is such that, even though it is poorly edited and more often quoted than studied, a latter day Saivite school goes as far as to oppose it to some extent to the authors of the Meykaṇṭa Śāstra, not only in terms of the yogin or the siddha versus intellectual philosophers, nor of the mystical experience against the rationalism of the pandits, nor as illumination against inference, but as monism against pluralism.33

  • 34 It is significant that the compilation (in 1307) of an anthology of Tēvāram hymns, the Tirumuṟait (...)

32Indirectly it is the Tēvāram nevertheless which is at the heart of this debate because it is between the Bhakti movement and the Kevalādvaita of Śaṃkara; the solutions favouring a synthesis are orientated towards the Viśiṣṭādvaita of Rāmānuja (11th c.) or towards the doctrinaire effort of Saiva Siddhānta from Meykaṇṭār to Umāpati. We should not therefore allow appearances to mislead us; the diffusion of the Tēvāram by the permanent institutions of daily recitations in the temples and the elaboration of doctrinaire Saivite texts in Tamil are phenomena which are not only contemporaneous but complementary as well. Perhaps the texts of Meykaṇṭār, or those from the 12th c. which predate them, the Tiruvuntiyār, the Tirukkaḷiṟṟupāṭiyār, and especially a text not collected in the Canon but interesting and very carefully written, the Ñāṉāmirtam of Vākīca Muṉivar, do not quote the Tēvāram directly but all these doctrinaire works flourished in the faith that the Tēvāram maintained amongst the mass of devotees.34

33Its popularity has given it the reputation of being easy to read which only leads to another equivocation: the familiarity acquired through the repetition of themes and through the liturgical character of the epithets is nothing but a psychological condition analogous to that aroused by the recitation of the dhyāna śloka; this serves to make the presence of Siva perceptible in the here and now but the work as a whole never departs from its outstanding literary quality. It employs several registers and may become more and more obtuse even to the point of there being urgent need for a commentary, as for one patikam that goes to the extreme in the play of homophones; a few others are built upon the play of numbers, another has thirty verses each beginning with a letter of the alphabet in its order etc. We must not forget the literary context; the long tradition of letters established, in our opinion, since Nakkīrar and Nallantuvaṉār, is continued and followed by texts such as the Kallāṭam, which intermingles the games of Siva with amorous situations taken from the Tirukkōvaiyār, or the works of Cēramāṉ Perumāḷ, anthologised in the Xlth Tirumuṟai, which are very high flown in their tone, thus giving the first instance of a poem of the ulā type (royal procession) applied to Siva, incorporating two distichs of the Kuṟaḷ and playing upon the transposition of akam themes in meeting up again with the literature of the Caṅkam. Cēramāṉ Perumāḷ never, however, achieved the popularity of his contemporary Cuntarar, who though not prolific left nonetheless a work whose astonishing variety is matched by its wide range of different melodies (paṇ). The later authors of the Xlth Tirumuṟai had neither the mastery of Cēramāṉ nor any popular appeal; they demonstrate what happens to an ancient literary style when the genius goes out of it.

34By contact, established, repeated and instituted amongst the mass of Saivites, the Tēvāram has maintained through the centuries an outstanding educative role and its presence in the oral tradition and in memories compensates somewhat for the lack of explicit references from written texts. More than these, the Tēvāram has assured the diffusion of the doctrine, the familiarity with the great myths and with the essentials of the ritual and, through its felicitous mixture of Tamil and Sanskrit terms, of a truly literary culture which awakens the sensibility to exercises in prosody or rhetoric, giving eventual access to forms of poetry more profane and less directly communicable. In the 19th c. the literature of the Caṅkam and the great narrative texts were nothing but a memory but we know from the biography of Sri Meenakshisundaram Pillai of Tiruchirappalli by U. Ve. Swaminatha Iyer and from the memoirs of the latter, how the Tēvāram remained alive for the tradition of letters; U. V. S. made a practice of reading and commenting upon some of its verses each day and, as his unpublished manuscripts show, dreamed of an edition and was collecting documentation to that end. His disciple, V. M. Subrahmanya Iyer spent the last years of his retirement at the French Institute in Pondicherry preparing a translation and the edition which we are now publishing. Outside of this tradition, which is also that of the Saivite maṭam, literature had no other refuge than the princely courts, for example those of Ramnad and that of Ettaiyappuram, which latter was more or less of a travesty remaining celebrated because of Subramania Bharati, where were cultivated the kāvya and the light modes. The balance then is held between court and maṭam, licence and prayer, eroticism and devotion, in short the usual ingredients of decadence with an additional alexandrine zest for learning.

35But, before that last stage, Tamil culture and Saivite faith found their mutual celebration in the Tēvāram: myth and ritual towards a more or less personal devotion at a given site and to a given musical air. It is noticeable that the mythological themes are exceedingly repetitive and at first sight seem to lack variety. The Tēvāram is a haunting evocation of two legends which may be the most powerful expression of the supremacy of Siva and of the relationship between the supreme Master and his flock, misled by an excess of self-confidence: the legend of Liṅgodbhavamūrti and that of Rāvaṇa, rebel demon redeemed. The next most frequent theme is that of the destruction of Tripura, three cities symbolizing, most apparently, the three mala: āṇava, karma and māyā from which there must be deliverance. These are the representations of Siva contained in the classical purana, many of which are sculpted in the temples as in the Kailāsanātha of Kanchipuram, and are amongst the most frequent evocations. Among them are those of Appar's dasapurāṇam (Tev. IV, 14) of which only half correlate with the exploits of Siva localized in eight places (aṭṭavīraṭṭāṉam), all situated in Tamil Nadu and mostly in Tancavur, which gives to the most striking manifestations of Siva a presence directly accessible to his devotees. This eagerness to bind Siva up in the land is naturally to be found in numerous, often very brief and sibylline, allusions to purely local myths about miraculous incidents which would appear to be connected with the individual history of one or another Nāyaṉār. But if this relation, between diverse mūrti of Siva and each sanctuary thus exalted to the status of Kailāsa itself, is found in seed form in every poem it is most often the Periya purāṇam version which really animates the story and gives a personal connotation to the relationship of a myth with a landscape. The majesty of Siva dominates the Tēvāram and seems not to accommodate anecdote very comfortably. This may be why, in these decidedly Tamil hymns, Murukaṉ has so little place, the son of Siva himself appearing only as the vanquisher of Tārakaṉ or murderer of Cūr and, as well, why the popular Tiruviḷaiyāṭal, or “games” of Siva, are not mentioned more often. Kandiah listed ten of these by Campantar, nine by Appar and seven by Cuntarar which, out of sixty-four, is singularly few and the parallel with Vaishnavite complaisance towards Vishnu-Krishna's līlā is rather striking. The same parallel with the texts of the Āḻvārs, with the exception of the first amongst them, may perhaps also underline the preference of Saivites for the cult worship of Siva over the sort of sentimental ardour expressed in some Vaishnavite songs. It remains to be seen however whether this impression would stand up to a statistical examination of the respective languages for the tone of the Tēvāram is very passionate in places and the ritual vocabulary is often common to both lines. We should, as well, avoid attributing to an “agamic culture” a great part of the sacrificial vocabulary which is, properly speaking, simply a part of the Vedic tradition. Each hymnist has his own preferences anyway. Ñāṉacampantar, a Smārta brahmin, predisposed to refer to the Vedas, is not unaware of the āgamas (III, 23, 6), nor of the aspects of the worship revealed by them (for example, I, 61, 5, III, 1, 1) nor of the pūjā to the lotus in the heart, called antaryāga in the āgamas (cf. 1, 132, 6). Cuntarar, an ādisaiva of Tirunavalur, makes multiple references to flowers for Siva puja, to materials required for the abhiṣeka of Siva and, more directly, to the dikṣā (VII, 38, 1), and to systems of meditation from the āgamas (VII, 30, 7) and speaks in technical terms of Agastya's puja to the sand liṅgam (VII, 65, 5). As for Appar, the only non-brahmin, a veḷḷāḷa, probably of the Pāśupata persuasion, who declares that, even when lost in error, he never forgot the daily worship of Siva (IV, 1, 6), his work is a small agamic encyclopaedia from which, bearing in mind the text by Bāṇa quoted above, we may remember one whole poem consecrated to the aṣṭapuṣpapūjā (V, 54) where he specifies (st. 7) that this usage conforms to textual prescription. He refers incidentally to the teaching of the āgamas given by Siva to the Goddess (V, 15, 4); we may, however, draw from his work an extremely precise and technical description of temple ritual and understand thereby how authors such as Maṟaiñāṉacampantar (16th c), who wrote the Civatarumōttaram a sort of Saivite compendium in the shape of an upāgama “adapted” from Sanskrit, would have been able to draw from the poems of the Tēvāram, without quoting them, the Tamil vocabulary best suited to their versified technical treatises which they wanted written in fine language.

