Pramāṇa · Search · About · API

Early South Indian Palaeography c11: Chapter III. A. Antiquity of Writing in India

Early South Indian Palaeography c11

Language
English

Contents

Chapter III. A. Antiquity of Writing in India

c11.1

Indian traditions of both the orthodox and hetrodox sects claim great antiquity for the art of writing in India. According to them it was created by Brahmā, the God of creation. The *Nārada Smṛti*, a redaction of the *Manu Samhitā* and a work on ancient Hindu Law usually assigned to the 5th century A.D. (referred to by Bāṇa, the court poet of Harṣa of Kanauj) states: "Had not Brahmā the creator created the written (literature), the best of eyes, this world could have never attained to its happy condition".¹ Bṛhaspati in his *Varttikā* on Manu mentions the same tradition when he says, "because in a period of six months memory is confused regarding a particular thing, in very early times the creator produced letters depicted on leaves".² Again Kālidāsa stresses the value of learning the art of writing and says, "by the proper grasp of the art of writing one reaches the vast treasure of literature, as one approaches the ocean through the mouth of a river".³

c11.2

¹ *S. B. E.*, XXXIII, pp. 58 ff.

c11.3

² > *Ṣaṇmasike tu samaye bhrāntiḥ samjāyate yataḥ*

c11.4

³ > *Liperyatha vad grahapeṇa vanyam nade*

c11.5

There are evidences in the Vedas themselves that the art of writing was known even during that age. The Ṛgveda mentions the names of metres like the *Gāyatri*, *Anuṣṭubha*, *Bṛhati*, *Virāja*, *Triṣṭubha*, *Jagatī* etc.⁴ The *Vājsaneyi Samhitā* of the Yajurveda mentions some additional metres like *paṅkti*, *dvipada*, *tripada*, *catuṣpada*, *ṣaṭpada* etc.⁵ The Atharvaveda contains reference to eleven metres.⁶ The names of metres and the technical terms

c11.6

⁴ X, 14. 16; 132. 3. 4.

c11.7

⁵ XI. 8; XIV. 19, XXIII. 33; XXVIII. 14.

c11.8

⁶ VIII. 9. 19.

c11.9

*Nākariṣyadyadi Brahmā likhitam cakṣuruttamam*
*tatreyamasya lokasya na bhaviṣyat śubhā gatiḥ*
See also Bṛhaspati's *Vārttikā* on Manu, *Ibid.*, p. 304.

c11.10

*dhatrakṣāṇi sṛṣṭani patrārūḍānyath pura*
(quoted in *Āhnikatattva.*)

c11.11

*mukheneva samudramāviśat* l
(*Raghuvamśa*, III, 28).

c11.12

regarding the composition could have been possible only after the systematisation of the rules of prosody; and this presupposes a knowledge of the art of writing. Reference is also made in the Vedic literature to high figures involving knowledge of complicated Mathematics which could have been possible only with the use of written numbers and Arithmetic. The Ṛgveda mentions that king Sāvarṇi gave away in alms one thousand cows with the number eight marked on the ears of each.⁷ This custom is also noticed by Pāṇini. The *Vajsaneyi Saṃhita* of the Yajurveda includes the astronomer (*gaṇika*) in the list of persons connected with the sacrifice *Puruṣamēdha*.⁸ The *Taittirīya Saṃhita* mentions the following numbers in the ascending order; *daśa* (10), *śata* (100), *sahasra* (1,000), *ayuta* (10,000), *niyuta* (1,00,000), *prayuta* (10,00,000), *arbuda* (1,00,00,000), *nyarbuda* (10,00,00,000), *samudra* (1,00,00,00,000), *madhya* (10,00,00,00,000), *anta* (1,00,00,00,00,000) and *prārdha* (10,00,00,00,00,000).⁹ References to high figures are found in the *Brāhmaṇa* literature also.¹⁰ The *Śatapatha Brāhmaṇa* mentions minute divisions of a day and night. It says that a day and a night together consist of 30 *muhūrtas*, one *muhūrta* of 15 *kṣipras*, one *kṣipra* of 15 *etarhis*, one *etarhi* of 15 *idānims* and one *idānim* of 15 *prāṇas*. It will be seen from the above that a day and a night contain 30 × 15 × 15 × 15 × 15 = 15,18,750 *prāṇas*.¹¹ One *prāṇa* is equal to 1/17 of a second. These calculations would have been possible only with written figures, and imply the existence and a good knowledge of the art of writing.

c11.13

⁷ *Sahasram me dadato aṣṭakarṇyaḥ*. X. 62. 7.

c11.14

⁸ *grāmaṇyam gaṇakamabhikrośakam tanmahase*. XXX. 20.

c11.15

⁹ IV. 40. 11. 4; VII. 2. 21. 1.

c11.16

¹⁰ *Pañcaviṃśa Brāhmaṇa*, XVIII. 3. *Śatapatha Br*., X. 4. 2; 22-25.

c11.17

¹¹ XII. 3. 2. 1.

c11.18

There are other pieces of evidence also which would suggest the antiquity of writing in India. The *Nirukta* of Yāska, a work on the etymology of words, contains the names of some of his predecessors, namely Audumbarāyaṇa, Āgrāyaṇa, Aruṇavābha, Aupamanyava, Gārgya, Gālava, Kāṭṭhakya, Kautsa, Carmaśiras, Taitiki, Varṣyāyani, Śākalya, Śatabalākṣa, Śākaṭāyana, Śākapūṇi and Śthaulāṣṭīvi. This list¹² of the predecessors of Yāska in the field of technical works on language takes them to a very early period as also the art of writing.

c11.19

¹² The list is only selective.

c11.20

The Upaniṣads contain references to *akṣaras* or letters.¹³ It is said they were not only pronounced but also written since they were associated with suffix *kāsa* (something to be made) and with *varṇa* (something to be coloured or painted).¹⁴ Some of the Upaniṣads mention *varṇa* (written letters) and *matras* (medial signs) together.¹⁵

c11.21

¹³ *Himkara iti tryakṣaram prastava iti tryakṣaram tatsamam* (*Chhandogya*, II. 10).

c11.22

¹⁴ *Chhandogya*, I, 13. ; II. 22. 3.

c11.23

¹⁵ *Varṇaḥ svaraḥ matra balam* (*Taittiriya*, I. 1).

c11.24

Some of the Āraṇyakas make minute differentiation between *ūṣman* (sibilants), *sparśa* (mutes), *svara* (vowels) and *antastha* (semi-vowels), as also between *vyañjana* (consonants) and *ghoṣa* (voice) and between *mūrdhanya* (cerebral) and *dautya* (dental). There is a discussion on *sandhi* (joining of letters). Further the composition of the sacred *akṣara*, *om*, is explained there as a combination of letters *a*, *u* and *m*.¹⁶ It must be remembered that the Upaniṣads, Āraṇyakas and the Brāhmaṇas which are all largely composed in prose constitute a voluminous literature, and could not have been handed down from generation to generation merely by rote. The existence of many technical terms on grammar, etymology and prosody in them also makes one think that there must have been some knowledge of writing even then.

c11.25

¹⁶ *Aitareya*, III, 2. 1; II, 2. 4; III, 2. 6; III, 1. 5.

