Prev | Home | Next

Although Sanskrit is the oldest language in India as furnished by the Vedic literature, the language of the earliest written records, viz., the Asokan edicts is Prakrit. As seen before, since the Indus or the Harappan script has not yet been successfully deciphered, we are not in a position to decide about the language of this script in spite of various views put forward by different scholars. Besides Prakrit, Asokan edicts are written in Greek and Aramaic languages also. One edict written entirely in Greek script and language and another edict written in Greek and Aramaic script and languages are found in Skandar in Afghanistan while a record in Aramaic script and language found in Taxila (now in Pakistan) is attributed to Asoka. All the edicts of Asoka in the Kharo#sht?hi# and Bra#hmi# scripts are written in the Prakrit language. Thus originally the epigraphical language of India can be said to be Prakrit and Sanskrit was used in the inscriptions only at a later period. Edicts found at Dhauli, Jaugad?a (both in Bihar) and Er-r-agud?i (in Andhra Pradesh) are written in what is called the Magadha dialect some of the characteristics of which are l representing Sanskrit r, s representing all the three slants s, sh and s, and the double consonants changing to single ones. And the retention of r and the occasional change of l to r met with in e western and north-western edicts found at Girnar, Sopara and Shahbaz-garhi and Mansehra (now in Pakistan)1. In the edict2 found at Delhi, r is changed to l but an interesting feature is the change of s into ch in the word chake (Sanskrit sakyah?). This feature is- noticed in the Gujarra (Madhya Pradesh) edict3 where r is retained except in the word Chilathitike. In the Sopara fragment edict4, it is found that besides the retention of r, the change of l to r occurs in the words Mangaram and phare for Sanskrit Mangalam and phale which may also be due to the mistake of the scribe. In the Minor Rock edicts and Rock edicts found at one and the same place, viz., Er-r-agud?i5 , a remarkable difference is noticed in the linguistic features between the two sets of edicts. The language of the minor edicts resembles the Magadhan dialect in that n? is changed to n and s and sh to s. But r is retained without being changed to l in words like sa#tireke, a#ra#dhetava, etc. The language of the rock-edicts resembles that of the Dhauli, Jaugad?a and Kalsi edicts but also shows some of the features found in the Girnar, Shahbazgarhi and Mansehra edicts. Here the consonant n? is used for Sanskrit n? and n. It has been suggested that the differences between the minor rock-edicts and rock-edicts at Er-r-agud?i were due to the fact that they were received separately on different dates and occasions6. A study of the linguistic features of these edicts does not support the view propounded by Dr. M.A. Mehandale that "though the edicts which were to be inscribed at various places in Asoka's empire were the same when they were issued from Magadha, they were altered actually at the time of inscribing them at various places so as to suit the requirements of the local dialects"7. The edicts at Er-r-agud?i as well as the Minor rock-edicts found in Karnataka exhibit Magadhan as well as other linguistic features. Just as there were no regional varieties in the Asokan Bra#hmi# script as pointed out above, we feel that there were no regional or local varieties in the language of the Asokan edicts found in various parts of the country.
After the period of Asoka, the use of the Prakrit language continued in the inscriptions for a few more centuries. In North India Prakrit was replaced by Sanskrit about the end of the 3rd century A.D. while it happened about a century later in South India. But even before this period some of the inscriptions, though written in Prakrit language were influenced by Sanskrit while some records were entirely couched in Sanskrit. The epigraph of the Kusha#n?a kings are found in a mixture of Prakrit and Sanskrit while the Mathura inscriptions of the time of saka so#d?a#sa belonging to the first quarter of the 1st century A.D. contain verses in classical Sanskrit like sa#rdu#lavikri#d?ita8. Even earlier than this, the Besnagar (Madhya Pradesh) pillar inscription of Heliodoras who was an ambassador from the Indo-Greek king Antialkidas at the court of king Bha#gabhadra of Vidisa#, belonging to the end of the 2nd century B.C., though written in Prakrit language exhibits some influence of Sanskrit9. And the Gho#sun?d?i# stone inscription from Rajasthan belonging to the latter half of the 1st century B.C. and of the time of king Sarvata#ta is couched in Sanskrit language10. The Junagadh (Gujarat) inscription of the saka king Rudrada#man and dated in 150 A.D. is one of the earliest prose epigraphs which is written in beautiful Sanskrit prose in the classical ka#vya style11. From the eastern part comes the Ayodhya inscription12 of Dhanade#va (latter half of 1st century B.C.) which is written in Sanskrit slightly influenced by Prakrit. Inscriptions of slightly later period (1st century to 3rd century A.D.) like the Kailvan (Bihar State) inscription13, the inscriptions of Kausa#mbi and the Bandorgarh records are written in Sanskrit slightly influenced by Prakrit, though the later inscriptions of Bandorgarh are entirely in Sanskrit. From the fourth century onwards, the Guptas came to power at Pataliputra and they were great patrons of Sanskrit language and literature and hence Sanskrit became the language of the inscriptions. It was during this period only that great poets like Bha#sa and Ka#lida#sa flourished.