36The evocation of the ritual is closely connected to its religious significance, to its symbolism and to Saivite theology. Even so, it would seem to be most often mentioned for its own sake, insistently: forms and varieties of worship, details of each phase, offerings, chants, prostrations, pradakṣiṇa-s, aspersions and anointings: (care is taken to distinguish cāttutal from aṭṭutal just as snāpana from lepana in the āgamas) preliminary purifications, festivals, processions and, most especially, details pertaining to flowers, garlands, incense, lamps and so on. Why this delight and indulgence in repetition? It is a matter of a specific form of conviviality, an arousing of the emotions by an active or imaginary participation in temple ritual and this lyricism is not less "personal", "profound" and "emotional" than that of Vaishnavite Bhakti; it simply comes out of a different psychic conditioning. The Sanskrit literary tradition of the viraha or the vipralamba śṛṅgāra and the original developments of the Tamil literature of the akam predisposes the Vaishnavite devotee to have, before Krishna, the soul of a gopi. A discipline of meditation and inner purification, represented and facilitated by the physical purification and the gestural of the rite, predisposes the Saivite devotee to the unconditional service, or Bhakti, of Siva. This is amalgamated at other times with the yogic method where the yogin makes of his body the temple itself and of his pure mind a crystal liṅga. This entails intense practice of the mantras of the names of Siva, the use and symbolic understanding of the sacred ash and the frequent recitation of the hymns themselves, as is confirmed by all the final verses of these songs which are consecrated to their own right usage. While, in purely ritualistic terms, the brahmin requires no temple, the Tēvāram centres Bhakti around the service of Siva in temple ritual which it charges with a powerful and emotional attraction. The theme so common amongst the cittar and, more generally, amongst the not particularly orthodox mystics, of the vanity of ritual acts, ceremonies and edifices of the official cult compared with unconditional surrender to the absolute god, is rarely present. We cannot but be aware of the place of these hymns, especially those of Appar, (the non-brahmin!) in the absolutely emphatic position of the Saivite temple in the Tamil landscape.

  • 35 We think first of George W. Spencer's article, "The Sacred Geography of the Tamil Shaivite Hymns" (...)
  • 36 And, nonetheless, very abundant. We should specify that the Tiruppatikakkōvai has as its subject t (...)

37We have mentioned, with regard to the usage of myths and legends, the preponderant part given by the Tēvāram to the sthala, or the sites. The vision of the Nāyaṉmār in the process of accomplishing their kṣetrayātrā or tala yāttirai, as the pilgrims of Tamil lands always going from one sacred spot to another, is not only that of the Periyapurāṇam but is very evidently the natural consequence of the content of the hymns themselves which we cannot dissociate today, in Saivite practice, from the pilgrimage route. All there is to say about the geographical anchoring of these hymns has been said in drawing from them the geopolitical map of the Cōḻa empire with speculations upon immanence and locality.35 The essential part played by the tīrtha is a phenomenon throughout Hinduism but the Tēvāram, though naturally not solely “Tamil” in that regard, is so specifically steeped in the use of Tamil place names that it could not be better defined than in the terms of A. K. Ramanujan quoted above. The sense of the earth appropriated to the service of Siva is reinforced by those hymns which are lists and inventories of sthala, grouped more or less according to type (i. e. place-name endings) or to type of sanctuary (see for example the Kṣētrakkōvai-t-tiruttāṇṭakam and the Aṭaivuttiruttāṇṭakam of Appar). This taste for versified catalogue is present in Tamil literature as early as the Kuṟiñcippāṭṭu and there is, moreover, no room for doubt about the authenticity of these poems. On the other hand, the systematic classifications attributed to Umāpati, the Tiruppatikkōvai and the Tiruppatikakkōvai, appear to be texts of mnemonic convenience or later listings placed under distinguished patronage. Manuscript 436(a), in the Tancavur Sarasvati Mahal library, which credits the Tirupatikakkōvai with five supplementary verses also in the form of a numerical inventory (the number of general poems and then of the patikam of each Tirumuṟai preserved and the total number of patikam supposed to have been composed by each author), cannot but reinforce this impression of an artificial and apocryphal literature.36 This is, all the same, the descriptive order of the Tirupatikkōvai, starting with kōyil "the temple" i.e. Chidambaram, Umāpati's city which is followed by the manuscripts, less numerous, but ancient all the same, and editions classed according to sites and called, for that reason, talamuṟai. The concentration of sites scattered along the north and south banks of the Kāviri and the exaltation of Tillai or of Cīkāḻi, the most often celebrated (in 71 poems) contribute enormously to the sentimental fervour of the Tamil people and to the integration of an immanent God into the agricultural landscape of the Vēḷāḷa, a sentiment boosted by the popularity of the Periya purāṇam which has the genius always to present the saints and their works in familiar Cōḻa country, dear to its audience and made sacred moreover by the local exploits of Siva himself.

38The most frequent and natural classification of the hymns is however paṇmuṟai, the order of melodies according to which they are sung. Everything culminates in the ecstasy of the music and even of the dance, the supreme expression of religious feeling consecrated by Siva's own dance. The wives of the ṛṣi following Bhikṣāṭana in the forest may be nothing but shadowy counterparts of Krishna's gopi-s but the participation of dancers in the body of processions and the association of the dance with the life of the temple and with the representation of religious episodes has certainly contributed to the emotional current in Saivite Bhakti in the aesthetic context of the Tēvāram; the festival crowns the ritual, and Appar, in several poems, has restored the “feminine" atmosphere of akam poetics or has evoked, for example regarding Tiruvātirai or Tiruvarur, the choreographic and musical environment of the temple which has its archetype in the troupes of gaṇa musicians and in the celestial dancers. The sublime union of poem with music and dance finds expression in the dance of Siva with Pārvati; Siva, as poet, musician and dancer is the perfect transcendent model that a hymnist can conceive of in which to worship and adore the god.

