Pramāṇa · Search · About · API

Elements of South-Indian Palæography, from the Fourth to the Seventeenth Century A.D. c6.s5.t2: B. The Method of Expressing Numerals

Elements of South-Indian Palæography, from the Fourth to the Seventeenth Century A.D. c6.s5.t2

Language
English

B. The Method of Expressing Numerals

c6.s5.t2.1

a. By words

c6.s5.t2.2

The earliest inscriptions found in S. India in which the date is referred to an era have it written at full length in words. After the seventh century the dates are *mostly* expressed by significant words, and after the tenth century this is *always* done. These significant words appear to be a device of the Indian astrologers, as the earliest examples occur in their treatises. The first complete list is that given by Albīrūnī (A. D. 1031); the following is from his list as translated by Woepcke¹ supplemented from Brown's "Cyclic Tables" and inscriptions. As no limits can be placed to a fanciful practice like this, I cannot give this list as complete; it is merely an attempt to make a complete list²).

c6.s5.t2.3

¹ "Mémoire" pp. 103-9.

c6.s5.t2.4

² This system was first explained by v. Schlegel. Here (as is so perpetually the case in Indian literature) we find that the present system has had predecessors. In the 'Jyotisha' (see Profr. Weber's ed. p. 6) ǎya = 4; yuga = 12; bhasamūha = 27; rūpa = 1. In the 'Chandas' similar expressions occur.—In the above list I give firstly those words given by Albīrūnī about which there can be no doubt; then others mentioned by Mr. C. P. Brown which I mark \*. Lastly I add terms not already mentioned, which I have found in inscriptions, and which I mark †. This system is also used in the Javanese inscriptions. See v. Humboldt's "Kawi-Sprache" i., pp. 19-42. Crawford "On Hindu Religion in Bali." *As. Res.* xiii., pp. 150-1. See also on the Javanese calendar, Gericke in "Verhandelingen" (xvi., pp. 65-80). "Iets over de Javaansche Tijdrekening"; and Cohen-Stuart's Essay in the "Tijdschrift voor Nederlandsch-Indie", 1850, i. pp. 215-324.

c6.s5.t2.5

*Cipher*. . . . . Çūnya; kha; gagana; viyat; ākāça; ambara; abhra; ananta\*; vyoma\*. 1 . . . . . Ādi; çaçin; indu; xiti; urvarā; dharā; pitāmaha; candra; çītāmçu; rūpa; raçmi; pṛithivī\*; bhū\*; tanu\*; soma†; nāyaka†; vasudhā†; çaçāṅka†; xmā†; dharaṇī†. 2 . . . . . Yama; Açvin; ravicandrau; locana; axi; Dasra; yamala; paxa; netra; bāhu\*; karṇa\*; kuṭumba\*; kara†; dṛishṭi†. 3 . . . . . Trikāla; trijagat; tri; triguṇa; loka; trigata; pāvaka; vaiçvānara; dahana; tapana; hutāçana; jvalana; agni; vahni\*; trilocana\*; trinetra\*; Rāma\*; sahodara\*; çikhin†; guṇa†. 4 . . . . . Veda; samudra; sāgara; abdhi; dadhi (?); diç; jalāçaya; kṛita; jala nidhi\*; yuga\*; koshṭha\*; bandhu\*; udadhi†. 5 . . . . . Çara; artha; indriya; sāyaka; vāṇa; bhūta; ishu; Pāṇḍava; tata; ratna\*; prāṇa\*; suta\*; putra\*; viçikha†; kalamba†; mārgaṇa†. 6 . . . . . Rasa; aṅga; ṛitu; māsārddha; rāga\*; ari\*; darçana\*; tarka\*; mata†; çāstra†. 7 . . . . . Aga; naga; parvata; mahīdhara; adri; muni; ṛishi\*; Atri\*; svara\*; chandas\*; açva\*; dhātu\*; kalatra\*; çaila†.