In South India, Prakrit was used in the inscriptions till the century A.D., though in a few records of the Ikshava#kus of Na#ga#rjunakon?d?a Sanskrit is employed. The inscription14 of Yajna-sa#takarn?i (2nd century A.D.) from Amaravati is considered to be the earliest Sanskrit inscription from Andhra Pradesh discovered so far. The earlier inscriptions (4th century A.D.) of the sa#lanka#yanas of the Telugu region are in Prakrit while their later records (belonging to the 5th century A.D.) are written in Sanskrit. In the Kannad?a -speaking area, the Chandraval?l?i inscription of Mayu#rasarman, regarded as the founder of the Kadamba dynasty of Karnataka, is written in Prakrit language and is assigned to the 4th century A.D. But all the later inscriptions of this dynasty are in Sanskrit language only.
The early copper-plate inscriptions of the Pallavas of Ka#nchi# belonging to the 4th century A.D. are written in Prakrit language and hence they are known as the Pallavas of the Prakrit charters as against the Pallavas of the Sanskrit charters of a later period (5th and 6th century A.D.). This Prakrit is of a literary style. While the entire texts in the Mayidavolu and Hirehad?agali plates are couched in Prakrit, their seals, however are in Sanskrit. A mangala passage in the Hirehad?agali plates is also in Sanskrit. Thus both in the records of the Kannad?a -Telugu areas and in the Pallava charters of the early period Prakrit prevailed upto the 4th century A.D. or so when it was replaced by Sanskrit.
In the cave inscriptions of Tamilnadu, the language of the southern Bra#hmi# inscriptions varying in dates from about the 2nd century B.C. to about the 2nd or 3rd century A.D. is said to be early Tamil. But there is no unanimity in this regard. In fact, the language of these records has exercised the minds of several scholars for the last sixty years or so and, though more cave-inscriptions have been discovered during this period, they have not been successfully deciphered and interpreted so far. H. Krishna Sastri and K.V. Subramanya Ayyar of the Government Epigraphist's Office were the pioneers in the study of these southern cave inscriptions. They thought that the script employed here followed that of the Bhat?t?iprolu casket inscriptions and it was K.V. Subramanya Ayyar who first suggested that the language of these records was Tamil in a paper presented by him at the Third All India Oriental Conference held at Madras in 1924. But his views as well as the interpretation of several letters were not acceptable to other scholars. In 1946, the official view of the Archaeological Survey of India was given as follows: "The exact nature of the language of these inscriptions is still open to question but they appear to be in early Tamil (as distinguished from the Tamil found in early Tamil literature, as well as modern Tamil) with a sprinkling of Prakrit15.
