Pramāṇa · Search · About · API

Early South Indian Palaeography c15: Appendix I: Some Inscriptions of the Early Centuries of the Christian Era

Early South Indian Palaeography c15

Language
English

Appendix I: Some Inscriptions of the Early Centuries of the Christian Era

c15.2

The epigraphical records of the Tamil country after the period of the Brāhmī labels discussed above become available only from the commencement of the seventh century A. D. The intervening period may be treated as a comparatively 'dark period' in the epigraphical field as in many other aspects in the history of the Tamil country. There are only a few streaks of light flashed by some stray records which can be definitely attributed to this period and found in this part of the counrty.

c15.3

Three varieties of script are found used in the inscriptions recovered in the Tamil country and assignable to the period commencing from the seventh century A. D. They are (i) the Grantha script employed for writing the Sanskrit portion and the Sanskrit letters occurring in the Tamil portion of the records, (ii) the Tamil script, also referred to as Grantha—Tamil used for writing the Tamil records in the northern part of the country by the Pallavas and their successors the Cōḻas and (iii) the Vaṭṭeḻuttu script used for writing in the southern part of the country, viz., the Pāṇḍya territory and the adjacent areas till their final conquest and annexation to the Cōḻa empire in the eleventh century A. D. It is to be noted in this connection that in writing words of Tamil origin occurring in the Sanskrit portions of the records only the Grantha or the Grantha—Tamil is used, while the Vaṭṭeḻuttu script is employed in writing the Tamil portion in such inscriptions.

c15.4

All these three varieties of scripts are derived from the Brāhmī. The particulars as to when and how these varieties separated from the parental Brāhmī and evolved as independent scripts are not known. But if a surmise is possible the evolution may be traced with the help of the few inscriptions of the intervening period. They are the Tirunāthar kuṉṟu epitaph and the labels on the rock at Tiruchirapalli. Fortunately the labels at the latter place are found recorded in all the three varieties.

c15.5

The earliest of these¹ is a short record of four lines containing in all twenty-nine characters. It is found engraved on a rock at Tirunāthar-kunṟu near Tirunaruṅkoṉṟai in the Jinji Taluk, South Arcot District. It reads

c15.6

¹ *Annual Report on South Indian Epigraphy*, No. 239 of 1904., *S. I. I.*, Vol. XVII, No. 262 and plate facing p. i. An eye copy of the record is published by T. A. Gopinatha Rao in the *Travancore Archaeological Series*, Vol. I, p. 231.

c15.7

*ai m pa t te ḻ = a ṉ = a*
*ca ṉa n no ṟ ṟa*
*ca n ti ra na n ti ā*
*ci ri ka r ni ci ti kai*

*Aimpattēḻ = anaśanan noṟṟa Candira—*
*Nandi āśirikar niśīidikai.*

c15.8

The *niśīidikai* of Candra-Nandi Ācārya who (died), having observed the fast (not taking food) for fifty-seven days. Another inscription assignable to a period earlier than the seventh century A.D. has also been recorded at the same place. So far it has not been possible to state definitely whether this record is to be treated as written in the Tamil or Vaṭṭeḷuttu script. Gopinatha Rao included this under both the varieties. Very likely the two scripts had not become separate once then. But some letters in the record clearly exhibit Vaṭṭeḷuttu forms. The letter *ai* is of the trident variety, already noticed in the Brāhmī lables at Aḷagarmalai.² It occurs only in the Vaṭṭeḷuttu while the form of this vowel in the Tamil script was derived from the Brāhmī form of a triangle, with a horizontal stroke to the left at the top (𑀏). The form of *ṇa* of this record is the same in the Vaṭṭeḷuttu and is also found in the Brāhmī labels from Pugaḻiyūr, while the form in the Tamil script is different. The letter *ma* has a flourish at the right in which the end of the line forming the loop is extended, just as in the Vaṭṭeḷuttu form which peculiarity is not found in the Tamil script. The symbol for the medial vowel *e* in the letter *te* and for the medial *o* in the letter *no* are also of the type found in the Vaṭṭeḷuttu and in the Tamil script. This record should therefore be considered the earliest in the Vaṭṭeḷuttu variety.

c15.9

² See *ante*, pp. 237-38.

c15.10

The date of the record presents another knotty problem. The earliest datable records in this variety of the script are the Madras Museum plates³ and the Vēḷvikkuḍi plates⁴ on copper and the Tirupparaṅkunṟam inscription on stone of the time of the Pāṇḍya king Parāntaka Neḍuñjaḍaiyan in the latter half of the eighth century A. D. Considering the Vaṭṭeluttu form of the letters in these records as well as the changes found in the evolution of the Grantha script and the form of the Brāhmī script of the inscription examined above, the Tirunātharkunṟu inscription may be taken as belonging to the fourth century A.D.⁴

c15.11

³ *Indian Antiquary*, Vol. XX, p. 57 and plate.

