Pramāṇa · Search · About · API

Early South Indian Palaeography c10: Chapter II: Origin of Writing in India

Early South Indian Palaeography c10

Language
English

Contents

Chapter II: Origin of Writing in India

c10.1

More than seventy years ago R. N. Cust, the then Honorary Secretary of the Royal Asiatic Society, London, published an article ‘on the origin of the Indian Alphabet’ in the Journal of the Society.¹ But he was not the first to attempt to solve the mystery of the origin of the Indian alphabet. Earlier still many others had paid their best attention to the problem, *i.e.*, even from the beginning of the last century. But the attempt of Cust reviewing the position as it then stood and formulating his own conclusions came in just when in India too regular and systematic survey of the archaeological and epigraphical material available had been begun. Since then many scholars have studied the subject from various angles and produced a vast literature. Still, there has been no unanimous opinion on the origin of scripts in India. Even on the Brāhmī script, the earliest known in India, there exists to-day much difference of opinion among scholars.

c10.2

¹ *J. R. A. S.* (N.S.), XVI (1884), pp. 325-359.

c10.3

Though Sir William Jones hazarded his theory on the origin of the Indian alphabet as early as 1806 and was followed in the attempt by some others, yet it cannot be said that the question received any importance or much attention till the Brāhmī inscriptions of Aśoka were deciphered by James Prinsep in 1837.² The study of his decipherment of the script is itself very interesting and reads very much like fiction.

c10.4

² *J. A. S. B.* III, pp. 7 and 485. (Plate 5.).

c10.5

During this long period of over a century and a half many theories, some of them very fantastic, have been suggested. A few of them still hold the field, while the majority of them have received no support at all.

c10.6

Theories concerning the origin of the Brāhmī script, varied as they are, may be conveniently divided into two main groups: one ascribing a foreign origin to the Brāhmī script *i.e.*, that it was borrowed from some foreign source and the other claiming an indigenous origin for it, *i.e.*, the script was invented and developed in India itself.

c10.7

Even among the protagonists of the theory of a foreign origin for the script there are wide divergences of opinion as to the country from which and the channel through which India borrowed the script. These different theories about its foreign origin may be reviewed here.

Greek Origin

c10.s1.1

James Prinsep was the first scholar who successfully unravelled the mysteries of the Brāhmī script. He was inclined to ascribe the alphabet of the Aśokan inscriptions to the Greek source. In this view he was followed by Otfried Mueller and some time after even by Èmile Senart.³ The same view also appears to have been shared by Raoul de Rochette, Goblet d'Alviella and others.⁴ Prinsep and Mueller seem to have based their view on the guess hazarded by Wilson that "*Aśoka's* Buddhists derived their letters from Greek or Phoenician models", taking the former of the alternatives suggested as their source.⁵ It may be true that there are resemblances between the Greek and the earliest Brāhmī characters. But the theory of Greek origin is beset with formidable difficulties relating to both chronology and philology. These have never been successfully answered. It is fairly certain that the Brāhmī script was used in India even before the days of Aśoka. Taking into consideration the various forms (recensions or varieties) of the Brāhmī script noticed in the inscriptions of Aśoka, it will not be incorrect to conclude that it should have been in vogue in the country for a long time to attain such variations by natural evolution. The Indians came into direct contact with the Greeks only long after their acquaintance with other peoples using alphabetic writing. If it was a question of borrowing they would have borrowed from other alphabetic systems and not the Greek which was itself a borrowed one. Diringer also says that "the main improvement of the Greek alphabet on the Semitic was the introduction of vowels, while the chief weakness of the Indian scripts is their unsatisfactory solution of vocalization."⁶ This would not have happened if it had been the result of any borrowing. This apart, as regards the resemblance between the Greek and Brāhmī characters, it actually seems to be the other way. There is no doubt that the Greek alphabet was deeply indebted to the Phoenician characters. This does not necessarily suggest a Phoenician origin for the Brāhmī script for it is even suggested by some scholars that the Phoenicians themselves were of Indian origin (the Vedic Paṇi) who very probably carried with them the art of writing and spread it in Western Asia.⁷ The Phoenician alphabet itself has been put forward as the parent of the Brāhmī script and this will be considered separetely below.

c10.s1.2

³ *Ind. Ant.*, Vol. XXXIII, p. 253, Isaac Taylor, *The Alphabet*, Vol. II, p. 304.

c10.s1.3

⁴ Diringer, *op. cit.*, p. 335.

c10.s1.4

⁵ 'The Origin of the Brāhmi Alphabet' by Dr. I. J. S. Taraporewala, *Proceedings of the Fourth All-India Oriental Conference*, p. 625.

c10.s1.5

⁶ Diringer, *op. cit.*, p. 335.

c10.s1.6

⁷ R. B. Pandey, *Indian Palaeography*, Vol. I, p. 40.

c10.s1.7

Hellenistic influence on the invention of the Brāhmī script has been suggested by Joseph Halevy and others.⁸ Halevy maintained that the Brāhmī was derived from a mixture of Aramaic, Kharoṣṭhi and Greek letters in the last quarter of the fourth century B.C. But we know that the Brāhmī had been in existence even before that period. In the words of Dr. Bühler "this theory may be at once eliminated as it does not agree with the literary and palaeographic evidences".⁹ Consequently the theory of the Greek origin of the Brāhmī script had long been given up.

c10.s1.8

⁸ *Journal Asiatique*, (1858), p. 268.

c10.s1.9

⁹ *Indian Palaeography*, (*Ind. Ant.*, Vol. XXXIII, App., p. 9.)

Chinese Origin

c10.s2.1

Terrein de la Couperie thought that the original source of the Brāhmī was to be sought in the Chinese picture-writing. But this suggestion cannot be accepted. As said earlier the Chinese writing is only in the ideographic form. It has never passed that transitional stage, and thus has never reached even the syllabic stage. But the Brāhmī is a fully developed script. Thus this fanciful theory need not be considered seriously. In fact nobody seems to have bestowed much thought to it. The theory was discarded long ago.

