Prev | Home | Next

As stated above, the history of decipherable epigraphy in India starts with the well-known records of the Mauryan emperor Asoka, Besides Bra#hmi#, Asoka's records are found in Kharo#sht?hi#, Aramaic and scripts. The Aramaic and Greek scripts are used in the edicts found in Afghanistan while one edict in Taxila (now in Pakistan) is supposed to be in Aramaic script. Kharo#sht?hi is a modification of the Aramaic script introduced in the Uttara#patha region of India when it was under the control of Persia before Alexander's invasion. Like Aramaic, Kharo#sht?hi was written from right to left and is used in the rock edicts of Asoka found at Mansehra and Shahbazgarhi (in Pakistan) in the Northwestern part of Asoka's empire. After Asoka, the use of Kharo#sht?hi script spread upto Mathura region and was prevalent in the ancient province of Ga#ndha#ra upto the 2nd century A.D. used by the Kusha#n?a kings in their records. South of Mathura, it is used very rarely in single inscriptions at Bharhut and while a .few letters mentioning the name of the scribe in the Asokan edict at Siddapur in Karnataka are also in Kharo#sht?hi script. After 3rd century A.D. this script went outside India and was in use in the area known as Chinese Turkestan for about a century. It did not survive in India longer possibly because it was not found suitable for writing in Prakrit and Sanskrit languages in which the early records are written. Therefore, it was only the Bra#hmi# script used by Asoka which survived in India for centuries and which became the mother of all the modern Indian scripts - both northern as well as southern - and also of the scripts of South-East Asia.
Now what is the origin of this Asokan Bra#hmi# script? Was it invented by Asoka and used by him in his edicts for the first time? Or was it evolved out of an earlier script, though we have no evidence of this earlier script at present? Or was it borrowed from foreign script and was modified to suit the genius of the Indian language? In other words, is Asokan Bra#hmi# script of indigenous origin or of foreign origin? The problem regarding the origin of this Bra#hmi# script has exercised the minds of several scholars during the last century and a half, ever since its successful decipherment by James Prinsep. Different scholars have expressed different views about the origin of the Bra#hmi# script. But before entering into the discussion of these views, let us briefly discuss about the name given to this script, viz., Bra#hmi#.
It has been noticed above that the Jaina works Samava#ya#n#gasu#tra and Pan?n?avan?a#su#tra and the Buddhist work Lalitavistara include, in the list of scripts mentioned therein, the name of the script Bhambhi#, i.e., Bra#hmi#. But we have no idea of the nature of this script since no material evidence of it has been found so far. So we cannot say whether the reference to the Bra#hmi# script in these works can be connected to the script used in Asokan edicts. Asoka himself does not say anywhere that he has engraved his records in the Bra#hmi# script. In the Chinese encyclopedia Fa-Wan-Shu-Lin (7th century A.D.), it is stated that the Bra#hmi# script, written from left to right was invented by Fan, i.e., Brahma#1. According to the Arab writer Alberuni (10th century A.D.) the Hindus had once forgotten the art of writing and through a divine inspiration it was re-discovered by Vya#sa, the son of Para#sara2. But this must have been a popular belief prevalent then and recorded by him. Raj Bali Pande suggests that the name Bra#hmi# indicates that it was invented for preserving Brahma, i.e., Ve#da by the Aryans and that it was mainly used by the Bra#hman?as whose duty it was to preserve the Vedic literature and to prepare copies of this literature for the future generations. This view is indeed fanciful and untenable. In Vedic and epic literature, the name Bra#hmi# connotes language and not script. The script employed by Asoka in his edicts was meant for the benefit of the Buddhist monks who used Prakrit language. Hence, the script used for Prakrit language of the followers of Buddhism and Jainism for several centuries, and in fact, they seem to be responsible for the origin and widespread use of this script in the country. It has been suggested that the Brahmins who preserved and transmitted the Vedic tradition might have developed a script of their own for their private use in their limited circle which might have died a natural death when the script employed by Asoka became popular amongst the majority of non- Brahmins, Some of the signs used by these Brahmins might have also been merged in the script of Asokan Bra#hmi# and did not survive after his period and since it might have been written only on perishable materials no evidence of the same is available. Of course, at a later period when the Vedic culture was revived, even the Brahmins adopted the developed Asokan Bra#hmi# script for their purpose3. And this state of affairs explains why the earliest Bra#hmi# script used by Asoka for Prakrit language does not possess the signs of Sanskrit language like visarga, jihva#nu#li#ya, upadhma#ni#ya etc., which were introduced into the script later on when it was' used for Sanskrit language also. However, at present, all scholars when they use the term Bra#hmi# in their study of Indian scripts, they mean the script used by the emperor Asoka in his records in the 3rd century B.C. It is also called the Mauryan Bra#hmi# script or Asokan Bra#hmi# script to distinguish it from the later Bra#hmi# of the early Kshatrapa, Kusha#n?a and Sa#tava#hana periods and also from the southern Bra#hmi# where some letters were modified to suit the genius of the Dravidian languages.
The above discussion would thus show that while there are literary references testifying to the knowledge of writing in India prior to this period of Asoka (3rd century B.C.) and even to the script named Bra#hmi#, there are no specimens of any of these scripts to enable us to decide anything about them or about their relationship with other contemporary or later scripts. Similarly, the attempts made by some scholars to connect the Asokan Bra#hmi# script with the Indus script on the basis of similarity of some signs are not convincing in view of the fact that there is a wide ga#p of more than one thousand years between the latest period of the Indus script and the Bra#hmi# script of Asokan period and the absence of any specimens of any script in India during this period. Moreover, as noticed above while discussing the Indus script, scholars have not yet come to a unanimous decision about the signs and their values and hence no connection of this script with any other script can be thought of in the present state of our knowledge.
