Pramāṇa · Search · About · API

Elements of South-Indian Palæography, from the Fourth to the Seventeenth Century A.D. c5.s3: § 3. The Vaṭṭĕluttu

Elements of South-Indian Palæography, from the Fourth to the Seventeenth Century A.D. c5.s3

Language
English

§ 3. The Vaṭṭĕluttu

c5.s3.1

This is the original Tamiḻ alphabet which was once used in all that part of the peninsula south of Tanjore, and also in S. Malabar and Travancore where it still exists though in exceedingly limited use, and in a modern form. It may, therefore, be termed the Pāṇḍyan character, as its use extended over the whole of that kingdom at its best period; it appears also to have been in use in the small extent of country below the ghats (South-Malabar and Coimbatore of the present day) which belonged to the Cera kingdom. As it was only gradually supplanted[^1) by the modern Tamiḻ character beginning about the eleventh century under the Coḷas, it is, therefore, certain that the Tölkāppiyam, Kuṟaḷ and all the other early Tamiḻ works were written in it, under the most flourishing period of the “Pāṇḍya” (or Madura) kingdom, or before the eleventh century when it finally fell under the Coḷas.

c5.s3.2

But though it is certain that the beginning of the Tamiḻ literature may be safely put about the ninth century, there is nothing to show that there was in any way a literature before that time. The legend of Agastya’s settlement in the South is, of course, historically worthless[^2), and though the three Dravidian kingdoms[^3) were undoubtedly ancient, we have nothing about their condition till Hiouen-Thsang’s visit to the peninsula about 640 A. D. He says of the inhabitants of Mo-lo-kiu-tch’a (Malakūṭa)[^4): “Ils ne font aucun cas de la culture des lettres, et n’estiment que la poursuite du lucre”[^5). He mentions the Nirgranthas or Digambara Jains (ascetics)[^6) as the most prominent sect in the South, and this corresponds with the actual

c5.s3.3

[^1): Tölkāppiyam, i., 14 mentions the V. forms of p and m. [^2): According to the Vīraśoḻiyam, a highly Sanskritized Tamiḻ Grammar of the 11th century, Agastya got his knowledge from Avalokita! But the author’s name Puttamittiran (Buddhamitra) shows his religion plainly enough. [^3): The Pāṇḍya kingdom (*e. g.*) is mentioned in the Açoka inscriptions (250 B. C.) by Ptolemy (vii., 1, 11, vol. ii., p. 143. ed. Nobbe) in the second century A. D. and by the Periplus in the third century A. D. The Mahāvamso (ed. Turnour) makes Vijayo (543 B. C.!) marry the daughter of the king of Dakkhina Madhura called “Panduwo” (p. 51), I do not find any subsequent mention of the Pāṇḍyas in this very monkish chronicle. I put the date of the Periplus at the third century A. D. following Reinaud. [^4): Malakūṭa is mentioned in the Tanjore inscription, and there can be little doubt that the Coḷa kingdom of Tanjore is to be understood by it and not Madura as has been supposed. [^5): “Voyages des Pèlerins Bouddhistes” iii., p. 121. [^6): I proposed the identification of the Nirgranthas with the Jains (in *I. A.* i., p. 310, *n.*) on the ground that in the Jain Atthapāhuḍaka (*i. e.* Ashṭaprābhṛitaka) Nirgrantha is constantly used as an epithet of the true (Digambara) Jains, and that, therefore, it could not be referred to the Brahmans as had always been done hitherto, and also on the ground of probability, as *e. g.* Hiouen-Thsang’s account (iii., p. 27) of the Nirgranthas is much more likely of Jains than of Brahmans; but I have since got additional information which makes my identification certain, and can leave no doubt that Jain ascetics are intended by the word ‘niggantha’ (nirgrantha), though the word is now not understood by the Jains.