Assyrian (Cuneiform) Origin

c10.s3.1

Scholars like Deecke and Canon Isaac Taylor have argued that the origin of the Brāhmī script should be sought in the Assyrian Cuneiform through the South Semitic (which was also the parent of the Sabaen or Himyaritic script) as an intermediate step. Rhys Davids also derived the Brāhmī from the same source. He suggested that the only hypothesis harmonising these discoveries is that the Indian letters were derived from the alphabet of neither the Northern nor the Southern Semites, but from the source to which those, in their turn, owed their origin *i.e.*, the pre-Semitic form of writing used in the Euphrates Valley¹⁰. But "this great authority on Buddhist literature is practically alone in his theory, which is unsubstantiated by any important evidence in its favour".¹¹ He has not shown what, the "pre-Semitic form of writing used in the Euphrates valley is, whether there is any convincingly sufficient resemblance between it and the Brāhmī *lipi*, and at what period approximately it was transplanted into India".¹² South Semitic itself has been considered by some scholars as the original source of the Brāhmī. It will be considered below in its proper place. But to connect the Brāhmī with the Assyrian Cuneiform is too fanciful to be seriously considered.

c10.s3.2

¹⁰ *Buddhist India*, p. 114.

c10.s3.3

¹¹ Diringer, *op.cit.*, 335.

c10.s3.4

¹² D. R. Bhandarkar, 'Origin of Indian Alphabet', *Proceedings of the First All-India Oriental Conference*, p. 311.

Semitic Origin

c10.s4.1

Most of the scholars advocating a foreign origin for the Brāhmī script hold the Semitic characters as its parent. But they are not agreed among themselves as to what branch of the Semitics produced or influenced the Brāhmī. There are at least three sources cited for its origin : (a) Phoenician, (b) South Semitic through the Sabaen of Yemen and (c) North Semetic through the Aramaean of Babylon.

Phoenician Origin

c10.s4.t1.1

Weber posulated the theory that the earliest Phoenician alphaket was the source from which the Brāhmī script was derived and in this he was supported by scholars like Benfey, Jenson and others.¹³ They tried to prove "that about one-third of the Phoenician letters were identical with the earliest forms of the corresponding Brāhmī signs; that another third were somewhat similar and the remaining third can be more or less harmonised." The earlier scholars objected to this theory mainly on the ground that practically all intercourse between Phoenicia and India ceased about 800 B.C. *i.e.*, more than five centuries before the date of the inscriptions of Aśoka. But it is very difficult to maintain this view any longer. This question is closely linked up with the origins of the Phoenicians as a people. The scholars of Tyre always thought, and it was accepted by the Greek historians also, that the Phoenicians reached the east coast of the Mediterranean from the east by sea.¹⁴ Again the Phoenician script itself is considered as a derivative of the Canaanite (North) branch of the Semitic. It is also said: "it seems probable that the Phoenicians had no influence whatever on the origin of the scripts of countries lying to the east of them".¹⁵

c10.s4.t1.2

¹³ Diringer, *op.cit.*, p. 335; Buhler, *Indian Palaeography*, pp. 9-11.

c10.s4.t1.3

¹⁴ Herodotos, II, 44.

c10.s4.t1.4

¹⁵ Diringer, *op. cit.*, p. 335.

South Semitic Origin

c10.s4.t2.1

According to Professor Deecke, Canon Isaac Taylor and in recent years Professor Sethe, the Brāhmī script descended from the South Semitic alphabet. Deecke thought that it was derived from the Assyrian Cuneiform characters through an ancient South Semitic alphabet which was also the parent of the Sabaean or Himyaritic script.¹⁶ Taylor derived it from a lost South Semitic alphabet, the predecessor of the Sabaean.¹⁷ He takes the ancient alphabet of Safa and Sabaea as his sources and gives a table in which the derivations are given.¹⁸ Among the most remarkable of his derivations may be mentioned those of the letter *va* from the Semitic *y* sound, of *gha* from the Semitic *kha*, of *j* from ṣ and of *c* from the deep glottal *qab*. Regarding these derivations Bühler comments as follows: "As the Hindus are very particular, even pedantic in matters connected with phonetics, and as the framers of the Brāhmī *lipi* have been careful with regard to the formation of many derivative signs, duly deriving *dha* from *da*, *pha* from *pa*, *bha* from *ba* and so forth, it seems incredible that they should have no regard for phonetic affinities in utilising the signs they borrowed ".¹⁹

c10.s4.t2.2

¹⁶ *Z. D. M. G.*, 31, pp, 598 ff.

c10.s4.t2.3

¹⁷ *The Alphabet*, Vol. II, pp. 314 ff; This has been restated with some modifications by F. Muller.

c10.s4.t2.4

¹⁸ *The Alphabet*, II, p. 320.

c10.s4.t2.5

¹⁹ *On the Origin of the Indian Brahma Alphabet*, being the Second Revised Edn. of his *Indian Studies* No. III, (1898,) p. 54.

c10.s4.t2.6

Quite apart from the phonetic problems the very real difficulty lies in pointing out how exactly the South Semitic script could have been the parent of Brāhmī. Taylor says: “In comparing the Indian and Sabaean forms it must be borne in mind that no south Semitic inscriptions have yet been discovered of a date sufficiently remote to supply the absolute prototype of the Aśoka letters. Of the inscriptions which accident has preserved, none probably is older than the middle of the second century B.C., *a period later by about a century than the earliest Indian Inscriptions* ", (*Italics ours*) and adds that “it must therefore be remembered that it is only possible to compare sister alphabets from a common but unknown source ”.²⁰

c10.s4.t2.7

²⁰ *The Alphabet*, II, p. 320.

c10.s4.t2.8

Bühler, who subjected the above theory to strong criticism, remarks: “Neither the Sabaean alphabet, nor its perhaps a little more archaic variety, the Libyanian or Thammudaean will serve the purpose, in spite of a general resemblance in the *ductus* and of a special resemblance in two or three letters. The derivations proposed by Deecke and Taylor do not fulfil the absolutely necessary conditions, and it will probably not be possible to obtain satisfactory results, even if all the impossible equations are given up, and the oldest Indian signs in every case are chosen for comparison. It would be necessary to assume that several Sabaean letters such as *Alaph*, *Gimel*, *Zain*, *Teth*, *Phe*, *Qoph* and *Resh*, which show strong modifications of the North-Semitic forms, had been again made similar to their prototypes on being converted by the Hindus into *a*, *ga*, *ja*, *tha*, *pa*, *kha* and *ra*. In other cases it would be impossible to show any connection between the Sabaean and the Indian signs”.²¹

c10.s4.t2.9

²¹ Bühler, *op. cit.*, p. 10.