c6.s5.t2.6

8 . . . . . Vasu; ahi; gaja; dantin; maṅgala; nāga; bhūti\*; ibha†; sarpa†(?). 9 . . . . . . Go; nanda; randhra; chidra; pavana; antara; graha\*; aṅka\*; nidhi†; dvāra†. 10 . . . Diç; āçā; kendu; rāvaṇaçara; avatāra\*; karma\*. 11 . . . Rudra; Īçvara; Mahādeva; axauhiṇī; lābha\*. 12 . . . Sūrya; arka; āditya; bhānu; māsa; sahasrāmça; vyaya\*. 13 . . . . Viçva; Manmatha\*; Kāmadeva\*. 14 . . . Manu; loka\*; Indra\*. 15 . . . Tithi; paxa\*; ahan\*. 16 . . . Ashṭi; nṛipa; bhūpa; kalā\*. 17 . . . Atyashṭi. 18 . . . Dhṛiti. 19 . . . Atidhṛiti. 20 . . . Nakha; kṛiti. 21 . . Utkṛiti; svarga\*. 22 . . . . Jāti\*. 24 . . . Jina\*. 25 . . . . Tattva.

c6.s5.t2.7

Albīrūnī (1031 A. D.) says that numbers beyond twenty-five were not noted in this way. The following, however, occur but in late documents only.

c6.s5.t2.8

27 . . . Naxatra\*. 32 . . Danta\*, Rada. 33 . . Deva\*. 49 . . . . Tāna\*.

c6.s5.t2.9

This list might be made much more extensive, as it is obvious that any synonyms of any word that can be used to signify a number can be used; *e. g.* any word signifying ‘moon’ besides those mentioned as equivalent to 1, may be used for the same purpose, and so with the others. The ordinary numeral words are commonly mixed with the words given above.

c6.s5.t2.10

In marking numbers by this system units are mentioned first and then the higher orders; *e. g.* Ṛishināgakhendusaṃvatsara is year 1087; guṇaçāstrakhendugaṇitasamva°= 1063; dahanādrikhendugaṇitasamva°=1073. It appears, however, that occasionally in recent inscriptions the words are put in the same order as the figures are written.

c6.s5.t2.11

From 600 A. D. up to 1300 nine out of ten inscriptions that bear dates, have them expressed in this style, which is, therefore, of the greatest importance.

c6.s5.t2.12

b. Expression of numbers by letters

c6.s5.t2.13

Three systems of this kind are known in India¹: that of Āryabhaṭa, which he used in his treatises on astronomy, and which does not appear to have ever been used by any one else or in inscriptions; that used in S. India (but almost exclusively in Malabar, Travancore and the S. Tamiḷ country), in which the date is given by a chronogram; and a third system in which the letters of the alphabet are used to mark the leaves of MSS.

c6.s5.t2.14

¹ The earlier system used by Pāṇini is not otherwise known. See my "Aindra Grammarians" p. 88 (based on Goldstücker's "Pāṇini's Place", p. 50).

c6.s5.t2.15

It is unnecessary to describe the first, as it is never used in inscriptions, and the text of Āryabhaṭa's work (once almost inaccessible) has been admirably edited by Profr. Kern (1874).

c6.s5.t2.16

The second system gives values to the consonants of the Sanskrit alphabet as follows²:

c6.s5.t2.17

² It was first explained by the late C. M. Whish (in pt. i. pp. 54 ffg. of the Transactions of the Madras Society). Mr. Whish was one of the first to pay attention to Sanskrit astronomy. He died at Cuddapah, April 13th, 1833. His scientific reputation is not so great as it might have been; for if he did not originate, he, at all events, gave circulation to some forgeries. His paper on this system of notation was translated by Jacquet (*J. Asiatique* 1835). On this method of marking dates see also Z. d. D. M. G. xvii., pp. 773 ffg. (by Profr. Weber).