c5.s3.4

remains of the early Tamil literature which are in fact Jain, but he would have hardly said what he does if the grammars and the Kuṟal then existed. The earliest apparent or probable mention of writing in S. India is the passage in the Periplus of the Red Sea which describes Cape Comorin. Among other facts the author mentions that "it is related (ἱστορεῖται) that a goddess bathed there". Considering that this log-book was composed in the third century, and that, therefore, the Greek is very late, it is quite possible that this word ἱστορεῖται may mean that the legend was written, and the earlier editors and translators of the text took it in this sense¹), but the passage is by no means beyond doubt in this respect²). The earliest Tamil Grammar by Agattiyan (Agastya) clearly refers to writing, if we may trust a quotation (preserved by a commentary on the Nannūl) which compares the relation between a letter and the sound it stands for, with the relation of an idol to the deity it represents. The age of this is unknown.

c5.s3.5

¹ See the edition in Hudson's “Geographi Græci Minores” vol. i. p. 33, where the passage is translated: “Literis enim memoriaeque proditum est deam olim singulis mensibus ibi lavari fuisse solitam”. The latest and more critical editor (C. Müller) has on the other hand: “Dea aliquando ibi commorata et lavata esse perhibetur”. (‘Geographi Græci Minores.’ p. 300 of vol. i. of Didot's Edition). It is therefore uncertain.

c5.s3.6

² I pass over the statement of Iambulus (“Diodorus Siculus,” ed. Dindorf, ii. 59 in vol. i., p. 222) as it is impossible to explain it by any Indian alphabet as yet known.

c5.s3.7

The Vaṭṭeḻuttu was gradually supplanted by the modern Tamil after the conquest of Madura by the Coḷas (eleventh century), and it appears to have entirely gone out of use in the Tamil country by the fifteenth century³). In Malabar it remained in general use up to the end of the seventeenth century among the Hindus, and 'since then, in the form of the Kōlēḻuttu, it is the character in which the Hindu sovereigns have their grants drawn up. The Māppiḷas of the neigbourhood of Tellicherry and in the Islands

c5.s3.8

³ I owe the fact of the existence of the Vaṭṭeḻuttu up to so recent a time in the S. Tamil country, to the Right Rev. Dr. Caldwell.

c5.s3.9

848. So ummaggāhimuho čaümuho sadadivāsaparamāü čālisarajjao jidabhūmi pučhaï samattigaṇaṁ |
C. Sa Kalkī unmārgābhimukhaç čaturmukhākhyaḥ saptativarshaparamāyushyaç čatvāriṁçadvarsharājyo jitabhūmiḥ san svamantrigaṇaṁ pṛiččhati.
849. Amhāṇaṁ ke avasā? ṇigganthā atthi! kīdisāyārā? ṇiddanavatthā bhikkhabhoji jahāsattham idi vayaṇe |
C. Asmākaṁ ke avaçā? iti. mantriṇaḥ kathayanti: nirgranthāḥ santi 'ti. punaḥ pṛiččhati: kīdṛiçākārā? iti. nirdhanavastrā yathāçāstram bhixābhojina iti mantriṇaḥ prativacanaṁ çrutvā—
850. Taṁ pāṇiüḍe ṇivaḍitapathamapiṇḍaṁ tu sukkam idi geyam ṇiyame sa jīvakade čattāhārā gayā muṇiṇo |
C. Teshāṁ nirgranthānāṁ pāṇipuṭanipatitaṁ prathamapiṇḍaṁ çulkaṁ iti grāhyaṁ iti rājño niyamena jīvena kṛitena tyaktāhārāḥ santo munayo gatāḥ.

c5.s3.10

The Niganthas (*i. e.* Nirgranthas) are frequently mentioned in the Pali “Dāṭhāvaṁso” (of the 12th cent.) as heretical enemies of the Buddhists who worshipped Vishṇu (see iii., 23); this answers to Jains, but certainly not to Çaiva Brahmans. The Nirgranthas are already mentioned in an Açoka edict.

c5.s3.11

used this character till quite recently; it is now being superseded by the modified Arabic character which has religious prestige on its side[^1).

