Pramāṇa · Search · About · API

Elements of South-Indian Palæography, from the Fourth to the Seventeenth Century A.D. c4: Chapter I. The Probable Date of the Introduction of Writing into India

Elements of South-Indian Palæography, from the Fourth to the Seventeenth Century A.D. c4

Language
English

Chapter I. The Probable Date of the Introduction of Writing into India

c4.2

THAT the art of writing was imported into India is now allowed by most Orientalists who can claim to be heard, but how and when this occurred is by no means clear[^1). The earliest written documents that have been discovered in India are the proclamations of the Buddhist king Piyadasi or Açoka which are written in two different characters; and the silly denunciations of writing in which the Brahmans have always indulged, render it excessively improbable that they had anything to do with the introduction of the art. The inscriptions of Açoka are of about 250 B. C., but it seems probable that writing was practised to a certain extent in Northern India nearly half a century before that period.

c4.3

Nearchus (B. C. 325) expressly states that the Brahman laws were *not written*[^2). Megasthenes a few years later (c. 302 B. C.), mentions that they had no written books, and that they did not know letters (grammata)[^3) or use seals, but he also mentions milestones at a distance of ten stadia from one another, "indicating the bye-roads and intervals"[^4). It is difficult, though not impossible, to suppose that these indications were made by the stones merely, and that there were not any marks on them to tell more than

c4.4

[^1): Kopp (in 1821) first suggested a foreign Semitic source of the Devanāgari alphabet. Dr. R. Lepsius followed in 1834: and then with much stronger arguments came Profr. A. Weber (Z. D. D. M. G. x. pp. 389 and ffg. "Indische Skizzen" pp. 127-150). He has always been the strongest supporter of this theory. But many consider it probable: Profr. Th. Benfey ("Orient und Occident" iii., 170); Profr. Max Müller (A. S. L. 2nd ed. p. 521). Profr. N. L. Westergaard ("Über den ältesten Zeitraum der Indischen Geschichte" p. 37) hesitates. He considers it likely that writing was, originally, in India *a* secret known to the traders only. I am not able to refer to Böhtlingk's article on the age of writing in India mentioned by Lassen. Profr. Pott ("Etymologische Forschungen, Wurzel-Wörterbuch" ii., 2 p. liii.) is not however satisfied (1870). Mr. E. Thomas (1866) suggested a Dravidian origin of the Indian alphabets. Profr. Lassen repudiates a foreign origin for the Indian alphabets (I. A. K. Vol. I. 2nd ed. p. 1008) altogether. Profr. Whitney ("Studies" p. 85) considers a Semitic origin probable. [^2): Frag. F. in "Reliqua Arriani et Scriptorum de rebus Alexandri". Ed. C. Müller, Paris, 1846 (p. 60.) [^3): "Megasthenis Indica" ed. Schwanbeck, Frag. xxvii. (fr. Strabo. xv. 1. 53-56) p. 113. [^4): Do. Frag. xxxiv. (from the same source). pp. 125-6.

c4.5

the mere position of the stones could do¹". The inscriptions of Açoka are also in themselves proofs that writing was about 250 B. C. a recent practice, for they present irregularities of every kind[^2). That these inscriptions are of a period immediately after the introduction of writing has been insisted on by Profr. Wassiljew, who also remarks that it is not long after their date that the Buddhists refer to their scriptures as written[^3).

c4.6

¹ It is however singular that, as yet, none of these milestones have been discovered. [^2): Thus in the third tablet we find añapitam, and in the fourth añapayisati, but in the sixth āñāpiº. The reduplication of consonants is universally omitted where it should be found (*e. g.* piyasa, janasa, ārabhisante, dukaram, svagam, dighāya, etc.). Nor is the orthography uniform; we find in the Southern inscriptions: etārisam and etādisam also. Again in the Southern inscriptions we have anathesu, but in the Northern (at Kapurdigiri) anathesu. Again the Southern inscriptions have both dasana and dasaṇa. The insertion of nasals before consonants is also excessively irregular. But this may perhaps be attributed more properly to the carelessness of the masons who carved the text on the rocks. The existence of inscriptions like the Açoka edicts proves that writing was more or less commonly understood, but it is impossible, looking at the above irregularities and the numerous others that occur, to suppose that writing was then used to express the minute distinctions that we find in the grammarians' rules. For other similar irregularities, see Profr. Bühler's "Three New Edicts" pp. 7, 9, 32, 34 etc. [^3): "Der Buddhismus" p. 30 (28). It is much to be regretted that this admirable work, which marks an epoch in Indian studies, is not known by an English translation. The author's immense learning has not prevented him from giving his results in the clearest way, and he has evidently worked without any prejudice. See also Haug, "Über das Wesen und den Werth des Wedischen Accents (4º. Munich, 1874) p. 18. [^4): u. s. p. 64, *a*. [^5): "Megasthenis Indica" ed. Schwanbeck Fr. I. 42 (p. 91). [^6) Do. Fr. xxxiv., 5 (p. 126). [^7) Do. Fr. xxvi., I (p. 112).