As stated above, there are several theories regarding the origin of the Bra#hmi# script. Otfried Mueller, followed by James Prinsep, E.Senart and others, thought that the Bra#hmi# script had its origin in the Greek alphabet when the Greeks came into contact with the Indians in the wake of Alexander's invasion of India5. Scholars like William Jones, A. Weber, A.C. Burnell, Isaac Taylor and George Buhler held the view that the Bra#hmi# script could be derived from the Phoenician (Aramaic), South-Semitic and North-Semitic scripts6. Bu×hler, a great expert on Indian Epigraphy and Palaeography, stated that twenty-two letters of the Bra#hmi# script were derived from North-Semitic script and some of them were found in early Phoenician records. To the objection that while Bra#hmi# script is written from left to right while the Semitic one was written from right to left, he maintained that originally Bra#hmi# was also written, like the Semitic script, from right to left only. And to support his view, he quoted the example of the Eran Coin legend written in the reversed style of the Asokan Bra#hmi# letters dha, O and ta which are also written in the reverse way. Some scholars later on who wanted to support Bu×hler's view of Semitic origin of the Bra#hmi# script and its original direction of writing from right to left quoted the instance of the Yer?r?agud?i (Andhra Pradesh) rock edicts7 of Asoka where some lines are written in boustraphedan style and the Duvegala (Sri# Lanka#) inscription8 which is written in the reverse fashion. But these, examples of the Eran? coin, Yer?r?agud?i rock-edicts and the Duvegala record are stray cases and are due to the mistake or whim of the individual scribe and do not help to determine the direction of writing. As shown by the large majority of the Asokan edicts in Bra#hmi# script, the direction of writing is only from left to right which is also the case in regard to all the scripts of India derived from this Bra#hmi# script.
Some scholars, as noticed before, trace the origin of the Bra#hmi# script to the Indus or Harappa script which was prevalent from about 2500 B.C. to 1500 B.C.9 But, as already pointed out, it is difficult to connect the two scripts in view of the wide gap of over one thousand two hundred years between the two scripts and also in view of the fact that the Indus script has not yet been finally and successfully deciphered. Moreover the consensus of scholars agree that Indus script was written from right to left.
General Cunningham was of the View that the Bra#hmi# script is of Indian origin, though he thought that the plan of the system was borrowed from the Greeks10. R. Shamasastri connected the origin of the Bra#hmi# script with some Sanskrit Tantric texts which contained signs and symbols which represented the De#vas and De#vanagara11. Among the other profounders of the theory of indigenous origin of the Bra#hmi# script may be mentioned John Dowson12, Jayaswal13, G.H. Ojha14, and R.B. Pandey15. On the assumption that before the advent of the Aryans in India, Dravidians occupied the entire country, scholars like Edward Thomas traced the origin of Bra#hmi# script to the Dravidian races of South India16. Recently, T.N. Subrahmanyam has supported this view17. But all the above views have not received the acceptance of most of the modern scholars. As regards the theory of Dravidian origin, it has been pointed out that the oldest Dravidian language possesses the signs, due to its phonetic characteristics, for the first and the last sound of each varga, viz., k and n#, ch and n, t? and n?, t and n and p and m while the Bra#hmi# script which must have been perfected by the Sanskrit and Prakrit grammarians possesses the signs for all the letters of the five vargas. Moreover, no script of the Dravidian languages earlier than the Asokan Bra#hmi# script has been discovered so far and hence it is difficult to accept the theory of Dravidian origin of the Bra#hmi# script.
Recently, T.P. Verma has suggested a fresh approach to the problem of the origin of the Asokan Bra#hmi# script18. He says that the birth of the Bra#hmi# script should be considered in the light of the culture renaissance that preceded the period of Asoka, i.e., from the time of Gautama Buddha. During this period, India, particularly north India, witnessed great cultural movements which had far-reaching impact on the social, religious, artistic, literary and linguistic history of the country. Many anti-Vedic preachers of parivra#jaka teachers came into existence, particularly in the eastern part of the country. Popular religions like Buddhism and Jainism which embraced all common people without distinction of caste or sex were introduced and the Vedic religion which addressed only to a select community received a setback. Sanskrit, which was the language of this select community and their monopoly, was not accessible to the bulk of the society. So the new religious teachers adopted the use of the Prakrit language for the benefit of the common people. This language gradually spread throughout the country and even outside and thus secured an honoured place like Sanskrit. And, under the influence of this strong cultural movement, a simple and precise script was created for the use of this popular Prakrit language. The credit of using this script and propagating it throughout the length and breadth of India goes to the missionary zeal of Asoka who wanted to record his edicts on permanent material, viz., stone by using a standard script. And since Prakrit has fewer sounds than Sanskrit, the script used by Asoka has also fewer symbols and when this script was later on adopted for Sanskrit also a few more signs were added. Similarly when Dravidian languages like Tamil, Telugu and Kannad?a used this script, some new signs to suit the genius of those languages were also introduced in course of time.
In the light of the above observations, it is not possible to with A.H. Dani's views19 that the creation of Bra#hmi# script is based on North Semitic script. He says that neither the Bra#hmi# script is the creation of the Indian grammarians nor can the Indian alphabet, i.e., varn?ama#la# be derived from this Bra#hmi# script. But, as pointed out above, the Asokan Bra#hmi# script was the creation of the Buddhist monks for the propagation of the teachings of Buddha in the Prakrit language for the benefit of the common people. And the Indian alphabet or varn?ama#la# was the creation of the Sanskrit grammarians and was in existence long before the period of Asoka. The Bra#hmi# script was based on this existing varn?ama#la# of the Sanskrit grammarians but contained fewer signs to suit the needs of the Prakrit language. Therefore, it is not correct to say, as stated by many European and Indian scholars, that Bra#hmi#r script was perfected by Sanskrit grammarians.
Now let us consider the nature of the Asokan Bra#hmi# script. Taylor says that it is "bold, simple, grand, complete, the characters are easy to remember, facile to read and difficult to mistake, etc."20 While Cust observes that "the Indian alphabet is a marvelous and magnificent phenomenon quite unrivalled in the world"21 . And, indeed, the arrangement of letters representing the symmetrical combination of the symbols to indicate different shades of sound together with the cerebral letters, aspirate consonants, anusva#ra and anuna#sika and the three sibilants s, sh and s make this script not only a perfect one but also a unique one in the world22.
