Indian Journal of History of Science, 51.1 (2016) 125-130 DOI: 10.16943/ijhs/2016/v51i1/48384
A Note on Grammatical Knowledge in Early Tamiakam
Y Subbarayalu*
(Received 15 October 2015; revised 24 October 2015)
Abstract The article seeks to examine the historical context of the formulation of Tolkāppiyam, the earliest available grammatical text of the Tamil language and the nature of its relation to the early Tamil literature, popularly called the Sangam literature. It discusses the structure, composition and linguistic features of the text, which provide insights into the methodological aspects of textualisation. Based on a fairly big database drawn from the traditional linguistic usages within the geographical limits of the Tamil speaking area, both from the literary texts and from colloquial practices, the author, Tolkappiyar, has produced this monumental work, treating in the first two sections the phonology, morphology, syntax and semantics of the Tamil language and in the third section with prosody and literary composition. A versatile scholar both in Tamil and Prākrit/Sanskrit, the author was familiar with the northern grammatical works of Pāini and others and was influenced to some extent by them. But the treatment is quite original in keeping with the genius of the Tamil language. The third section, by its elaborate treatment of the tinai classification and its rich semiotic codes, provides the very key to understand the Sangam poetry. And in this respect it has differed remarkably from the Sanskrit grammatical tradition, which relegated the matters relating to literary composition to separate texts called alankāra-śāstra. The treatment of the subject matter in Tolkāppiyam is partly analytical and partly paradigmatic whereas that of Pāini is highly analytical. Key words: Grammatical tradition, Literary composition, Pāini, Phonology, Sanskrit, Semantics, Standard and colloquial usages, Syntax, Tamil, Tinai classification.
1. INTRODUCTION (pāyiram) to this work by one Panmapāran, Among the South Indian/Dravidian supposed to be a contemporary of the author languages, Tamil has the earliest literature, the so- Tolkāppiyan (Tolkāppiyar is polite form). The called Sangam literature, comprising nine Preface mentions that the work, named after the anthologies, which were made sometime before author, was made public in the court of a Pandya the fifth century AD. Most of the individual songs king in the presence of a brāhmaa scholar/critic included in these anthologies seem to have been Atankōāsān. Apart from this preface, the composed in the first couple of centuries AD and antiquity of the work may be inferred from the thereabouts by several bardic poets who lived in reference to this work in the 8th or 9th century commentary on Ka aviyal (otherwise called various parts of Tamil.akam (present day Tamil Nadu and Kerala together). Tolkāppiyam (hereafter Akapporu) a grammatical work on love theme TK), the earliest extant Tamil grammatical work, alone said to be written after some centuries later is usually treated as part of this early literature, than TK. The introduction to this commentary, though it is not recognized so by some for various rather legendary in nature, refers to the existence reasons (Shanmugam, 1989; Zvelebil, 1975). The of three successive literary academies called name Tolkāppiyam is mentioned in a brief preface Sangams along with the names of some poets and
* 37-1, Bharathi Park, 7th Cross, Saibaba Colony, Coimbatore - 641 011; Email: [email protected] 126 INDIAN JOURNAL OF HISTORY OF SCIENCE literary works associated with them. Though the section being divided into 9 chapters (iyal). Each account of the first two academies is steeped in chapter has a number of chūttiram (Tamil form of myth, there is some quasi-historical basis for the sūtra), otherwise called as nūpā in Tamil which third one as the literary works said to be associated are in stanzaic form, and range in length from one with this tally more or less with the existing corpus to fifty-nine lines. The total number of stanzas is of ‘Sangam’ literature. TK is said to have been a nearly 1600. Since the meaningful splitting of standard (grammatical) treatise (nūl) for the poets some stanzas differs from one commentator to of both the second and third academies. In any another, the exact number is difficult to arrive at case the Sangam legend is not helpful to get some present. concrete date for TK. The above legend also says The first section called Euttu (literally that the work Akattiyam written by an Akattiyar alphabet) with 483 stanzas deals with phonology was another treatise used in all the three and morphophonemics. Interestingly there are four academies. Legends that gained currency in some stanzas which describe the characteristic shapes commentaries on TK, dating from the 12th century onwards, assert that Akattiyar (identified with the of letters standing for the vowels and consonants. This description would apply exactly to the forms sage Agastya) was the teacher of Tolkāppiyar as well as Panampāran who wrote the preface to TK. found in the developed stage of the Tamil-Brāhmi But either of the disciples does not refer to the script when a diacritical dot was introduced for teacher in any way. All the imaginative and differentiating the pure consonant from the vocalic exaggerated legends which associated Akattiyar consonant of Aśokan Brāhmi (Mahadevan, 2003, with Tamil and treated him as the author of a pp.230–31). This section also has rules regarding premier Tamil grammar are found to be late and the generation of speech sounds, the sandhi useless for the present discussion (Vaiyapuri Pillai, (puarchchi in