39The profound harmony of the religious conception and the artistic vision does not have as its goal the satisfaction of aesthetic experience; the dynamism of the Saivite faith invariably brings us back to another concept which runs all through the art as through the ritual, that of service, and to the ideal of Bhakti defined by the behaviour of the sixty-three saints presented by the Periya purāṇam who have in common that they are the servants (aṭiyār) of Siva. First drawing up the list of sixty-two, Cuntarar proclaims himself the servant of each and so deserving of being the sixty-third, less because of the power of his songs than because they teach the service of Siva through that of his true devotees. Thus the believer has, through the Tēvāram, the experience of a double truth, the essential metaphysical truth of Saiva Siddhānta where the relationship of God to the world and with souls is one and where the relation of identity is never separate from the relation of difference (tāṉāy vēṟāy uṭaṉāṉāṉ, Tev. 1,11,2), and the historical truth of Tamil Saivism, in as much as it is a distinct community strongly conscious of its identity and its own cultural heritage.

40It is this heritage that the Tēvāram represents: at once a ceremonial and quasi-hieratic song and inexhaustible food for private piety and popular belief, a Tamil "Veda" for Saivites, a Veda which remains at a distance from sectarian philosophical texts, is proof against scholarly commentaries but is certainly a "Veda" in the sense that it is the object of a tradition and of an oral transmission as much and more than of a written one. Before approaching the latter with the problems of the edition, the other aspect of the Tēvāram's tradition should be stressed: its diffusion and the preservation of its presence in the minds of the masses by the community of singers, ōtuvār who, since the beginnings of the Cōḻa empire, have been responsible for musical recitation in temples and at festivals.

  • 37 Synopsis of Poems to Siva, The hymns of the Tamil Saints, a manuscript by Indira Viswanathan Peter (...)
  • 38 Apart from Indira V. Peterson's work, Elisabeth Barnoud (Sethupathy) in 1984 completed a first fie (...)

41The lavishness which the inscription in the great Tancavur temple suggests must not mislead us; the number and quality of performers was most probably modest in the great majority of temples but the tradition has been maintained and is still in practice in today's temples, supported by transmission from master to disciple, helped by religious foundations and enriched by the expansion of temples and maṭams and with, therefore a relatively stable economic status. This stability no longer exists and the efforts of the remaining schools are nowadays threatened: entailing some kind of temple culture and an apprenticeship of a minimum of five years of specializing in Tamil and vocal music, the traditional recital of the Tēvāram is in grave danger of disappearing along with its cultural environment. The ongoing lustre of some artists does not obscure the economic conditions and the precarious status of ordinary officiants. It is therefore high time that their knowledge was recorded, and this is being done to some extent but the question of the continuity of their tradition remains open, notwithstanding a series of ill-defined controversies. The original music of the Tēvāram has been reinterpreted through modern musical systems and thus retranslated so it has probably lost some of its distinctive characteristics. The evidence of U. Ve. Swaminatha Iyer is not very encouraging for, according to him, the state of paṇ in the Tēvāram, as taught today goes back no further than to a disciple of Muttucami Diksitar, Kilvelur Cokkalinkam Decikar (according to Nalluraikkōvai II p. 121). No ancient notation has been preserved and our knowledge of the medieval music is largely speculative in spite of some theoretical texts we do have; the identification of the ancient instruments mentioned still remains problematical even if the analogies with the modern situation allow for inferences to be drawn. Lastly, the collective memory of the ōtuvār does not have the esteem of the pandits and variant readings, if there are any, have not been collected. The practice of singing in the temples is, it is true, repetitive and limited; rare are the artists capable of memorizing the lesser known texts. This gloomy appraisal must not lead however to our forgetting the vital role of these recitations in the maintenance of the living tradition of the Tēvāram. Indira V. Peterson has cause to emphasize, in the synopsis of the book she is publishing on the text, that “in the absence of a formal tradition of commentary, the ōtuvār's spontaneous, improvisatory performances, involving creative elaborations and rearrangements of the verses, becomes an important interpretive activity, celebrated as such by Tamil Saivites”.37 We believe that it is in fact indispensable to accompany the canonical edition of the text by the study of its life in the oral tradition and in religious practice. Unfortunately the documentation on this point is difficult to collate other than by direct enquiry which, in present times, is hardly feasible.38

42The manuscript tradition is already being presented in the form of a salvage operation. According to a known model, the text we have cannot be other than the one Nampi Āṇṭār Nampi was able to wrest from the termites and which comes from the much longer original text locked away in the temple at Chidambaram. This is the legendary version of the Periya purāṇam but the much later Tirumuṟaikaṇṭa purāṇam, attributed to Umāpati, gives the original figures of sixteen thousand "patikam" composed by Campantar, forty-nine thousand by Appar and thirty-eight thousand by Cuntarar who has himself, in fact, referred to (VII, 65, 2) "Ēḻeḻunūṟṟirupaṉuval" of Appar, that is to say, four thousand nine hundred Paṉuval but might be understood as forty-nine thousand verses, for no one has been able to determine clearly whether pāṭṭu, patikam, paṉuval etc. designate verses or poems of ten verses. Nampi himself mentions the sixteen thousand "songs" of Campantar on two occasions (A. P. Tiruvulāmālai v. 63 and A. P. Tiruttokai v. 21). The figure for Cuntarar is not otherwise known. Regarding these theoretical indications, verse 26 of the Tirumuṟai kaṇṭa purāṇam gives the very realistic actual figures of three hundred and eighty-four, three hundred and seven and one hundred poems for Campantar, Appar and Cuntarar, respectively. These figures conform to a reasonable extent with those we have and the complementary cross-checking with the reckonings of sites and songs in the stanzas, whether authentic or not, of the different kōvai already mentioned, confirm that we have most probably had approximately the same text since it first took its canonical shape.

43There are surprises to be sure. We only knew, for instance, of three hundred and eighty-three poems by Campantar when, in 1917, the first version of the newly discovered Tēvāram of Tiruviṭaivāy was published at Tiruvārūr in an eight page brochure (10xl5cm, 4"x6") thanks to Ti. Mu. Cuvāmināta Upāttiyāyar. It came from an inscription engraved in characters of the 12th c. at the temple of Tiruviṭaivācal near Tiruvārūr (Madras ARE No 8 1918). Republished in Centamiḻ (t. XVII. 1918-19 pp. 169-172) by Somasundara Desikar, then reexamined and corrected by T. G. Aravamuthan in the Quarterly Journal of the Mythic Society (t. XXIV, 1933-34 pp. 266-275 and 343-360), the text, whose metre is the same as I 30 and II 36, is generally held to be authentic and is included in recent editions. We would therefore be in possession of the three hundred and eighty-four poems by Campantar, yet it continues to be astonishing that this one, engraved on stone in the 12th c. and constituting thereby by far the most ancient " manuscript", was not found in any manuscript discovered, even though the site is featured in one very early list of saints' places, the Kṣētrattiruveṇpā (st. 7, under the name of Iṭaivāy) of Aiyaṭikaḷ Kāṭavar Kōṉ of the 7th century.

  • 39 This negative conclusion correlates with the article that E. N. Tanikācala Mutaliyar has published (...)
  • 40 The seriousness and quality of his works renders the tradition barely credible if not as homage to (...)

44We shall be more cautious when it comes to another patikam published by the journal Cittāntam in 1932 (Vol. V fascicle 11). The poem is dedicated to the site at Tirukkiliyaṉnāvūr and attributed to Campantar in its envoi verse and in an eleventh verse or colophon. It would have been found in an old palm leaf manuscript. We must quote the statement by V. M. Subrahmanya Iyer about this text whose exact origin is not known, "It is my candid opinion that this decade falls very short of the standard of the verses of Ñānacampantar. The veḷḷippatikam on Maṟaikkāṭu which is an interpolation is a thousand times better than these ten verses".39 This is the judgement of an expert who, in addition, had the best critical advice from his master U. Ve. Swaminatha Iyer. As for the veḷḷippatikam, it is a patikam dedicated to Vētāraṇyam which is not to be found in any manuscript though collected in some editions (perhaps to bring the number of poems by Campantar to a round three hundred and eighty-four) without any concealment of its apocryphal nature. The expression veḷḷippatikam would designate an interpolation possibly made by Veḷḷiyampala [vāna] Tampirāṉ, a brilliant representative of the tradition of the Dharmapuram math in the 17th c. and considered to be the author as well of interpolations in the Periya purāṇam, the Kamparāmāyaṇam etc.,40 a sort of literary game originated, to the prejudice of the Caṅkam corpus, by a Jain poetess by the name of Kanti, a generic term for Jain nuns.

