them is the theory that it was derived from the oldest north-Semitic alphabet, which prevailed from Phoenicia to Mesopotamia, and may, it is held, have been introduced into India by traders at some time about 800 B.C. It is, however, admitted that the earliest known form of the Brāhmī is a script framed by Brāhmaṇs for writing Sanskṛit. Also, the theory is largely based on a coin from Ēraṇ, in the Saugar district, Central Provinces, presenting a Brāhmī legend running retrograde from right to left; from which it is inferred that that was the original direction of this writing, and that the script eventually assumed the other direction, which alone it has in the inscriptions, after passing, like the Greek, through a stage in which the lines were written in both directions alternately. But we can cite many instances in which ancient die-sinkers were careless, wholly or partially, in the matter of reversing the legends on their dies, with the result that not only syllables frequently, but sometimes entire words, stand in reverse on the coins themselves; moreover, the Ēraṇ coin, being one of the earliest known Indian coins bearing a legend at all, may quite possibly belong to a period before the time when the desirability of working in reverse on the dies presented itself to the Indian die-sinkers. In all the circumstances, the evidence of the Ēraṇ coin cannot be regarded as conclusive; and we require some inscription on stone, or at least some longer record on metal than a brief legend of five syllables, to satisfy us that the Brāhmī writing ever had a direction different from that which it has in the inscriptions. Further, if there is any radical connexion between the Brāhmī and the Semitic alphabet indicated above, so many curious and apparently capricious changes must have been made, in adapting that alphabet, that it would seem more probable that the two scripts were derived from a joint original source. In view of the high state of civilization to which the Hindus had evidently attained even before the time of Chandragupta, the grandfather of Aśōka, it must still be regarded as possible that they were the independent inventors of that which was emphatically their national alphabet. The Brāhmī alphabet is the parent of all the modern Hindu scripts, including on one side the Nāgarī or Dēvanāgarī, and on the other the widely dissimilar rounded forms of the Kanarese, Tamil̤, Telugu, and other southern alphabets; and the inscriptions enable us to trace clearly the gradual development of all the modern forms.
The great classical Indian language, Sanskṛit, is not found in any inscriptional records of the earliest times. It is not, however, to be supposed therefrom that the use and cultivation of Sanskṛit ever lay dormant, and that there was a revival Languages. of this language when it did eventually come to be used in the inscriptions; the case simply is that, during the earlier periods, Sanskṛit was not known much, if at all, outside the Brāhmaṇical and other literary and priestly circles, and so was not recognized as a suitable medium for the notifications which were put on record in the inscriptions for the information of the people at large.
In Northern India, the inscriptions of the period before 58 B.C. present various early Prākṛits, i.e. vernaculars more or less derived from Sanskṛit or brought into a line with it. From 58 B.C., however, the influence of Sanskṛit began to manifest itself in the inscriptions, with the result that the records present from that time a language which is conveniently known as the mixed dialect, meaning neither exactly Prākṛit nor exactly Sanskṛit, but Prākṛit with an intermixture of Sanskṛit terminations and some other features; and we have, in fact, from Mathurā (Muttra), a locality which has yielded interesting remains in various directions, a short Brāhmanical inscription of 33 B.C. which was written wholly in Sanskṛit. The mixed dialect appears to have been the general one for inscriptional purposes in Northern India until about A.D. 320. But a remarkable exception is found in the inscription of Rudradāman, dated in A.D. 150, at Junāgaḍh in Kāṭhiāwār (mentioned on a preceding page), which is a somewhat lengthy record composed in thoroughly good literary Sanskṛit prose. Also, the extant inscriptions of the descendants of Rudradāman—(but only four of their records, ranging from A.D. 181 to 205, are at present available for study)—are in almost quite correct Sanskṛit; and this suggests that, from his time, the language may have been habitually used for inscriptional purposes in the dominions of his dynasty. That, however, is only a matter of conjecture; and elsewhere pure and good Sanskṛit, without any Prākṛit forms, appears next, and is found in verse as well as in prose, in the two inscriptions from Ēraṇ and Allahābād, referable to the period about A.D. 340 to 375, of the great Gupta king Samudragupta. From that time onwards, as far as our present knowledge goes, Sanskṛit, with a very rare introduction of Prākṛit or vernacular forms, was practically the only inscriptional language in the northern parts of India. We can, however, cite a record of A.D. 862 from the neighbourhood of Jōdhpur in Rājputānā, the body of which was written in Māhārāshṭrī Prākṛit.