c5.s3.12

The ultimate origin of the Vaṭṭeḷuttu is again a difficult problem in Indian Palæography. In the eighth century it existed side by side and together with the Grantha[^2); it is, therefore, impossible to suppose that the Vaṭṭeḷuttu is derived from the S. Açoka character, even if the conclusive argument of the dissimilarity between the phonetic values of many of the corresponding letters be neglected[^3). Again the S. Açoka character would have furnished a more complete representation of the Tamil phonetic system than either the Vaṭṭeḷuttu or the modern (Grantha) Tamil alphabet does[^4); it must, therefore, follow that the alphabet was formed and settled before the Sanskrit grammarians came to Southern India, or we should find as accurate a representation as they effected for Telugu and Canarese. The Tamil grammarians, however, evidently found the language already written when they began their labours, and thus this part of their grammars is comparatively imperfect[^5). Again as the Vaṭṭeḷuttu is an imperfect alphabet, it cannot be the origin of the S. Açoka character; for, if it were, the evidence of the extension and adaptation must be far greater than it is. It is plain that many of the aspirated letters in the S. Açoka character are formed from the corresponding unaspirated letters, but if that alphabet were formed from the Vaṭṭeḷuttu, it would have shown traces of a similar formation in the letters **g**, **j**, **ḍ**, **d** and **h** for which there are no forms in the Vaṭṭeḷuttu. But these letters appear to be primitive in the S. Açoka character. The only possible conclusion, therefore, is that the S. Açoka and Vaṭṭeḷuttu alphabets are independent adaptations of some foreign character, the first to a Sanskritic, the last to a Dravidian language. There are, however, resemblances between the two that point to a common Semitic origin; and these extend perhaps to two-thirds of the

c5.s3.13

[^1): See No. ii. of my "Specimens of South-Indian Dialects". [^2): Cfr. the grants to the Israelite and Christian communities in Travancore. These were first attempted (like most branches of S. Indian archæology) by Dr. John Leyden (see his Life by Morton, Calcutta, 1823, p. 52). Even in a treatise like the present it may not be out of place (as so little is now known about this distinguished scholar) to give an unpublished piece of information about him, viz., his epitaph at Batavia. It runs: "Sacred | to the memory of | John Casper Leyden, M. D. | who was born | at Teviotdale in Scotland, and who died | in the prime of life | at Molenvliet near Batavia | on the 28th August 1811 | two days | after the fall of Cornelis | | The poetical talents and superior literary | attainments of Dr. Leyden rendered him an | ornament of the age in which he lived— | His ardent spirit and insatiable thirst after | knowledge, was perhaps unequalled: | And the friends of science must ever | deplore his untimely fate— | His principles as a man were pure and spotless | And as a friend he was firm and sincere. | —Few have passed through this life | with fewer vices or with a greater | prospect of happiness in the next". I owe this copy to the kindness of Dr. Stortenbecker, the first Colonial Secretary at Batavia; when I was there, I was unable to seek out Leyden's tomb, but I doubt not that it is well cared for by the Dutch. In Java, at least, cemeteries are not allowed to be desecrated and become the abomination and disgrace that they so often are in British India. [^3): See Appendix A. [^4): See Appendix A. [^5): The Telugu and Canarese grammars explain the respective phonetic systems by a steady reference to that of Sanskrit; the Tamil grammars do not refer to the Sanskrit at all in this way.

c5.s3.14

Vaṭṭeḻuttu letters; the others differ totally, yet several of these sounds (**ḷ**, **ḻ**, **ṟ**) exist in the other Dravidian languages, and distinct letters have been invented to express them. Thus the Tamil-Malayāḷam **ḻ** is expressed by ழ, the Canarese identical letter by ೞ. Again the Telugu-Canarese **ṟ** is expressed by ఱ, whereas the same letter in Tamil is written ற, so the Telugu-Canarese and Tamil **ḷ** which are identical in sound are written quite differently. There is also a peculiarity in the popular Tamil way of naming the letters; in Sanskrit (excepting repha=**r**) names of letters are formed by adding -kāra to the letter in question; in Tamil -na is affixed to short and -vēna to long syllables, every consonant being named with some vowel following it¹).