c11.26

There are definite references to the art of writing in some ancient works. For instance, the *Aṣṭadhyāyi* of Pāṇini contains a few words which have some bearing on writing in general. Pāṇini has been assigned to the fourth century B.C. by Max Muller and Bühler and to the eighth century by Goldstucker. The words bearing on writing are, for instance, *lipi* (script), *lipikāru* (writer or scribe),¹⁷ *yavanani* (Greek script),¹⁸

c11.27

¹⁷ 13. 2. 21. Bühler thinks that the words " *Dipi* and *Lipi* are probably derived from the old Persian *Dipi*, which cannot have reached India before the conquest of the Panjab by Darius about B. C. 500, and which later became *Lipi*". (*Indian Palaeography*, p. 5.) Pandey has well pointed out that the word is not derived from *dipi*. See his *Indian Palaeography*, p. 11, fn. 1.

c11.28

¹⁸ 4-1-49.

c11.29

*grantha* (a book),¹⁹ *svarita* (a mark in writing)²⁰ etc. Pāṇini mentions the practice of the marking of the ears of cattle with the signs of figures 5 and 8 as also religious symbols like the *svastika*, 卐.²¹ Further he refers to a number of earlier writers on Sanskrit Grammar, among them being Apiśali,²² Kaśyapa,²³ Gālava,²⁴ Gārgya,²⁵ Cakravarman,²⁶ Bhāradvāja,²⁷ Yāska,²⁸ Sākalya,²⁹ Sākaṭāyana³⁰ and Sēnaka.³¹ All this shows the long history of Sanskrit grammar, before Pāṇini. This could have been possible only if the art of writing had been known to the earliest of them. Curiously enough Max-Muller writes, "I maintain that there is not a single word in Pāṇini's terminology which presupposes the existence of writing",³² and thus ignores the above references. He has also overlooked the fact that a work on advanced and developed grammar itself presupposes writing. In this connection it may be mentioned that the languages or dialects of the words without a definite knowledge of the script has not been known to possess a codified grammar.

c11.30

¹⁹ 1-3-75; 3.87; 4-3-116.

c11.31

²⁰ 1-3-11.

c11.32

²¹ 6-3-115.

c11.33

²² 1-2-25.

c11.34

²³ 6-3-61.

c11.35

²⁴ 8-3-20.

c11.36

²⁵ 6-1-130.

c11.37

²⁶ 7-2-63.

c11.38

²⁷ 2-4-63.

c11.39

²⁸ 8-3-19.

c11.40

²⁹ 3-4-111.

c11.41

³⁰ 5-4-112.

c11.42

³¹ 6-1-123.

c11.43

³² *History of Sanskrit Literature*, p. 262.

c11.44

The *vedāṅga-śikṣa* (the science of correct pronunciation), *kalpa* (ritual, procedure or litany), *nirukta* (etymology), *vyākaraṇa* (grammar), *chhandas* (prosody) and *jyōtiṣa* (astronomy), all of which constitute very ancient branches of the technical literature of India, also lead one to think that the art of writing was not unknown in India in ancient times.

c11.45

The *Sūtra* literature which has by common consent been assigned to the period between the eighth and the third centuries B. C. also bears evidence of the knowledge of writing in the country. For example the *Vasiṣṭha Dharma Sūtra* mentions written documents for purposes of legal evidence.³³

c11.46

³³ XVI, 10-14-15.

c11.47

The *Arthaśāstra* of Kauṭilya, which belongs to the fourth century B. C., makes definite mention of writing. Among such references in it are the following:

c11.48

i. Having gone through the tonsure ceremony the student should learn the alphabet and writing.³⁴

ii. In the fifth (*nālika*) the king should correspond in writs (*patra sampreṣaṇena*) with the assembly of his ministers.³⁵

iii. With signs and writings he should send his spies.³⁶

iv. The writer should be prompt in composing, elegant in writing and able in reading documents.³⁷

c11.49

³⁴ 1-5-2. *Vṛtta caula karma lipisaṁkhyānam copayuñjīta.*

c11.50

³⁵ I-19-6. *Pañcame mantripariṣadā patra sampreṣaṇena mantrayet.*

c11.51

³⁶ I. 12, 8.

c11.52

³⁷ II. 10-2.

c11.53

The epics, the *Rāmāyaṇa* and the *Mahābhārata* contain a number of terms relating to writing, such as *likh*, *lekha*, *lekhana*, *lekhaka* etc. The two works are usually assigned to a period before the Mauryan age, though not in its present form. Some are inclined to think that there are some interpolations in the work. Still Bühler himself thinks that "though the testimony of the Epics can, therefore, only be used with due reserve, yet it is undeniable that their terms regarding writing and writers are archaic".³⁸

c11.54

³⁸ *Ind. Ant.*, XXXIII, App., p. 4.

c11.55

Literary works in classical Sanskrit such as epics, *smṛtis*, *kāvyas*, dramas and works dealing with polity, morals and technical subjects contain innumerable references to the art of writing in India. Though such of them as are posterior to the days of Aśoka may not be of much value as evidence of the earlier existence of writing in India, the Brahmanical literature of the country assignable to the period before the rise of Buddhism are of much value for a study of the problem of the antiquity of writing in India.

c11.56

Besides the early Brahmanical literature of the country, the early Jaina and Buddhist literature also bears testimony to the knowledge of writing in India from very early times. The early Jaina works, the *Samavāyaṅga Sutta*, the traditional date of which is considered to be about 300 B. C., and the *Paṇṇavaṇa Sutta*, traditionally ascribed to about 168 B. C. mention eighteen kinds of scripts as then in use in the country. The latter work divides the Āryas into nine classes, the sixth of which is the *bhaṣāriya*, 'Aryan by language', who speak the Ardhamāgadhi (the Prākṛt in which Mahāvīra is said to have preached the Jaina religion and with whom the Brāhmī script is customary).³⁹ The *Bhagavatī Sūtra*, another Jaina work, has *Namo Bhambhīya liviye* (Salutation to the Brāhmī script).

c11.57

³⁹ Pischel, *Prakrit Grammar*, Introduction, Tr. into English by Bata Krishna Ghosh, *Indian Culture*, Vol. II, (1935 - 36), p. 474.

c11.58

Among the Buddhist works the *Mahavastu* gives a list of thirty names as the varieties of script prevailing in the country. The *Lalitavistara*, a work in Sanskrit dealing with the life of the Buddha,⁴⁰ describes how the Buddha as a boy went to the *lipi śala* (a school where writing was taught) and how Viśvāmitra, his teacher, taught him letters on a writing board of sandal wood and with a golden pen.⁴¹ The book further states that sixty-four kinds of scripts were in use in the country and gives their names. The numbers eighteen, thirty, sixty-four etc., are considered auspicious and some scholars hold the view that these numbers should not be taken seriously. Anyhow, these clearly show that many varieties of scripts were in use in the country even in those days.

c11.59

⁴⁰ It is not easy to fix correctly its date. But as the work was translated into the Chinese in 308 A.D. it may be taken to belong to a period at least one or two centuries earlier.

c11.60

⁴¹ Ch. X.