T.V. Mahalingam, Iravathan Mahadevan and T.N. Subrahmanyam have studied these Tamil?-Bra#hmi# cave inscriptions in detail and have also come to the conclusion that the language of these records is Tamil, though they differ in their readings and interpretation. T.V. Mahalingam has made a detailed study of this problem and has come to the conclusion that his analysis would show that the language of these records have only the first letter (varga-prathama), viz., k., c, t?, t and p of the Sanskrit yarn?ama#la# while the other letters as well as s, sh and h are absent16. In addition, some new symbols are also found in these records which are not noticed in the records employing Asokan Bra#hmi# script. Some of these symbols represent long i#, n, l, l?, r-, nn which are peculiar to the genius of the Tamil language. Thus, the cave inscriptions of the South following the Asokan script in general, have introduced a few characteristics and deviations to suit the genius of the local language, viz., Tamil. This view of T.V. Mahalingam is shared by the other scholars mentioned above, though they differ in their reading and interpretation of the records.
As stated above, Prakrit was the language of the earliest inscriptions, i.e., those of Asoka, throughout the country and also for some time after the period of Asoka and Sanskrit replaced Prakrit in the inscriptions by about the end of the 3rd century A.D. in North India and by about the end of the 4th century in South India17. During this intervening period, we get some records written in Prakrit influenced by Sanskrit in the earlier period and in Sanskrit influenced by Prakrit in the later period. And, in exceptional cases, some later inscriptions like the Ghat?iya#la inscription18 of Pratihara Kakkuka dated 862 A.D. is written in Prakrit language. The Dhar (Madhya Pradesh) inscription of the time of Bho#ja (c. 1000-55 A.D.) contain the text of the Prakrit poem Ku#rmasataka ascribed to Bho#ja.
Some of the early inscriptions which are couched in Sanskrit (though sometimes with slight influence of Prakrit) are the Mathura inscription 19 of saka king so#d?a#sa from the north, the Gho#sun?d?i record20 of Ga#ja#yana Sarvata#ta from western India, the Ayodhyaya epigraph21 of Dhanade#va from eastern India and the Na#ga#rjunakon?d?a inscriptions22 of the Ikshava#ku ruler Ehuvala sa#nta#mu#la. These records range in date from the 2nd half of the 1st century B.C. to the end of the 3rd century A.D. But, as already pointed out, the Junagadh inscription of Rudrada#man dated 150 A.D. is couched in beautiful Sanskrit prose of Ka#vya style.
From the 4th century onwards, with the rise of the Guptas, Sanskrit became the predominant language of Indian epigraphs. It was adopted as the court language by the rulers of this dynasty and some of them like Samudragupta have been stated to be proficient in that language. Some of the celebrated poets like Bha#sa and Ka#lida#sa flourished during the Gupta period and the great epics Ra#ma#yan?a and Maha#bha#rata as well as of the Pura#n?as are stated to have assumed their final form during this period. The contemporary rulers of the Guptas in Central India and parts of the Deccan like the Va#ka#t?akas, the Kadambas and Gangas of Karna#t?aka and the Pallavas in South India also employed Sanskrit as the court language as found in their inscriptions. As regards the sa#lanka#yanas in the Andhra region, while their earlier records (4th century A.D.) are written in Prakrit language, the later ones (5th century A.D.) are couched in Sanskrit language. Thus Sanskrit became the epigraphic language of the country from the 4th century onwards, replacing Prakrit. The Ta#l?agunda pillar inscription 23 of the time of the Kadamba king sa#ntivarman (5th century A.D.) composed by poet Kubja and the Aihole inscription24 of the Ba#da#mi Cha#lukya king Pulike#sin II (634 A.D.) composed by the poet Raviki#rti are fine specimens of classical Sanskrit found in early inscriptions of Karnataka.

Regional Languages
Sanskrit continued to be the language of inscriptions of all parts of India till late medieval period. In the meanwhile the regional languages also began to be used in the inscriptions. We have already noticed that the language of the cave inscriptions found in Tamil Nadu is considered to be an early form of Tamil language and as such is the earliest Dravidian language to be used in inscriptions. At a later period the copper-plate charters of the Pallavas, the Cho#l?as and the Pa#n?d?yas are written in both Sanskrit and Tamil languages, though some of the Pallava grants like Vunnaguruvayapalem plates25 of Pararnesvaravarman I (7th century A.D.) and Reyjru plates26 of Narasimhavarman II (8th century A.D.) are written entirely in Sanskrit. The Kuram plates27 of Pallava Parame#svaravarman, and the Bahur plates28 of Nripatun#gavarman (9th century A.D.) are written in Sanskrit and Tamil languages. The larger Leiden plates29 of the Cho#l?a king Ra#jara#ja I (l0th-11th century A.D.) are written partly in Sanskrit and partly in Tamil while the smaller Leiden plates30 of Kulottun#ga I (11th-12th century A.D.) are written entirely in Tamil. While early Pa#n?d?ya inscriptions are couched partly in Sanskrit and partly in Tamil languages31, the later records, of the dynasty are written only in Tamil language32.