Tamil) involving vowels and 1988, pp.46-48; Zvelebil, 1975, pp.61-67). consonants in different combinations, casal/non- casal relations, and so on. Though TK is quite popular now among Tamil scholars and is adored as a great work, there The second section on chol (‘word’), in is as yet no critical edition of the text in the real about 460 stanzas, deals with morphology, syntax sense. It was mostly recovered part by part from and semantics. It has syntactical rules in forming the several commentaries on it from 1847 onwards discourse/sentences, cases, case-variation, classes and pieced together. The earliest commentary of nouns and verbs, particles (iai-chol) and covering the entire text was made by Iampūraar qualifiers (uri-chol). The correct usages are said in the eleventh or twelfth century. The other to be derived both from the colloquial usage important commentaries, which, however, cover (vaakku) and the practices in poetry (cheyyu). only some portions each, are those by Pērāchiriyar The nouns are classified according to tiai (c. 13th century), Nachchinārkkiniyar (c. 14th (human/non-human) and gender (pāl) in addition century), Sēnavaraiyar (13th-14th century), to number (single, plural and epicene). Verbs are Deyvachchilaiyār (c. 16th century). There are a few conjugated for three tenses, two numbers and three more. The so many commentaries vouch for the persons. popularity of TK, in spite of the subsequent The third section is on poru, literally appearance of some simplified grammatical texts meaning/matter, here the subject-matter of poetry. in Tamil. This section is the longest with 660 and odd TK has three major sections or books stanzas and deals with poetical themes, aspects of called atikāram (adhikāram in Sanskrit), each rhetoric, prosody and traditional usages. The two A NOTE ON GRAMMATICAL KNOWLEDGE IN EARLY TAMIAKAM 127 major poetical themes are akam (interior aspects) until then were oral compositions. covering pre-marital love and wedded love, and What is the necessity of writing this pu am (exterior aspects) covering warfare, grammatical work? What is the social and cultural panegyrics, contemplation on the meaning of life, context? What was the database of the work? The and so on. The first two chapters, respectively on answers partly lie in the work itself. In several akam and pu am, discuss the respective themes places in the work the author refers, either on the basis of tiai1 or landscape classification explicitly or otherwise, to cheyyu or composition that was peculiar to the early Tamil poetry. The (of poetry/prose) as the ultimate goal of the next three chapters (3 to 5) elaborate on the love grammar. Thus there are only four kinds of words theme. Chapters 6 and 7 treat respectively (iya-chol, tiri-chol, tichai-chol and vaa-chol) that sentiments (meyppāu in Tamil equivalent to bhāva go to make cheyyu (2:9:1). In fact the penultimate in Sanskrit) and simile (uvamai/upamā). Then chapter in the third section is entitled as cheyyu- follow the chapters (8 and 9) on prosody (cheyyu ) iyal (3:8). Among the different components of and on traditional usages (marapu). composition, he starts with speech sound (māttirai/ The date of the composition of TK has been mātra), letters (euttu), and so on (3:8:1). In the hotly debated among scholars taking extreme subsequent stanza (3:8:2) he says that the first two positions. On the basis of the Sangam legend components have been discussed above, meaning mentioned earlier, this work was given a hoary his treatment of letters in the first chapter of the antiquity by some scholars. Some would place it first section. By the way this is actually a case of before the date of Pāini to assert that Tamil had a cross-referencing across the sections. more ancient literature and grammatical tradition There is no doubt that TK dwells upon the than Sanskrit. Another vexing problem is to decide structure of Tamil language in the first two sections the place of TK with reference to the early Tamil –– phonology, various parts of speech, sentence anthologies, whether it preceded or followed them. pattern, and so on. However, throughout the work On the basis of certain linguistic features, it was his interest was directed to inculcate the thought by some that TK should have preceded prospective composers in the right kind of the said literature. But some other analytical language use to make their composition. This studies of the same features would suggest that aspect has been emphasized by more than one TK need not be anterior to the anthological scholar (Kailasapathy, 1968, Shanmugam, 1989, literature. The clinching evidence is the date when Vijayavenugopal, 2009). The tiai classification a proper writing system was available to the with its diverse semiotic codes is used as the basic author. A grammatical work like TK could not concept to explain and appreciate the exact import come into existence without a proper script. At of the earlier poetry and thereby give guidance to the earliest, such a script was available only in his contemporaries and future generations in the the second century AD or thereabouts. That is, as composition of literary texts. noted above, the mature form of the Tamil-Brāhmi script with the pui-marked consonant characters In several places he refers to the views of and with the differentiation of short and long forms other scholars either predecessors or of ‘e’ and ‘o’. Moreover this would be also the contemporaries without mentioning their names juncture when the anthology making started by while explaining his rules, implying the existence the process of writing the Sangam poems, which of some earlier indigenous grammatical tradition.