c11.61

The works included in the *Vinaya piṭaka* commend the art of writing (*lekhana*) to the monks on account of its innocence.⁴² The *Mahāvagga* mentions *lekha* (writing), *gaṇana* (arithmetic) and *rūpa*⁴³ which formed the curriculum of studies in the primary schools in ancient India. The *Jatakas* which contain stories of the earlier lives of the Buddha refer to *phalaka* (wooden writing-board) and *vaṛṇaka* (wooden pen) as used for writing.⁴⁴ They also mention writing in connection with official and private letters,⁴⁵ royal proclamations⁴⁶ documents relating to family affairs,⁴⁷ moral and political maxims,⁴⁸ usury and bonds⁴⁹ and manuscripts.⁵⁰ The *Suttanta* is a book of sermons on the conduct of the *bhikṣus*. Among the prohibitions for them was the *Akkharikā* (*Akṣarikā*), a game that was played by children and in which one had to recognise letters written with fingers either on one's back or in the sky.⁵¹

c11.62

⁴² *Bhikkupacittiya*, 2. 2.

c11.63

⁴³ *Ibid.*, 65. 1.

c11.64

⁴⁴ Ch. X.

c11.65

⁴⁵ *Kathaka Jataka*; *Kama Jātaka*.

c11.66

⁴⁶ *Rūru Jātaka*.

c11.67

⁴⁷ *Kaṇha Jātaka*.

c11.68

⁴⁸ *Kurudhamma Jātaka*.

c11.69

⁴⁹ *Rūru Jātaka*.

c11.70

⁵⁰ Bühler, *Indian Studies*, III, 120.

c11.71

⁵¹ *Brahmajala sutta*, 14; *Samaññaphalasutta*, 49. Terms like *lekha*, *lekhaka*, *likhati*, *akkhara*, *chindati* etc., and also the writing materials of wood, bamboo, *paṇṇo* (leaves), *suvaṇṇapaṭṭa* etc. are mentioned in early Buddhist literature. Bühler opines that these refer to the primitive character of writing, that is incision of signs on hard materials like stone. (*Ind. Ant.*, XXXIII, App. p. 5). Only the word *chindati* suggests incision, which may apply to writing on palm leaves (*talapatra*). But the writing on *bhūrjapatra* (birch bark) was done with the aid of inks. Further it is also to be remembered that even for writing on stone, the signs are first traced on stone by ink or chalk or some visible solution, and then the masons inscribe the signs following the tracing.

c11.72

Foreigners who visited India in ancient times also testify to the antiquity of writing in India. The earliest of them were the Greeks, some of whom, on account of their stay in the country for some time, had occasions to observe closely the general conditions of the people in all their aspects. One of them was Nearchus, one of the generals of Alexander during his invasion of India. He says that the people of the Panjab knew the art of manufacturing paper out of cotton and tattered cloths, obviously for writing purposes.51a Q. Curtius, another Greek writer refers to the inner tender bark of certain trees used for writing. It was probably the *bhūrjapatra* or birch-bark.⁵² Megasthenes, the ambassador of Seleukos Niketor at the court of Candragupta Maurya, writes in his *Indica* that for the convenience of travellers mile stones were fixed on the roads at intervals of ten *stadia* so that they could know the distance between rest-houses.⁵³ He also refers to the current practice of expounding *varṣaphala*, *i.e.* the good and bad prospects of the coming year according to an almanac as also the preparation of horoscopes for individuals. These would clearly prove that the art of writing was known and practised in India in the fourth century B. C. They do not hint anywhere that writing in India was of foreign origin.

c11.73

51a Strabo, XV. 717. Bühler takes this to mean that they wrote letters on well-beaten cotton cloth (*Ind. Ant.*, XXXIII, *App.* 6).

c11.74

⁵² Mc Crindle, *History of Alexander's Invasion of India*, VIII. 9.

c11.75

⁵³ *Indica*, 91, 125–26.

c11.76

Certain other foreign sources also indicate the tradition that the art of writing was indigenous to the country. The *Fa-Wan-Shu-Lin*, a Chinese encyclopaedia says that the Brāhmī script, written from left to right, was invented by *Fan* (Brahmā).⁵⁴ The celebrated Chinese pilgrim Hiuen Tsang also refers to the very early invention of writing in India.⁵⁵ Alberuni, the Arab scholar, records a slightly different story. He says that the Hindus had once forgotten the art of writing and that through a divine inspiration it was rediscovered by Vyāsa, the son of Parāśara.⁵⁶ According to him the history of the Indian alphabets would begin with the commencement of the *Kaliyuga* in 3101 B. C. This tradition got currency evidently due to the fact that Vyāsa is believed to have collected the Vedas and classified them into four and composed the eighteen Purāṇas and the *Mahābhārata*.

c11.77

⁵⁴ Babylonian and Oriental Record, I, 59. quoted by Bühler, *Op. cit.*, p. 1.

c11.78

⁵⁵ Beal, *Si yu ki*, I, 77.

c11.79

⁵⁶ Sachau, *Alberuni's India*, I, p. 171.

c11.80

It may be noted that Megasthenes and Hiuen Tsang were minute observers of the customs and manners of the people and recorded their observations. Alberuni, an Arab full of Semitic sympathies would not have missed to record the Semitic origin of the Brāhmī script, if he had got even the slightest hint about it. On the other hand he says that the script was invented in India itself. His statement is quite in conformity with the literary traditions about the origin of the script preserved in the country.

c11.81

Thus all the evidence, traditional, literary, circumstantial and inferential points to the great antiquity of the art of writing in India from the remotest past and its indigenous evolution. It is, however, unfortunate that such antiquity has not been substantiated by the discovery of specimens of writing going back to such antiquity. In fact the earliest datable records so far discovered are the royal edicts of the Mauryan Emperor Aśoka (273-36 B.C).

c11.82

It is true that there are some records and coins bearing writing assigned to the pre-Aśoka period, *viz.*, fourth and fifth centuries B. C. But scholars are not unanimous in their dating of these records, some of them bringing down such writing to as late as the first and second centuries B. C., in the post-Aśoka period. The following are the important ones among them :

i. The Eran Coin Legend

c11.t1.1

A coin with a legend has been found at Eran near Jabbalpore.⁵⁷ The legend is believed to run from right to left. Bühler uses this as one of his strong points for postulating the theory of a Semitic origin for the Brāhmī script. The forms of some of the letters in the legend are considered to be earlier than those in the Aśokan inscriptions.⁵⁸

c11.t1.2

⁵⁷ Cunningham, *Ancient Coins of India*, p. 101.

c11.t1.3

⁵⁸ *Indian Palaeography*, *Ind. Ant.*, XXXIII, App. p. 8.

ii. The Taxila Coin Brāhmī Legend

c11.t2.1

On palaeographic and numismatic grounds it is placed in the fourth century B. C.⁵⁹

c11.t2.2

⁵⁹ Cunningham, *op. cit.*

iii. The Mahasthān Stone Plaque Inscription

c11.t3.1

This inscription was found in the Bogra district in Bengal. It records an endowment to the *Pañcavargīya* (order of the) Buddhist monks. It is also considered to be pre-Aśoka.⁶⁰

c11.t3.2

⁶⁰ *Ep. Ind.*, XXI, p. 85; *Ind. Hist. Quarterly*, 1934, pp. 57. ff. Luders, *List*, No. 931.