Next to Tamil, Kannad?a is found used in the inscriptions dating from about the 6th century A.D. onwards. The Halmid?i (Belur Taluk, Shimoga District) inscriptions33 and the Vaishnava cave inscription34 at Ba#da#mi (Bijapur District) in Karnataka State are considered to be the earliest epigraphs written in Kannad?a language. While all the copperplate grants of the early Cha#lukyas of Ba#da#mi are written in Sanskrit language, most of their stone inscriptions consisting of private records are in Kannad?a language. Similar is the case with the records of the other imperial dynasties of Karna#t?aka like the Ra#sht?raku#t?as, the later Cha#lukyas, the Kal?achuris, the Ya#davas, Hoysal?as, etc., with a few exceptions. Thus the British Museum plates35 of the Ra#sht?raku#t?a king Govinda III dated saka 726 or 804 A.D. is written in Kannad?a language. For a long time this inscription was considered as the earliest copperplate grant36 in Kannad?a language. But recently another copper-plate grant, belonging to the Al?upa ruler Al?uvarasa II, has been discovered at Bel?man?n?u in South Kanara District of Karnataka State. Though it is not dated, it is assigned, on palaeographical grounds, to the 8th century A.D. and thus it becomes the earliest copper-plate inscription in Kannad?a language discovered so far. There are many inscriptions which are written partly in Sanskrit and partly in Kannad?a and even in the records which are composed in Kannad?a language only, the invocatory or the benedictory verses at the beginning and the imprecatory verses at the end are in Sanskrit language. It is also interesting to note that the inscription37 of Jinavallabha, brother of the famous Kannad?a poet Pampa, discovered a few years ago, is written in three languages, viz., Sanskrit, Kannad?a and Telugu.
Telugu language is used in inscriptions belonging to the 6th or 7th century A.D. while some Telugu place-names are mentioned in earlier records. The kalamalla inscription38 of Erikal-Muthura#ju Dhanañjaya assigned to the last quarter of the 6th century A.D. is considered to be the earliest record completely written in Telugu. This and other records39 of the Re#na#n?d?u Cho#l?a rulers from the Anantapur and Cuddapah districts of Andhra Pradesh furnish the earlier stone inscriptions written in Telugu language, the Madras Museum plates40 of Ballayacho#d?a of the Telugu Cho#d?a family belonging to the middle of the 9th century A.D. furnishes the earliest copper-plate inscription written in Telugu language.
It is only from 15th century A.D. onwards that Malaya#l?am language appears in the inscriptions, although in an earlier Tamil inscription41 of the 13th century A.D., Malaya#l?am influence is noticed. Thus the Attingal inscription42 of 1452 A.D. and the Tonnal inscription43 of 1474 A.D. are written in Malaya#l?am language.
Amongst the new Indo-Aryan languages used in inscriptions, Mara#t?hi is found used in early records of the 11th century A.D. and the earliest epigraph in which this language is used is the Dive Agar copper-plate inscription44 dated saka 982 or 1060 A.D. 45 Mara#t?hi language became popular in the inscriptions of the sila#ha#ras and the Ya#davas of Devagiri in the Marat?hi-speaking area.
The Oriya language begins to appear in inscriptions from 13th century A.D. onwards, though the influence of this language in records written in Sanskrit language appears as early as 10th century A.D. 46 Some copper-plate grants are partly in Sanskrit and partly in Oriya languages while the Veligalani grant47 of the Gajapati king Kapile#svara (15th century A.D.) is written in Oriya, Sanskrit and Telugu.