North Semitic Origin

c10.s4.t3.1

This theory deriving the Brāhmī script from the North Semitic alphabet was originally suggested in 1806 by Jones, in 1811 by Von Seetzen, in 1821 by Kopp, in 1834 by Lepsius and was elaborated in 1856 by Weber and by the end of the last century by Bühler.²² They expressed the view that the Indian alphabet is in no respect an independent invention of the people of India, who, however, *elaborated to a marvellous extent (Italics ours)* a loan which they had received from others and said that the idea of representative vowel and consonant sounds by symbols of a pure alphabetic character was derived from Western Asia beyond any reasonable doubt.22a But they were not able to fix up the exact branch which could be considered as the parent for Brāhmī. It was Bühler, however, that subjected all their theories to severe criticism and finally advocated the north Semitic origin. He says: "These difficulties disappear with the derivation of the Brāhmi from the oldest North Semitic alphabet which shows the same type from Phoenicia to Mesopotamia. The few inadmissible equations which Weber's earlier attempt contains may be easily removed with the help of recently discovered forms".²³ Since then, this theory has found acceptance with many scholars and the other theories appear to have been almost discarded and not seriously considered. For instance, Diringer says: "All historical and cultural evidence is best co-ordinated by the theory which considers the early Aramaic alphabet as the prototype of the Brāhmī script. The acknowledged resemblance of the Brāhmī signs to the Phoenician letters also applies to the early Aramaic letters, while in my opinion there can be no doubt that of all the Semitics, the Aramaean traders were the first who came in direct communication with the Indo-Aryan merchants."²⁴

c10.s4.t3.2

²² *Z. D. M. G.*, Vol. X, pp. 389 ff; *Ind. Skizzen*, pp. 125 ff.

c10.s4.t3.3

22a See also A. H. Dani, *Indian Palaeography*, pp. 23-8.

c10.s4.t3.4

²³ Bühler, *op.cit.*, p. 10.

c10.s4.t3.5

²⁴ Diringer, *op.cit.*, p. 336.

c10.s4.t3.6

But this theory of North Semitic origin of the Brāhmī script has not been accepted by some scholars, both Indian and foreign. The main arguments in favour of the Semitic origin of the Brāhmī script are:

c10.s4.t3.7

1. The absence of the specimens of writing before the fifth century B.C. in India.

c10.s4.t3.8

2. The resemblance between the Semitic and the Brāhmī characters and 3. The supposed original direction of the Brāhmī from right to left.

c10.s4.t3.9

The argument has been slightly amended as follows as a result of the excavations at a number of Harappan sites which have disclosed the earliest highly developed culture in India namely the 'Indus Valley Culture'. But it is felt that "the early Indian writing was pictographic; no alphabetic writing can be derived from pictographs; the earliest known alphabets are Semitic; hence, the Brāhmī (semi-alphabetic) could be derived only from the Semitic sources."

c10.s4.t3.10

The resemblance between the Semitic and the Brāhmī characters is the principal reason for attributing a Semitic origin to the Brāhmī.

c10.s4.t3.11

Before proceeding further it is necessary to know something about the North Semitic alphabet. It consists of twenty-two letters or symbols, which correspond roughly to the first twenty-two letters of its descendant, the Greek alphabet. Their names as preserved in Hebrew are *‘aleph, beth, gimel, daleth, he, waw, zayin, kheth, teth, yod, kaph, lamed, mem, nun, samekh, ‘ayin, pe, sade, qoph, resh, shin, taw*. Most of these names end with a consonant, though some end with a vowel. It is said that these twenty-two letters represent only eighteen sounds. It is also said that the early distinction between some letters in the North Semitic alphabet (for example between *samekh* and *shin*) was lost at a later stage. Some scholars attribute this fact to the use in later times of Aramaic in which for example *samekh* displaced *shin*. It has several characters for one sound. It does not make any distinction between the long and short vowels. Even after the introduction of the vowels on account of the influence of the Greeks the consonants and the vowels of the Semitic alphabet cannot coalesce; rather the vowels are written after the consonants. Phonetically the Semitic alphabets are a jumble rather than a system. For instance, just after *a* (*alef*), which is guttural, we have *b* (*beth*) which is labial. The Semitic alphabet, it must not be forgotten, is one of mere consonants, and originally it did not contain vowels. The Semitics could, if necessary, dispense with vowel signs. The Hebrew even now writes *k-t-b* to indicate any word having a meaning connected with 'writing', although the word would never be read *ktb*, but *katab* (he wrote), *koteb* (he is writing), *ektob* (I shall write), and so forth, according to the sense of the sentence; whereas in an Indo-Āryan tongue, a word written with mere consonants would have many meanings or no meaning at all (*e.g.*, in English *c-t* could mean 'cat', 'cut', 'cot', 'city', 'cute', 'act', 'acute' or no meaning at all).²⁸

c10.s4.t3.12

²⁵ Pandey, *Indian Palaeography*, Part I, p. 45.

c10.s4.t3.13

²⁶ Diringer, *op.cit.*, p. 218.

c10.s4.t3.14

²⁷ *Ibid.*, p. 219.

c10.s4.t3.15

²⁸ Diringer, *op.cit.*, p. 336.

c10.s4.t3.16

In the table giving the derivation of the Brāhmī script, Bühler maintains that twenty-two letters of the Brāhmī alphabet were derived from the North Semitic alphabet some of them from early Phoenician, a few from Mesa's stone inscription and five from the 'script on the weights from Assyria.' The remaining letters of the Brāhmī were also derived from the borrowed signs by introducing certain devices.

c10.s4.t3.17

Key to the Table of Bühler deriving the Brāhmī script from the North Semitic Alphabet.