iv. The Piprahwa Buddhist Vase Inscription

c11.t4.1

It was discovered at Piprahwa in the Basti District in Uttar Pradesh. It records the dedication of a relic casket containing the portion obtained by the Śākyas out of the mortal remains of the Buddha and it can be assigned to c. 483 B. C., the date of his *nirvāṇa*.⁶¹

c11.t4.2

⁶¹ *Journal of the Royal Asiatic Society*, 1898, pp. 387. ff.

v. The Sohgoura Copper Plate Inscription

c11.t5.1

This was discovered in 1894 in the Gorakpur District in Uttar Pradesh. The inscription which is in Prākṛt runs to four lines and records the provision of grains and fodder during periods of famine.⁶²

c11.t5.2

⁶² *Ep. Ind.*, XXII, p. 2; *Ind. Hist. Quarterly*, X. pp. 54. ff. Luder's, *List*, No. 932.

vi. The Badhi Inscription

c11.t6.1

The inscription was discovered in the Ajmer District. It reads *Vīrāya Bhagavate Catusite Vase* (dedicated to Lord (Mahā) Vīra in his 84th year). It is assigned to 443 B. C. (527—84.)⁶³

c11.t6.2

⁶³ Ojah, *Pracina Lipimala*. The inscription is preserved in the Rajputana Museum, Ajmer.

c11.t6.3

As said above there is no unanimity of opinion among scholars with regard to the dates of these inscriptions.63a

c11.t6.4

63a D. C. Sircar, *Select Inscriptions*, Vol. I, pp. 82-86; also David Diringer, *The Alphabet*, pp. 332 and 338.

vii. The Bhaṭṭiprolu Relic-casket Draviḍi Inscriptions

c11.t7.1

These inscriptions engraved on the relic caskets found at Bhaṭṭiprolu⁶⁴ in the Brāhmī script of the Drāviḍi type are unique. Bühler who has edited them in the *Epigraphia Indica* assigns to them a date slightly earlier than the period of Aśoka.⁶⁵

c11.t7.2

⁶⁴ In the Krishna District of Andhra Pradesh.

c11.t7.3

⁶⁵ *Ep. Ind.*, Vol. II, pp. 323 ff.; D. C. Sircar, *op. cit.*, Vol. I, pp. 215. ff.

c11.t7.4

Scholars who are unwilling or hesitant to assign high antiquity to the art of writing in India mainly base their conclusions on the *argumentum ex silentium*, ignoring some of the literary references to it cited above giving later dates to some others. But one thing they cannot ignore, namely the internal evidence furnished by the inscriptions of Aśoka. They are engraved on rocks, pillars of stone or the walls of caves in the then main scripts of the country, Brāhmī and Kharoṣṭhi, and distributed over a wide area throughout the length and breadth of the country from the Himalayas in the north to Siddhapura, Yerraguḍi and Rājulamandagiri (roughly a little to the south of the Tungabhadrā) in the south and from Girnar in the Kathiawar in the west to Dhauli and Jaugadha in the east. A study of the palaeography of these records indicates the following characteristics.

c11.t7.5

i. The majority of the letters in these inscriptions have different forms, which must have developed under varying conditions in different regions, and periods by different peoples. For instance, the letter *A* vowel alone has more than ten forms (Bühler).

c11.t7.6

ii. At least two main varieties of scripts, northern and southern are noticed, though other regional varieties are also traceable.

c11.t7.7

iii. Again it is seen that each letter has two forms, the monumental (generally angular and carved with due attention being paid to its aesthetic aspect) and the cursive (with a tendency to the formation of curves as in ordinary writing). This would be possible only when the letters had been in use for a long time to enable them to be easily recognised and identified.

c11.t7.8

iv. Some letters show advanced forms indicating that they were changing in their basic forms in the course of their evolution.

c11.t7.9

Bühler says: "The existence of so many local varieties, and of so very numerous cursive forms proves in any case that writing had a long history in Aśoka's time and that the alphabet was then in a state of transition."⁶⁶ Several centuries should have elapsed for the development of the several varieties of scripts to be used in the inscriptions of the period of Aśoka. Aśoka himself says that he was engraving his edict on stone so that it may last long.⁶⁷ This implies that writing was then made on perishable materials also. Further the Mauryan Emperor prescribed a number of religious texts for daily use and recital by monks and laity alike.⁶⁸ These works, it is certain, were not engraved on stone but written on common materials such as leaves, barks or paper which were all frail and perishable. The use of such perishable material also explains why specimens of earlier writing are not found in the country. The few specimens so far recovered are only either on stone or metal which will not easily perish. Further it is the practice to discard old manuscripts after they are copied afresh for future generation. By this process even the script would be changed, only the one prevailing at the time of copying being used.

c11.t7.10

⁶⁶ *Ind. Ant.*, XXXIII, App. p. 8.

c11.t7.11

⁶⁷ Rock Edict II, Khalsi.

c11.t7.12

⁶⁸ Bhabru Edict.

c11.t7.13

The system of education that obtained in ancient India laid emphasis on learning direct from the teachers by mouth personally, thus memorising the text.⁶⁹ The Vedas had to be pronounced correctly and it was believed that incorrect pronunciation would kill the *Yajamān*. This could be achieved only by oral learning direct from a teacher qualified for the same. Even in secular literature memorising of the texts was considered necessary. These do not, however, mean that written texts were not used at all, as will be evident from the view of Bṛhaspati referred to earlier.

c11.t7.14

⁶⁹ Ṛg. Veda, VII, 103. 5.

c11.t7.15

Bothling in his English introduction to the edition of the *Mānavakalpa Sūtra* prepared by Goldstucker says that in his opinion, though writing was not used for the propagation of literature (which was done orally) it was employed at the time of composing new works.⁷⁰ The mention in the *Mahabharata* that Gaṇeśa wrote the text to the dictation of the author Vyāsa as he went on composing is quite in conformity with the above. Roth was strongly of the opinion that the art of writing must have been known very early in India for works like the *Pratiśakhyas* of the Vedas could not be composed without its help.⁷¹ Bühler himself says: "there is nothing to bar the conjecture, repeatedly put forward that even during the Vedic period manuscripts were used as auxiliaries both in oral instruction and on other occasions. An argument in favour of this conjecture, it is now possible to adduce the indisputable fact that the Brāhmī alphabet has been formed by phonologists or grammarians and for scientific use''.⁷²

c11.t7.16

⁷⁰ p. 69.

c11.t7.17

⁷¹ Quoted by Ojah, *Pracinalipimala*, p. 15.

c11.t7.18

⁷² *Ind. Ant.*, XXXIII, App. p. 4.