45We have three hundred and twelve poems by Appar; this, being five more than the declared number, opens the door to a host of indefinite questions on the possible interpolations but there is nothing decisive. Furthermore, the commentary by Namaccivāyattampirāṉ (1678) to the twelfth poem of Arunanti's Irupāhvirupatu (1254) contains a verse explicitly attributed to the Tēvāram which we assign to Appar in the belief that there is a prosodic and semantic resemblance to a poem by the latter dedicated to Tiruvitaimarutur (V, 14.10); in considering that the following poem in the same metre as the isolated verse is incomplete, since it contains only six verses, we may be tempted to see, in the text quoted by Namaccivāya, one of the lost stanzas but it must rather pertain to a general hymn and not to that of the site for no site is named in it. Here again, no manuscript has offered us any new hypothesis.

  • 41 See the history of Kalarirrarivār Nāyanār st. 85. The other poem already consecrated to Nākaikkārō (...)

46Lastly, in the case of Cuntarar we have, if we may so put it, a patikam too many since the site at Nākaikkāronam, alias Nākappaṭṭiṇam, was sung twice and the second poem, not included in any printed edition, was found in seven of the manuscripts consulted (1, 11, 14, 24, 25, 26, 38), amongst which are some of the best. It seems reasonable to accept the authenticity, and all the more so since, to the satisfaction of purists, the Periya purāṇam puts forward the idea of two different visits by Cuntarar to that place.41 On the other hand, a manuscript attributes to Cuntarar the paternity of a curious patikam of thirteen verses dedicated to Surya which is a crude, later imitation. This will be dealt with in an appendix to the last volume of this edition.

47Still, it is remarkable that the manuscript tradition is so constant. Several poems or verses are similarly incomplete in all the manuscripts consulted. A single verse by Appar (VI, 33, 9), of which we knew only two lines, was able to be completed, thanks to a manuscript of Mu. Arunachalam (25) which contained the third line and another manuscript from the Sarasvati Mahal Library (36) which has provided us with the fourth. Stanza 1568, however, which also contains only two verses, could be amended thanks to manuscript no. 11412 in the University of Kerala, but only by four fifths of the third line and by a fifth of the fourth. The tradition, put together on the evidence of M. Rajam, and according to which the greater part of the available manuscripts come from only two copy agencies, situated in bygone days at Vedaraniyam and at Tirunelveli, appears quite likely and a comparison of manuscripts has revealed no shade of appreciable linguistic difference but shows only minor variations, often ordinary copying errors. The respect of the world of letters for the text is such that no one has suggested completing the poems whose lacunae are evident. As for emendations posed conjecturally, about thirty suggestions from V. M. Subrahmanya Iyer have been confirmed by information found in the manuscripts. The detailed discussion of some variant readings is given in the Tamil introduction.

48The respect accorded to the letter of the text and the eminently meritorious character given to the copying of it and to its financing does not suffice to explain the homogeneity of the textual tradition. There are additional structural reasons for that. Compared with the Vaishnavite Nālāyirattiviyappirapantam, the compilation of the Tēvāram seems slacker. Texts classified according to site or to musical mode have no chronological order of composition and are not connected one to another, whereas about half the Vaishnavite corpus is linked according to the principle of the antāti which sets the succession, each verse beginning with the last term of the preceding one. The composition of poems of the mātal type or Tiruveḻukūṟṟirukkai may not have been rearranged very much. Lastly, the abundance of commentaries has contributed to the fixing of the arrangements of these poems and created precise textual constraints. At first sight nothing like this applies to the Tēvāram where the internal evidence for such an arrangement is practically nonexistent. The composition of the patikam is always fairly stereotypical, however, to the extent that some authors have seen rigidity in it. The hymns of Campantar are generally made up of eleven verses, the first seven given over to a description of the site and to the eulogising of Siva, the eighth to the episode in which Rāvaṇa attempts to lift Kailaca, the ninth to the legend of liṅgodbhavamūrti, the tenth to invective directed against Buddhist and Jain heresy, the eleventh to a sort of detailed envoi which made someone say of Campantar "taṉṉaip pāṭiṉāṉ", (that is: he sang of himself), making his own panegyric at the same time as announcing the fruits attendant upon the recitation of his songs. This schema, which is subject only to 9% to 12% variation, is thus very stable. We have given the exceptions in detail; these are most often due to the omission of one or several sections or to the grouping of two or three in the same verse. It would be going too far to consider every infraction of the rule as standing for a lacuna in the text. The argumentation of the Tamil introduction tends to show that most of the shorter poems have certainly been written like this and should not be considered as being incomplete; the idea of a unit of ten verses is elastic to some degree where the Aḻvār as well as the Nāyaṉmār are concerned. The songs of Appar have more variation in their composition; usually of ten verses, at times of eleven, many are however very short, forty-two having less than nine stanzas; it is the prosody that characterizes them as well as the story of Rāvaṇa overcome by Siva appearing in the last verse which is not normally in the form of an envoi. Cuntarar, who often talks about " decade" has only sixty-four poems with ten verses, thirty-one with eleven and three with twelve, the others being either much shorter or incomplete. He ordinarily signs the last verse of a poem, clearly in echo of Campantar, but what is specifically his is the repetition of the last line of the first verse and the introduction of repetitive schemata such as the invocation of the god of the site, a procedure which facilitates memorizing and contributes in a structural way to the preservation and identification of the texts.

49Two types of classification are known from the manuscripts, classification by musical mode or paṇ, which is the most common and which alone corresponds to the grouping into seven tirumuṟai, and classification by site according to an immutable geographical order within which the other traditional order finds its place; this order has been set since Umāpati, or at least since the texts attributed to him.

50The first complete edition of the Tēvāram was produced between 1864 and 1866 by a pioneer of the Tamil edition, Capāpati Mutaliyār of Kāñci, with the financial assistance of Cupparāya Ñāniyār. The text of Campantar (1864) and that of Cuntarar (1865) were provided by the maṭam at Tiruvāvaṭutuṟai and that of Appar (1866) by a manuscript from the maṭam at Dharmapuram, supposed to be copied from the complete version of the Tēvāram engraved on copper at the behest of a prince called Maṇavirkōn-Kalinkarkōn under Kulōttuṅka I (cf. Peruntokai, No 1088), the plates having been preserved at Chidambaram and then at Tiruvārūr where the one of Appar would have been copied on palm leaves for the maṭam of Dharmapuram. It is interesting that it should be the Tēvāram of Appar that should be thus honoured with such an affiliation, unique in the history of the Tamil edition. That first edition was sold out in ten years and the texts of Campantar and Cuntarar reprinted in 1875; a second edition of Appar was not published till 1879 by Ciṟṟampala Ñāniyar from the text of the maṭam of Umāpati Civācāriyār. The first complete edition, in one volume, of a version classed according to site, tala muṟai, was compiled from numerous manuscripts from the maṭams at Maturai, Tiruvavatuturai etc., by Tiruñāṉacampantam Pillai, alias Ramaswami Pillai of Maturai in 1881. It must not have been reprinted but another edition tala muṟai was published by Centilvēl Mutaliyār in 1894 (second edition in 1905) and a third by Cāminātap-p-paṇṭitar in 1911 with a long introduction.