c15.12

⁴ *Epigraphia Indica*, Vol. XVII, p. 291 ff.

c15.13

In the cavern in the rock at Tiruchirapalli, described above, a few of the stone pillars of the beds show traces of obliterated writing which have been assigned palaeographically to the fifth century A.D. One of these is said to contain the letter *cira*. Evidently the name Cirāppaḷḷi or Tiruccirāpaḷḷi owes its origin to this label.⁵

c15.14

⁵ "At Trichinopoly which was visited during the year, some interesting discoveries were made. Its ancient name as found in the hymn of Jñānasambandha in the *Dēvāram* is Chirāppaḷḷi and the same occurs also in the long verse inscription of about the 11th century A. D., engraved in the Pallava cave on the hill. This name was in vogue for several centuries in inscriptions as well as in literature, until the time of the Vijayanagar rulers, in a few of whose records, however, the incorrect form 'Tiruchchirappalli' was sometimes used and this has given rise to the modern anglicised name 'Trichinopoly'. Though the word *paḷḷi* has several meanings it appears to have in this case special reference to its association with the Jaina and Buddhist religion, ancient vestiges of which have been found here." (*ARE.*, 1937-8, pt. II, para 3, p. 78).

c15.15

On the ledge of the rock leading to the cavern there is a label of four characters deeply cut and repeated at least in four places.⁶ This has been read as *Kamṭṭuhu*, evidently meant for *Kamṭṭugu* and interpreted as a Sanskritised Telugu word meaning 'enemy', suggesting 'that like many other similar titles of Mahēndravarman, this may also have been his *biruda*'. But the reading *Kamṭṭuhu* does not seem to be acceptable (sustainable), both palaeographically and etymologically⁷. If the script in which the label has been incised is taken as Grantha then the reading should be either *Kaṭamṭuhu* or *Kamṭaṭuhu* according to the place assigned to the *anusvāra*. In the early records of South India as also in those of North India, the *anusvāra* is written over or on the top of the previous letter. Thus in the Uruvapalli plates of Pallava Yuvamahārāja Viṣṇugopa the word *jitam*⁸ is written with the *anusvāra* dot on the letter *ta*. But this has been subsequently changed to the top of the succeeding letter, evidently treating it as the pure consonant *m*, thus making it appear more as a ligature. Not realising this some mistakes have been committed by earlier epigraphists, eg. the name Maṅgiyuvarāja, viz., Maṅgiyuvarāja of the Eastern Cāḷukya dynasty has been read as Maginduvarāja and the name of the place in the Pulimtura plates of the Viṣṇukuṇḍin Mādhavavarman has been read as Pulībūmra⁹. It is the latter practice that appears to have been followed in writing the labels at Tiruchirapalli. This will become clear from the label *Amīthananda* discussed later. It may also be mentioned here, that to avoid this confusion in the later grants of the Pallavas the *anusvāra* is found written not on the top of the letters, but as a small circle independently in the proper place, next to the previous letter.¹⁰

c15.16

⁶ 7. It also appears to have been incised in a fifth place, where only the two last letters are now visible.

c15.17

⁸ *Indian Antiquary*, Vol. V, p. 50.

c15.18

⁹ *Epigraphia Indica*, Vol. XIX, p. 258.

c15.19

¹⁰ Vide in the Kaśākkuḍi plates of Nandivarman Pallavamalla II. (*South Indian Inscriptions*, Vol. II, Plate XI).

c15.20

The formation of the first letter *ka* and of the symbols of the medial *u* attached to the letter *ṭu* and the contention that the label has been written in the Grantha script. In the Grantha of the period the vertical stroke of *ka* is doubled either completely or at least in the lower portion. But it is only a single stroke and not a doubled one. Similarly the symbol for the medial *u* usually found in the Grantha script of the period is a vertical stroke at the bottom of the character, also doubled in most cases. But here the mark to denote the medial vowel is in the shape of a crescent. The fourth and last character which has been read as *hu* and taken to represent the *visarga* (*ḥ*) is not seen written in this form anywhere. The hook-like curve on the left at the beginning of this letter is not found for the letter *ha* in any other record of the period.¹¹ We do not also know of any other instance where the *visarga* has been written as *hu*. The dot on the second letter has been taken as an *anusvāra*. In Sanskrit as also in Telugu, the hard consonant occurring after an *anusvāra* is not doubled. Again when consonants are doubled it is usual to have both the aspirated and unaspirated ones and not of the same variety as found here. Under the circumstance it is difficult to treat the record as having been written in the Grantha script.

c15.21

¹¹ Especia Nos 137, 138 and 140 of 1938.