c4.7

On the other hand Nearchus is also represented as stating that the Indians wrote letters on a sort of cotton cloth or paper[^4).

c4.8

Again, passages in Megasthenes have been understood by Schwanbeck to imply the use of writing at the period when he visited India. These are: (1) some passages which describe the proclamation at the beginning of the year of a sort of astrological calendar for the coming seasons[^5); again, (2) the statement that births were considered for astrological purposes[^6). But it is obvious that such usages afford but a faint presumption that writing was necessarily employed to enable them to be practised. There are many savage tribes still existing which are utterly ignorant of writing, and nevertheless do exactly the same things. Thus the description given by Megasthenes might apply to the 'Medicine men' of America, and the Fetish priests of parts of Africa at the present day who are utterly ignorant of any art at all like writing. The Aztecs who, at the best, had only an imperfect hieroglyphic character, were great astrologers. Megasthenes also mentions (3) songs in honour of gods and deceased persons[^7); but there is no necessity to assume that these were written. The (4) milestones that he describes, I have already mentioned. On the other hand it is expressly stated by Megasthenes that the Indians had no *written* laws, and strangely enough this is quoted by much later writers like Strabo, who must have been able to correct this statement if wrong at their time.

c4.9

The next point for consideration is: whence did these two alphabets come that we find in use in India in the third century before our era?

c4.10

During several centuries before that time, the natives of India had opportunities of becoming acquainted with many different systems of writing then current in the West and in Persia.

c4.11

The Phœnicians who voyaged for Solomon came to Southern India at least, and exported from thence peacocks which were called in Hebrew by a Tamiḷ name[^1). The Persians about 500 B. C. conquered India (that is probably, the Punjab and part of India Proper or Northern India), under Darius; and in the inscriptions at Persepolis and Naksh-i-Rustam India occurs as the 21st and 13th province, respectively, of that monarch's empire[^2). According to Herodotus India was the 20th satrapy, and paid as tribute 360 talents of gold. To pay such a very large sum a great extent of the country must have been subject.

c4.12

Still earlier conquests by Semiramis and Sesostris[^3) are mentioned, but the former is certainly mythical[^4), and the latter rests on the assertion of Diodorus Siculus alone. As his statement is not, as yet, corroborated by Egyptian monuments, little weight can be attached to it, but that the Egyptians traded with India, and that from very early times can hardly be doubted.

c4.13

Thus, before the conquests of Alexander, the natives of India had ample opportunities to learn the art of writing from others, or to invent a system for themselves, and thus it must be held that they copied, for there has not been found as yet the least trace of the invention and development of an independent *Indian* alphabet[^5), while of the two characters in which the inscriptions of Açoka were written, the northern has been conclusively identified (by Mr. E. Thomas) with an Aramaic original, and a number of letters in the Southern alphabet point clearly to a similar source. I shall also show, further on, that there is a third

