Home | Biodata | Biography | Photo Gallery | Publications | Tributes
[Back to Epigraphy List]


Origin of the Tamil-Brahmi script
The Institute of Traditional Cultures, University Buildings, Madras-600005, 1983.       
Gift Siromoney

The modern Tamil script is the result of a process of evolution1 beginning with the earliest form of script called the Tamil-Brahml scipt2 (This script is referred to as Damili by Nagaswamy and as Tamil by Krishnan). We shall refer to it as TALBI for short. The stages of evolution, though generally understood by epigraphists, are not known to the educated layman in Tamil Nadu. Some of us are making special efforts to make this knowledge more easily available.3 In this paper we shall raise some questions of general interest regarding TALBI.

What is the relationship between the Tamil-Brahmi script (TALBI) and the Asokan Brahmi (ASHBI) script? What are the possible origins from which the shapes of the different letters developed? Was there an earlier Tamil script which was replaced by the Tamil-Brahmi script for committing Tamil to writing?

Mayilai Seeni Venkatasamy4 has expressed the view that Tamil must have had a script of its own prior to the introduction of the Tamil Brahmi script. However excavations have so far not revealed any other script belonging to the prehistoric period of 1500-500 BC. The Harappan civilization belongs to the period of 2500-1500 BC. and the gap of 1000 years has not been satisfactorily bridged. Except for some graffiti markings on burial urns we do not have any material evidence at the moment to support the theory that there was writing in the Tamil language prior to 500 BC. The same argument holds good for Sanskrit or Pali or Prakrit languages.

How old is the Asokan. Brahmi script? Emperor Asoka who ruled from 269 to 232 BC had to his credit inscriptions in the Greek, Aramaic, Kharoshti and the Brahmi scripts. The Brahmi script of Asoka, therefore, is at least as old as the third century BC. It could have evolved slowly from an earlier script or it could have been invented during a fairly short period. It could have also been adapted/expanded from a contemporary script and the process could have been quite complex.

In order to clarify some of our thinking we shall look at four main aspects of the problem of Brahmi inscriptions. First we shall deal with the basic signs of the alphabet and ask ourselves the question as to the origin of these basic signs. We are not even defining what these basic signs are but it is a useful question. Secondly there is the question of relation between pairs of signs whose sound values are related. In some cases there is close similarity in sound values between letters which have similarity in shape. Thirdly there is the problem of representing vowels occurring in the medial position using special vowel markers. Fourthly we have to consider the method of representing pure consonants. In other words we have to consider (i) basic signs, (ii) related signs, (iii) vowel-markers and (iv) systems of representing pure consonants.

Let us take up the question of the basic signs. We omit in this discussion those derived signs which are obviously related in shape to the basic signs in the Brahmi script. Without much difficulty some of the signs of the script can be treated as basic signs even if there is room for some difference of opinion about the details. Where did these basic signs come from? Some would claim that the signs were directly taken from some Semitic alphabet, others would claim that each sign evolved over hundreds of years from the Indus script. A more recent view is that these basic signs were invented.5 We have proposed that the Brahmi script (TALBI and ASHBI) was invented from simple geometric shapes like the square and a circle and possibly based on certain geometric patterns found along with the Tamil-Brahmi inscriptions of Madurai region and the Brahmi inscriptions of Sri Lanka8. This view is in conformity with a Tamil tradition9 that the shapes of letters were derived from geometrical shapes. An inventor could have also made use of other scripts known to him and assigned suitable sound values to his signs.

We shall now discuss the aspect of related signs. Both in ASHBI as well as in TALBI there are pairs of signs which are similar in shape as well as in sound value. The sign for PHA in ASHBI is obviously derived from the signs for PA. So is the sign for THA (circle I with dot in the middle) derived from the sign for THA@ (circle). In TALBI the sign for one NA can be derived from another NA and sign for one LA can be derived from another by simply adding a horizontal line bent down on the right. It is fairly clear that there is some kind of design in these related signs and they could not have evolved over the years ending up accidentally with such close similarity in shape. Whether it be ASHBI or TALBI there was some design, some invention with reference to the related signs. Even those who do not approve of the theory of invention of the Brahmi script have to agree that some of the signs were invented. If only some of the signs were invented then it weakens the case for wholesale borrowing of a foreign script. There is no reason why an inventor should not assign signs which appear similar to sounds which are closely related.