The decipherment of the Asokan Bra#hmi# script has an interesting story. It was long forgotten by the Indians. Firoz Shah Tughlak (1351 -1388 A.D.) shifted two pillars containing Asokan edicts from Topra and Meerut in Uttar Pradesh to Delhi in 1356 and invited a number of Sanskrit Pandits to read the writings on the pillars. But nobody could decipher the script on them. A similar attempt was made by the Mughal emperor Akbar (1556 - 1605 A.D.) but without success. Only after the advent of the British rule in India, scientific efforts were made to decipher early writings when the Asiatic Society of Bengal was founded by William Jones in 1784. Though some of the later inscriptions including the Gupta records were deciphered earlier, it goes to the credit of James Prinsep to find out the key to the decipherment of the Asokan Bra#hmi# script with the help of the short inscriptions from Sa#nchi#. With great patience and preseverence he worked at it for quite a few years and wistfully gazed at the unknown script which concealed the mystery of India's thought and culture and at last, in 1837, found the key to it and successfully deciphered the entire script.
The Asokan Bra#hmi# script consists of seven vowels including nasal am# and thirty-three consonants including semi-vowels. The special feature of this script is that the vowel a is inherent in each consonant letter while other vowels are indicated by separate strokes attached to it. On this account some scholars are inclined to describe this script as syllabic-cum-alphabetic. There is no pure consonant as such which later on was indicated by a diminutive form of the required consonant letter only. Another characteristic of this Bra#hmi# script is that it is written from left to right which direction is adopted by all the derivative scripts also. Earlier scholars like Buhler thought that there were regional varieties of Asokan Bra#hmi# but after a detailed study Upasak has shown that there were no regional varieties in this script and the differences in the letters noticed are due to the individual stylistic characteristics either in engraving or in writing the draft23. The different varieties of letter are found in the same region. Dani also supports this view and calls it an imperial script in spite of minor differences24. Asokan Bra#hmi# inscriptions are found engraved on rocks, pillars, caves and in one case on a slab. The pillars are well polished in accordance with the high standard of Mauryan art and hence the records engraved on them are beautiful and well preserved. The letters of Asokan Bra#hmi# are straight and angular in general appearance, though some letters are round and cursive. The lines run parallel to each other from left to right and some sort of punctuation was observed by leaving some space between two words. When there are more than one edict oh a rock or pillar each edict starts as a fresh line, normally. Only in the case of the Er?r?agud?i rock edicts, the lines sometimes are written from right to left which, as pointed out above, is due to the mistake or fancy of the engraver.
Upasak postulates basic and derivative forms of letters in the Asokan Bra#hmi# script25. Thus, amongst the vowels while a, i and u are basic a#, e, o are evolved ones. This distinction between the basic and evolved letters is more pronounced in regard to the vowels than the consonants. As regards the method of producing the letters, the writer used the simplest of the tools, viz., a longish piece of chalk, charcoal or haematite so that the lines drawn are uniform in width which we notice in the Asokan edicts. The actual shape of the letters depend on the writer's method of producing the letters. Two stages are involved in this writing. First the -writer called lipikara in the edicts of Asoka who writes the inscription on the stone by chalk, charcoal or haematite and then the letters are cut or engraved by the stonecutter. While the writer was a literate person, the engraver was probably an illiterate person. While the varieties in some letters are due to the mannerisms of the writer, some differences are due to the method of engraving also. But ultimately the shape of the letters is due to the tools employed by the writer who is responsible for the style of writing system. Since the writing tools, as stated above, for the Asokan edicts consisted of a longish piece of chalk, charcoal or haematite, the lines formed were straight and of equal width and there were no 'serifs' or head-mark in these letters as thought of by scholars like Buhler. This 'serif' or head-mark, is found later on due to the use of new tools of writing as pointed out by Dani.

The Scripts of Post-Asokan Period
As stated before, the Asokan Bra#hmi# script did not have any regional varieties and hence it was called an imperial script. And the different forms of each individual letter are due to the vagaries of the engraver or the manner of engraving. It is a significant fact that this standard script prevailed throughout the vast empire of Asoka upto the Karnatak region in the south. A few records engraved soon after the period of Asoka, though retaining the features of the script of his edicts, are not so well executed. These are the Nagarjuni cave inscriptions of Dasaratha, grandson of Asoka, the Maha#stha#n (now in Bangladesh) inscription, the Priprahva and Sohgura records26. Some scholars thought that Priprahva and Sohgura inscriptions even belonged to pre-Asokan period but it is now generally agreed by the scholars that they too are post-Asokan. To this period may be assigned the script of the Bhat?t?iprolu (Andhra Pradesh) casket inscriptions. The peculiarity of this script is that while ka# of Asokan Bra#hmi# script forms ka here, ka# is formed by adding another vertical stroke. Some scholars connect the Bra#hmi# script of the Tamil Cave inscriptions with this script. The script of all these records represent the writing soon after the period of Asoka and hence may be called post-Asokan script.

Except the records of Dasaratha, the other inscriptions of the post-Asokan period are entirely different in content and style than those of the Asokan edicts. The uniform Mauryan standard of script as well as style of writing prevalent during Asoka's period was no longer valid and we begin to. get inscriptions showing a variety of purpose and they are not only royal but also private records inscribed at the instance of the individuals. This, coupled with the change in the technique of the writing materials, naturally has given rise to different styles of writing in different parts of the country. Though these scripts developed out of Asokan Bra#hmi# script, the evaluation of the script in different regions did not take place at the same rate and in the same manner and hence we get a variety of scripts developed in different areas out of the parent Asokan Bra#hmi# script. Dani calls these scripts as Provincial scripts27.
The most notable features in the script of the post-Asokan period are: the addition of vowels, long i#, long u# and ai devised by adding one more dot to i and one more stroke to u and e respectively. Similarly we have the use of visarga in Sanskrit words indicated by adding two dots, one above the other, at the end of the word. Important changes are introduced in the formation of the conjunct letters, besides devising symbols to indicate au and r?i in Sanskrit words. The Besnagar inscription of Heliodorus (2nd century B.C.), where the angular forms of pa, va, sa la and ha are noticed; Ghosun?d?i#, Bhilsa, Barli and Hathigumpha inscription (1st century B.C.) where the equalization of the verticals in letters like pa and sa are found, are some of the notable examples of the records belonging to the post-Asokan period. The script of the casket inscriptions from Bhat?t?iprolu in the Krishna district of Andhra Pradesh, which presents certain peculiar features of the Bra#hmi# script, is considered by scholars like Bu"hler as earlier than the date of the above mentioned records and nearer to the date of Asokan edicts, i.e., earlier than 200 B.C. But D.C. Sircar and others suggest that the script is not much earlier than 100 B.C. According to some scholars this script of Bhat?t?iprolu inscriptions is the origin of the southern Bra#hmi# inscriptions.