B. Antiquity of Writing in South India

c11.s1.1

It has been noticed above that the Jaina works, the *Samavāyaṅgasutta* and the *Paṇṇavaṇasutta* mention that eighteen varieties of scripts were prevalent in the country in early times.¹ The Buddhist work *Lalitavistara* refers to sixty-four scripts as having been in use in the country in the days of the Buddha.² Those found in the Jaina list are in Prākṛt while those found in the Buddhist work are in Sanskrit.

c11.s1.2

1. The eighteen scripts are:
1. Bambhi (Brāhmī)
2. Yavanānī or Yavaṇāliya (Greek script)
3. Dosapuriya (Dosapurisa)
4. Kharotthi (Kharoṣṭhi)
5. Pukkharasariyā
6. Bhogavaiyā
7. Pahārāiya (Paharaiyā)
8. Uya-amtarikkhiyā (Uyamitra Kariyā)
9. Akkharapaṭṭiyā (Akkharapuṃṭṭhiyā)
10. Tevanaiyā (Veṇaiyā)
11. Gi (ni) nhaiyā (or ṇhiṇattiyā)
12. Amkalivi (Amkalikkha)
13. Gaṇitalivi (Gaṇiyalivi)
14. Gamdhavva-livi
15. Ādamsalivi (Āyasalivi)
16. Māhesarī (Māhassari)
17. Dāmili (Dravidian or Tamil)
18. Polimdi (Paulindi, belonging to the Pulindas).

c11.s1.3

The sixty-four scripts are:

c11.s1.4

| 1. | Brāhmī | 12. | Draviḍalipi |
| 2. | Kharoṣṭhi | 13. | Kanārilipi |
| 3. | Puṣkarasārī | 14. | Dakṣiṇalipi |
| 4. | Aṅgalipi | 15. | Ugralipi |
| 5. | Vaṅgalipi | 16. | Saṃkhyālipi |
| 6. | Magadhalipi | 17. | Anulomalipī |
| 7. | Maṅgalyalipi | 18. | Ūrdhva-dhanurlipi |
| 8. | Manuṣyalipi | 19. | Daradalipi |
| 9. | Aṅgulīyalipi | 20. | Khasyalipi |
| 10. | Śakārilipi | 21. | Cīnalipi |
| 11. | Brahmavallilipi | 22. | Hūṇalipi |

c11.s1.5

Many of the names contained in the lists are identical. Of the several scripts mentioned in them the following four seem to be important and historically true:

c11.s1.6

1. Brāhmī or Bambhī
2. Kharoṣṭhī or Kharotthī
3. Drāviḍī or Dāmilī
4. Yavanānīya or Yavaṇālīya.

c11.s1.7

In addition to the above, the name of another *Puṣkarasāri* or *Pukkharasāriya* is interesting. Pāṇini’s grammar, Āpastambha’s *Dharmasūtras* and some other works mention *Puṣkarasādi* or *Pauṣkarasādi* as one of the many ancient teachers of the *Dharmaśāstras* and *Vyakaraṇa*. It is possible to infer from this that *Puṣkarasādi* was the alphabet newly invented or an existing one modified by Puṣkarasādi, who was the founder of the school or by a member of the school founded by him.³

c11.s1.8

³ There is also a work called *Puṭkaram* (i.e. *Puṣkaram*) in Tamil verse dealing with certain tantric or yogic practices and signifying mystic values to some of the letters of the alphabet. Very likely the system might have been one dealing with the mystic values of the alphabet.

c11.s1.9

| 23. | Madhyakṣaravistāralipi | 44. | Prakṣepalipi |
| 24. | Puṣpalipi | 45. | Sāgaralipi |
| 25. | Devalipi | 46. | Vajralipi |
| 26. | Nāgalipi | 47. | Lekha-pratilekhalipi |
| 27. | Yakṣalipi | 48. | Anudrutalipi |
| 28. | Gandharvalipi | 49. | Śāstrāvartalipi |
| 29. | Kinnaralipi | 50. | Gaṇāvartalipi |
| 30. | Mahoragalipi | 51. | Utkṣepāvartalipi |
| 31. | Asuralipi | 52. | Vikṣepāvartalipi |
| 32. | Garuḍalipi | 53. | Pādalikhitalipi |
| 33. | Mṛgacakralipi | 54. | Dviruttarapada-sandhilikhitalipi |
| 34. | Cākralipi | 55. | Daśottarapada-sandhilikhitalipi |
| 35. | Vāyumarulipi | 56. | Adhyāhāriṇīlipi |
| 36. | Bhaumadevalipi | 57. | Sarvarutsamgrahaṇīlipi |
| 37. | Antarikṣalipi | 58. | Vidyānulomalipi |
| 38. | Uttarakurudvīpalipi | 59. | Vimiśritalipi |
| 39. | Uparagauḍalipi | 60. | Ṛṣitapastaptalipi |
| 40. | Pūrvavidehalipi | 61. | Dharaṇīprekṣaṇalipi |
| 41. | Utkṣepalipi | 62. | Sarvauṣadhaniṣyandalipi |
| 42. | Nikṣepalipi | 63. | Sarvasārasaṃgrahaṇīlipi |
| 43. | Vikṣepalipi | 64. | Sarvabhūtarudgrahaṇīlipi |

c11.s1.10

Of the four scripts mentioned above *Yavanānīya* is evidently the script of the Yavanas or Greeks and identical with *Yavanānī* mentioned by Pāṇini and explained as such by his commentator Patañjali.⁴ Kharosthī is the name of the script written from right to left and used in the north-western parts of India (more or less in the region covered by modern Pakistan) before the fourth century A.D. after which it disappeared completely. Brāhmī is the name of the script which was prevalent practically throughout India including the region where the Kharoṣṭhī was used simultaneously and running from left to right. All the modern scripts of India are derivatives of this script by evolution. The inscriptions of Aśoka are the best examples of this variety.

c11.s1.11

⁴ 4-1-49. It is explained by Kātyāyana as *Yavanalipyām*. Patañjali says: *Yavanalipyāmiti vaktavyam Yavanānī lipiḥ*.

c11.s1.12

The remaining one, the *Draviḍī* or *Damili* script should, as suggested by its very nomenclature, be the script used for writing the early Dravidian language. The existing Tamil script as well as the one that has disappeared, the Vaṭṭeḷuttu, used for writing the Tamil language in the inscriptions are themselves derivatives of the Brāhmī script. No specimens of any other independent script which can be shown as having been used for writing the Tamil language once upon a time and since disappeared have yet been recovered. Dr. Bühler thinks that the variety of the Brāhmī script incised on the relic caskets recovered from Bhaṭṭiprolu in the Krishna District in the Andhra Pradesh showing peculiarities not met with anywhere else represents the *Draviḍī* type. Before examining this claim and the special features exhibited by this type, it is necessary to know what the Tamil sources themselves say about the art of writing.

c11.s1.13

Tamil is a very old language claiming great antiquity. Even during the earliest known period of the history of South India we find that it was a highly developed language with its grammar fully worked to the minutest detail. But unfortunately the earliest extant literature is not much, though there are indications to show that there was a vast literature, stray references to which are found in later works. The extant works are principally some anthologies of poems by different authors. It is not clear whether these poems were stray pieces or parts of bigger works, the names and nature of which are unknown. There are also the twin epics, the *Śilappadikāram* and the *Maṇimēkalai* and also the celebrated *Tirukkuṟaḷ*. Scholars are not agreed about the age of these works, some holding that these works collectively known as the Śangam works are assignable to the first few centuries of the Christian era, a few dating them even earlier in the pre-Christian era, and some others being inclined to push the limit to a later period, *i.e.*, the fifth or sixth century A. D. All of them recognise that there are certain poems, as in the collection of the *Puṟanāṉūṟu*, which prove that the extant Śangam works clearly bear evidence of the impact of Sanskrit literature and also of the culture associated with it, *i.e.*, the Vedic and Āryan culture.