c10.s4.t3.18

| S.No. | | | |
| 1. | Aleph | a | ā |
| 2. | Be (Beth) | ba | bha |
| 3. | Gimel | ga | — |
| 4. | Daleth | da | da, ḍa, ḍha |
| 5. | He | ha | — |
| 6. | Waw | va | u, ū, ō |
| 7. | Zayin | ja | jha |
| 8. | Kheth | gha | — |
| 9. | Teth | tha | ṭa, ṭah |
| 10. | Yod | ya | — |
| 11. | Kaph | ka | — |
| 12. | Lamed | la | ḷa |
| 13. | Mem | ma | — |
| 14. | Nun | na | ṇa, na, ṅa |
| 15. | Samekh | ṣa | sa |
| 16. | ‘Ayin | e | i, ī, ai |
| 17. | Pe | pa | pha |
| 18. | Ṣade | ca | cha |
| 19. | Qoph | kha | — |
| 20. | Resh | re | — |
| 21. | Shin | śa | — |
| 22. | Taw | ta | — |

c10.s4.t3.19

An examination of the above table will show that:
1. No less than ten letters, *i.e.*, nearly half the total had to be given ‘intermediate’ forms as Bühler calls them; in other words they are merely hypothetical.
2. Only one Brāhmī letter *ga* may be identical with its supposed proto-type: and
3. Of the rest only five, *viz.*, *a*, *t*, *th*, *l* and *ś* may be reasonably regarded as probable derivatives and even of these the *la* has come through an intermediate form.

c10.s4.t3.20

In addition to the above the following points also should be noted:

c10.s4.t3.21

“In deriving one script from the other, even if similarities could be proved between the shapes of these two scripts evidence would still be lacking that one descended from the other, unless the likeness of the signs belonging to the two systems corresponds with the similarity or identity of their phonetic values. In this connection it may be noted that the Minoan script as well as the Cypriote syllabary contained many signs resembling the early Greek letters, but one was not derived from the other”28a The derivation given by Bühler in the above Table does not satisfy this general principle in many cases. His methods of derivation are difficult of acceptance; and if they are accepted, the Brāhmī characters can be derived not only from the Phoenician and the Aramaic but also from any known character of the world.

c10.s4.t3.22

28a Diringer, *op.cit.*, p. 329.

c10.s4.t3.23

It will be seen that the following signs appearing in both the Semitic and the Brāhmī alphabets indicate different values unconnected with each other.

c10.s4.t3.24

| From | Semitic Phonetic value | Brāhmī Phonetic value |
| I | Zayin | ṇa |
| + | Tan | ka |
| △ | Daleth | e (vowel) |
| ○ | ‘Ayin | tha |
| 𐡋 | Lamed | pa |
| 𐡒 | Qoph | cha |
| 𐡊 | Kaph | a (vowel) |

c10.s4.t3.25

But curiously enough Bühler ignores this similarity. He derives the symbols for these Brāhmī letters from some other Semitic originals; and similarly the above Semitic alphabets have been treated as the parents of some Brāhmī letters, all of them supposedly based on the similarity in the written form.

c10.s4.t3.26

It is generally considered that the Kharoṣṭhi script also was derived from the same North Semitic alphabet. But they present quite a different picture. For a better understanding of the problem, a perusal of the table given by Bühler showing the derivation of the Kharoṣṭhi from the North Semitic script will be necessary.

c10.s4.t3.27

Key to the Table of Bühler deriving the Kharoṣṭhi script from the North Semitic alphabet.

c10.s4.t3.28

| S.No. | Original North Semitic | Sign borrowed for Kharoṣṭhi | Derivative |
| 1. | Aleph | a | i, u, e, o |
| 2. | Be (Beth) | ba | bha |
| 3. | Gimel | ga | gha |
| 4. | Daleth | ḍha | dha, da, ḍha |
| 5. | He | ha | — |
| 6. | Waw | va | — |
| 7. | Zayin | ja | jha |
| 8. | Kheth | śa | — |
| 9. | Teth | — | — |
| 10. | Yod | ya | — |
| 11. | Kaph | ka | — |
| 12. | Lamed²⁹ | la | — |
| 13. | Mem³⁰ | m | ṃ |
| 14. | Nun | na | ṇa, ña |
| 15. | Samekh | sa | — |
| 16. | ʽAyin | — | — |
| 17. | Pe | pa | pha |
| 18. | Ṣade | ca | cha |
| 19. | Qoph | kha | — |
| 20. | Resh | ra | — |
| 21. | Shin³¹ | ṣa | — |
| 22. | Taw | ta | tha, ṭa, ṭha |

c10.s4.t3.29

The table shows that (i) out of the twenty letters borrowed, no less than eight are identical with the Semitic forms, *viz.*, *ca*, *da*, *na*, *pa*, *ba*, *na*, *va*, and *śa*.

c10.s4.t3.30

(ii) nine others, *viz.*, *ka*, *kha*, *ga*, *ja*, *ma*, *ya*, *la*, *ṣa* and *ha* may be reasonably regarded as derived from the Semitic prototypes and

c10.s4.t3.31

(iii) there are no assumed 'intermediate' forms.

c10.s4.t3.32

A comparison of these two tables will show that the Brāhmī script derives from twenty-two letters of the North Semitic alphabet while Kharoṣṭhi only from twenty, the two letters *Teth* and '*Ayin* having been excluded.

c10.s4.t3.33

The derivation of the cerebrals and dentals in these two systems also deserve special notice. The Brāhmī and the Kharoṭṣhi alphabets, for both of which the same parentage is attributed by scholars, differ very much from each other. The marked characteristics of the Kharoṣṭhi script are as follows:

c10.s4.t3.34

1. The letters are generally thin and long, more or less slanting to the right with appendages attached to their upper parts.
2. They are very irregular in shape and size and many forms are confusingly similar.
3. The script is clearly cursive.
4. There are very few looped forms.
5. The script always runs from right to left.

c10.s4.t3.35

On the contrary the following are the chief characteristics of the Brāhmī script as given by Bühler himself:

c10.s4.t3.36

1. The letters stand upright.
2. They are made of regular lines and loops.
3. The letters hang down unlike in Kharoṣṭhi which are top-heavy.
4. Most of the vowel signs are added at the top in the shape of horizontal lines.

c10.s4.t3.37

²⁹ The original Semitic has been turned upside down.

c10.s4.t3.38

³⁰ The derivative represents the *anusvāra*.

c10.s4.t3.39

³¹ The original Semitic has been turned upside down.