In this connection, we have to consider the script of the inscriptions found in the caves of Tamil Nadu in the districts of Tirunelveli, Madurai, Ramanathapuram, Tiruchirapalli and Coimbatore. There is also one inscription found in a cave in the Nellore district of Andhra Pradesh28. These are the earliest inscriptions of South India discovered so far during the last seventy-five years or so. They are found engraved on the beds of the caverns or on the brows of the overhanging boulders or on walls and are short ones consisting of only one or two lines with a few letters. Their decipherment and interpretation, including the dating of the script, has exercised the minds of the scholars for quite a long time. Some records were also discovered, at a later date, engraved on potsherds recovered from Arikamedu near Pondicherry and from Uraiyur near Tiruchirapalli in the course of the excavations conducted there. The dating of these records varies from 3rd century B.C. to the 1st century A.D. Since they are short epigraphs, they appear to be label or dedicatory records. The script in which they are written resembles mostly the Bra#hmi# script of the Asokan and the post-Asokan periods. But there are a few symbols which are peculiar to these records only and are not found in the normal Bra#hmi#ml script. These symbols, according to T.V. Mahalingam29, represents the sounds long i#, n, n?, l, l? and r? which are characteristic of the Tamil language.
Dani thinks that these records of the southern caves should be dated in the 1st century A. D30. He says that the inscription from Sittannava#sal is unique in that it copies certain features of northern letters which are not known even in Amara#vati. The form of la found here occurs in the Saka Kshatrapa records of Mathura. But the form of i with a dot on either side of a verticle appearing in this inscription is not seen at this time outside South India. It is used for long i# in North and West India in the later period. Dani is inclined to think that this form of i# is introduced into the North from South India, where it is used both for short i as well as long i#. Ma is a new form resembling the Roman letter U with a crossbar in the middle and is called the Dravidian of South Indian Bra#hmi# ma. The letters in the other cave inscriptions of South India, though some of them are cruder, resemble those of the Sittannava#sal inscriptions and hence their date, according to Dani, falls in the first half of the 1st century A.D. He supports his view by the evidence of the excavated material of the Arikamedu potsherds which contain some inscriptions and which are stratigraphically dated in the 1st century A.D. According to T.V. Mahalingam, however, the date of the southern cave inscriptions ranges from 2nd century B.C. to 1st century A.D. since some of them belong to the earlier period while others to the later period, thereby showing a chronological development31. He also suggests that some of these records adopted the Dra#vid?i# style while others did not and followed the Asokan Bra#hmi# style.
The "above discussion would show that after the period of Asoka (3rd century B.C.), the art of writing, which was confined to the imperial court of Asoka and which had a standard script throughout his vast empire, came to be used by private people also. This resulted in the script gaining popularity amongst the common people. The mode of expression also changed in the post-Asokan inscriptions. The Maha#stha#n, Sahgaura Piprahwa and Ra#mgarh hill epigraphs belong to this category, though the inscriptions of Dasaratha, Grandson of Asoka, found in the Na#ga#rjuni hill cave are royal records of the Asokan type. But there is a change in the face of the letters of these records, though the general formation is similar to Asokan edicts, which is partly due to the tools of writing used and partly to the different materials on which they were inscribed. T.P. Verma calls the script of these epigraphs as belonging to the transition period and from the point of view of development of the script while some letters are more advanced and cursive, others are more primitive and angular33. The same scholar considers the next stage covering the period of 200 B.C. to 100 B.C. as the period of experiments on the basis of his study of the script on the coin legends of this period. He compares the palaeography of these coin-legends with that of the Besnagar pillar inscriptions of Heliodorus. But it may be pointed out that the palaeography of the legends on the coins is not a reliable guide to determine the development of the scripts in view of the fact that the material on the coin legends is extremely limited for instituting comparison with the script of a normal inscription on stone or copper-plate and moreover, the changes in the letters on these coin-legends take place in a very slow manner. Hence, we have to leave the palaeography of the coin-legends out of consideration in our study of the development of Indian scripts.
Thus chronologically speaking, according to T.P. Verma34, the inscriptions of Sa#nchi# Stu#pa I, the Besnagar Garuda pillar inscription, the Na#na#ghat cave inscriptions of Nayanika#, the inscriptions found at Ghosun?d?i and Barli - all belong to the early part of the 1st century B.C., while Sa#nchi# inscriptions of Stupa II and the Bodh Gaya# railing records belong to the later part of this century35. He further suggests that the development of Indian scripts upto this period was rather slow and the tools and technique employed were the same as those of Asoka for the Asokan Bra#hmi# and the changes noticed are mainly due to the fact that the art of writing had become more popular and was practiced by private people also.
With the advent of the Saka Kshatrapas and the Kusha#n?as as political powers in India, the writing system assumed a definite change which, according to modern scholars like Dani and T.P. .Verma, is due to the use of new writing tools and the techniques36. Dani states that the new writing tool of this period consisted of a sharp-edged pen. Due to this and also due to the fact that more and more private individuals, besides the royal persons, took to the mode of writing, a noticeable change in the development of the scripts appears from this period onwards. The use of the edge pen made the letters in their vertical portion thick at the top and gradually thin towards the bottom, giving a dagger like shape to the letters something like the effect of the reed-pen used in recent times in some parts of the country. The same writing instrument was also responsible for the flourish found in the curves of the letters. The thick top of the letters caused by this writing tool was mistaken by some scholars as the introduction of the 'serif' or top ma#tra# in the development of the script but, as stated above, this thickened top is due to the use of the new writing tool. Such thickness in the top of the letters is noticed in the case of ka, ra, la, va, etc.