In Southern India we have an instance of the mixed dialect in the
Nāsik inscription, referable to A.D. 257 or 258, of the Ābhīra king
Iśvarasēna, son of Śivadatta, which has been mentioned on a preceding
page. With the exception, however, of that record and of the
few which are mentioned just below, the inscriptional language of
Southern India appears to have been generally Prākṛit of one kind
or another until about A.D. 400, or perhaps even somewhat later.
rit figures first in one of the records at Nāsik of Ṛishabhadatta
(Ushavadāta), son-in-law of the Kshaharāta king Nahapāna, which
consequently gives it almost as early an appearance in the south
as that which is established for it in the north; but it is confined
in this instance to a preamble which recites the previous donations
and good works of Ṛishabhadatta; the record passes into Prākṛit
for the practical purpose for which it was framed. Sanskṛit figures
next, in an almost correct form, in the short inscription of not much
later date at Kaṇheri, near Bombay, of the queen (her name is not
extant) of Vāsishṭhīputra-Śrī-Śātakarṇi. It next appears in certain
formulae, and benedictive and imprecatory verses, which stand at the
end of some of the Prākṛit records of the Pallava series referable to
the 4th century; but here we have quotations from books, not
instances of original composition. We have a Sanskṛit record, obtained
in Khāndēsh but probably belonging to some part of Gujarāt,
of a king named Rudradāsa, which is perhaps dated in A.D. 367.
But the next southern inscription in Sanskṛit, of undeniable date, is a
record of A.D. 456, belonging to the Vyārā subdivision of the Baroda
state in Gujarāt, of the Traikūṭaka king Dahrasēna. The records of
the early Kadamba kings of Banawāsi in North Kanara, Bombay,
exhibit the use of Sanskṛit from an early period in the 6th century;
and records of the Pallava kings show it from perhaps a somewhat
earlier time on the other side of India. The records of the Chalukya
kings present Sanskṛit from A.D. 578 onwards. And from this latter
date the language figures freely in the southern records. But some of
the vernaculars, in their older forms, shortly begin to present themselves
alongside of it; and, without entirely superseding Sanskṛit
even to the latest times, the use of them for inscriptional purposes
became rapidly more and more extensive. The vernacular that first
makes its appearance is Kanarese, in a record of the Chalukya king
Maṅgalēśa, of the period A.D. 597 to 608, at Bādāmi in the Bijāpūr
district, Bombay. Tamiḷ appears next, between about A.D. 610 and
675, in records of the Pallava king Mahēndravarman I. at Vallam in
the Chingalpat (Chingleput) district, Madras, and of his great-grandson
Paramēśvaravarman I. from Kūram in the same district.
Telugu appears certainly in A.D. 1011, in a record of the Eastern
Chalukya king Vimalāditya; and it is perhaps given to us in A.D.
843 or 844 by a record of his ancestor Vishṇuvardhana V.; in the
latter case, however, the authenticity of the document is not certain.
Malayālam appears about A.D. 1150, in inscriptions of the rulers of
Kēraḷa from the Travancore state. And on the colossal image of
Gommaṭēśvara at Śravaṇa-Beḷgoḷa, in Mysore, there are two lines of
Marāṭhī, notifying for the benefit of pilgrims from the Marāṭhā
country the names of the persons who caused the image and the
enclosure to be made, which are attributed to the first quarter of the
12th century: this language, however, figures first for certain in a
record of A.D. 1207, of the time of the Dēvagiri-Yādava king Siṅghaṇa,
from Khāndēsh in the north of Bombay.
Bibliography.—The systematic publication of the Indian inscriptions has not gone far. Cunningham inaugurated a Corpus Inscriptionum Indicarum, by giving us in 1877 the first volume of it, dealing with the records of Aśōka; but the only other volume which has been published is vol. iii., by Fleet, dealing with the records of the Gupta series. The other published materials are mostly to be found here and there in the Journals of the Royal Asiatic Society of London, its Bombay branch, and the Asiatic Society of Bengal, in the Reports of the various Archaeological Surveys, and in the Indian Antiquary, the Epigraphia Indica and the Epigraphia Carnatica; and much work has still to be done in bringing them together according to the periods and dynasties to which they relate, and in revising some of them in the light of new discoveries and the teachings of later research. The authority on Indian palaeography is Bühler’s work, published in 1896 as part 2 of vol. i of the Grundriss der Indo-Arischen Philologie und Altertumskunde; an English version of it was issued in 1904 as an appendix to the Indian Antiquary, vol. xxxiii.
- (J. F. F.)
III. Greek Inscriptions
Etymologically the term inscription (ἐπιγραφή) would include much more than is commonly meant by it. It would include words engraved on rings, or stamped on coins, vases, lamps, wine-jar handles, &c. But Boeckh was clearly right in excluding this varia supellex from his Corpus Inscriptionum Graecarum, or only admitting it by way of appendix. Giving the term inscription a somewhat narrower sense, we still include within it a vast store of documents of the greatest value to the student of Greek civilization. It happens, moreover, that Greek inscriptions yield the historian a richer harvest than those of Rome. Partly from fashion, but partly from the greater abundance
- The legends on coins form part of numismatics, though closely connected with inscriptions.
- The amphorae which conveyed the wine and other products of various localities have imprinted on their handles the name of the magistrate and other marks of the place and date. Large collections have been made of them, and they repay inquiry. See Dumont, Inscriptions céramiques (1872); Paul Becker, Henkelinschriften (Leipzig, pt. i. 1862, pt. ii. 1863); Hiller v. Gaertringen, I.G. xii. 1065-1441.