c6.s5.t2.18

| k | kh | g | gh | ṅ |
| :-: | :-: | :-: | :-: | :-: |
| 1 | 2 | 3 | 4 | 5 |
| c | ch | j | jh | ń |
| 6 | 7 | 8 | 9 | 0 |
| ṭ | ṭh | ḍ | ḍh | ṇ |
| 1 | 2 | 3 | 4 | 5 |
| t | th | d | dh | n |
| 6 | 7 | 8 | 9 | 0 |
| p | ph | b | bh | m |
| 1 | 2 | 3 | 4 | 5 |

c6.s5.t2.19

| y | r | l | v | ç | sh | s | h | ḷ |
| :-: | :-: | :-: | :-: | :-: | :-: | :-: | :-: | :-: |
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |

c6.s5.t2.20

The order of the letters is from right to left, in double letters the last pronounced consonant *only* counts, and vowels have no value. Thus Vishṇu = 54; badhnāti annam sasarpi = 17,750,603. As might be supposed, the use of this method brought numerous grammatical errors.

c6.s5.t2.21

The peculiarity of this system is that it allows dates to be expressed by words with a connected meaning. This system was commonly in use in the fifteenth century³, but, apparently, not long before then. The oldest specimen of this notation (1187 A. D.) is in Shaḍguruçishya's commentary on the Ṛigveda Anukramaṇika⁴. It is now much used for remembering rules to calculate horoscopes, and for astronomical tables. The resemblance to the Semitic chronograms is complete. This method is also used in a kind of anukramaṇī which exists for the Ṛig-, Yajur- and Sāmavedas, but apparently in S. India only. These lists of contents (for they are no more) must be modern¹.

c6.s5.t2.22

³ *Indian Antiquary*, ii., pp. 361-2, and other inscriptions.

c6.s5.t2.23

⁴ *I. S.* viii., p. 160 *note*.

c6.s5.t2.24

¹ “Catalogue” p. 49. “Index to Tanjore MSS.” p. 4.

c6.s5.t2.25

The third system is only applied to numbering the pages of MSS.; it was used a good deal in Malabar, and also occasionally in the Telugu country, but not to any extent in MSS. written in this century. It is also known in Ceylon and Burmah. By this system the consonants (with short a, and in their usual order) stand for 1, 2, etc. up to 34, and then they are repeated with long ā, *e. g.* kā = 35, khā 36 and so on. By the addition of the other vowels the series may be continued to a considerable length. Another system (used by the Buddhists and Jains in N. India) uses syllables in an apparently arbitrary manner; this is (so far as I am aware) unknown in S. India. I have already² given the chief of these.

c6.s5.t2.26

² See p. 65.

c6.s5.t2.27

In MSS. one often finds an abridged way of writing numbers, *e. g.* 20 || 1 || 2 etc. for 20; 21; 22 etc. And this has been suspected with reason to exist in some inscriptions. It was done (according to Albīrūnī) in reckoning by the ‘Lokakāla’.

c6.s5.t2.28

This formidable number of eras and complicated calendars might seem to encourage hopes of an accurate chronology, but such hopes are entirely delusive. The exact length of a king’s reign is seldom given in years and days, but fractions of years are taken as years. Again, Hindu kings in S. India often nominated and consecrated their successors, and the length of the reign is sometimes reckoned from this event; an approximation, not certainty, is then, all that is to be hoped for. The most important information likely to be soon available respecting Indian eras is to be hoped for in the edition and translation of Albīrūnī’s works already begun by Profr. Sachau. But it must not be forgotten that Albīrūnī himself found the greatest confusion in respect not only of Indian eras, but also of the beginning of the year, and that even he could not solve all the difficulties he detected (Reinaud, “Fragments”, pp. 139, 145). Hiouen Thsang³ long before this had occasion to notice the confusion that prevailed. From what is now known respecting Indian chronology, there can be little doubt that originally a number of local eras and calendars were used, and that these have been gradually superseded for the most part by the more precise eras and calendars of the astronomers, and in recent times by the ‘Lokakāla’.

c6.s5.t2.29

³ “Pélerins Bouddhistes” ii., p. 493.

Cites (4)