c4.14

[^1): That the Hebrew *tuki* is the Tamiḷ *tōgai* seems to be determined. The identification is finally due to Dr. Caldwell ("Comparative Grammar" p. 66) and is in every way satisfactory. The remaining foreign terms in the same Hebrew passage appear however to have not been fairly considered as yet, and all proposed identifications of "almug" or "algum" would present the greatest difficulties. What has been proposed is to be found in Profr. Max Müller's "Lectures on the Science of Language" I. pp. 224-5. The word Tukiim has been last discussed by M. Vinson in Hovelacque's "Revue de Linguistique" VI. fasc. 2, very fully. That it cannot be derived from *çikhin*, the ō shows. [^2): On the Empire of Darius see Menant "Les Achéménides" pp. 167-9. Kossowicz ("Inscriptiones Palæo-Persicæ Achæmenidarum" pp. 72-3 and 76-7.) translates the passages as follows: (Inscription of Persepolis) "**2.** Edict Darius rex: Voluntate Auramasdae hae *sunt* provinciæ, quas ego tenui cum isto Persiae populo .... mihique tributum afferebant: Susiana .... India" etc. (Inscription of Naksh-i-Rustam) 3. Edict Darius rex: "Hae *sunt* provinciæ quas ego cepi extra a Persia (extra Persiam). Ego eas meae ditionis feci, mihi tributum afferebant quodque eis a me edicebatur hoc obsequentissime faciebant, lex quae mea est, haec *ab iis* observabatur: Media . . . . . Indi" etc. The original Persian word is 'Hi(n)dus'. [^3): Chabas "Etudes sur l'antiquité historique" p. 94. Thothmes iii. (? 1500 B. C.) penetrated to 'the country of elephants', but by 2500 B. C. there was regular intercourse with S. Arabia. (Brugsch "Histoire d' Égypte" I. p. 81.) [^4): La Legende de Semiramis, par F. Lenormant (1872). p. 11 etc. [^5): Max Müller, Sanskrit Grammar (2nd edition) p. 3.

c4.15

alphabet used only in S. India, the Vaṭṭeḷuttu or old Tamiḻ alphabet, which must also have been derived from the same or a Semitic source; but which is apparently, not derived from, nor is the source of the Southern Açoka alphabet though in some respects very near to it. Perhaps the most important proof of the Semitic origin of these two last alphabets is the imperfect system of marking the vowels which is common to them both. They have, like the Semitic alphabets, initial characters for them, but in the middle of words these letters are marked by mere additions to the preceding consonant. In the Vaṭṭeḷuttu it is difficult to avoid the conclusion that the initial **i** and **u** are anything more than the consonants **y** and **v**. These points are intelligible only on the supposition that the Indian alphabets are derived from the Phœnician, which was formed to suit languages in which the vowels are subsidiary to the consonants, a condition which is not met with either in the Sanskritic or Dravidian languages. The character in which the Northern Inscription of Açoka (at Kapurdigiri) is written, is from right to left, like all the Semitic characters; and the character of the Southern Inscriptions which runs in the contrary direction, yet shows traces of once having been written the same way¹.

c4.16

¹ The Southern Inscriptions of Açoka have *e. g.* yv where vy must be read, (*e. g.* in katavyo) and the v is put under the right end of the y. Again the vowel e precedes the consonant which in reading it must follow. The peculiar way of marking r to be read before or after the consonant above which it is marked (as was first pointed out, I believe, by Profr. Westergaard) appears to me also to point to the same conclusion. So also the marks which qualify the sign for 100 in the cave character, and which are affixed to the *right* side of the sign. [^2) ]: In the Journal of the R. Asiatic Society, New Series V. pp. 420-3, see p. 420 *n.* "The Aryans invented no alphabet of their own for their special form of human speech, but were, in all their migrations, indebted to the nationality amid whom they settled for their instruction in the science of writing: (4) The *Devanāgarī* was appropriated to the expression of the Sanskrit language from the pre-existing Indian Pāli or *Lāt* alphabet which was obviously originated to meet the requirements of Turanian (Dravidian) dialects." Mr. Thomas goes on to connect the advance of Sanskrit Literature and Grammar "with the simplified but extended alphabet they (*i. e.* the Aryan invaders of India) constructed in the Aryan provinces out of a very archaic type (! ?) of Phœnician, and whose graphic efficiency was so singularly aided by the free use of birch bark." On p. 423 he appears to consider that the Dravidians were taught by Scythian invaders who preceded the "Vedic Aryans". It is not clear if Mr. Thomas considers that the primitive alphabet which he assumes to have existed, was invented in India or an importation. [^3) ]: Below, App. A. Now in the cuneiform inscriptions of the Achæmenides ḍipi is the term used for those edicts. Thus in the Behistan inscription of Darius we find (iv. 15) "tuvm kā hya aparam imām ḍipim vaināhy." 'Thou whoever beholdest afterwards this *writing*! It seems to me, therefore, not unlikely that lipi has been introduced into India from the Persian dipi. Both Kossowicz and Spiegel refer ḍipi to the Sanskrit √() lip, but I see (by a note) that Dr. Hincks took this word to be Semitic. I have lost the reference, so cannot give his derivation, but the root *ktb* will occur to every one. With an admittedly Semitic ultimate origin of the Indian alphabets, it is natural to expect a foreign term for the art of writing, and I would, therefore, suggest that lipi is not a derivative of √() lip, but, a corrupt foreign term. The *primâ facie* derivation from √() lip assumes that 1. writing is indigenous to India, and 2. that it originally began there with marks not scratched on a hard substance but painted on the prepared surface of a suitable stuff; both which assumptions are strongly negatived by facts. (*contra* Pott's W. W. v. pp. 180-1). On lipikara=maker of inscriptions, see M. Müller, in Ṛigv. iv., p. lxxiv., *n*.