c10.s4.t3.40

To these characteristics may be added the following :

c10.s4.t3.41

5. The script is written from left to right 31a.

c10.s4.t3.42

31a The alleged specimens of Brāhmi writing indicating a right to left direction, the *boustrophedon* style, are discussed later on,

c10.s4.t3.43

6. And finally, the script is distinctly monumental and donative.

c10.s4.t3.44

It may be seen, thus, that "it does not appear very probable that two such fundamentally different alphabets could have been derived from the same source, especially when we bear in mind that they were both needed for the same purpose and practically for the same language too. To overcome this difficulty it has been suggested that while the Kharoṣṭhi came overland through Gandhara, the Brāhmī came by sea through Babylon *via* the Persian Gulf". A further suggestion is that one was a clerk's alphabet also used by the merchants while the other was elaborated by the learned pandits.

c10.s4.t3.45

Comparative table of the Derivatives of the Brāhmī and Kharoṣṭhi alphabets as given by Bühler.

c10.s4.t3.46

| Letter | Derivation for Kharoṣṭhi | | Derivation for Brāhmī | |
| i ī | Derivative from | Aleph (a) | Derivative from | ‘Ain (e) |
| u ū | ,, | ,, | ,, | Waw (o) |
| e | ,, | ,, | ,, | ‘Ayin |
| ai | ,, | ,, | ,, | ‘Ayin (e) |
| gh | ,, | Gimel (g) | ,, | Kheth |
| ṭ | ,, | Taw (t) | ,, | Teth (th) |
| ṭh | ,, | ,, | ,, | ,, |
| ḍ | ,, | Daleth (d) | ,, | Deleth (dh) |
| ḍh | ,, | ,, | ,, | ,, |
| th | ,, | Taw (t) | ,, | Teth |
| d | ,, | Daleth | ,, | Daleth (dh) |
| dh | ,, | Daleth (dh) | ,, | ,, |
| ś | ,, | Kheth | ,, | Shin |
| ṣ | ,, | Shin | ,, | Samekh |
| s | ,, | Samekh | ,, | ,, (ṣ) |

c10.s4.t3.47

Scholars postulating the North Semitic source to both the Kharoṣṭhi and the Brāhmī are unable to derive the same letter in both from the same original source. In Kharoṣṭhi all the vowels were derived from the letter *aleph*, the first in the Semitic alphabet. But in the Brāhmī only the first two *i.e.*, *a* both short and long, were derived from the letter *aleph*; the vowel *e* from the deep glottal Semitic letter ‘*ayin* and the vowels *i*, *ī* and *ai* were in turn derived from this letter *e*, while the vowels *i*, *u* and *o* were derived from the Semitic letter *waw*. Similarly the consonant *gha* of the Kharoṣṭhi was derived from the Semitic letter *gimel*, while that of the Brāhmī has been derived from the Semitic *kheth*. Further phonetically the originals and the derivations differ very much. In Sanskrit the vowel *e* is the derivative (*guṇa*) of the vowel *i*; and while it is so, to say that the letter of the vowel *i* was derived from that of *e* would certainly be not acceptable. To say that the accurate and even “pedantic” phoneticians of India evolved the Brāhmī from the Semitic in complete disregard of the phonetics clearly makes a very large demand on our credulity.

c10.s4.t3.48

In the Kharoṣṭhi the letter *ta* was derived from the North Semitic *taw* and the letters *tha*, *ṭa* and *ṭha* were in their turn, from the original derivative *ta*. Similarly the letter *da* was derived from the North Semitic *daleth* and the *dha*, *ḍa* and *ḍha* were in their turn from the derivative *da*. Only the dental unaspirated sounds *ta* and *da* were represented in the North Semitic alphabet. Their aspirated form and the corresponding aspirated and unaspirated forms of the cerebrals as well were derived from the respective dentals. This definitely shows that there was some principle in the derivation.

c10.s4.t3.49

In the Brāhmī also, the letter *ta* was derived from the Semitic *taw*; it had no derivative in its turn. The aspirated form of *tha* was derived from the Semitic *teth* and the corresponding aspirated and unaspirated forms of the cerebral *ṭa* and *ṭha* were in their turn obtained from the derivative *tha*. All the four letters of the Kharoṣṭhi derived from the Semitic *daleth* were derived in the Brāhmī also from the same letter. But the order was different. The aspirated form *dha* was the main derivative and the others, *da*, *ḍa* and *ḍha* were obtained from the main form *dha*. Phonetically it is anamolous and incongruous to derive the unaspirated from the aspirated. The usual process is the reverse, *viz.*, the aspirated is obtained from the unaspirated; and a perusal of the Brāhmī script will also bear out the fact. In the original Semitic both *Teth* and *Deleth* are unaspirates.

c10.s4.t3.50

Further, from a comparison of the Brāhmī forms of the cerebrals and dentals, it would appear that the cerebrals were the main forms and that the dentals are derived from them. This will be very clear in the cases of *ṭha* and *tha*, where the difference is the addition of a dot to the former to distinguish it from the latter. This would indicate that the Semitic *teth* is a later form derived from the Brāhmī. The same will also apply to the Semitic form ‘*Ayin*, which has the same form as the Brāhmī *e*. Very likely these two were introduced in the Semitic through the influence of the Brāhmī. We shall then have to conclude that the Brāhmī far from deriving from the Semitic actually influenced the Semitic at least to a small extent.

c10.s4.t3.51

Bühler recognises the following peculiarities of the Indian alphabet in his attempt to derive the Brāhmī from the North Semitic alphabet and the principles according to which the conversion took place:

c10.s4.t3.52

1. The letters are set up as straight as possible; and with occasional exceptions in the signs of *ṭa*, *ṭha*, and *ba* they are made equal in height.

c10.s4.t3.53

2. The majority consists of vertical lines with appendages mostly to the foot, and occasionally both at the foot and at the top, or rarely in the middle, but there is no instance in which there has been an appendage to the top alone.

c10.s4.t3.54

3. At the top of the letters appear mostly the ends of verticals, less frequently short horizontal strokes, still more rarely curves on the tops of angles opening downwards, and quite exceptionally in *ma* and in one form of *jha* two lines rising upwards. In no case does the top show several angles placed side by side, with a vertical or slanting line hanging down, or a triangle or a circle with a pendent line”.

c10.s4.t3.55

And he deduced the principles of the derivation of the Brāhmī from the North Semitic alphabet on the basis of the following tendencies of the Indian alphabet :

c10.s4.t3.56

1. A certain pedantic formalism ;
2. A desire to frame signs suited for the formation of regular lines ; and
3. An aversion to top heavy characters.