The presence of vowel-markers is an important aspect of the Brahmi alphabet. A short vertical stroke attached to the top right side of a consonantal sign represents the vowel I when it occurs in the medial position. Short strokes placed in other positions denote the other medial vowel signs. Such an use of vowel-markers is an inherent part of the Brahmi system. It is difficult to claim that a system was also borrowed from a Semitic script which is not known to have had any such developed system. In many Semitic systems the medial vowels were not represented but had to be guessed from the context and the writing was from right to left. In Brahmi the writing is from left to right and we have a developed system of vowel-markers which are arranged in such a way that there is not much ambiguity between the different simple and compound signs. Continuing our discussion on the system of vowel-markers we shall compare ASHBI with TALBI and see whether there are any differences. There seems to be one uniform method used in the ASHBI system whereas more than one system seems to be present in the TALBI inscriptions. In ASHBI a short horizontal stroke on the top right side of a consonantal sign would represent the long medial A. Such a uniformity of representation is lacking in the TALBI system. In some TALBI inscriptions the sign represents only the short medial A and in some others which appear to be a little later, the same sign represents the long medial A. Such a difference will be discussed in the context of the antiquity of the TALBI system a little later. The system of vowel-markers is unique to Brahmi and I know of only one other script, viz., Ethiopian, where a slightly different system is used to represent vowel-markers.

The fourth aspect is the method of representing pure consonants. Surprisingly the system used in the TALBI inscriptions is quite different from the one used in ASHBI inscriptions. In the ASHBI system to denote a pure consonant the sign representing the next sound was placed below the first sign. Each consonantal sign represented a consonantal syllable with an inherent A and the placing of a sign below another indicated that the top sign represented a pure consonant. In some cases the bottom sign represented the pure consonant and the combined sign had to be read from bottom to top. This system has an obvious drawback when the last letter of a word is pure consonant. In ASHBI the pure consonant M was represented by a small circular mark and the language of the Asokan inscriptions did not normally use other consonants in the terminal position. So far we have not discovered any TALBI inscription which uses this Asokan system of representing pure consonants. Surprisingly Tolkappiyam refers to the shape of the letter M as the shape of a pulli which was most probably a small circular mark as in the TALBI inscription of Anaimalai.8

Pure consonants were denoted in two different ways in TALBI inscriptions. In some inscriptions a consonantal sign without a vowel-marker denoted a pure consonant. This is quite different from the ASHBI system in which each consonantal sign represented a consonantal syllable with an inherent A. In some other TALBI inscriptions a consonantal sign accompanied by a consonant-marker in the form of a pulli represented the pure consonant. The pulli was also used in some TALBI inscriptions to discriminate a short E from a long E both in the initial and the medial positions. It was also used to distinguish between the short and long O's. In ASHBI inscriptions no distinction is made between the long and the short forms of E as well as O. In many TALBI inscriptions no distinction is made between the short and the long forms of E as well as O. These two sounds occur in Tamil language much less frequently than the vowels A, I and U. In many TALBI inscriptions no distinction is made between the long and the short A in the initial position. Neither is any distinction made between the initial long and short I.

What is the relationship between the TALBI and the ASHBI scripts ? The shapes of letters which are common to both the scripts are more or less the same and therefore the two scripts could not have originated completely independent of each other. Some scholars have held that the two scripts descended from some common source. Many have held that TALBI originated from ASHBI and a few others have suggested that TALBI is more ancient than ASHBI.9

The problem is getting more interesting now due to two reasons. The first is the possibility of ASHBI (or TALBI) having been invented.10 The second is the proposal by Mahadevan11 that there are three systems of writing ancient Tamil which we shall call TALBI-I, TALBI-II and TALBI-P, the pulli system which is described in Tolkappiyam. Other systems have to be reconstructed from the available data.