c4.17

Mr. E. Thomas[^2) ]has lately propounded a theory that the Southern Açoka alphabet is originally Dravidian, and then adapted to the N. Indian languages. This could only be the case if we assume the Vaṭṭeḷuttu to be the prototype, but as this is an imperfect expression of the Dravidian sound-system[^3) ], it cannot be an indigenous invention, and the theory presents many other objections. One insuperable difficulty is the entire absence of traces of any alphabet having existed in S. India before the Vaṭṭeḷuttu, and that all written monuments now known to exist prove a gradual invasion of the South by Buddhist and Brahmanical civilizations which brought more complete alphabets (derived from the Southern Açoka character) with them in historic times, and meeting the old Tamiḻ alphabet or Vaṭṭeḷuttu gradually supplanted it. It is especially remarkable that this last never had separate signs for the sonant letters (g etc.) which must have existed if Mr. Thomas's theory is correct, but though as I shall afterwards prove, the Tamiḷ language had these sounds in the third century after our era, the earliest monuments do not exhibit any marks or letters for them.

c4.18

Very few Sanskrit books are nowadays even supposed to belong to a period when writing did not exist in India, and the only early ones that appear to mention writing are the grammars attributed to Pāṇini and to others. But the age of these works is by no means clear and even if it be supposed that the Mahābhāshya (or great commentary on Pāṇini by Patañjali) has not been since worked over again and again and tampered with (a supposition it is for intrinsic reasons, very difficult to avoid), this commentary would only prove the existence of Pāṇini's Sūtras in the second century before our era, a time when writing was certainly in common use in India.

c4.19

Pāṇini implicitly mentions (according to the Mahābhāshya) the writing of the Yavanas. It has not yet been fully determined what was intended by this term, nor is it clear whether it was in use in India or not It can mean either Persian or Greek

c4.20

[^1): Profr. Goldstücker considered Pāṇini to have lived before Buddha ("Pāṇini's Place" pp. 225-227) chiefly on the ground that the sūtra viii., 2, 50 ("nirvāṇo 'vāte) does not provide for the peculiar Buddhist sense of nirvāṇa, and that therefore it is subsequent to Pāṇini. The same identical sūtra, however, occurs in the Grammar attributed to Çākaṭāyana (iv., 1, 249), and is explained by the commentator (Yaxavarman) in a manner that makes it appear as if Goldstücker's interpretation were too strict—avāte kartari | nirvāṇo muniḥ | nirvāṇaḥ pradīpaḥ | ‘avāta’ iti kim | nirvāto vātaḥ | nirvātam vātena |

c4.21

Profr. Benfey ("Geschichte d. Sprachwissenschaft" p. 48 *n.* 1) puts Pāṇini's Grammar at about 320 B. C. The latest authority is Profr. Aufrecht who says ("Annual Address" by A. J. Ellis Esq. as President of the Philological Society, 1873, p. 22): "Sanskrit Grammar is based on the grammatical aphorisms of Pāṇini, a writer now generally supposed to have lived in the fourth century B. C. at that time Sanskrit had ceased to be a living language." Cfr. Whitney "Studies" pp. 75-7. Lassen I. A—K i., 866; II. p. 477 (2nd ed.) puts Pāṇini at 330 B. C. If his date be put a little later, many difficulties would disappear. See my "Aindra Grammarians" p. 44. [^2): The passages (text and C. Mahābhāshya) are: (P. iv., i. 49) “Indravaruṇabhavaçarvarudramṛḍahimāraṇyayavayavanamātulācāryāṇām ānuk.” On this sūtra the Mahābhāshya (Benares edition, p. 27 of ch. iv. in Vol. iii.) remarks: “Himāraṇyayor mahattve” | ‘Himāraṇyayor mahattva’ iti vaktavyam | mahad dhīmam himānī | mahad araṇyam araṇyānī || “yavād doshe” | ‘Yavād dosha’ iti vaktavyam | dushto yavo yavānī || Yavanāl lipyām | ‘Yavanāl lipyām’ iti vaktavyam | yavanānī lipiḥ || etc.