51Different editions multiplied, introducing for example for Cuntarar in 1908, an alphabetical index of incipit and in 1913 the first unbroken commentary by Rāmānanta Yogi. For the quality of its readings and the accumulation of documentation for its edition we have chosen as basic text the one-volume complete edition of Iḷamurukaṉār (1953). This reference edition shows the condition of the text today with a noticeable return to a more critical reading of the manuscripts, most of the intermediate editions being nothing but reprints.

  • 42 The classified manuscripts are almost all on palm leaf; it is in the notebooks of the ōtuvār that (...)

52What can we hope to add, and will we find partial solutions to the aggravating ambiguity of an immutable text edited with variants? To establish the text, we have given equal credence to the manuscripts and the first printed editions; based on manuscripts inaccessible today, their authority is on a par with that of the available manuscripts, some of which are even more recent. The fifty or so manuscripts consulted, belonging to a dozen different public and private collections, represent only a percentage of the manuscripts in circulation but it is the largest sample ever put together and is sufficiently representative of the group for solid conjectures to be based upon it.42 In staying with the most complete series, we have had to ignore a mass of versions of isolated patikam. Even so, nearly half the reference sources are incomplete and not dated, or dated only according to the usual cycle of sixty years which is not always decisive. The Tamil editors hesitate to give sixty more years to one manuscript which presents the songs in the order of the sites because they too often feel that this is a modern editorial arrangement and forget that it is in this direction that the Periya purāṇam already orientates its listeners and that all the poems which enumerate and classify the places of saints have been, for centuries, attributed to Umāpati and have been congesting libraries. A good half dozen of our manuscripts follow that arrangement and some of these are of ancient aspect; the printed edition does not precede them.

53To the extent that we can trace its origin, the tradition of the Tirunelveli manuscripts is represented by nearly two thirds of the manuscripts consulted; apart from those whose place of origin is explicit, there are those dated from the Kollam era and those which owe allegiance to the maṭam at Tiruvāvaṭutuṟai whose chief priests invariably came from Tirunelveli up to the end of the last century. Those dated according to the Salivāhana, or the Tamil cycle of sixty years, can most easily be linked, we believe, with the Vedaranyam tradition but these secondary distinctions are not supported by any characteristics in the calligraphy that might encourage the defining of a group or the tracing of a strict affiliation. The usual principles of textual criticism are hardly pertinent to a text which is also very much diffused. On the other hand there is, to this day, no mean as would lead to establishing a census of varieties of the ōtuvār tradition. In short, we tend to consider this edition as the pre-critical stage, indispensable to the development of a stricter textual study, yet based on principles other than the codicology and especially on the metrical analysis and formulae, the handling of which demands defining, around the core, the thresholds of identity and of analogy which all the rhetoric of the hymns plays upon.

54Meanwhile we have looked for the most reliable text in terms of the absolutely determinant metre, of the grammatical integrity (before the sandhi is exploded, its usages are respected) and of all the knowledge that could be gathered of the styles and literary language of the authors. The Tamil introduction illustrates, with more than a hundred examples, the method and mastery of the editor. The list of more than twelve hundred emendations supplied to the reference edition indicates that progress has been made. The result is an edition which is not definitive but "authorized", authoritative as well in its readings as in its interpretations and, in short, a more reliable reference tool than recent editions give us. The last volume will provide a body of complementary information: besides the variants, alphabetical lists of the incipit etc. there will be lists of place names, identified as far as possible, of references to the epigraphy, as well as literary notes on the sites and on the puranic allusions, on the text of some reference documents and commentaries, ancient and modern, and on the more obscure patikam, highflown exercises in rhetoric of the ēkapātam, eḻukkūṟṟirukkai, yamakam and mālaimāṟṟu types, clarification of the prosody etc. and naturally a lexicon of difficult expressions and terms.

55This edition gives the text of the Tēvāram, with the sandhi split up according to the slightly revised principles of the classic series published elsewhere by S. Rajam and given at length in the Tamil introduction, since this grammar for the reduction of the sandhi is useful only to Tamil scholars. No apology is needed today for this gesture which is anti-prosodic and therefore anti-musical and which would previously have been sacrilege. M. S. Rajam's detractors nowadays accept the reprinting of his editions for a wide public and we are proud and happy to render thanks to him for having entrusted us with the responsibility of bringing to completion the project on the Tēvāram that he had conceived and for having put at our disposal the first steps of his work. All the other editions on the market today follow the metre, thus allowing for the chanting or declaiming of the Tēvāram. Our edition is intended to be an aid to reading and understanding it and the considerable typographical effort involved in this version, that is to say, besides the introduction of punctuation, the significant uses of spaces between words and of the hiatus, represents a more economical equivalent to a continuous commentary on each verse. Experience has shown us that this step towards the reader, who is nowadays as likely to be Indian as western, induces the complementary effort of a more attentive reading which is half the work of translation and diffusion.

This presentation of Tēvāram was written, first of all, to introduce to readers the new edition prepared by the FIP and to pay it its due. The review written by Indira Peterson has properly emphasized all the pros and cons of such a publication, half way between an impossible “critical” edition, and a very useful tool for all interested scholars. She also published, almost simultaneously with our edition, an excellent selection of translations and her general introduction confirmed and complemented all the aspects of our own approach. It had the additional interest of competently underlining the importance of the musical tradition, a crucial point upon which we should perhaps have insisted even more.
The publication in 1990 of Songs of the Harsh Devotee, a translation by David Shulman of Cuntarar’s Tēvāram (Dpt. of South Asia Regional Studies, Univ. of Pennsylvania), which was also appreciative of our work, did not compel us to revise it yet. At most, we might have entered into a rather gratuitous debate on the text itself, as D. Shulman did not consider the inclusion of the second poem on Nākaikkārōnam in his corpus; incidentally, both of us would have been able to claim the authority of Cēkkiḻār for our contrasting decisions.
While the text established by T. V. Gopal Iyer is in the traditional paṇmuṟai order, V. M. Subrahmania Iyer had been working on the translation which was available for consultation at the IFP on the talamuṟai 1968 edition by the Tiruppanantal maṭam, a worn out copy copiously annotated and corrected by his own hand.
Therefore the first merit of the CD-Rom elaborated between 1998 and 2007 by Jean-Luc Chevillard was that it offered an interface between the text published by TVG and the tentative English rendering of VMS, now qualified as a pretranslation, which was at last made available to all. The CD-Rom also contains a table of “incipit” of all the hymns and provides linguists and scholars with a complete concordance, the editor invites us to use as a pre-dictionary.
We can only congratulate him for all these preliminaries which supplement and will eventually replace our “pre-critical” edition.
When reading Judith Torzök’s article on “Siva le fou et ses dévots tamouls dans le Tēvāram” in South Indian Horizons (FIP, 2004) one may measure easily the usefulness of that CD-Rom, while the footnotes to the same article, show how the work of VMS required a revision which he could not do in his lifetime, and how unfair it would have been to make a book of it, thus giving a final touch to an unfortunately unachieved task.
Additional information provided by this CD-Rom is unfortunately not as “scientific” and reliable as its philologico-linguistic corpus. Maps, and to a lesser extent pictures, of temples may give us some random information. However, the original project of a complementary third volume presumed a joint approach to the sites, the archaeological approach by R. Nagaswamy and the traditional one by TVG. Such a combination turned out to be unrealistic and the proper historical enquiry into whatever is relevant on each site remains to be accomplished: archaeology, iconography, epigraphy, not to mention the most recent history.
Similarly, the audio samples do not introduce the listener to the enormous technical literature devoted to endless controversies on the singing of the Tēvāram through the ages and to the various forms of tradition which survive economic pressure. A systematic diachronic study of Tamil poetry putting together prosody, rhythm and melody is yet to be made; the first attempt by the M. A. thesis of V. N. Muthukumar at Berkeley University is definitely to be encouraged.