c15.22

The label is to be taken as written in the Tamil script, and in the Tamil language, and read as *Kaṭṭuṇā* taking the dot on the second letter as indicating the pure consonant according to the rules of the Tamil grammar *Tolkāppiyam* referred to earlier.¹²

c15.23

¹² See *ante.*, p. 121.

c15.24

The word is made up of *Kaṭṭu* + *uṇ* + *a* negative, and may be interpreted as ‘one who is not bound; or ‘one who has transgressed the bond (of attachment)’. cf. *Kaṭṭuṇḍān* ‘one who has been bound’; ‘*Kaṭṭuṇṇi*’ ‘one who suffers himself to be bound.’ Thus the term *Kauṇṭṭa* is something like *vītarāga* and indicates one who is free from the bonds of worldly attachment.¹³

c15.25

¹³ The word *Kaṭṭu* also means control or regulation and thus may be taken as indicating ‘one who is not bound by regulations.’ We know that Pallava Mahēndravarman took pride in assuming such titles. If so it is possible that *Kaṭṭuṇā* is one of his titles in pure Tamil. We know some of his titles like *Citrakārapuli* (tiger among artists), *Kūṟṟambu* etc. though sounding like Tamil are really mixed with Sanskrit and Telugu. The term *Kaṭṭuṇā* would appears to be the first title of Mahēndravarman in pure Tamil.

c15.26

As mentioned earlier, this label is found engraved in four different places, along the precipitous approach to the cavern. In three places, a different label in early script is also engraved faintly below the word, while the fourth has a similar label inscribed above the word. One such¹⁴ is *Kaiyviḷakku*. It is to be noted that as in the previous label here also the letters *ya* and *ka* (second and fifth characters) have dots on them to denote that they are pure consonants. The addition of a superfluous *y* after *kai* is a peculiarity met with sometimes in the Tamil inscriptions of the period. The word *kaiyvilakku* means hand lamp or *karadīpam* which at once brings to our mind the word *Dīpaṅ-kara* found in the Buddhist literature as the name of one of the Bodhisattvas. Dīpāṅkara is considered as the patron deity of sea-farers and it is said that in times of distress, and when they are lost in the sea, *Dīpaṅkara* will appear to them with light and guide them to safety. Thus the label may be taken as meaning the torch or becon leading a person to eternal bliss. Evidently it denoted the name of a teacher.

c15.27

¹⁴ No. 138 of 1937-38.

c15.28

The above two are the Tamil labels inscribed on the rock in the Tamil script.

c15.29

Another¹⁵ of the labels inscrided below the word *Kaṭṭuṇa* is written in the Grantha script and has been read as *gatadoṣa*. The photograph of the impression shows above the last character *sa* a little to the right a dot which may be taken as indicating the final *m*. Thus the label would read *gatadoṣam* in Prākṛt and evidently stands for *gatadoṣam* in Sanskrit, which should properly be *gatadoṣaḥ*. The final *m* in the word is most probably due to the influence of the Tamil language, and is not the *visarga(ḥ)* required for a Sanskrit word. The term, would mean one whose fault or sin has gone i.e. a pure soul. The word *doṣa* also means night or darkness; and as such it is possible to take the term as indicating 'one who has dispelled darkness and thus brought light.'¹⁶

c15.30

¹⁵ No. 138 of 1937-38.

c15.31

¹⁶ The early Kadama king Kṛṣṇavarman II is known to have the title Doṣarās'i. In his Kirukuppaṭūr grant (line 7) he is called Doṣarāśi-rāgāṅkitaḥ (R. S. Panchamukhi, *Digest of the Annual Report on Kannada Research in Bombay Presidency for the year 1940-41*, p. 4; In his Bennur plates he is called Doṣarāśivarma, (*Epigraphia Carnatica*, Vol. V. Intron., p. III and Bl 245).

c15.32

The third label¹⁷ also in the Grantha script has been read as *Amitananda*. The third letter which has been read as *ta* does not have the loop and looks more like *tha*. Thus the label has to be read as *Amithananda*. Evidently this also is in prākṛt and stands for the Sanskrit form *Amithyananda* made up of the two words *amithya* and *ananda* with the long letter *ā* shortened as *a*.

c15.33

¹⁷ No. 137 of 1937-38.

c15.34

The word *mitha* is the Prākṛt form of the Sanskrit *mithya*, meaning falsehood, illusion. The term would then mean the Bliss of Disillusion or the Bliss that is not false. Very likely this also was the name of some person.

c15.35

The fourth label,¹⁸ consisting of three letters, written above the term *Kaṭṭuṇa* is also in the Grantha script and has been read as *pra[yu]lai*. The middle one is very peculiar and the last one may be read as *le*. As at present, the reading of this label cannot be explained. Above it are seen portions of a diagram having some squares intended for playing some sedantic game like the chess.