T.P. Verma thinks that the script of the Kshatrapa inscriptions of Mathura, Pabhosa cave inscription, Bharhut Gateway inscription, Ayodhya inscription of Dhanadeva and Hathigumpha inscription of Kharavela as belonging to the first half of the 1st century A.D. and represent the script of the pre-Kusha#n?a period37. But some of these inscription e.g., the Pabhosa cave inscription and the Hathigumpha inscription of Kharavela are regarded by scholars like D.C.Sircar as belonging to the end of the 1st century B.C. which view seems to be correct since the palaeography of these records appears to be earlier than that of the other epigraphs.
With the advent of the Kusha#n?as of Kanishka I's line, we notice further development in the script. The number of private and individual records increases as compared to the official records. As stated by Dani, there was not one uniform Kusha#n?a style throughout their empire. Different styles developed in different localities since local writers were employed in the art of writing the inscriptions of that particular locality. And T.P. Verma broadly classifies two main varieties of script of this period, viz., the latter half of the 1st century A.D. till the end of the 2nd century A.D. He calls them as western variety found in the inscriptions of Mathura and its surrounding region including Sa#nchi# while eastern variety was found in the inscriptions of the Kausa#mbi region. It may be pointed out that while official records were neatly written since the skilled hands of the royal scribes were employed, the private records were written by skilled as well as by unskilled hands and hence exhibit variety. A distinguishing feature in the script of this period is in regard to the head-mark, viz., the line head-mark in the place of the tapering one of the earlier period. In fact, from the 1st century A.D. onwards, the development of the head-marks is noticed during the next two or three hundred years. These head-marks, deliberately made by the writers, are distinguished as (1) the line head-mark, (2) the block head-mark, (3) notched head-mark, (4) solid square head-mark, (5) hollow square head-mark and (6) solid triangular head-mark38. Other note-worthy features in the development of the script of this period are the tailed variety of ma, the looped variety of sa and the hooked variety of ha noticed in the inscriptions of the period of the Kusha#n?as and later from northern India. These features are more noticeable in the inscriptions of the Gupta period so much so that some scholars called the style with these features as the eastern or northern variety of Gupta script while the style with the older forms of ma, sa and ha as the western or southern variety of the Gupta script. But it must be stated that in as much as these special features of ma, sa and ha begin to appear even in the Kusha#n?a period, it will not be correct to say that they are the characteristic features of the script of Gupta period (4th century A.D.), though they became more popular during the Gupta period39.

The Gupta Script
The next stage in the development of the script of the inscriptions from North India is noticed in the records of the kings of the Gupta dynasty and their contemporaries who belonged to the period 4th to the 6th century A.D. Hence this script is commonly called the Gupta script, a nomenclature given by earlier scholars like Bühler, Burnell and Fleet. D.C. Sircar would prefer to call this script as late Bra#hmi#40. Dani, however, studies the script of this period under the title proto-regional scripts41. The Gupta script is considered to be the successor of the Kusha#n?a script in North India and some of the characteristic features of this script like the looped variety of sa and box-type of ma ae already found in the inscriptions of the Kusha#n?as, as pointed out above. But these features, along with some others, became more common in the inscriptions of the Gupta kings and their contemporaries. Along with old forms, some new forms of letters also begin to appear. Some of these features are: the initial i, along with the old form consisting of three dots, has two new forms - one having two dots one above the other with a vertical bar to the right and the other denoted by two dots in horizontal lines and a curved stroke placed above them. The initial u rounds its angle. The right arms of ga, ta, bha and sa are prolonged. Na has a loop and tha has optionally a line instead of a dot in the centre. La and ha have developed a hooked form while pa. and ba have a notch on the left side.
The Gupta script is divided into two main varieties, viz., northern and southern. The northern variety is further divided into northeastern variety and northwestern variety while the sub-varieties of the southern variety are the western variety and central variety. The chief characteristics of the north-eastern variety are the looped form of sa, cursive ha and la, box-type of ma, etc., while the older forms of these letters are found in the north-western variety. In the southern variety, the letters like a, i, ka and ra develop hook at the bottom, has a tailed la, rectangular form of cha, etc. In the central variety, the characteristic feature is the box-head in the place of top-strokes. The north-eastern variety is met with in the Allahabad pillar inscription of Samudragupta42, Gad?hwa inscription43 of Kuma#ragupta and the Kahaum inscription44 of Skandagupta while the north-western variety is found in the Mathura pillar inscription45 of Chandragupta-II and the Udayagiri cave inscription46 of Kuma#ragupta I. The western variety of the southern variety is noticed in the Sanchi inscription47 and the Juna#gad?h inscription48 of Skandagupta while the central variety is met with in the Eran? stone inscription49 of Samudragupta.
As already pointed out, Dani deals with the script of the inscriptions of the Gupta period in a different way. He considers that the Allahabad inscription of Samudragupta and the Udayagiri cave inscription of Chandragupta-II follow the Kausa#mbi# style of writing because the so-called features of looped sa and box-type ma found in them were already noticed in the Magha inscriptions of the rulers of Kausa#mbi#50. Hence he calls it Kausa#mbi# style of writing. The inscriptions from this period onwards show a tendency to develop regional styles of writing. Thus some inscriptions at Udayagiri were following the Malwa style while others at Mathura were following Mathura style of writing. Similarly, the style of writing found in the Allahabad inscriptions of Samudragupta is not found throughout North India. Hence, according to him, these styles of writing were getting localized and getting affiliated to various cultural regions. Hence, he studies the scripts of this and the following periods under the title 'proto-regional scripts'51.

Scripts of the Post-Gupta Period
From the 6th to about the 10th century A.D., the inscriptions in North India were written in the script which was called by different names. Since the Setters begin to incline from right to left resulting in the formation of acute-angle on the right-end of some letters this script was called acute-angled script by Bu"hler while Fleet named it as Kut?ila52. This acute-angle is noticed in letters like pa, ma, ya and sa. The head-mark of the letters looked like a nail and hence the script was also called nail-headed script. According to Dani, this formation of nail-head is due to the use of the pen as a writing material53. The use of this writing tool assumed the shape of a triangle which was called a 'wedge' besides nail-head. The use of this pen is also responsible for the ornate shape of medial vowels in the letters. The same script is also named Siddhama#trika#, a term used by A1-Bi#ru"ni# in the 11th century A.D. It is also called 'early Na#gari#' because it was the origin of the later Na#gari# or De#vana#gari# script. This acute-angled script is found used in the Bodh-Gaya# inscription54 of Maha#na#man dated 588-89 A.D., some seals of the Maukhari rulers55, Madhuban and Baushkera plates of Harsha56 and the Aphsad inscription57 of Adityase#na, belonging to the 6th-7th century A.D.