c5.s3.15

¹ The order I have given to the Vaṭṭeḻuttu corresponds with that of the Tamil alphabet, and is that of the Sanskritizing grammarians. There is, however, a sūtra in the Nannūl which appears to me to indicate that this was not the case when the grammarians began their labours. It runs: “*Śiṟappiṇum iṇattiṉuñ šĕrind’iṇd’ ammuḍaṇaḍattaṟāṉē muṟaiy āyum*” (ii., 18). *i. e.* “The series of letters beginning with ‘a’ (and) arranged according to their priority and relationship, is *here* their order”. iṇḍu=here (atra), *i. e.* in this grammar. If this order were the usual one, this explanation would have been unnecessary: I am unable to find any trace of this other arrangement of the Tamil alphabet. The Kuṟaḷ (i., 1) mentions ‘a’ as the first letter. The Nannūl (ii., 71) directs -ayaram for the names of consonants, -karam for short vowels, -kān for the foreign ai and au, and -kāram for the long vowels etc. This is clearly an imitation of the Sanskrit. Again the same work (ii., 43) mentions the tŏllai vaḍivu or ‘old forms’ of the letters.

c5.s3.16

There is another peculiarity in the Vaṭṭeḻuttu system of writing, which might appear also common to the early Prakṛit (Cave) inscriptions in India, and to those of Ceylon; but there is, in reality, a difference. This peculiarity consists in the writing of consonants which follow one another without the intervention of a vowel, on the same line, not perpendicularly, as is done in the other Indian alphabets. Instances of this occur in the Vaṭṭeḻuttu documents in numbers in every line, and it is the chief feature in this system as compared with those of the other Indian alphabets. In the Cave inscriptions this practice does not really exist; the duplicated letters (which constitute the bulk of the instances of consonants directly following one another) are either not marked, or marked by a **o**²). In the Ceylon inscriptions, the practice is to omit reduplicated letters, and instances of the Sanskrit system occur³).

c5.s3.17

² See above p. 13.

c5.s3.18

³ I gather this from remarks by Mr. Rhys Davids and Dr. P. Goldschmidt.

c5.s3.19

The neglect to mark the duplication of consonants is a primitive Semitic practice⁴); but in Pahlavī (so far as can be ascertained) the duplication of consonants was marked, as we find in the Vaṭṭeḻuttu; one example is, beyond doubt, viz, the duplication of **n**.

c5.s3.20

⁴ The signs now used to supply this defect are all of modern origin.

c5.s3.21

It is thus evident that the Vaṭṭeḻuttu differs greatly from the Canarese and Telugu alphabets; but if one compares the forms of **ī**, **k**, **t**, **r**, and even **a** and **ā**, in both, it is hardly possible to avoid the conclusion that they are derived from the same source. That an alphabet should have been imported independently into Northern India (probably Gujarat) and also into the Tamil country much about the same time, seems strange; but it is nevertheless most likely, considering the circumstances of foreign trade with India as reported by the classical authors. The Periplus, for example, mentions a large trade with Āriakē, *i. e.* Bombay and the country of the Prakrit-speaking peoples; there is then a gap, and again large trade with Limurikē. Now this is simply the Western Tamil country or Malabar¹, and between the two provinces there was the Pirate Coast which preserved its evil name till within recollection of many. There would be no trade there, and the Western and S. Western Coast would thus be in fact distinct countries. Again there could not have been any communication by land, for Fa-hian (400 A. D.) mentions the Deccan as uncivilized and inaccessible; it is, therefore, more likely that the S. Açoka character and the Vaṭṭĕḻuttu are totally distinct importations, than derived the one from the other.