The French and English versions both originally appeared in Tēvāram, Hymnes śivaites du pays Tamoul, Institut Français d’indologie, Pondichéry, 1984.


1 This presentation is neither a substitute for nor a synopsis of the Tamil introduction. Both imply reference to a history of Tamil literature, for example, K. V. Zvelebil, Tamil Literature, (Handbuch der Orientalistik) Leiden, 1975 and to an exposition of caiva cittānta; an orthodox and popular Tamil compendium, the Viḻā malar, published on the occasion of the first caiva cittānta world conference, Dharmapuram, May 1984, representing the average knowledge of today's reader of the Tēvāram; the specialist will read K. Veḷḷaivāraṇaṉ, Paṉṉiru Tirumuṟai Varalāṟu, Annamalai University, 2 Vol., 1962 and 1969.

2 For example, L. Siegel, Sacred and profane dimensions of Love in Indian tradition as exemplified in the Gitāgovinda of Jayadeva, Delhi, 1978, and, essential to our subject, Friedhelm Hardy, Viraha-Bhakti, The early history of Kṛṣṇa devotion in South India, O.U.P., 1983.
Excellent introduction to the Tamil Alvārs with bibliography, in A. K. Ramanujan, Hymns for the Drowning, Poems for Viṣṇu by Nammāḻvār, Princeton University Press, 1981.

3 Kāraikkālammaiyār - Chants devotionnels tamouls PIFI, 1982 (new ed.) see postface p. 102, see the previous in this vol.

4 Burton Stein, Arjun Appadurai, R. Kennedy etc. And lastly, C. J. Fuller, Servants of the Goddess, C.U.P. 1984 and Rudolph. L. I. (ed.) Cultural Policy in India, Delhi, 1984.

5 To quotations commonly given add S.I.I. XXIV 262 (Srirangam), of the 8th year of Vira Somesvara, where the donor, is an Iṟaiyaṇṇaṉ "of the Tēvāram [of the queen] Sōmaladēviyār." It is to be remembered as well that the tevāranāyakam (for example, in A.R. 97 of 1932, dated 1057) designates the supervision of the temples and not the service of hymns as is too often understood due to anachronism.

6 See A.R.S.I.E. 1936-37 pp. 60-61; Ep. Ind. XXIX pp. 199-201, ARIE for 1972-73 (published in 1983), intro. p. 11 and inscription B 273 of Veralur (Madhurantakam tk., Chingleput Dt.) in Tamil of the 9th c.

7 Ch. II., especially st. 20 ff., even if it refers to an altar rather than to a lectern or tabernacle.

8 The simplest formula has cymbals (tāḷam) to beat the measure and the drum in the form of an hourglass (uṭukkai). Two other instruments may be added, kālam and cēkaṇṭi(kai), horn and gong, then conches (caṅku) and a wide variety of drums. The vocabulary of the hymns and that of the inscriptions more or less cross-check but several of the instruments no longer in use are difficult to identify exactly in spite of research carried out in temples. See vol. Ill and, for the modern period, the thesis of Elisabeth Barnoud-Sethupathy, Le chant du Tēvāram dans les temples du pays tamoul, Au confluent de la Bhakti sivaïte et de la musique tamoule, Univ. of Paris III, 1994.

9 See the epigraphic documentation volume III of this edition and the article by B. G. L. Swamy, "The four Saivite Samayacaryas of the Tamil Country in Epigraphy", Journal of Indian History, L, Part I, (April 1972) pp. 95-128, which instances the perusal of 350 inscriptions but, nonetheless, draws doubtful conclusions from this study.

10 Article quoted n. 9 above and, in the last instance, in the Bulletin of the Institute of Traditional Cultures - Madras 1977 July to December, pp. 131-159, the report of a seminar on "Recent Light on Medieval Saivism", where the thesis of B. G. L. Swamy, with the bibliography of his articles, is discussed at length. The most detailed presentation appeared in the same periodical, Jan-June 1975, under the title, "the Date of the Tēvāram trio: an Analysis and Re-appraisal", (pp. 119-179). The lastbook by K. K. Pillai, Studies in the History of India with special reference to Tamil Nadu, Madras, 1979, refutes B. G. L. Swamy's thesis (see pp. 464-468), as does R. Nagaswamy in Studies in Ancient Tamil Law and Society, Madras 1978, pp. 28-29, but if the thesis of B. G. L. Swamy contains errors (especially in his article of 1975) it is still regrettable that all his detractors support their refutations, as we shall see later, with the authority of a legendary text, the Periya purāṇam, accepting more or less tacitly the life spans attributed to the three saints (and to Māṇikkavācakar) by an old anonymous poem: Campantar 16 years, Appar 81, Cuntarar 18, and Māṇikkavācakar 32! Though it may be an essential part of the story the mythical statement does not have to be historically true whatever may be the place of myth in cultural history.
Swamy and his adversaries disagree about the exact date of Śaṃkara (between the 7th and 9th centuries) and on the authenticity of the Soundarya laharī; what is at stake is the reference to verse 76 where a Drāviḍa Śiśu, having drunk of the divine milk, becomes eminent in the art of Tamil poetry. We see here a reference to Campantar, i. e. to the story of his childhood according to the Periya purāṇam, but, before drawing chronological conclusions, it is important to remember an analogous legend about Śaṃkara himself and that, from the 12th century onwards, Campantar was considered as an avatar of Murukaṉ (see the Takkayākapparaṇi, VI, 169-221): such parallels restrict to an even greater extent the place of factual history in this mythical context.
Lastly, there is too great a tendency to reconstruct the past of medieval Tamil Saivism from the inscription of Malkāpuram (Guntur dt.) A. R. 94 of 1917 dated 1261 and well known to the social history of the movement. See the nāgari edition and translation by J. Ramayya Pantulu in the Journal of Andhra Historical Research Society IV (1929-30) pp. 147-162 and S.I.I. X N ° 235 for a version in Telugu script.

11 Early Chola Art, Part One, Bombay, 1966 and Early Chōḻā Temples, New Delhi, 1971. For another view, see, for example, G. Hoekveld Meiyer Kōyils in the cōḻamaṇṭalam, Typology and Development of early Cola temples, an art-historical study based on geographical principles, thesis of the University of Amsterdam, 1981.

12 Work quoted above n. 2., p. 107.

13 Placed in the reign of the Cōḻa king Abhaya Kulacēkara, the recovery of the Tēvāram hymns in a storeroom in the temple at Chidambaram is obviously legendary and a definite identification of the sovereign is not possible. The theses veer between Āditya I and Kulōttuṅka I. Tamil historians seem to prefer a dating era anterior to that of Rajaraja I, in fact the 10th c, but the debate remains open and the internal evidence sustained, for example by T. P. Meenakshisundaram, cannot be absolutely conclusive, (cf. K. Zvelebil, Tamil Literature, p. 134 n. 21).