Some of the characteristics of this script are: the vowel-length in the initial a is invariably denoted by a curved stroke at the bottom; besides the old form of initial i indicated by three dots, two new forms are found, one in which a curved stroke is drawn in place of the lower data and the other in which two dots are placed below a horizontal bar; a new shape of ka is developed in which the end of the vertical line shoots up to the left side and meets the cross-bar; and tripartite ya is found in most of the earlier records, while the bi-partite form is met with in the later records.
The Siddhama#trika# or the early Na#gari# script was introduced in the western parts of the Deccan (Maharashtra) in about the 8th century A.D. replacing the southern or Telugu-Kannad?a script which was prevalent there. It is found, further south, used in a few inscriptions like the pat?t?adkal (Karnataka) inscription58 of the Cha#lukyas of Ba#da#mi, the Alampur (Andhra Pradesh) inscription59 of the same dynasty and the Sa#l?uvankuppam (Tamil Na#d?u) inscription of the Pallavas, all belonging to the 8th century A.D.
From 8th century A.D. the Siddhama#trika# script developed into the Sa#rada# script in the Ka#smir and Panjab, the proto-Bengali or Gaud?i# in Bengal and Orissa.
The later phase of the Siddhama#trika# script which was prevalent in the other parts of North India is found in the Nalanda inscription60 of Yasovarmade#va, the copper-plates61 of Dharmapa#la and De#vapa#la. It is also found in some Ra#sht?raku#t?a records like the Multai plates62. These records can be assigned to the 8th century A.D. In these inscription, we find that the headline of the letters is sometimes longer. The other features are: initial a# is indicated also by a little vertical line starting from the top-line; ka is usually of the looped variety; a new shape of ja is met with in which the central stroke curves downwards. The older form is also found in the Ra#sht?raku#t?a records; ma has also a new shape where there is a loop in the left corner and a tail at the right-end and ya is bi-partite with a tail at the right-end.
The Gaud?i# or proto-Bengali script which is derived from Siddhama#trika was prevalent in Bengal, Bihar, Assam, Orissa and Nepal between the 10th and 14th century A.D. In Bihar and Nepal, this script was later on replaced by Na#gari# script while in Bengal, it developed into Bengali script which was also used in Assam. In Orissa, the Oriya script fully replaced the Gaud?i script about the 14th century A.D., though the influence of Oriya script is found in earlier period. Some records from Ganjam and Srikakulam region exhibit the influence of Siddhama#trika#, Gau?di¯ and Telugu-Kannad?a scripts during the period 7th to 11th century A.D. These records belong to the Eastern Ganga dynasty and this admixture of the scripts is called Kalinga or later Kalinga script.

Na¯gari¯ or De¯vana¯gari¯ Script
The Siddhama¯trika¯ or early Na¯gari¯ script gave rise to the eastern Na¯gari¯ script in Uttar Pradesh and Madhya Pradesh from the 10th century A.D. onwards while the western Na¯gari¯ script was developed in Rajasthan, Gujarat and Maharashtra. The name De¯vana¯gari¯ is now used for writing Hindi all over India and also Sanskrit. But this name De¯vana¯gari¯ appears to have come into vogue during the last two or three hundred years only. The word Na¯gara, however, is used for a script in Malwa region by Albi¯ru¯ni¯ in tile 11th century A.D. According to some scholars, the name Na¯gari¯ for the script is to be derived from the word Nagara which stood for Pa¯t?a¯liputra, the capital of Magadha country in ancient period. We cannot however, be certain about the origin of the words Na¯gari¯ and De¯vana¯gari¯.
From 10th century A.D. onwards Na¯gari¯ script was used widely in northern India and it gave rise to many local scripts during medieval and modern periods. The Sa¯rada¯ script which was prevalent in the Ka¯smir valley from the 8th century A.D. onwards was replaced by Na¯gari¯ script in the late medieval period. By the 11th century A.D., Na¯gari¯ was used in Maharashtra, Gujarat, Madhya Pradesh and uttar Pradesh in the inscriptions of the Sila¯ha¯ras, Cha#l?ukyas, Parama¯ras and Prati¯ha¯ras. The earliest use of the Na¯gari¯ script is found in the Kanheri (Maharashtra) inscription63 of Sila¯ha¯ra Pullasakti dated in 851 A.D. In North India, it appears for the first time in the copperplate grant64 of Vina¯yakapa¯la. The Na¯gari¯ script is also found in the copper-plate grants of the Ra¯sht?raku¯t?as, the Cha¯l?ukyas of Kalya¯n?a, in the stone inscriptions and copper-plate grants of the Ya¯davas of De¯vagiri, the Sila¯haras and the Vijayanagara kings who ruled in Karnataka from the 8th century to the middle the I6th century A.D. While some of the early copper-plates of the Ra#sht?raku#t?a kings are in proto-Na¯gari¯ or early Na¯gari¯ script, the later copper-plates are in regular Na¯gari¯ script.
The main characteristic of this Na¯gari¯ script is the substitution of the straight top-strokes for the wedges or head-marks. By about 12th century A.D. the Na¯gari¯ script has developed the letters which are not much different from the modern script. Though the general pattern of the Na¯gari¯ script is common, there are slight variations in different regions where it developed. Thus there are two types of a, ja, jha, etc., as found in the inscriptions. Similarly, the Na¯gari¯ script used in the Jaina manuscripts would differ from the script found in the epigraphs. The Na#gari# script used in the records of the Cha#lukyas of Kalya#n?a and the Vijayanagara kings in Karnataka is called Nandi-na#gari#. The Na#gari# script of the North India only is generally classified into Na¯gari¯ of - (1) Central India, (2) Gujarat, (3) Rajasthan and (4) Ganges Valley, there is overlapping of the special features of one region in the script of the other region. The use of Nandina¯gari¯ script is mostly confined to the Karnataka region. According to some scholars, its name Nandina¯gari¯ is to be derived from Nandina#gara, a place identified with modern Nanded in Marathva#d?a# region of Maharashtra. But according to others, the name of the script is due to the Jaina teachers whose names end in -nandi like Simhanandi, Jat?a#nandi, Ma#ghanandi, etc., who introduced this script in Karnataka when they came from the north. We also know of a sangha called Nandisangha in Karnataka. This Nandina¯gari¯ script consists of the special Na¯gari¯ letters of a, jh, n?a, etc. But they are no longer used in Karnataka after the introduction of printing.