The use of Hindi language is traced to 11th century A.D. on the basis of a Jaina image inscription48 found at Shyopur in Madhya Pradesh. But most of the other Hindi records, which also come from Madhya Pradesh region, belong to the medieval period of 15th or 16th century A.D. only.
Gujara#ti# is used in the records of the 15th century A.D. hailing from Kathiawar region while a few inscriptions of earlier date from the same area are written in Sanskrit and Gujara#ti#49.
In the Bengali-speaking region, the epigraphic records written in. Sanskrit and seldom in Bengali language. During the late medieval period, the copper-plates50 of Tripura king Govindama#n?ikya (15th century A.D.) are mainly written in Bengali language.


1. D.C. Sircar, Indian Epigraphy, p.40.
2. Ep. Ind., Vol.XXXVIII, pp.1 ff.
3. Ibid., Vol.XXXVI, pp.205 ff.
4. Ibid., pp.211 ff.
5. Ibid., Vol.XXXII, pp.1 ff.
6. Ibid., Vol.XXXII, p.4.
7. Cf. G.S. Gai, Proceedings of the Seminar on Prakrit Studies (Poona), 1970, pp.121-23.
8. Ep. Ind., Vol.II, p.200.
9. ASI, A.R., 1908-09, p.126.
10. Ep. Ind., Vol.XVI, p.27.
11. Ibid., Vol.VIII.
12. Journ. Anc. Ind. Hist., Vol.IV, pp.7 ff.
13. Ep. Ind., Vol.XVI.
14. Journ. Anc. Ind. History, Vol.IV (1971), pp.7-8.
15. Ancient India, No.2, p.109.
16. Early South Indian Palaeography, pp.134-38.
17. D.C. Sircar, Indian Epigraphy, pp.39-40.
18. Bhandarkar's List, No.31.
19. Ep. Ind.. Vol.IX, p.247.
20. Ibid., Vol.XVI, p.27.
21. Ibjd., Vol.XX, p.57.
22. Ibid., Vol.XXI, p.62.
23. Ep. ind., Vol.VIII, pp.81 ff.
24. Ibid., Vol.VI, pp.1 ff.
25. Ep. Inci., Vol.XXXII, pp.91 ff.
26. Ibid., Vol.XXIX, pp.89 ff.
27. SIL, Vol.I, pp.144 ff.
28. Ep. Ind., Vol.XVIII, pp.5 ff.
29. Ibid., Vol.XXII, pp.213 ff.
30. Ep. Ind., Vol.XXII, pp.267 ff.
31. Cf. SIL, Vol.III, pp.441 ff.
32. An. Rep. Ind. Ep., 1946-7, No.A.30.
33. Mys. Arch. Rep., 1936, pp.72-81.
34. Ind. Ant., Vol.X, p.59.
35. Ep. Ind., Vol.XXXIII, pp.327 ff.
36. Journ. Ep. Soc. Ind., Vol.4, pp.91 ff.
37. Prabuddha Karna#t?aka (Kannad?a), Vol.53 (No.4), pp.73 ff.
38. Ep. ind., Vol.XXVIl, pp.221 ff.
39. Ep. Ind., Vol.XXVII, pp.220 ff.
40. Journ. Ind. Hist., Vol.XV, p.254.
41. A.R. Ep., 1958-59, No.13.464.
42. Trav. Arch. Ser., Vol.VI, p.80, No.64.
43. Ibid., pp.34-35, No.23.
44. Ep. Ind., Vol.XXVIII, pp.121 ff.
45. It is observed that the Marmuri grant of Cha#lukya family (Journ. Bomb. Hist. Soc., Vol.II, p.214), dated 974 A.D. contain specimens of Mara#t?hi language but this record considered spurious.
46. D.C. Sircar, Ind. Ep. (1965), p.58.
47. Ep. Ind., Vol-XXXIII, pp.275 ff.
48. An. Rep., Arch. Dept., Gwalior State, V.S.1992, No.39
49. D.C. Sircar, Ind. Ep. (1965), p.56.
50. A.R. Ep., 1951-52, Nos.A 13 ff.