If the hypothesis that ASHBI (or TALBI) was invented along with the system of vowel-markers is acceptable then we have only two possibilities, viz., that either ASHBI precedes TALBI or TALBI precedes ASHBI (We are overruling a simultaneous invention of both the scripts at one centre in India). Even those scholars who do not subscribe to the hypothesis of invention can join in the discussion by starting at the stage of the introduction of the vowel-marker system.

Let us consider the TALBI-I system of writing. The syllable KA is written in the form of a cross with a horizontal vowel-marker on its top right side. This is in contradiction to the method in Tolkappiyam which says that all consonant signs without pulli will represent the corresponding syllables with medial A. However Tolkappiyam allows for lengthening of vowels by adding additional vowel signs. This could support the existence of the TALBI-I system. It is still puzzling however why Tolkappiyam does not speak explicitly about the different systems if they had existed before it was written.

In the second system, viz., TALBI-II, the system described in Tolkappiyam is followed except for the pulli. A sign in the form of a cross will represent both the pure consonant K as well as the syllable KA. In the third system, viz., TALBI-P, there is no ambiguity between the pure consonant (marked with a pulli), the syllable with the short medial A (no pulli) and the syllable with the long medial A (with an additional vowel-marker attached to the sign on the top right hand side). Even in the representation of the initial vowel signs a pulli is introduced to distinguish the short E and O from the corresponding forms representing long vowels. For the other initial vowel signs a short horizontal stroke is used to indicate the lengthening of the vowel signs. If the Tamil system of writing is equated with TALBI-P system then of course the ASHBI script will appear to be older than TALBI script but there are other possibilities.

One can get away with saying that TALBI- system is a figment of imagination and all that the scribes had done was misspelling of words including those in the Mangulam inscriptions which refer to Nedunchezhian. However there is another system of writing which is very similar to TALBI-I where the long medial vowels are indicated by a long horizontal stroke bending down on the right. This is the Bhattiprolu system, which we can call BATBI. I am quite convinced that in the TALBI inscriptions many systems are mixed up and words are misspelt. In spite of that one can have a discussion on the relationship between the five systems, viz., ASHBI, TALBI-I, TALBI-II, TALBI-P and BATBI.

Let me first state my present position. Not because this is the only solution to the problem but mainly to emphasize that it is one of the possible solutions. TALBI-I was invented first for writing Tamil language. It could have had an early as well as late phase. ASHBI was designed on the basis of TALBI-I using the same geometrical compound signs. The letter I was modified into three little circles instead of three horizontal strokes. New signs for the voiced consonants were invented. ASHBI spread all over India including the present Mysore region, Andhra Pradesh and Sri Lanka. TALBI-I was slowly given up in Tamil Nadu and people changed over to TALBI-II whereas in Andhra Pradesh TALBI-I had already developed into BATBI. In TALBI-II there was no way of representing a pure consonant unambiguously and the TALBI-P came into vogue. However the scribes often ignored the pulli since they could make out the correct value from the context.

On the contrary if we assume that ASHBI came first and TALBI-I was an adaptation of ASHBI, the question is why did they not follow simply the ASHBI system of writing for Tamil and just invent additional signs for the special sounds? By using TALBI-I the writings would not be intelligible to the officers from the Mauryan court and the Jain and the Buddhist monks of north India who visited Tamil Nadu. ASHBI system is closer to TALBI-II system. It is quite likely that the ASHBI system was not followed in Tamil Nadu since they had already a TALBI-I system. Therefore they modified the ASHBI system but did not follow the system of writing one letter below the other. It is not possible to visualize a simple transition from ASHBI to TALBI-I and then from TALBI-I to TALBI-II.

There are many other possibilities. TALBI-I and TALBI-P could have existed from the earliest times and all the three systems could have existed side by side. If there were so many systems in TALBI that were experimented upon in Tamil Nadu then it would support the hypothesis that the original home of the Brahmi scripts was Tamil Nadu.