South Indian Scripts
Grantha, Tamil? and Vat?t?eluttu Scripts:
We have already dealt with the script of the cave-inscriptions found in Tamil Na#d?u. These inscriptions are mostly short records and they do not mention the name of any dynasty which ruled in that region during that period, viz., 2nd century B.C. to 3rd century A.D. From about the 7th century A.D. onwards we begin to get a number of inscription belonging to the Pallavas, the Cho#l?as and the Pa#n?d?yas. These records are written in three varieties of scripts called (1) the Grantha script, (2) the Tamil script and (3) the Vat?t?eluttu script. The Grantha script is used in writing the Sanskrit records and also the Sanskrit portion in Tamil records while the Tamil script is noticed in the Tamil inscriptions found in the northern part of Tamil Na#d?u belonging to the Pallavas and their successors the Cho#l?as. In these records, for some Sanskrit words or letters, the Grantha script is used and hence the script is sometimes called the Grantha-Tamil script. The Vat?t?eluttu script was prevalent in the southern portion of Tamil Na#d?u and also in Kerala till late period, though in some parts it was replaced by the Tamil script as early as 11th or 12th century A.D. In bilingual inscriptions employing Sanskrit and Tamil, for writing words of Tamil origin occurring in Sanskrit portion, the Grantha or the Grantha-Tamil script is used while the Vat?t?eluttu script is used for writing the Tamil portion65. The Grantha script has all the letters like the Na¯gari¯ script or the Telugu-Kannad?a script. But the Tamil and the Vat?t?eluttu scripts possess only those letters which require to represent the genius of the Tamil language. Thus they do not possess the symbols to represent s, sh, s, h, etc. While some scholars regard Vat?t?eluttu script as the cursive form of the Tamil script, others consider that all the three scripts, viz., the Grantha, Tamil and Vat?t?eluttu are independent and directly derived from the Bra#hmi# script66. In the Malayalam region, the Vat?t?eluttu script developed into a still more cursive script called Ko#leluttu during the period 14th and 15th centuries A.D. At the same time, the modern Malayalam script developed out of the Grantha Script. It may also be pointed out that some inscriptions are written in mixed scripts, i.e., there are a few Vat?t?eluttu letters in an inscription written in Tamil script or a few Tamil letters in a record written in Vat?t?eluttu script.

Telugu-Kannad?a Script:
The early form of the Telugu-Kannad?a script is found in the inscriptions of the early Kadambas of Banava#si and the early Cha#lukyas of Ba#da#mi in the west and the sa#lanka#yana and the early Eastern Cha#lukyas in the east of the Kannad?a and Telugu speaking areas during the period 4th to the 7th century A.D. During the period 7th to 10th century A.D. further developments of this script is noticed in the inscriptions of the Cha#lukyas of Ba#da#mi and the Imperial Ra#sht?raku#t?as in the west and in the records of the Eastern Cha#lukyas in the eastern part. A characteristic feature of this script is the tendency to make the letters round so that the letters like ka and ra are closed on the left, na has sometimes the looped form, ba is opened at the top and the ma#tras are occasionally attached below the consonants. This script continued to be used in the Kannad?a and Telugu speaking areas, during later period also, with certain modifications and more rounded forms. Since the same script was prevalent in this entire region, used both in the Kannad?a same as well as in the Telugu inscriptions, it was called as Telugu- Kannad?a script. It was only after 15th century A.D. or so that we begin to notice the modern characteristics of Kannad?a script and the Telugu script so that they could be distinguished from each other.

Box-Headed Script:
In Central India, i.e.. in Madhya Pradesh, in the Eran? inscription of Samudragupta, the Udayagiri inscription of Chandragupta II, the copper-plate grants of the kings of the Sarabhapura and Ko#sala, the records of the Va#ka#t?akas which belong to the period 4th to the 7th century A.D. show a special feature in that the heads, of the letters have small squares which resemble small boxes and hence the script used in them is called 'box-headed'. This script is used in some inscriptions of the Kadambas and Pallavas in the southern region also.

General Remarks
It will be noticed that all the scripts described above, which were prevalent in North India as well as in South India were derived from the common script, viz., the Asokan Bra#hmi# script of the 3rd century B.C. They have developed different characteristics in different parts of the country during the several centuries of their history. In South India, what is known as Tamil- Bra#hmi# inscriptions were also derived from the same Asokan Bra#hmi# script with some modifications to suit the genius of the Tamil language. This feature is noticed even in the later inscriptions of the sa#tava#hana, sa#lan#ka#yana, Vishn?ukun?d?in, early Gan#ga, early Kadamba, early Cha#lukya and early Pallava inscriptions in south India where some Dravidian place-names are written by adopting suitable letters in the usual script. Some of these letters represent the sounds like l?, l- and r- while in Tamil inscriptions two letters are used to represent the sound n when it occurs initially or medially. It may also be pointed out, in this connection, that the Telugu- Kannad??a script does not indicate the distinction between short e and long e# and short o and long o# till very late period. This distinction, which exists in Telugu and Kannad?a languages, has to be understood from the context only.