c11.s1.14

The early literary works in Tamil contain references to the art of writing in South India. The *Tirukkuṟaḷ* opens with the couplet⁵ :

c11.s1.15

⁵ I. 1.

c11.s1.16

*agaramudaleḻuttellām adi*
*bhagavan mudarrē yulagu*

c11.s1.17

"All the letters (of the alphabet) begin from the letter *a* (*agaram*); the word begins from the Lord, the first creator, Ādi (Bhagavān).

c11.s1.18

Another couplet compares the two, the letters and the numerals (writing and arithmetic), to the two eyes of a man.⁶

c11.s1.19

⁶ In the *Śilappadikāram* it is mentioned that the articles of merchandise imported at the port of Kāvērippūmpaṭṭiṉam were marked with *Kaṇṇeḻuttu* : be noted that though *Kaṇṇeḻuttu* may be strained to be taken to be visual writing, even alphabetic writing also may be taken to be visual writing. The *Jīvakacintāmaṇi*, admittedly a late work, refers to a letter written in the *Karanteḻuttu* (X. 1667). This has been taken by some to mean some sort of secret writing or *gūḍhalēkha*. But from the gloss supplied by Naccinārkkiniar it does not appear to be a separate kind of letter or alphabetic system, not even a secret code or cipher system. Naccinārkkiniyar says:

c11.s1.20

*eṇṇenbav ēṉai yeḻuttenbav ivviraṇḍum*
*kaṇṇeṉba vāḻum uyirkku.*

c11.s1.21

*irupadiṉāyiraṉ kaṇṇeḻuttup paḍuttana*
*kaipuṉai s'akaṭamum.* (XXVI. 135-36.)

c11.s1.22

In another place royal scribes are called *Kaṇṇeḻuttāḷar*

c11.s1.23

*kaṇṇeḻuttāḷar kāval vēndan*
*maṇṇuḍai muḍangalam maṇṇavarkkaḷittangu.* (XXVI. 170-71.)

c11.s1.24

M. Raghava Ayyangar explains *Kaṇṇeḻuttu* as visual writing and suggests that it was some form of picture writing (*Ārāiccittoguti*, pp. 140-41). More probably it means only superscription. The *Aṟumpadauraikkārar* says that the *Kaṇṇeḻuttāḷar* were the *Tirumuga-meḻuduvōr*, those who superscribed addresses on letters. It may also The *Tolkāppiyam* is the earliest of the extant Tamil grammatical works. According to tradition it was composed by Tolkāppiyar, one of the disciples of Agattiyar, who is considered to be the same as Agastya of the North Indian tradition and associated with the spread of Aryan culture in the south and the spread of Indian culture in foreign lands, particularly in South-East Asia. Agastya is said to have learnt the Tamil language from God Śiva himself and taught it to twelve of his disciples of whom Tolkāppiyar was one, the other eleven being Ataṅ-kōṭṭāśān, Turāliṅgan, Śempūtcei, Vaiyāpikan, Vāippiyan, Panampāranār, Kaḻāramban, Avinalan, Kākkaipāḍiniyan, Naṟrattan and Vāmanan. Of them Panampāranār has written the *pāyiram* or the introductory stanza (preface) to the *Tolkāppiyam*. It is said therein that the *Tolkāppiyam* was written by Tolkāppiyanār after a careful study of the earlier treatises on Tamil grammar on the model of *Aindram* (*Aindiram niṟainda*) dealing with the Tamil language current both in literature and usage in the Tamil country extending from Tirupati in the north to Cape Comorin in the south, and that it was first read in the court of a Pāṇḍya king,⁸ for recognition in the immediate presence of the grammarian Ataṅ-kōṭṭāśān.⁹

c11.s1.25

⁸ Naccinārkkiniyar, says that the king was Nilantarutiruvir Pāṇḍya (*Tolkāppiyam*, *Naccinārkkiniyar Eḻuttadikāram Pāyira uṟai*, p. 9.) He was variously known as Vaḍimba'ambaninṟa Pāṇḍya, Neḍiyōn, Pāṇḍyan mākīrtti etc. (See T. V. S. Pandarattar, *Pāṇḍiyar Varalāṟu*, p. 5.)

c11.s1.26

⁹ Mayilainātar one of the commentators on the *Naṉṉūl* says that Avinayiṉar, evidently another of the twelve disciples of Agastya, wrote a treatise on Tamil grammar and that it was commented upon by Rāja pavittira Pallavadaraiyan. But neither the treatise nor the commentary is available now. The names of the co disciples are known to us only from stray references found in the works of others.

c11.s1.27

*eḻuttai karantu eḻutinān* . . . . .
*avveḻuttarindālum piṟar vāśilādapaḍi eḻudinār.*

c11.s1.28

The *Tolkāppiyam* is a compact and methodical treatise. Apart from the statement by Paṇampāraṉār in the *pāyiram* that the author had studied the earlier treatises (*muṉnūl kaṇḍu*), there are also internal evidences in several places in the work itself indicating the existence of previous works on the subject. Some of the *sūtras* have close parallels (which may be taken even as translations from one or the other) in the (Vedic) *Prātiśākhyas*, the *Nirukta* of Yāska, the *Śikṣa* and the *Aṣṭādhyāyi* of Pāṇini or those of his predecessors in Sanskrit. Similarly there are strong differences also. It is said that when the *Aṣṭādhyāyi* gained popularity, it ousted all the eight treatises on Sanskrit grammar. Paṇampāraṉār says that Tolkāppiyar was well versed in *Aindram* (*Aindiram niṟainda Tolkāppiyam*). It is considered that the *Aindra* was one of the eight schools of Sanskrit grammar, ousted by Pāṇini. But curiously enough, Pāṇini, who mentions several previous authors, nowhere mentions Indra as one of them. The only reference which connects Indra with Sanskrit grammar is found in Patañjali's *Mahābhāṣya* where it is said that he studied Sanskrit grammar word by word under Bṛhaspati for one thousand celestial years and was not able to complete it.¹⁰ There is striking correspondence between the *Kātantra* and the *Tolkāppiyam* in respect of arrangement of topics and use of technical terms such as *vibhakti* (*vēṟṟumai*), *dīrgha* (*neḍil*), *hrasva* (*kuṟil*) etc. From this Dr. Burnell infers that the *Tolkāppiyam* follows the *Kātantra* and the *Prātiśākhyas*.¹¹ But it must be noted that the *Kātantra* grammar is associated with Śarvavarman, the minister of the Śātavāhana king mentioned in the *Bṛhat-kathā* and said to belong to the first century after Christ.¹² If really Tolkāppiyar had followed the *Kātantra* school in his grammar, certainly Paṉampāraṉār would have indicated that as *Kātantiram niṟainda* and not as *Aindiram niṟainda*.

c11.s1.29

¹⁰ *Bṛhaspatiḥ Indrāya divyaṁ varṣa-sahasraṁ pratipadoktānāṁ śabdānāṁ śabdapārāyaṇaṁ pravacanāntaṁ jagāma*.