c4.22

The other Grammar gives the substance of this sūtra in several (Çākaṭāyana I., 3, 52-57):—
52. Mātulācāryopādhyāyād ān ca |
53. Varuṇendramṛḍabhavaçarvarudrād ān |
54. Sūryadevatāyām |
55. Āḍ | (This allows sūryā also).
56. Yavanayavāl lipidushṭe |

c4.23

On this last sūtra Yaxavarman's C. runs: Yavanayavābhyām yathākramam lipau dushte cā 'rthe striyām ānpratyayo bhavati yavanānām lipiḥ yavanānī | yavanānyā | dushto yavo yavānī | yavānyā || 57 Himāraṇyād urau | etc. This is really modern.

c4.24

The word lipi (which occurs in a sūtra of Pāṇini—iii., 2, 21, corresponding to Çak. iv., 3, 132, *i. e.* divāvibhāniçāprabhābhāskarārushkartrantānantādināndilipibalicitraxetrajñaghābāhvahardhanurbhaktasankhyāt taḥ ||) is in some respects remarkable. The Açoka edict (where it first occurs) is called a dhammalipi and is said to be lekhitā or lekhāpitā. As in every case writing originally consisted of scratches or incisions on a hard substance (bricks were used in Assyria; bamboos in China, and stone in Egypt *primitively*), one would expect instead of a word from √() lip (=smear), a derivative of √() likh (=scratch); especially as the last is always used in India to express the act of writing on *any* substance (*e. g.* in the Mānavadharmaçāstra).

c4.25

writing. If the date of Pāṇini is put before 350 B. C., the first would be the probable meaning, as has been assumed by Profr. Goldstücker¹); if later than that, it could not possibly mean anything but Greek, for which Profr. Weber has decided²).

c4.26

¹ "Pāṇini's Place" p. 16. "It would seem to me that it denotes the writing of the Persians, and probably the cuneiform writing which was already known, before the time of Darius, and is peculiar enough in its appearance, and different enough from the alphabet of the (17) Hindus, to explain the fact that its name called for the formation of a separate word."

c4.27

² "Indische Studien" iv., 89. In the Berlin "Monatsbericht" for Dec. 1871, p. 616 *n*. he says: "der Name . . . Yavana . . . ist übrigens jedenfalls wohl schon vor Alexander's Zeit, durch die früheren Perser-Kriege nämlich, in denen ja auch Inder als Hülfstruppen gegen die Griechen mit im Felde standen, den Indern bekannt geworden." Profr. Westergaard is also of opinion (Über den ältesten Zeitraum p. 33) that Greek writing is intended, and no one can doubt that this is the correct view.

c4.28

But Pāṇini's sūtras show that writing was known in his time, and many expressions render it impossible to doubt that he used writing, and that to express minute details³); and one of his sūtras (vi., 3, 115) shows that the figures for eight and five were then used for marking cattle. That writing must soon have come into general use in India for literary purposes cannot be doubted, for without it, it is impossible that the systematic *prose* treatises which form so large a share of the Sanskrit literature, could ever have been composed⁴).

c4.29

³ "Pāṇini's Place" pp. 34-61 Profr. Westergaard appears to have arrived independently at the same conclusion.

c4.30

⁴ Cfr. Haug's "Essays on the Religion etc. of the Parsees" p. 129. "In the fragments of the Ancient Literature as extant in the Zend-Avesta, nowhere a word of the meaning 'to write' is to be found. That is merely fortuitous; because systematical books on scientific matters can never be composed without the aid of writing." *cfr*. Whitney, "Studies" p. 82.

c4.31

In all the earlier Sanskrit works there is very little, if any, reference to writing, and the preference for oral teaching exhibited by them is very marked; in fact the Brahman seem to have regarded the writing of any of their sacred or grammatical works as a deadly sin. But in the mediæval treatises it is evident that this most useful of arts had gained recognition in spite of priestly fanaticism and exclusiveness. Thus the earliest Sanskrit treatise on prosody which is attributed to Piṅgala contains nothing that can be held to imply the use of writing; the later imitation which describes the Prakṛit metres, however, contains a sūtra which proves the use of writing at the time it was composed⁵); so also does the recent (13th century) grammar, the Mugdabodha.