1 It may be noted the term tiai is used in TK in two different senses, ‘human’ and ‘region or landscape’, perhaps in the extended sense of ‘inhabited landscape’. 128 INDIAN JOURNAL OF HISTORY OF SCIENCE
Also he points in certain places how his rules are between TK and Sanskrit works and suggested that different from those of the Sanskrit scholars. His TK is modelled on those works. But he also made central discussion is mostly confined to the the following pertinent statement: ‘ Tolkāppiyar traditional usages within the Tamil-speaking has worked out a beautiful Tamil grammar on the country (tami kūu nallulukam). According to the models of Sanskrit Prātiśākyas, Yāska’s Nirukta, preface mentioned above this area is said to be Pāini’s śika and Pāini’s Aādhyāyi … … bounded by the Venkatam (hills) on the north and without doing the least violence to the genius of Kumari on the south. TK himself defines the the Tamil language’ (Sastri, 1934, p. 3). More normal Tamil word (iya-chol) as that current significant is the reappraisal of K. Meenakshi who within the area of standard Tamil (chen-tami) and made a detailed comparison of the first two whose meaning causes no ambiguity to anybody. sections (euttu and chol) of TK with Aādhyāyi. Elsewhere he also mentions the land within the She has drawn our attention to the fact that most four boundaries under the rule of the three of the previous Tamil scholars who commented (meaning the traditional three Tamil kings: Chēra, on the Sanskrit sources to TK depended more on Chōa and Pāya). He also refers to the words secondary studies than caring to look into the found in dialects current in the ‘twelve’ areas originals and hence made sometimes even adjoining to the standard Tamil region. It is also misleading statements, for instance in the TK’s to be stressed that he was not averse to the treatment of cases. She has stressed three inclusion of words from the northern tongue important differences between TK and Pāii’s (Prākrit/Sanskrit) provided the northern letters Aādhyāyi. Pāini’s treatment is highly analytic (meaning aspirates and voiced stops) are properly whereas TK’s is both paradigmatic and analytic. assimilated into Tamil. Majority of TK’s rules are concerned with sentence pattern while that of P ini is mostly concerned This brings us to the final problem. That ā with correct word formation (Vijayavenugopal, is, to what extent the knowledge base of TK is 2009). In A dhy yi there is no parallel thing to indebted to Sanskrit grammatical tradition. It is ā ā TK’s third section on poru dealing with literary A.C. Burnell (1875) who first suggested composition. This is a subject usually treated in similarities between TK and the Sanskrit work Alankāraśāstras, not in regular grammatical works Kātantra in the arrangement of sections, the nature in the Sanskrit world. The real purpose of P ini of technical terms used and the treatment of cases. ā to write his grammar seems to have been to He also suggested that the latter work belonged preserve the purity of bhāa, the language of his to a Pre-Pāinian grammatical school by name times, spoken by the elite in north-western parts Aindra. But subsequent more exhaustive of India. TK did not have any such avowed purpose comparative studies point to the fact that Kātantra towards the language per se. Meenakshi of course is just a popular, simplified work following admits the possibility of influence of the Sanskrit Pāini’s Aādhyāyi. This was made most probably grammatical tradition on TK, especially when a in the third century AD under the patronage of the strong grammatical tradition had existed much Sātavāhana court. Meenakshi avers that Burnell’s before TK’s time. conclusions are superficial and did not take into account the vast differences between the two There is a difference in saying ‘influenced works (Meeanakshi, 1997, pp. 445–52, 456). P.S. by’ and ‘modelled after’. If we say TK Subrahmanya Sastri, a scholar both in Sanskrit and modelled his grammar after Skt grammars, it is tantamount to accepting Tamil, in his fairly detailed English commentary TK’s dependence on Skt grammatical to TK observed in several places the parallels texts. But the internal evidences prove to A NOTE ON GRAMMATICAL KNOWLEDGE IN EARLY TAMIAKAM 129