c11.s1.30

¹¹ *Aindra school of Sanskrit Grammarians*, p. 8.

c11.s1.31

¹² *Systems of Sanskrit Grammar*, p. 83.

c11.s1.32

The absence of any reference associating Indra with any grammatical work has led some scholars to doubt the correctness of the statement. But in Tamil there are some clear indications of such association. Sēnāvarayar in his commentary on *Sūtra* 75 of the *Śolladikāram* of the *Tolkāppiam* quotes the following:

c11.s1.33

*ēliyiṉ muraiya tetirmuka vēṟṟumai*
*vēṟena viḷambāṉ peyaratu vikāramen*
*ṟōdiya pulavaṉā mulaṉoru vagaiyal*
*intira neṭṭāṉ vēṟṟumai yeṉṟaṉaṉ.*

c11.s1.34

This is also quoted by Śankara Namaśśivāya Pulavar in his commentary on the *Nannūl*. According to this there was also another scholar who had stated that there were only seven cases, taking the vocative *viḷivēṟṟumai* case as only a variant of the nominative case (*peyar alladu mudal vēṟṟumai*), but that Indra had taken that as the eighth case. All the extant grammars in Tamil, including the *Tolkāppiyam*, speak only of eight cases. In Sanskrit, Pāṇini mentions only seven cases, not treating the vocative case as a separate one. M. Raghava Ayyangar thinks that the scholar mentioned in the above quotation and who recognised only seven cases should be taken to be Pāṇini.¹³ The above statement is ascribed to Agastya. But the language and the reference to the vocative case as the eighth case, even though Tolkāppiyar mentions it as *viḷivēṟṟumai*, do not indicate such an early date. Anyway the reference to Indra as the author of a grammar is interesting.¹⁴

c11.s1.35

¹³ *Ārāiccit toguti*, p. 84.

c11.s1.36

¹⁴ *Mudal vēṟṟumai* and *eṭṭām vēṟṟumai* are considered by Dr. P. S. Subrahmanya Sastri as late terms. (See his *History of Grammatical Theories*, p. 109.)

c11.s1.37

It is generally said that the *Tolkāppiyam* is followed by the Śaṅgam works, with regard to grammatical usages. But we find therein many usages which are not sanctioned by the rules contained in the *Tolkāppiyam* and even prohibited. We also find a large number of usages sanctioned by the *Tolkāppiyam* fallen out of use and some even viewed with disfavour. There are a few words which were current during the days of the *Tolkāppiyam* but subsequently fell out of use. Among them were, for instance, *mūḍu* and *kaḍama* which meant the female of the sheep or goat.¹⁵

c11.s1.38

¹⁵ *Mūḍum kaḍamaiyum yaḍalapeṟa* (*Tolk. Poruḷadikāram, Sūtra*, 619.)

c11.s1.39

We also find some usages in the *Tolkāppiyam*, which even in the days of its author, were becoming rare and which, though not proper, should be taken as correct on account of their long usage, eg., *āṇkutirai* (male horse) being called *sēval*.¹⁶

c11.s1.40

¹⁶ See M. Raghava Ayyangar "*Tolkappiyappaḻayavaḻakkugaḷ*" in his *Āraiccit toguti*.

c11.s1.41

These would indicate that some time, obviously a few centuries, must have elapsed between the period of the *Tolkāppiyam* and of the Śaṅgam works. According to tradition the *Tolkāppiyam* was written during the period of the second Śaṅgam (*iḍaiccangam*), while the majority, if not the entire body of the works known as the Śaṅgam works, belong to the third (*kaḍaiccangam*). Considering the scholastic orthodoxy and conservatism of the scholars one may not be far wrong if the period of the *Tolkāppiyam* is fixed somewhere about the second half of the first millennium B.C. if not earlier.

c11.s1.42

The grammatical works do not generally make any reference to written forms of the alphabet. The same is the case with the Tamil grammatical works which simply state that all letters are of the age-old forms (of distant past).¹⁷ But curiously enough the *Tolkāppiyam*, the earliest Tamil grammar, refers incidentally to the written forms of certain letters. Among them a few deserve mention.

c11.s1.43

¹⁷ *Nannūl*, *Eḻuttadikāram*, *Uruvam*, 98.

c11.s1.44

*The addition of dots*: Tolkāppiyar says that all consonants have dots added to them. (*meyyin iyaṟkai puḷḷiyōṭu nilaiyal*).¹⁸ In fact he refers to the consonants in some places merely by the appellation of *puḷḷi* or dot (eg.), *ya ra ḻa veṇṇumpuḷḷi*: the consonants *y r* and *ḻ*; *ta ṟa la ḷa veṇṇumpuḷḷi*: the consonants *ṭa ṟa la* and *ḷa*.¹⁹ Again he states that *e* and *o* have dots added to them (*ekara okarat tiyaṟkaiyum aṟṟē*: The nature of *e* and *o* are the same.²⁰ This follows the *sūtra* relating to the dots for consonants mentioned above. Thus the *nature* referred to in this *sūtra* is the addition of a dot mentioned in the previous *sūtra*.

c11.s1.45

¹⁸ *Eḻuttadikāram*, *Sūtra*, 15.

c11.s1.46

¹⁹ *Sūtras*, 29, 23 etc.

c11.s1.47

²⁰ *Sūtra*, 16.

c11.s1.48

Tolkāppiyar also says that the further shortened *m* has a dot within (the body of its form): *makarak-kuṟukkam*.²¹

c11.s1.49

²¹ *Sūtra*, 14.

c11.s1.50

He further says that the quantity of *m* is shortened to quarter of a *mātra* (time taken for one wink of the eyes or one snap of the fingers)²² when it follows some consonants. And in another *sūtra* he says *uṭpeṟupuḷḷi yuruvāgummē*²³, that the symbol of *m* is with a dot within. It follows from this that the pure consonant *m* which is half a *mātra* has already a dot, and to distinguish the shortened *m* from the consonant, another dot is added and placed within the form, as there is already one dot by the side. This *sūtra* incidentally conveys the idea that the dots were regularly placed outside the symbols as is generally found in inscriptions.²⁴

c11.s1.51

²² *Sūtras*, 7 and 13.

c11.s1.52

²³ *Sūtra*, 14.

c11.s1.53

²⁴ Naccināṟkkiniyar says that the internal dot is placed not inside *makarakkuṟukkam* but *makaram*. viz. *mi* (மி) would be written as (௴). But it seems that Naccinārkkiniyar is wrong for the context in which the *sūtra* *uṭpeṟupuḷḷi uruvāgum mē* occurs shows that the author means only *makarakuṟukkam* and not *makaram*.

c11.s1.54

According to the *Tolkāppiyam* the Tamil alphabet consists of thirty primary sounds (twelve vowels and eighteen consonants) and three secondary ones. This is mentioned in the first *sūtra*. The secondary ones are explained in the next *sūtra*. These are *kuṟṟiyalukaram*, further shortened *u* *kuṟṟiyalikaram* further shortened *i* and *āytam* ∴. Tolkāppiyanār uses the term *muppaṟ puḷḷi*, in this connection.²⁵

c11.s1.55

²⁵ In the second *sūtra* of *nūṇmarapu-eḻuttadikāram* (*Tolkāppiyam*) *muppaṟpuḷḷi* refers only to the *āytaeḻuttu* which is written in the form of three dots ∴. Further *puḷḷi* refers only to *meyyeḻuttu*. Evidently *kuṟṟiyalugaram* and *āyta* are not *meyyeḻuttu*, (*Sūtra*, 2).