c4.32

⁵ "Prākṛit Piṅgala" I., 2. Diho samjuttaparo bindujuo etc. Here bindu can only refer to a written mark o. It is explained by Laxmīnātha (in his "Piṅgalārthapradīpa"): 'bindujuo' binduyuktaḥ sānusvāraḥ.

c4.33

That a literature of considerable extent can exist without being written has been conclusively shown by Profr. Max Müller in his "Ancient Sanskrit Literature," but it could not possibly include scientific and systematic treatises, though the oral transmission of long epics is quite probable[^1).

c4.34

The foregoing facts will, I think, prove that the art of writing was little, if at all, known in India before the third century before the Christian era, and as there is not the least trace of the development in India of an original and independent system, it naturally follows that the art was introduced by foreigners.

c4.35

I have already mentioned the numerous indications that point to a Semitic original of the Indian alphabets, and which are generally received as sufficient; the immediate original is, however, as yet uncertain. Three probable sources may be suggested. The first is that the Indian alphabet came direct from Phœnicia, and was introduced by the early Phœnician traders[^2). The second is that the original of these alphabets is to be sought in the modified Phœnician alphabet used by the early Himyarites of Arabia, and this has been lately put forward as an ascertained and certain fact[^3). As a third possibility I would suggest that the Indian alphabets may be derived from an Aramaic character used in Persia or rather in Babylonia.

c4.36

As regards the first possibility, it seems altogether inconsistent with the evidence regarding the scanty use of writing in the fourth century B. C. already given; for, as Phœnician communications direct with India must have ceased full five-hundred years, if not more, before that date, it is almost incredible that the art should not have arrived at perfection as applied to the Indian languages in that time, and have been in common use; but this is (as has been already shown) far from being the case. Again it is difficult to understand how the forms of the letters could be retained with so little modification for such a long period as this view would require; for, from the date of the inscriptions of Açoka (250 B. C.), documents with undisputed dates show that changes were marked rapid, and the progress of adaptation no less so[^4).

c4.37

[^1): Cfr. Grote's "History of Greece", ii., pp. 144-148 on the long period during which the Homeric poems were recited before they were committed to writing. [^2): "Orient und Occident" iii., p. 170. "Dass es einen uralten Zusammenhang zwischen Indien und dem Westen gab, wissen wir mit Entschiedenheit durch König Salomon's Ophirfahrten. Sicherlich waren diess nicht die ältesten. Die Phönicier waren gewiss schon lange vorher Vermittler des Handels zwischen Indien und dem Westen und wie sie, höchst wahrscheinlich, die Schrift nach Indien brachten, mochten sie und vielleicht Ægypter selbst auch manche andre Culturelemente hinüber und herüber bewegt haben." [^3): By F. Lenormant ("Essai sur la propagation de l'alphabet Phénicien" Vol. I., pt. I., Table vi.) The author makes the "alphabet primitif du Yémen" the source of both the Himyaritic and Māgadhi (!!) alphabets. [^4): It is also worthy of notice that all the Southern Açoka Inscriptions from Gujarat to Ganjam (in the Bay of Bengal) are in precisely the same character. This looks as if the art of writing had then first spread over Northern India from the place where it was first used, perhaps Gujarat. In the course of a few hundred years, however, the alphabets used in Gujarat and Bengal had already become so different as to be very little alike in appearance.

c4.38

As regards the second possibility, that the southern Açoka alphabet came from the Himyarites, the great difficulty is to show that the people of S. W. Arabia were in a position to furnish India with the elements of an alphabet so early as the 4th century B. C. It is very remarkable that the Himyaritic character was written from left to right, and that this was an innovation made by the people of Arabia is proved by the boustrophedon Himyaritic inscriptions that have been recently discovered¹. The difficulty of the direction of the southern Açoka character being from left to right would disappear if the Himyaritic character be assumed to be the original; but it remains to be proved that the civilization of S. W. Arabia had advanced so far already in the fourth century before the Christian era, as to be able to furnish India with a system of writing. It must also be recollected that the Himyaritic alphabet did not mark the vowels, as its derivative, the Æthiopic alphabet does. It is to be hoped that the intrepid explorer M. Halévy will be able to clear up the very interesting question of the date of the Himyaritic civilization.