c10.s4.t3.57

He thinks that " the last peculiarity is probably due in part to the circumstance that since early times the Indians made their letters hang down from an imaginary or really drawn upper-line and in part to the introduction of vowel-signs most of which are attached horizontally to the tops of the consonants. Signs with the ends of verticals at the top were, of course, best suited for such a script. Owing to these inclinations and aversions of the Hindus, the heavy tops of many Semitic letters had to be got rid of by turning the signs topsy-turvy or laying them on their sides, by opening the angles, and so forth. Finally, the change in the direction of writing necessitated a further change, inasmuch as the signs had to be turned from the right to the left as in Greek."³²

c10.s4.t3.58

³² Bühler, *op. cit.*, pp. 10-11.

c10.s4.t3.59

This line of argument presupposes the following factors : (1) the Indians were already using some sort of writing, (2) they were familiar with the Semitic script ; (3) their method or system of writing was defective in many ways ; (4) at the same time they found the Semitic alphabet satisfactory, though not fully representing all the required sounds of their (sic. the Indian) language and (5) they adopted the Semitic alphabet making the necessary changes and creating new symbols for the non-Semitic, but Indo-Aryan sounds. For any borrowing, whether it be language, word, or any aspect of culture or anything else by one country or nation from any other it is necessary that (1) the borrowing nation or people are quite familiar with the borrowed material ; and (2) they have realised their utility, superiority and also the necessity of such material for being borrowed. Again when any foreign material is borrowed it is adopted wholesale in the first instance and then modified gradually to suit the local requirements, taste or inclinations. They would not happen in a single day, but would take a long time. Thus if the Brāhmī script had been derived from the Semitic alphabet after making the necessary changes, the Semitic script also must have been in use in this country in its original form. Only after several generations of constant use the form best suited to express a particular sound evolves finally to hold the field. Thus one would naturally expect to find specimens of such writings in this country for the theory of the Semitic origin of the Brāhmī script to be true. But nothing of the kind has been found so far.

c10.s4.t3.60

One other argument of the protagonists of the theory of the Semitic origin is that originally the direction of the Brāhmī script was from right to left. This was based on very meagre and doubtful data. At the time when Bühler wrote in his ‘Indian Studies’ and finally published in his *Indian Palaeography* the specimens of Brāhmī written from right to left were similar to the Brāhmī letters in the legend on the coin discovered at Ērān in the Jubbalpur District considered as pre-Mauryan and assigned to the fourth century B.C.³³ He has also found a few traces in the edicts of Aśoka of the writing from right to left in the forms of some letters, in the *o* of Jaugada and Dhauli versions and in the rare *dha* of Jaugada and Delhi-Sivalik versions. Again he says that in the Drāviḍī type of the Brāhmī found on the relic caskets of Bhaṭṭiprolu, the signs for the three letters, *d*, *dh* and *bh* are in the position of the writing running from right to left. According to him even the Ērān coin “probably dates from the time when the Brāhmī was written both from the right to the left and from the left to the right” and that “the time when the Brāhmī was written *boustrophedon* probably lies somewhat before the Maurya period, since the Aśoka edicts show only a few traces of the writing from right to left.”³⁴ Since then the version of the Minor Rock Edict of Aśoka at Yerraguḍi in the Kurnool district of Andhra Pradesh has been discovered.³⁵ This record contains twenty three lines of writing of which eight, namely lines 2, 4, 6, 9, 11, 13, 14 and 23 are from right to left.

c10.s4.t3.61

³³ Cunningham, *Coins of Ancient India*, pp. 99-102.

c10.s4.t3.62

³⁴ *Ind. Ant.*, Vol. XXXIII, App. p. 8.

c10.s4.t3.63

³⁵ *Arch. Sur. of India*, *Ann. Rep.*, 1928-29, pp. 161-67.

c10.s4.t3.64

Diringer, a strong supporter of the theory of Bühler, says that "there is thus sufficient evidence of the existence of an earlier Brāhmī script written from right to left, followed—as in the development of the early Greek script—by a transitional system of writing in *boustrophedon* style''.³⁶

c10.s4.t3.65

³⁶ Diringer, *op. cit.*, p. 339.

c10.s4.t3.66

However, Diringer's statement cannot be regarded as satisfactory and conclusive in deciding the question concerning the direction of writing of the original Brāhmī. Many scholars have assumed that the peculiarity or the Ērān coin, as suggested by Hultzsch was due to a fault in the matrix from which the coin had been cast.³⁷ Hultzch thinks that it is "a mistake of the engraver of the dye, who like the die sinker in the case of a certain coin of Holkar of the last century may have forgotten that he ought to reverse the legend on the die itself''. Fleet has also pointed out that "we have one instance of such remissness in ancient times in a coin of Rapla Rajuvula, the reverse of which presents a monogram formed of the Greek letters *E* and *y* facing the wrong direction.³⁸ Yet another is in the legend on a bronze stamp for making seals, where the engraver omitted to reverse the syllable *śū*.³⁹ Thus the data relied on by Bühler cannot be a sure indication of the direction of writing. That is why Hultzch and Fleet, two great authorities on Indian epigraphy, especially of South India and the Deccan, did not agree with the conclusions of Bühler.

c10.s4.t3.67

³⁷ *Ind. Ant.*, Vol. XXVI, p. 336.

c10.s4.t3.68

³⁸ Gardner, *Catalogue of Coins of the Greek and Scythian kings of Bactria and India* p. 67, no. 5.

c10.s4.t3.69

³⁹ *J. R. A. S.*, 1901, p. 98, plate No. 9.

c10.s4.t3.70

Similarly the forms of some letters of Aśoka's edicts with a right to left direction should also be attributed to the fault of the engravers. These are not stray ones. A few other examples may be cited. In the Brāhmī label inscription at Kiḻavaḷavu, described and discussed later in the book the letters are incised upside down, thus showing that the engraver had great convenience in working from the upper part of the cave than from the front. But one of the letters in the record *ta* is found in the proper position when seen from the ground, instead of being upside down as the rest. Again the Brāhmī label record at Kuṉṉakkuḍi, also discussed elsewhere in the book, presents another peculiarity. The individual letters are incised in the reverse position, the engraver working from the upper part of the cave, thus appearing upside down when seen from the ground in front, even though the record runs from left to right when seen from the floor. Thus the record through its reflection in a mirror is seen to run from right to left. But it appears to run from left to right to the person standing on the floor only with the letters reversed and upside down.⁴⁰ Among the nearly fifty such Brāhmī labels found so far in the Tamil country this is the solitary record of the kind. This can only be taken as due to the fault or vagary or idiosyncray of the engraver and not as indicating the method of writing from right to left.