The date of Tolkappiyam is closely linked with the date of the introduction of the TALBI-P system but it is difficult to date the TALBI inscriptions in general and the TALBI-P writings in particular. The contents of the TALBI inscriptions and the variations in the shapes of letters raise many interesting problems. What is the period of Nedunchezhian and Athiyaman Nedumananji ? Were they contemporaries of Asoka or did they belong to a period which is 400 years after Asoka ? We know that the Cholas, Pandyas and Satyaputras were ruling the neighbouring kingdoms of Asoka. Do we know the personal names of any of these kings? It is generally held by students of history that Senguttuvan, Karikalan, Nedunchezhian and Adiyaman Nedumananji belong to the second century A.D. If that is right then the recently discovered Jambai inscription of Adiyaman Nedumananji must belong to the second century A. D. In that inscription Adiyaman is referred to as Satyaputo and the title is spelt in exactly the same way as in the second rockcut inscription of Asoka at Girnar. The shape of letters are almost the same as the letters of Asokan edicts but is quite different from the letters of the well-known legend on the coin of a Satakarni king of the second century A.D.12

Scholars have tried to date the TALBI inscriptions on the basis of the development of the shape of the letters without taking into consideration the fact that there is a lot of variation in the shape of letters in the Asokan inscriptions13 as well as in the Brahmi inscriptions of Ceylon. There is no strong basis for dating TALBI inscriptions on the basis of the shapes of letters since we do not have securely dated inscriptions. Very often scribal characteristics are mistaken for period characteristics. We have to take a fresh look at the late dates given for Adiyaman and Nedunchezhian in the light of the recent epigraphical discoveries.

There is still no general agreement on the origin of the TALBI script and I have presented in this paper a theory describing the complexity of the problem. Fresh excavations need to be undertaken in the Madurai region and these may provide useful material that may lead to a better understanding of the origin and development of the Tamil-Brahmi script.

1. Gift Siromoney, S. Govindaraju and M. Chandrasekaran, Tirukkural in Ancient Scripts, Tambaram, 1980.
2. I. Mahadevan, "Corpus of the Tamil-Brahml inscriptions", Seminar on Inscriptions, Madras, 1968, pp. 57-73.
3. Gift Siromoney, S. Govindaraju and M. Chandrasekaran, Kalvettu El@uttukalai Karka Oru Putuvali, Tambaram.
4. Mayilai Seeni. Venkatasamy, Sangakalattu Brahmi Kalvetteluttukkal, Madras, 1981.
5. Gift Siromoney and M. Lockwood, "The invention of the Brahmi Script", Madras Christian College Magazine, XLVI (1977), pp. 31-33; Souvenir of the Fourth Annual Conference of the Epigraphical Society of India, Madras, 1978, pp. 42-50.
6. S. Paranavitana, Inscriptions of Ceylon, Vol. I, 1970.
7. Tolkappiyam, Eluttu Atikaram, Sutra I (Commentary of Nachinarkiniar)
8. Emmanuel Jebarajan and Gift Siromoney, "On the occurrence of the Pulli in the Tamil-Brahmi Inscription of Anaimalai", New Dimensions in the Study of Tamil Culture, Palayamkottai, 1978, pp. 8-12.
9. Gift Siromoney, "The Origin of the Tamil Script", Tamil Studies, Vol. II (1982), pp. 8-23.
10. S, P. Gupta and K. S. Ramachandran (Eds.), The Origin of the Brahmi Script, Delhi, 1979.
11. I. Mahadevan, Tamil Brahmi Inscriptions, Madras, 1970.
12. R. Nagaswamy, "A bilingual coin of a Satavahana", Seminar on Inscriptions, Madras, 1968, pp. 200-202.
13. E. Hultzch, Corpus Inscriptionum Indicarum, Vol. I, Oxford, 1925.

Go to the top of the page

Home | Biodata | Biography | Photo Gallery | Publications | Tributes