c11.s1.56

Tolkāppiyar again says that the symbols of all consonants without dots represent the vowel consonants with the inherent *a* and that the symbols for the vowel consonants in respect of other vowels (i.e., the vowels other than *a*) are different (having additional symbols).²⁶

c11.s1.57

²⁶ See M. Raghava Ayyangar's article on '*Tolkappiyanārum puḷḷi eḻuttukkaḷum*' in his *Ārāiccittoguti*.

c11.s1.58

It may now be examined how far the grammatical rules laid down in the *Tolkāppiyam* were followed in actual writing in the Tamil country in the early centuries of the Christian era. The script, language and grammar followed with regard to Tamil and Prākṛt in the period give us some idea about it.

c11.s1.59

It has been seen earlier that the Buddhist and Jaina literatures mention the names of several scripts prevalent in the country in early days and that the *Draviḍi* or *Damila* is one of them. Bühler thinks that it "is partly an independent variety of the Brāhmī which recently has become known through the relic vessels from the *stūpa* at Bhaṭṭiprolu in the Krishna District".²⁷ The same scholar while editing the inscriptions incised on these relic caskets points out that "the characters which mostly resemble those of Asoka's inscriptions" show also "peculiarities met with nowhere else".²⁸ Among them particular mention has been made of the notation of the medial and final vowels, which according to him show two remarkable peculiarities:—

c11.s1.60

²⁷ *Indian Antiquary*, Vol. XXXIII, (1904), App. p. 2.

c11.s1.61

²⁸ *Ep. Ind.*, Vol. II, p. 323.

c11.s1.62

"*a*) The short *a* is invariably marked by the horizontal stroke to the right of the consonant which denotes long *ā* in the Mauryan alphabet except when an *anusvāra* follows. The latter limitation is probably due to the circumstance that the *anusvāra* was considered equal to *aṁ* in which form it is invariably given in the native lists of *mātṛkas* or alphabets.

c11.s1.63

*b*) The long *ā* is usually marked by a horizontal stroke and a vertical hanging down from its end".

c11.s1.64

He also adds that "in other respects the notation of the medial vowels mainly agrees with that used in Aśoka's inscriptions".²⁹

c11.s1.65

²⁹ *Ibid.*, p. 324.

c11.s1.66

In the Aśokan alphabet, or as a matter of fact in all the scripts in the north, the symbol of a consonant (without the addition of any other sign for the medial vowel) has the short *a* inherent; and a horizontal stroke to the right at the top is added to denote *a* long. Thus in Aśokan Brāhmī + is to be read as *ka* with the short inherent *a* and [Figure:] as *ka* with the medial vowel *ā* long. But in the alphapet of the Bhaṭṭiprolu records the latter form represents the *ka* with the inherent short *a* (and not the long one as elsewhere); and another symbol, a vertical stroke hanging at the right end is added to indicate the letter *a* long. Thus *kā* is written as [Figure:]. So in the scheme of writing found in these records, a consonant without the addition of any symbol of a medial vowel would denote a pure or vowel-less consonant without the inherent *a* short, though no such usage is found there, while in the Aśokan and other alphabets such a pure consonant is not usually written as a separate character, but as a joint character (*dvandvākṣara*) with the succeeding vowelled consonant.³⁰ Bühler thinks that "the device is no doubt of later origin and has been invented in order to avoid the necessity for ligatures".³¹

c11.s1.67

³⁰ No pure or vowelless consonant has been found to occur in the Bhaṭṭiprōlu records and this is probably due to the peculiar nature of the Prākṛt language used in these records.

c11.s1.68

³¹ *Ind. Ant.*, *op. cit.*, p. 15.

c11.s1.69

The alphabet used in the Brāhmī inscriptions discovered in the Tamil country follows the above scheme of Bhaṭṭiprolu records. This becomes clear from the way in which symbols for the medial vowel *ā* long are added to the consonants in these records. First of all is to be noted that no ligature or *dvandvākṣara* is found in these Brāhmī inscriptions. And in the addition of symbols for the medial vowel of *ā* long two methods have been adopted. While a majority of them have the usual horizontal stroke added to the right at the top, there are also a few having two such symbols added. These are found as two parallel horizontal strokes to the right at the top, instead of a horizontal and a vertical one as at Bhaṭṭiprolu.³²

c11.s1.70

³² Evidently the scheme adopted in the Brāhmī inscriptions in the Tamil country is an improvement of the method adopted in the Bhaṭṭiprolu records and as such these records have to be assigned to a period later than those of Bhaṭṭiprolu.

c11.s1.71

It will thus be evident that even in these records the same principle followed in the Bhaṭṭiprolu inscriptions has been adopted. And accordingly, the consonants without the addition of any symbol should be considered as a pure or vowel less consonant (without the inherent short *a*); consonants with the addition of a single symbol for the medial vowel denote those with the *a* short or the inherent *a* and those with the addition of two such symbols denote vowelled consonants with the *ā* long.

c11.s1.72

This method of notation of the medial vowels *a* short and *ā* long is peculiar to the Brāhmī script of South India and not found elsewhere. As surmised by Bühler this was obviously to avoid ligatures in the expression of pure consonants.

c11.s1.73

As said earlier, the *Tolkāppiyam* specifically lays down that the written form of the character of a pure or vowel-less consonant will have a dot added to it.³³ This principle is followed in writing Tamil even to this day.³⁴ Thus the pure consonant *k* without the inherent *a* should, according to the *Tolkāppiyam*, be written in the Brāhmī of those days as ⨥ But the dot above the letter which has not been unknown to the Brāhmī script will denote entirely a different value in the Aśokan alphabet. The dot represents the *anusvāra* and therefore this symbol has to be read as *kam*.

c11.s1.74

³³ See *ante* p. 117.

c11.s1.75

³⁴ In epigraphy this is observed only on rare occasions especially in the records of the earlier period. The *puḷḷi* or *virāma* is generally conspicuous by its absence, more so in the inscriptions of later periods.

c11.s1.76

One other peculiarity is also seen occasionally in the Brāhmī inscriptions found in the Tamil country. Sometimes vowelled consonants are not written as one single letter, ie., consonants with the addition of symbols indicative of medial vowels, but as two separate letters with the pure consonants and the basic vowel forms side by side. Thus to denote *yu* we have the symbol for *ya* first and the vowel *u* as 𑀬 𑀉 instead of the usual form 𑀬𑀼 with the medial *u* added to the bottom. Similarly we find *nu* written as 𑀦 𑀉 with the two symbols of *na* and *u*, *ta* written as 𑀢 𑀅 with the symbols of *ta* and *a* etc. These instances strengthen the principle enunciated above that in this scheme the consonant letter without the addition of any vowel sign is to be vowel-less consonant.

First 191 of 251 passages. Read on

Linked texts (60)