14 We have analysed the creative steps of Cēkkiḻār and his manner of integrating the hagiographic tradition into existing literary work apropos the life of Kāraikkālammaiyār, published with a translation by Julien Vinson, in our new edition of Chants devotionnels tamouls of Kāraikkālammaiyār which appeared in this collection in 1982; see postface pp. 99-102, see the previous article in this vol.

15 K. R. Venkataraman, Devi Kāmākshi in Kānchi (A short historical Study) 2nd ed., Tiruchirapalli, 1973, Addenda pp. 64-75. The thesis renewed by A. Sreedhara Menon in his Survey of Kerala History, Trivandrum, 1967, or in Social and Cultural History of Kerala, Delhi, 1979, (p. 181 and 187) had already been repudiated in a fashion as definitive as it was summary, by S. Vaiyapuri Pillai, History of Tamil Language and Literature (beginning to 1000 A.D.), Madras, 1956, p. 112 n. 4. The "enterprising scholar" who is not named may well be the revered Ellamkulam Kunjam Pillai.

16 Verse 9: “I am the servant of the servants of Kaḻaṟciṅkaṉ, Kāḍava Prince, the Eminent who watches over the whole world surrounded by the sea". Let us note that the distinction established by V. Ramamurthy between Pallavas and Kāḍavas as two different lines, main and collateral, suggests one more hypothesis; still another should also be reexamined according to which Perumāṉ "the Eminent" may perhaps refer to a royal title, why not the one wrested away by the Ganga King Sri Puruṣa from a Kāḍava prince who might also have been Nandīśvaravarmaṉ (716-778). See V. Ramamurthy “The Pallavas and the Kadavas”, in Śrīnidhih, Perspectives in Indian Archaeology, Art and Culture. Shri K. R. Srinivasan Festschrift, Madras, 1983, pp. 333-338 and especially p. 336.

17 For example, K. R. Venkataraman see work quoted, n. 15. For an opposite view, see R. Nagaswamy, see work quoted, n. 10, pp. 13-15, a systematic refutation of the internal contradictions of the thesis. On the Calukyas see, lastly, K. V. Ramesh, Cālukyās of Vātāpi, Delhi, 1984, purposely silent on the "synchronization" of Parañcōti, due to the absence of historical documents (personal information).

18 Abundant literature and controversy between R. Nagaswamy, partisan of the synchronization and of the conversion of Mahendra (see work quoted, pp. 20-29) and Michael Lockwood "The philosophy of Mahendravarman's Tiruchirappalli Epigraph" in Studies in Indian Epigraphy Vol. 3, Mysore, 1976. Excellent critical study, unfortunately unpublished, by K. R. Nanjundan Rock-fort Inscriptions at Tiruchirapalli, 23 pages. We thank the author for having let us see his manuscript.

19 See note 17.

20 F. Gros, Le Paripātal, Texte tamoul, introduction, traduction et notes, 1967; Jean Filliozat, Un texte de la religion kaumāra, the Tirumurukāṟṟupaṭai, 1973; PIFI Nos. 35 and 49.

21 See work quoted above, note 2.

22 Arumugam Kandiah, A Critical study of early Tamil Saiva Bhakti literature with special reference to Tēvāram, Ph.D., Univ. of London, 1973, 2 vols., unpublished. Two chapters are particularly interesting, one given over to the treatment of akam and the other to the tirukkaṭaikkāppu, envoi verses of the hymns. The author, based in Colombo, has instituted a restricted separated distribution of each of these chapters, (ex. roneoed).

23 F. Gros, work quoted, introduction pp. XX-XXIV, referred to by K. Zvelebil, Tamil Literature, p. 43 and n. 27. Our hesitation to formulate a definitive conclusion, which K. Zvelebil would seem to be reproaching us with, is due to a certain amount of interpretation which remains conjectural in the reading and cannot be solved by astronomical calculations. We do not exclude the possibility that the compilations of all the Anthologies may be later (see our article on “La littérature du Caṅkam et son public” in Purusartha No VII, Inde et littératures, Paris 1983, see the article “Caṅkam literature and its public” in this vol.); the problem concerns rather the most ancient strata and the necessary spreading of the Anthologies over a period of several centuries of which the periodization remains very arbitrary.

24 Most recently, K. V. Sarma, "Spread of Vedic culture in ancient South India", The Adyar Library Bulletin, Vol. 47, Madras, 1983, pp. 1-14 which tends to place what he calls "the North-South acculturation in India" well before the first centuries A.D. This view is supported by the local administrative bilingual vocabulary attested to in inscriptions recently discovered at Pulankurichi (Dt. of Ramanathapuram), of the 5th century (cf. N. Sethuraman article in Śrīnidhih, pp. 285-291).

25 I. K. Sarma, The Development of early Saiva Art and Architecture (With special reference to Andhradesa), Delhi, 1982; the material studied, stretching from the 2nd c. B.C. to the 6th c. A.D., is of capital importance. For examples of brick structures preserved in Tamil Nadu, see in the Bulletin of the Institute of Traditional Cultures, Madras, 1980, Jan.-June, pp. 130-133 and 144-145, and more generally the whole exposition by M. Arunachalam on " Indian Architecture, Tamil Sources" pp. 121-170.

26 Work quoted p. 63.

27 Daṇḍin, Avantisundarīkathā, Trivandrum Sanskrit Series no 172, 1954 pp. 12-13, and Avantisundarīkathāsāra, pp. 4-5 ed. M. Ramakrishna Kavi, Madras.

28 Bāṇa, Kādambari, Nirnaya Sagar ed., Bombay, 1948, pp. 459-464. On the aṣṭapuṣpikā mentioned by Bāṇa, see the Śaivāgamaparibhāṣāmañjarī of Vedajñāna, VIII 67-7la, in Bruno Dagens', Le florilège de la doctrine sivaite, P.I.F.I. No 60, 1979, p. 312 and references given ad. loc., of which Somaśambhuppaddhati in P.I.F.I. No 25, Vol. I, pp. 292-293.

29 A good demonstration of the Tamil perception of the role of the agamas and of their place in the Saivite culture of southern India in M. Arunachalam, The Śaivāgamas, Gandhi Vidyalayam, 1983. Also see Jan Gonda, Medieval religious literature in Sanskrit, Wiesbaden, 1977.

30 V. Ramamurthy, art. quoted above n. 16, pp. 334 and 337, connects him with the collateral lineage of Bhimavarman (580-600) who may have been his father and who had become a devotee of Siva in his old age, according to a Kannada tradition.

31 The flow is not one way. Kashmiri authors like Rāmākaṇṭha or Nārāyaṇakaṇṭha seem well acquainted with the texts of the South and one of the former's texts on “the right of entry into Siva temples”, speaks of the singing of Dravidian poems (drāviḍaiḥ stotrapaṭalaiḥ, line 35) which probably stand for the hymns of the Tēvāram, in the mahāmaṇṭapa by the sat-śūdras, i.e. those who have received dīkṣā, see P. S. Filliozat “Le droit d'entrée dans les temples de Siva au XIe siècle” Journal Asiatique CCLXIII, 1-2, Paris, 1975, p. 103 ss. Well attested to in the later Cōḻa period, the North-South cultural exchanges and the migrations of specialists in ritual or doctrine had started much earlier.

32 Nanjundan, K. R. Tirumular aur Gorakhnath (in Hindi), Ph. D. thesis, University of Meerut, 1972.

33 T. N. Arunachalam, There can only be one final conclusion in Saiva Siddhantam (according to Tirumular), Dharmapura Adhinam's Reply to the Hawaii Saivites, Dharmapuram, 1984, and There has always been only a pluralistic Saiva Siddhanta philosophy, a reply to Hawaii Saivism's brochure on Monism and Pluralism, by B.S.V. K. Palasundaram, Kuala Lumpur, 1984. The present orthodox tendency is to emphasize that the "Realistic pluralism" of Saiva Siddhānta was the dominant conception in Kashmir as well as in the South before Samkara and that it is his monistic propaganda which has spoilt everything.