c10.s4.t3.71

⁴⁰ See *Journal of the Madras University*, vol. xxx, No. 1. pp. 5-7 where the inscription has been edited by the author.

c10.s4.t3.72

Thus the specimens so far known to us are sporadic and very few in number in comparison with the large number of contemporary inscriptions written from left to right. It is also to be noted that these irregular forms of letters appear only in the records discovered on the boundaries of the Aśokan empire. Even in the manner of writing from left to right, two methods, one proceeding clockwise and the other counter-clockwise appear to have been adopted according to the genius of the people concerned to suit their convenience. It is well known that the adoption of the latter or the counter-clockwise method led to the derivation of the cursive varieties, the Vaṭṭeḷuttu and the Telugu-Kannaḍa scripts. It is thus quite possible that the irregular form of some letters in the Aśoka's edicts in the transposed position was influenced by the second method.

c10.s4.t3.73

Regarding the Yerraguḍi version of the Minor Rock Edict of Aśoka it is true that more than half of the inscription appears to be written in *boustrophedon* or alternating lines. But this style has not been followed correctly; lines 7 and 8 and also from 15 to 22 run from left to right without alternation, while lines 13 and 14 similarly run from right to left without such alternation. This only shows that the *boustrophedon* style was known here during the time of Aśoka. Or it may be that the scribe was merely trying a new experiment. Moreover in the lines running from right to left only the positions of letters are changed and not their forms, which shows that it is a forced and artificial way of writing and that it has no bearing on the origin of the Brāhmī script.

c10.s4.t3.74

Such innovations are also not unknown to Indian epigraphy. The stone inscription referring to Pallava Simhavarman and incised in the Telugu-Kannaḍa script of the 11-12th century A.D. at Amarāvati in the Krishna District,⁴¹ begins at the bottom and proceeds upwards. From this instance one cannot postulate the theory that writing, at one time, commenced at the bottom and proceeded upwards at least in South India, if not in the whole country. At best these can be taken to be only due to the vagaries or the fancies of the engraver.

c10.s4.t3.75

⁴¹ *S. I. I.*, Vol. I. p. 25; *Ep. Ind.*, Vol. XVIII, p. 148.

c10.s4.t3.76

Thus the link discovered by Bühler in the two sets of specimens in the chain of arguments which were to prove the derivation of the Brāhmī alphabets from the Semitic alphabets which were written from right to left appears to be non-existent.

c10.s4.t3.77

The foregoing discussion will show clearly that there are formidable difficulties in accepting the theory of the Semitic origin for the Brāhmī script designed and formed according to the phonetics and highly complicated principles of grammar. The advocates of the Semitic theory are not unaware of this. For instance, Diringer says: “We do not assume that the Brāhmī is a simple derivative of the Aramaic Alphabet. It was probably mainly the *idea* of alphabetic writing which was accepted.” This is only a round about way of pleading for a view that is not sustainable.

Numerals

c10.s5.1

Besides these arguments against the Semitic origin of the Brāhmī there are a few other reasons which disprove the theory. Though by themselves they are not weighty enough to explode the theory, when considered along with other arguments set forth above, they strengthen the case against the acceptance of any foreign origin for the Brāhmī script.

c10.s5.2

⁴² *Op.cit.*, p. 336.

c10.s5.3

Generally the numerals and the letters go together, and hence they are considered as the two eyes.⁴³ The origin and development of the numeral forms are closely related to the origin and development of letters, and both of them are component parts of the art of writing.

c10.s5.4

⁴³ *Eṇṇum eḻuttum Kaṇṇeṇattagum* (*Koṉṟaivēndan*, 7) *Eṇ eṇba ēṉai eḻutteṇba ivviraṇḍum* *Kaṇ eṇba vāḻum uyirkku.* (*Kuṟaḷ*, *Poruḷadikāram*, *Kēḷvi*, 2).

c10.s5.5

The numeral forms are already found in the inscriptions of Aśoka written both in the Brāhmī and the Kharoṣṭhi scripts. But both of them differ materially. The system of numeral notation found in the Kharoṣṭhi inscriptions in India consists of fundamental signs ; and all numbers are expressed by grouping them.

c10.s5.6

The fundamental signs of the Kharoṣṭhi numerals are : (a) one, two and three vertical strokes for 1, 2, and 3; (b) an inclined cross ressembling the mathematical symbol of multiplication for 4 ; (c) a sign similar to the Kharoṣṭhi *A* for 10 ; (d) a double curve, looking like a cursive combination of two 10's for 20. (e) a sign resembling the Brāhmī letter *ta* or *tra* for 100 to the right of which stands a vertical stroke, where by the whole becomes equivalent to 'one hundred' (I C). (See *Bühler's Tables*, Plate I, Col. XIV for illustrations).

c10.s5.7

The numbers lying between these elements *i. e.*, the intervening numbers are expressed by groups and the additional ones are invariably placed on the left in consonance with the direction of writing running from right to left. Thus number 5 is represented by the symbols 4 and 1 ; 6 by 4 and 2 ; 8 by two 4's placed side by side ; 50 by two symbols of 20 and one of 10 ; 60 by three of 20 ; 70 by three of 20 and one of 10. To express the numerals 11 to 19 and 21 to 29 etc., groups formed of the signs for 10 and 1 to 10 and 9 and 20 and 1 to 20 and 9 and so forth were used. The higher numerals beyond 100 were expressed according to the same principle. Thus 103 is formed by grouping the symbols for 100 and 3. The sign for 200 consists of the symbols for 100 preceded on the left (*i. e.*, in the direction of writing) by two vertical strokes, thus becoming II C. The number 273, *eg.*, would be written thus:

c10.s5.8

II CXX XX XX XIII.

c10.s5.9

(In the Roman numerals given above the principle of diminishing applicable to the figure to the left of the basic figure (V-5, IV-4) does not apply. Thus I C will mean one hundred and one and not ninety nine).