c5.s3.22

¹ The Periplus and Ptolemy have λυμορική, but as the Peutingerian Table, the Ravenna geographer and Guido have Dimirice; there can be no doubt that the copyists have mistaken Λ for Δ, an exceedingly easy error in Greek. Dimurikē is thus Tamil + ikē; now Malayāḷam was called Tamil formerly, and at the time of the classical writers (beyond doubt) the languages in no way differed. It is thus impossible to identify Dimurikē with Canara, (as was done by Vincent[^1]) following Rennell for quite illusive reasons), but it must be taken to mean S. Malabar, and the three great ports Tundis, Mouziris and Nelkunda (Nincylda) are Kaḍal(t)uṇḍi (near Beypore), Muyirikkōḍu (Kishaṅkoṭṭa opposite to the site of Cranganore) and Kallaḍa (inland from Quilon up a large river). The Vaṭṭĕḻuttu must, therefore, have been imported at one of these places. The reasons for this new identification would take too much space here, and must be given elsewhere. (See Col. Yule's remarks on the map of Ancient India in Smith's "Ancient Atlas", 1875.)

c5.s3.23

¹ "Commerce and Navigation of the Indian Ocean", ii. p. 456.

c5.s3.24

What was this source? There is quite as much reason for supposing a Semitic original in this case, as in that of the S. Açoka character, resemblances to some of the Phœnician and Aramaic letters being equally apparent in both². Of all the probable primitive alphabets with which a comparison of the Vaṭṭĕḻuttu is possible, it appears to me that the Sassanian of the inscriptions presents most points of resemblance³. The number of letters also in both, narrowly agree. At present the difficulty is to find certain and dated examples of the Aramaic characters used in the early centuries B. C. and also similar specimens of the Vaṭṭĕḻuttu; there is also the difficulty of deciding which of the many derivatives from the Phœnician alphabet, but of which it is possible this S. Indian alphabet may have been formed, was actually used for this purpose.

c5.s3.25

² I must, however, point out that Profr. Max Müller is not satisfied in respect of the S. Açoka character (Sanskrit Gr. p. 3). He quotes Prinsep's "Essays" by Thomas, ii., p. 42.

c5.s3.26

³ The development of the Pahlavī from the early Aramaic character is traced by M. F. Lenormant in the "*Journal Asiatique*" for August and September 1865 (pp. 180-226). The resemblance between some of the Vaṭṭĕḻuttu letters and the corresponding Proto- and Persepolitan Pahlavi forms (as given by F. Lenormant) is very striking. Cfr. a; Pahlavī d with t; i; l (r); m; n; p; k; s with *s* etc. There is ample evidence as to trade between Persia and the W. Coast of India.

c5.s3.27

Another remarkable feature in the Vaṭṭĕḻuttu is the system of marking the secondary vowels. This is intermediate between the systems of the Northern and the Southern Açoka alphabets, and thus connects both. I was led by this striking fact to suggest in an article on the Vaṭṭĕḻuttu¹ that the Northern alphabets had, in this respect, copied from it. At present it appears to me that it is best to consider the Açoka alphabets and the Vaṭṭĕḻuttu as independent; the evidence afforded by the few facts that are satisfactorily known in respect of these characters is too imperfect to allow of more precise conclusions being drawn. Vaṭṭĕḻuttu is the modern Malayāḷam name of this character, and means ‘round hand’ apparently to distinguish it from the Kōlĕḻuttu or ‘sceptre hand’; it appears to be the best name for this alphabet as it prevents all confusion with the modern Tamiḻ.

c5.s3.28

¹ In the *Indian Antiquary* Vol. i., p. 229. This article is, I believe, the first to call attention to this alphabet. Specimens of the character occur in the preface to Rheede’s “Hortus Malabaricus” (1678), and in Fryer’s “New Account” (1698) p. 33. The author gives it as Telugu, but the specimen on p. 52 is Telugu and not Malabar (Tamiḻ) as he states; he has made a mistake between them.

Linked texts (6)