c4.39

¹ The discovery was originally made by a French traveller some years ago, but has been only recently confirmed. (Letter by von Maltzan in the Allg. Zeitung for March 1st 1871, pp. 10-11.)

c4.40

The possibility and probability that the Indian alphabets are derived from an Aramaic type used in Persia, seems not to have been yet considered. The Persian² or Assyrian cuneiform characters cannot be thought of as a direct source, though the last remained in use up to the first century of our era³ for many purposes; but it is certain that a cursive Aramaic character was already long used, before (in the third century A. D.) it became (in the form of Pahlavi) the most generally used character for the official languages of Persia. The researches of Fresnel and Layard brought to light bricks with inscriptions in cuneiform and also in Egyptian and Semitic characters⁴, and these go back, probably, to the time of the Achæmenides⁵. Whichever of these three probable sources of the Indian alphabets may be accepted, there is a difficulty which seems to have escaped the notice of palæographists—the origin of the manner of indicating vowels in the body of a word. All the three primitive Indian alphabets possess this peculiarity with comparatively unimportant differences, but, though the system closely resembles the vowel points used by the Semitic races, it seems that there is not the least evidence for believing that it was used by these last earlier than at a time when it was already in use in India. This problem is, perhaps, the most important that awaits solution out of the many regarding Indian palæography.

c4.41

² J. Oppert ("*Journal Antique,*" viième Série, iii. pp. 238 ffg.) has shown that this is the invention of Cyrus, and has given the principle of derivation from the Babylonian cuneiform.

c4.42

³ Oppert in "Mélanges d'Archéologie Égyptienne et Assyrienne" fasc. I. p. 27.

c4.43

⁴ Cfr. Renan, "Histoire des langues Sémitiques," p. 115 etc.

c4.44

⁵ Spiegel, "Grammatik der Huzvaresch Sprache", p. 26: "die späteren Alphabete Erāns verrathen einen Semitischen Ursprung, und mögen daher vielleicht aus einem früheren aramäischen Alphabete stammen, das. bereits unter den Achämeniden neben der Keilschrift im Gebrauche war." So Sayce ("Principles of Comparative Philology" p. 196): "Clay contract tablets in the British Museum, written in Assyrian cuneiform with Aramaic dockets, indicate that from the reign of Tiglath Pilaser (B. C. 745) downwards Aramaic was the language of commerce throughout the Assyrian world." Lenormant "Essai" I., 1. pl. ix., gives specimens of this character which is very near the Phœnician.

c4.45

A cursory inspection of the alphabet used in the Southern Açoka inscriptions will satisfy any one accustomed to such enquiries, that the character from which it is derived did not comprise a sufficient number of letters, and that new signs were made by altering some of the old ones¹. This is, in itself, sufficient proof that the Indian alphabet was adapted, and not an indigenous invention. Other facts also point to an adaptation from a Semitic character. It is possible (if the Phœnician origin of the S. Açoka character be admitted) to fix the period when it must have occurred within certain, though wide, limits. The late illustrious scholar Viscount E. de Rougé has (in his masterly treatise “Mémoire sur l’origine Égyptienne de l’alphabet Phénicien”) shown that the Phœnician alphabet was derived from Egyptian signs about the 19th century B. C.² Another not less eminent Egyptologist has shown that the tribute brought to Thothmes III. (17th century) proves that the Phœnicians had then commercial intercourse with India³. About the 17th century B. C. is, then, the earliest period at which it is possible to fix the introduction of the alphabet into India. But, again, though the changes in the Phœnician alphabet were, so far as is now known, of a very slight character even during several centuries, it is yet possible, even with the scanty information available, to trace some progress in development, and it is evident⁴ that the source of the S. Açoka character must rather be sought in the forms current in Phœnicia in or about the 5th century B. C. or even later than in the earlier forms. The N. alphabet is, on the contrary, nearer to the older forms, but it in no way concerns the people of S. India. Thus all known facts tend to prove that the earliest date of the introduction of the Phœnician alphabet into India in what became the S. Açoka character, cannot have been earlier than 500 B. C. and was probably not earlier than 400 B. C. At present, all available information points to a Phœnician-Aramaic origin of the Indian alphabets, but the information is too scanty to justify a more precise inference. Writing was, certainly, little used in India before 250 B. C.