the contrary. His description of Tamil communication. The other chapters in the last language reflects his independent section (poruatikāram) on similes/comparison approach despite his profound knowledge and on conventional usages of words if taken in Sanskrit and Prākrit particularly in the sūtras where he makes statements together with the related things explained in the comparing his method of description with chapters on iai-chol (particles) and uri-chol that of brāhmis (antaar) meaning (qualifiers) in the second section (chol-atikāram)) Sanskrit. go a long way in understanding the poetic diction Meenakshi, 1997, pp.459–60. of the ancient poems. ‘Thus one may conclude Some scholars consider some chapters in that the last part in Tolkāppiyam really focuses on the third section (like that on kaavu, kapu, the body of literature called ancient Tamil poems meyppāu and uvamam) as showing much (also identified as Cakam poems) by a deep and influence from Sanskritic works. Thus John penetrating analysis.’ Ralston Marr has made a detailed comparison of the sentiments (meyppāu) mentioned in TK:3:6 BIBLIOGRAPHY with those mentioned in Bharata’s Nāya Śāstra Burnell, A.C. On the Aindra School of Sanskrit and other Sanskrit works and suggested that TK Grammarians: Their Place in the Sanskrit and is much dependent on the latter for his ideas in Subordinate Literatures: Basel Mission Book and the field of dramatic theory (Marr 1985, pp. 56- Tract Depository, Mangalore, 1875. 68). But he has also noted there is no one to one Ganesh, K. N. Space-time, Event and Expression in Early correspondence. In this connection the comments Tamil Texts: Historicising Tolkāppiyam, Indian of Vaiyapuri Pillai, a profound and critical Tamil Historical Review, 38.1 (2011):1–22 scholar on these points are worth mentioning: Hart, George, The Poems of Ancient Tamil: Their Meliu and These chapters on sentiments and figures Their Sanskrit Counterparts, Oxford University Press, of speech are no doubt based upon works New Delhi, 1999. like Bharata’s Nāya Śāstra; but the Kailasapathy, K. Tamil Heroic Poetry, Kumaran Book treatment shows a rare inwardness, a House, Chennai, 2002 (Reprint of 1968 edn). brilliant expository power and crystal clear formulation peculiar to the author. Mahadevan, Iravatham. Early Tamil Epigraphy from the His sub-sections on prosody and on Earliest Times to the Sixth Century A.D., Cre-A and literary usage are master-pieces of their Harvard University, Chennai 2003. kind. His deep knowledge of the works Marr, John Ralston. The Eight Anthologies: A Study in Early of the earlier grammarians, his Tamil Literature, Institute of Asian Studies, Madras, thoroughness on the mechanistic side of 1985. prosody and his accuracy in ascertaining the usage of words have not been Meenakshi, K. Tolkāppiyam and Ashtadhyayi, International approached by any grammarian since his Institute of Tamil Studies, Chennai, 1997 time. Meenakshisundaram, T.P. A History of Tamil Language, Vaiyapuripillai, 1988, pp.52–53 Deccan College, Pune, 1965. Commenting on the chapter on prosody Shanmugam, S.V. Language Development and Language (cheyyuiyal), G. Vijayavenugopal (2009) also Attitude (Sangam Period) (In Tamil), Manivasagar observed that it not only deals elaborately with Publishers, Chennai, 1989. the forms of literary compositions but also with Subrahmanya Sastri, P.S. History of Grammatical Theories other essential features like context, the speaker, in Tamil and Their Relation to the Grammatical the hearer, time and other literary techniques, Literature in Sanskrit, Kuppuswami Sastri Research resembling very much the theory of Institute, Chennai, 1934, Reprint 1997. 130 INDIAN JOURNAL OF HISTORY OF SCIENCE
Subrahmanya Sastri, P.S. Tolkāppiyam – Collatikaram with Semiotics of Ancient Tamil Poetry’, Between an English Commentary, Annamalai University, Preservation and Recreation in Tamil Tradition of Annamalainagar, 1945. Commentary: Proceedings of a Workshop in Honour Vaiyapuripillai, History of Tamil Language and Literature, of T.V. Gopal Iyer, (ed.) Eva Wilden, French Institute NCBH, Madras 1988 (1956). of Pondicherry, 2009, pp. 133–43. Vijayavenugopal, G. ‘Tolkāppiyam: A Treatise on the Zvelebil, K.V. Tamil Literature, E.J. Brill, Leiden, 1975.