34 It is significant that the compilation (in 1307) of an anthology of Tēvāram hymns, the Tirumuṟait tiraṭṭu (99 hymns, 26 by Campantar, 63 by Appar, 10 by Cuntarar) is attributed to Umāpati himself; the preponderant place of Appar, a theme of our presentation, is justified here by the doctrinal and ritual elements which prevail in his poems. The classification of the anthology in 10 chapters suggests the parallel with the Tiruvaruṭpayaṉ, another work by Umāpati. It is sometimes said too that the Civañāṉapōtam of Meykaṇṭār is the elucidation of Campantar's hymn 312, called Tiruppācuram, but the problem is complicated by a confrontation with the version according to which it is nothing but a translation of the Rauravāgama.

35 We think first of George W. Spencer's article, "The Sacred Geography of the Tamil Shaivite Hymns" which appeared in Numen 17, 1970, pp. 232-244, then of the recent development of a significant literature on Indian pilgrimages, from the classic thesis by Surinder Mohan Bhardwaj, Hindu places of pilgrimage in India (A study in Cultural Geography) Berkeley, Univ. of California Press 1973, to E. A. Morinis Pilgrimage in the Hindu Tradition, Delhi, 1984, and a current programme at the French Institute of Pondicherry on the site of Tiruvaṇṇāmalai. See, lastly, two articles by Indira V. Peterson on the Tēvāram, first, "Lives of the Wandering Singers: pilgrimage and poetry in Tamil Saivite hagiography" in History of Religions, Vol. 22 no 4, Chicago, 1983, which allots a great deal of space to the Periya purāṇam, and especially "Singing of a place: pilgrimage as metaphor and motif in the Tēvāram Songs of the Tamil Saivite Saints" in J.A.O.S. 102, 1, 1982, pp. 69-90, containing excellent and well translated extracts. Immanence and localization refer us to David Shulman, Tamil Temple Myths, Princeton, 1980, and to questions raised in this regard by Rajam Ramamurthi "On the Themes of Divine Immanence and Localization" in Tamil Civilization, Quarterly Research Journal of the Tamil University, Thanjavur, Sept. 1983, pp. 73-76, with an interesting attempt to redefine the idea of aṇaṅku and a most pertinent conclusion: "a site was so independent of the divinity-aṇaṅku, Teyvam, kaṭavuḷ or muruku – that it could precede or survive their presence". The same thought occurs to us in relation to the role of the sites whose true force does not change and rests on the positions of major tīrtha, irregardless of the ephemeral movements of the sects who temporarily conquer them, as Bruno Dagens says, illustrating this with examples, in the historical introduction to his thesis Entre Alampur et Srisailam, Recherches archéologique en Andhra Pradesh, volume I, Pondicherry, P.I.F.I. No 67, 1, 1984. The authors of the Tēvāram happily chose to sing the tīrtha and the god who transcends them and not the sectarian variations which came after them and added nothing to their lustre. This is also what Cēkkiḻār intuited.

36 And, nonetheless, very abundant. We should specify that the Tiruppatikakkōvai has as its subject the number of patikam given over to each tala (civastalam) but that the manuscript 436 (a) which gives a 19 verse version is mistakenly called Tiruppatikkōvai by the Descriptive Catalogue of Tanjore (Vol. II, ed. 1925; see also no 437 (a) which would be the same text), for the term is used for its generic value as in the Catalogue of the Swaminatha Iyer library which, in no 1408, describes an identical manuscript but cites it in the wrong order: verses 1 and 29 (sic, for 19) are verses 17 and 14 in the Tanjore ms.. The same catalogue, under the no 1414 designates as the kṣēttirakkōvai the text which is really the Tiruppatikkōvai, that is to say, the enumeration of the sites (pati) and which contains 15 verses in this ms. even though the printed versions are of only 14 verses, the same as for the Tiruppatikakkōvai, which adds to the confusion. The situation is further complicated by the addition, still under the title of kṣettirakkōvai, of the Civañāṉappaḵṟoṭai veṇpā, also attributed to Umāpati and detailing, along with the names of the sites, that of the divinity who resides there. Lastly, the Vaipputtalappāṭṭu, no 1407, lists in 7 verses the Vaipput-talaṅkaḷ or sites occasionally named in the Tēvāram but without a hymn dedicated to them.

37 Synopsis of Poems to Siva, The hymns of the Tamil Saints, a manuscript by Indira Viswanathan Peterson, which will shortly be published in the United States [Princeton University Press, 1989]. It is in the form of an anthology of the Tēvāram preceded by a comprehensive introduction on the poems and their role in Tamil Saivism. The author who is a musician as well as an indologist, quite rightly insists upon the sung tradition which she studied in situ in 1978.

38 Apart from Indira V. Peterson's work, Elisabeth Barnoud (Sethupathy) in 1984 completed a first field work report under our direction on Le Chant du Tevaram dans les temples du pays Tamoul as the result of an extensive investigation of the territory conducted amongst about a hundred ōtuvār. We referred above to her more comprehensive study in 1994. This work describes the music as well as the practice of traditional concerts and such teaching as it survives, especially in the Tēvārap-pāṭacālai of Dharmapuram, last remnant of an educational system on its way to becoming extinct. She endeavours to translate the Tamil musical experience and its modes of transmission into western terms. The effort of the “Sampradaya” Association at Madras to institute musical sound archives and to attempt to preserve the most authentic tradition is also to be cited in scientific documentation eager to present the Tēvāram in its total cultural context and its identity, which appears as quite distinct from the rest of the music known as "Karnatic".

39 This negative conclusion correlates with the article that E. N. Tanikācala Mutaliyar has published in the next volume (V, 12) in Tamil, in the same journal Cittāntam, under the title "Research on a new patikam by Tiruñāṉacampantar", the conclusions of which are against the authenticity but according to ill-founded metrical arguments; the patikam may thus retain the benefit of the doubt.

40 The seriousness and quality of his works renders the tradition barely credible if not as homage to his talent, this is why some authors consider that Veḷḷippāṭal or °-pāṭṭu, °-patikam only designate an apocryphal text in the absence of more precise indications.
We should add that the metre employed for the Veḷḷippatikam is known to the Tēvāram of Appar and not to that of Campantar which statement remains to be interpreted. Furthermore, another hymn by Campantar is already dedicated to the closing of the doors of the temple at Vedaranyam and its authenticity is confirmed by Cēkkiḻār; the second poem serves a double purpose, one which the Canon seems to exclude, since it is the reason given as well for the rejection as "veḷḷippāṭṭu" of the second poem of Cuntarar consecrated to the Nākaikkārōṇam site. However see note 41 below.

41 See the history of Kalarirrarivār Nāyanār st. 85. The other poem already consecrated to Nākaikkārōṇam is patikam 46.

42 The classified manuscripts are almost all on palm leaf; it is in the notebooks of the ōtuvār that we may find a manuscript tradition on paper. However, amongst the notebooks recently collected by the French Institute of Indology and which belonged to Centinātaiyar, the author of an excellent Tēvāram vētacāram, published in Madras in 1917, is featured an incomplete copy on watermark paper dated 1835, which retains some variants adopted by the most ancient editions but not found in the manuscripts consulted.

© Institut Français de Pondichéry, 2009

Licence OpenEdition Books

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search