c4.46

¹ ^1) Mr. Thomas has proved this clearly by his figures on p. 422 of the fifth volume of the New Series of the *R. As. Society’s Journal*. The letters čh, ṭh, ḍh, th, ph show their origin very clearly.

c4.47

² ^2) p. 108. This is now contested by Deecke who considers that the Phœnician alphabet is derived from the cuneiform syllabary. Z. D. D. M. G. xxxi. His attempt is, however, according to so competent an authority as Profr. Sayce (“*Academy*,” xi., p. 557) far from successful. He admits, that the Indian alphabets came from the Phœnician.

c4.48

³ ^3) Chabas, Etudes, p. 120.

c4.49

⁴ ^4) See plate iii. bis, iv., in Lenormant’s “Essai” (i).

c4.50

In considering the question of the age and extent of the use of writing in India, it is important to point out that the want of suitable materials in the North at least, before the introduction of paper, must have been a great obstacle to its general use. The best material for writing on to be found in India is the palm leaf; either of the Talipat (*Corypha umbraculifera*), or of the Palmyra (*Borassus flabelliformis*). But the former appears to be a recent introduction from Ceylon into S. India, and it is there by no means common even on the West Coast, and is hardly known elsewhere. The palmyra also appears to have been introduced from Ceylon or Tinnevelly into the rest of the Peninsula; it is by no means common out of the South¹. The materials mentioned at an earlier date (excluding lotus leaves and such fancies of poets) almost preclude the existence of MSS. of books or long documents. The 'bhūrjapatra' which is understood (apparently on philological grounds—I cannot find out what tree furnishes this singular tissue) to mean the bark of the birch-tree, could not have been available everywhere in large quantities, nor would it be very suitable². The supposition of those who with Whitney and Böhtlingk assert that writing was, in India, long used only *esoterically* for composition and the preservation of texts, while the instruction was entirely oral, is, on these grounds almost certainly correct.

c4.51

¹ Voigt. "Hortus Suburbanus Calcuttensis" p. 640. Roxburgh, however, states that it is "common all over India". (*Flora Indica*, III. p. 790.) It requires the leaves of *many* trees to make an ordinary *grantha*. Palm leaves (there called *lontar*) were and are used for writing the Kawi or Old Javanese in Java and Bali. When I was in Java I scarcely saw half a dozen of these palms. Cfr. Junghuhn "Java" I. p. 188.

c4.52

² MSS. written on this substance are said to be in existence, but I have not seen any. Cfr. Schlegel's Rāmāyaṇa I. pp. xv-xvi. A famous MS. on this substance is that of the Paippalāda çākhā of the Atharvaveda found lately in Cashmere. Dr. Bühler (*Journal Bombay As. Soc.* No. xxxiv. A. vol. xii. p. 29) calls the tree *Baetula Bhojpatra* and shows that the bark is common enough in Cashmere. However in considering a question such as this, it is necessary to remember that only indigenous products deserve mention, for in the earlier times commercial facilities did not exist. The oldest MS. yet found in India is on talipat leaves.

c4.53

Arrian³ (quoting Megasthenes) calls the palmyra palm by its proper name (tāla)⁴, but its leaves are not mentioned anywhere by classical writers as affording writing materials used in India. Pliny⁵ indeed mentions palm leaves as used for this purpose, but he refers the practice to Egypt before the discovery of papyrus.

c4.54

³ "Indica" ed. Dübner, ch. VII., 3 (p. 209).

c4.55

⁴ In S. India the palmyra is called 'tāla'; the talipat, 'çritāla.'

c4.56

⁵ Ch. XIII., 21.

c4.57

Paper was probably introduced by the Muhammadans; in all parts of India it appears to be called by some corrupt form of the Arabic name 'kāgat'. Its use in S. India is at all events very recent, and even now scarcely ever occurs except among the Mahrāṭī colonists. I have seen a Telugu MS. of a Sanskrit work written about the end of the 17th century, and Paulinus à St. Bartholomæo notices MSS. on paper of the Bhāgavata (in Travancore 18th century); but the bigoted Hindus of the South still consider this material to be unclean and therefore unfit for writing any book with the least pretence to a sacred character¹).

c4.58

¹ L. Varthema (who travelled from 1503-1508) remarks that Paper was in his time used in Pegu 'not leaves as at Calicut' (ed. of 1517 f. 61).

Linked texts (6)