c10.s5.10

From the few numeral signs found in the inscriptions of Aśoka at Shahbasgarhi and Mansehra (*Bühler's Table*, Col. XIII) it is found that the system of numeral notation in Kharosthi differed at least in one important point from the above. The inclined cross or multiplication symbol representing 4 is not found in them. The number 4 is expressed by four parallel vertical strokes and 5 by five such strokes.

c10.s5.11

The Kharoṣṭhi numerals have for a very long time been taken as of Semitic origin; and the latest view is that they have been borrowed from the Aramaic branch, just like the letters of the script.

c10.s5.12

It may be mentioned in this connection that the above system of numeral notation is found used in all the Semitic branches, including the prefixing of the signs for 1 and 2 to the 100; but the inclined cross or the 'multiplication' symbol used to express number 4 in the later Kharoṣṭhi inscriptions is found only in the Nabataean inscriptions incised after the beginning of the Christian era and is used there only rarely for the expression of higher units. The late occurrence of this sign both in Indian and Semitic inscriptions is significant.

c10.s5.13

The system of numeral notations in the early Brāhmī both in inscriptions and in coin legends is different. In this system numbers 1 to 3 are expressed by horizontal (not vertical as in Kharoṣṭhi) strokes or cursive combinations of such. Units 4 to 9, tens 10 to 90, 100 and 1000 each by a separate sign; the intermediate and higher numbers by groups or ligature of these fundamental signs.

c10.s5.14

For figures consisting of tens and units of hundreds and so forth, the symbols for the higher numbers are placed either unconnected to the right or vertically below the higher ones. In the inscriptions and on most coins we find the first method followed. The second is found on a few coins. This method might have been followed on account of the exigencies of space. But this principle is found adopted in the pagination of all manuscripts. To express 200 and 2000, one short stroke is added to the right of the symbol of 100 and 1000 respectively. Similarly 300 and 3000 are formed by the addition of two strokes to the same elements. Ligatures of the symbols of 100 with the signs for 4 to 9, and of 1000 with those of 4 to 7, stood respectively for 400 to 900 and 4000 to 7000 and the smaller figures placed to the right of the larger ones. (*Bühler’s Tables*, Plate IX).

c10.s5.15

Regarding the origin of the Brāhmī numerals, scholars originally thought that they were of indigenous origin.⁴⁴ But Burnell thought differently. He pointed out that the general agreement of the principles of the Indian system as well as the resemblance of signs from 1 to 9 of the Brāhmī symbols correspond to those of the Demotic branch of the Egyptian writing.⁴⁵ Bayley, though accepting that the principles of the Indian system have been derived from the hieroglyphic notation of the Egyptians, tried to show that the majority of the Brāhmī symbols have been borrowed from the Phoenician, Bactrian and Akkadian figures of letters, while for a few a foreign origin is not demonstrable.⁴⁶ Bühler thinks that there are great difficulties *inter alia* by the assumption that the borrowings were from four or five different, partly very ancient and partly more modern sources; but at the same time he says that the comparative table of the Egyptian and Indian signs, and the agreement in their methods in making the hundreds, induced him to adopt with certain modifications the view of Burnell.⁴⁷ He thinks it probable that the Brāhmī numeral symbols are derived from the Egyptian Hieratic figures.⁴⁸ But he himself remarks that this derivation however still presents difficulties and cannot be called certain. He is also aware that the varying forms in the Aśokan edicts show these movements to have had a longer history before the third century B.C.

c10.s5.16

⁴⁴ *Ind. Ant.*, Vol. VI, p. 143.

c10.s5.17

⁴⁵ *Elements of South Indian Palaeography*, p. 65, *n.* 1.

c10.s5.18

⁴⁶ ‘Genealogy of the Modern Numerals’ *J.R.A.S.* (*NS*)., Vol. 14, pp. 335 ff; 15, pp. 1 ff.

c10.s5.19

⁴⁷ *Ind. Ant.*, Vol. XVI, p. 35.

c10.s5.20

⁴⁸ See *ante*, p. 65.

c10.s5.21

If the origin of the letters and the numerals of the Brāhmī is traced to some foreign source one would naturally expect that both of them are traced to the same foreign source as in the case of Kharoṣṭhī. But the letters have been taken as the derivatives of the North Semitic and the numerals of the Egyptian.

c10.s5.22

This complicated and incompatible position is evidently the result of a thoroughly wrong approach to the problem from the very beginning. The numeral symbols were taken as the letters, usually a *mātrikā* or a ligature, with phonetic values of such symbols and find an explanation for the use of such letters. And in this scholars were rather influenced by the analogy of the Semitic scripts.

c10.s5.23

In the Semitic scripts, the individual letters have numerical values also in addition to the phonetic ones. Of the twenty-two original letters of the script, the first ten bear the values from 1 to 110 and then the next nine letters by tens from 20 upto 100, and the last three letters by hundreds upto 400. Later on six more letters were added by the Arabs and the letter numerals were upto 1000. These numerical values of the letters, given by the famous Arabic mnemonic commencing with *abjad* etc., is retained in almost all branches, derived from the Semitic. In Arabic and Persian these letter numerals are used chiefly to record the dates of events.

c10.s5.24

It is true that this method of indicating the numbers with certain letters assigning them numerical values is found in India also. There are three such systems, and not a single method as in the Semitic, and in each of them the numeral value attached to individual letters is different. But none of these three resembles the Semitic system in any way.

c10.s5.25

The three Indian systems of letter-numerals are the *kaṭapayādi*, *bārakhaḍī* or *siddhamātrikā* and the *akṣarapallī*.

c10.s5.26

The *khaṭapayādi* is the system in which each of the consonant letters of the Sanskrit alphabet beginning with *ka*, *ṭa*, *pa* and *ya* is assigned the numeral value consecutively from 1 to 9 and 0. This is found used both in inscriptions and in literary works to record the dates of events when such portions of the compositions are in verse form. This is not generally used when the composition is in prose.⁴⁹

c10.s5.27

⁴⁹ See T. N. Subrahmanyan, *South Indian Temple Inscriptions*, Vol. III, Pt. i, pp. 243—44.

c10.s5.28

The *siddhamātrikā* or *bārakhaḍī*, also known as the *dvādaśakṣarī*, is the system in which each of the vowels is assigned the number consecutively from 1 to 12. This is found used only in the pagination of manuscripts. There are also some other forms of this system.

First 140 of 331 passages. Read on

Cites (16)