Scripts Abroad
The Bra#hmi# mi script is also the origin of the scripts found in Sri# Lan#ka# (Ceylon), Burma, and South-eastern Asian countries like Ja#va#, Suma#tra, Cambodia. This Bra#hmi# script traveled to Sri# Lan#ka# probably during the period of Asoka himself and has developed into the modern Sinhalese script after passing through various stages. Similarly, the present alphabets of Burma, Thailand, Indonesia, Laos, Annam, Cambodia, etc., are all derived from Bra#hmi# script. The inscriptions in the areas of Ja#va#, Suma#tra# and Cambodia, found during the period 4th to 6th century A.D. are couched in beautiful classical Sanskrit language and in versified form. Some scholars attribute the script of these records as influenced by the Pallava-Grantha script while others think that the Gupta script of the Bengal region has influenced them. But, according to the latest view, the alphabet of these records closely resemble that of the records of the Kadambas and the early Cha#lukyas of Ba#da#mi; and since the use of saka era is also found there, it is suggested that the script traveled to those parts from the west-coast of Karnataka67.

FOOTNOTES

1. Cf. ind. Ant., Vol. XXXIII, Appendix 6, p.1.
2. Beal, Si yu ki, I, p.77.
3. T.P. Verma, Pra#chi#na Bha#rati#ya Lipisa#stra (1970), pp.19-20.
4. Ind. Ant., Vol.xxxv, p.253.
5. Ind. Ant., Vol.XXXV, p.253.
6. Taylor, The Alphabet, Vol.II, pp.304 ff.; Vol.I, pp.335-346; Diringer, The Alphabet, p.335, Burnell, Elements of South-Indian Palaeography, p.9; Buhler, Indian Palaeography, Ind. Ant., Vol.xxxIII, App. p.11.
7. Ep. Ind., Vol.xxxIII.
8. D.C. Sircar, Select Inscriptions (1962).
9. Marshal, Mohenjodaro and the Indus Valley Civilization, Vol.II, p.427; D.C. Sircar, The Inscriptions of Asoka (1957), p.25.
10. Cunningham, Inscriptions of Asoka, pp.51 ff.
11. Ind. Ant., Vol.xxxv, p.255.
12. JRAS, (New Series), Vol.XIII (1881), p.112.
13. JBORS, Vol.VI (1920), pp.188 ff.
14. Indian Palaeography (Hindi), p.30.
15. Indian Palaeography,, (English), Pt.l, p.50.
16. JRAS, (New Series), Vol.XIII (1881), p.115.
17. South Indian Temple Inscriptions, Vol.III, Pt.II, Appendix.
18. The Palaeography of the Bra#hmi# Script (1971), pp.4 ff.
19. Indian Palaeography, pp.22 ff.
20. The Alphabet, Vol.II, p.289.
21. JRAS, (New Series), Vol.XVI (1884), p.325.
22. C.S. Upasak, Hist, and Pal, of Mauryan Bra#hmi# Script, p.15.
23. C.S. Upasak, History of Mauryan Bra#hmi#, p.29.
24. Dani, Indian Palaeography, pp.48 and 51.
25. C.S. Upasak, History of Mauryan Bra#hmi#, pp.29 ff.
26. D.C. Sircar, Select Inscriptions (1962), pp.77, 79, 81 and 82.
27. Dani, Indian Palaeography, pp.50-51.
28. T.V. Mahalingam, Early South Indian Palaeography, p.161.
29. Early South Indian Palaeography, pp.135-38.
30. Indian Palaeography, pp.73-74.
31. South Indian Palaeography, p.130.
32. Ibid.
33. Palaeography of Bra#hmi# Script, p.40.
34. Palaeography of Bra#hmi# Script, p.40.
35. Dani, Indian Palaeography, p.52.
36. T.P. Verma, Palaeography of Bra#hmi# Script, p.82.
37. Palaeography of Bra#hmi# Script, pp.80 ff.
38. Dani, Op. cit., p.81.
39. G.S. Gai, Journ. Ep. Stud., Vol.VI, p.16, Note 19.
40. Journ. Anc. Ind. Hist., Vol.IV, p.113. According to him, the early Bra#hmi# script in North India covers the period 3rd cent. B.C. to 1st cent: B.C. and middle Bra#hmi#, 1st cent B.C. to 3rd cent. A.D., while in South India early Bra#hmi# continued upto 1st cent. B.C in northern parts but upto 3rd cent. A.D. in southern parts, middle Bra#hmi# from 1st cent. A.D. to 4th cent. A.D. and late Bra#hmi#mf from 4th cent. A.D. to 6th cent. A.D.
41. Ind. Pal., pp.108 ff. He divides the proto-regional scripts into four main divisions, viz.
(1) Northern scripts, (2) Scripts of Gujarat, Rajasthan and Central India, (3) Deccani scripts and (4) Scripts of South India.
42. CII, Vol.III (1888), Plate I.
43. Ibid., Plate 1VC
44. Ibid., Plate IXA.
45. Ep. Ind., Vol.XXI, Plate facing p.8.
46. CII, Vol.III (1888), Plate XXXVIIIA.
47. Ibid., Plate IIIB.
48. Ibid., Plate VIII.
49. Ibid., Plate IIA.
50. Ind. Pal., p.100.
51. Ibid., p.108.
52. CII, Vol.III (1888), p.201.
53. Ind. Pal., p.113.
54. CII, Vol.III (1888), Plate XII.
55. Ibid., Plate XXX, A and B.
56. Ep. Ind., Vol.I, p.67 ff. and Plate and Ibid., Vol.IV, pp.208 ff. and Plate.
57. CII, Vol.III (1888), Plate XXVIII.
58. Eg. Ind., Vol.III, pp.1 ff. and Plate.
59. Ibid., Vol.XXXV, pp.121 ff. and Plate.
60. Ibid., Vol.XX, Plate facing p.43.
61. JASB, 1908, Plate IV and Ep. Ind., Vol.XVII, Plate between pp.320-21.
62. Ind. Ant., Vol.XVIII, Plate between pp.234-35.
63. CII, Vol.VI (1977), pp.1 ff and Plate 1.
64. Ind. Ant., Vol.XV, Plate facing p.140.
65. T.V. Mahalingam, Early South Ind. Pal., p.299.
66. Ibid. A few inscriptions belonging to the period 4th to the 6fh cent. A.D. are written Grantha, Tamil and Vat?t?eluttu scripts, cf. Ibid., p.307.
67. Cf. Journ. Anc. Ind. Hist., Vol.IV, pp.132-33.