c15.36

¹⁸ No. 133 of 1937-38.

c15.37

In one place on the rock leading to the cavern is inscribed a label¹⁹ in Grantha script which has been read as *Tamcara[ka]*. Krishnamacharlu writes about it as follows :²⁰

c15.38

¹⁹ No. 135 of 1937-38.

c15.39

²⁰ *Annual Report on South Indian Epigraphy*, 1937-38, pt. ii, p. 79.

c15.40

" The word *Tamchahara[ka]* can be interpreted as a title of Mahēndravarman and to mean 'he who captured *Tañcha* (Tanjore).' From the Vēḷūrpāḷaiyam plates we know that Simhaviṣṇu, the father of Mahēndravarman, claimed to have conquered the Chōḷas; and in support of this fact it may be pointed out that Kañjanūr in the Tanjore District bore the surname Simhaviṣṇuchatur-—vēdimangalam in the Chōḷa times (No. 265 of 1907), testifying to its connection with the king of this name. As Pallava influence began to be felt in the Chōḷa country only from this period, it is possible that Mahēndravarman who may have participated in this southern expedition in the company of his father had adopted this title."

c15.41

This appears to be reading too much from the label. The last letter which has been read tentatively as *ka* has a loop or circle at the bottom and as such it is definitely not possible to read it as *ka*. It looks more like *ta* or *tha*. At the beginning before the letter *ta* there are traces of another letter which may be read either as *ta* or *tha*. The letter which has been read as *ca* looks more like *dha* and the *anusvāra* is indicated by a dot above. After the letter *ha*, two dots, one above the other, are seen placed, thus indicating the *visarga ḥ*. If these dots are ignored, then the space between *ha* and the following one, which has been read as *ra* will be disproportionate to what we find for the others. Thus the whole label will read:

c15.42

*bha* or *ta ta ṁ dha ha ḥ ra ta*
*bhataṁdha haḥrata* or *tataṁdha hahrata*.

c15.43

Is it possible to take the label as *Bhadanta Har(i) ta* i.e. the venerable *Harata* or *Harita*?²¹ And it is also doubtful whether the method of replacing the nasals with the *anusvāra* was in use at that early period and in this part of the country. As such it would be better to refrain from any attempt to interpret this label at this stage and leave it as it is.

c15.44

²¹ Even supposing that the label reads *Tamcaharaka* i.e. the capturer of Tañjāvūr, then it would indicate only Simhaviṣṇu who is really credited with the victory and not his son Mahēndravarman who is not associated with the victory in any place.

c15.45

In another place, we have a label²² written in the Vaṭṭeḷuttu script reading *Ceṉataṇṭaṉ*. The same label is found engraved again close by. Of these two labels, one is written in a single line while the other is written in two lines. Faint traces of a dot on the second letter *na* can be seen in the photograph of the impression. If that forms part of the inscription, then the label would be *Ceṇṭaṇṭaṉ* (*Śēnataṇḍan* or *Śentaṇḍan*). The suffix of third person singular masculine *aṉ* indicates that the label represents the name or title of a male person.²³ The labels

c15.46

²² No. 136 of 1937-38.

c15.47

²³ The term *Cenataṇṭaṉ* is made up of two words, *ceṉa* + *taṇṭaṉ*. The *first* word *ceṉa* is evidently the Sanskrit word *Seṇa* and the second *daṇḍa*. Both mean army. *Seṇa* will become either *Cēṉā* or *Cēṉai* in Tamil. On the other hand we know that *Cēṉaṉ* was used as a title of some Jains: c.f.

c15.48

(Sambandar, *Tiruvalavay padikam*). Appar who was the head of the monastery of Pāṭalipuram before his reconversion to S'aivism was known then as Dharmasēna. If the word in this label is intended as a title like that, it will come only as a suffix, and not as a prefix as is found in the label. The other word *daṇḍa* means a staff as well as obeisance. Sanyasis used to carry staff. The Vaiṣṇavas were known as *tridaṇḍins* and Smārtas *ēkadaṇḍins*. The word Markarin also indicates one who carries a staff, *daṇḍin*. Probably the word is equal to *daṇḍin* or Markarin.

c15.50

Plate—29. Tirunātharkunru evidently contain the names of the religious ascetics who were probably living there professing a religion different from the Vedic, either the Buddhist or Jain or the Ājīvika religion.

c15.52

Palaeographically these inscriptions may be assigned to the seventh century A.D. The rock-cut temple for Śiva excavated in the rock and the long inscriptions in Sanskrit incised there clearly show that Mahēndravarman took a special interest in the place and in the excavation of the Śiva temple there. If so it is doubtful if teachers of other religions inimical to Saivism would have found it possible to have their abode in the same rock and very near the temple. They should have been there earlier, and these labels also should have been incised earlier. They may be assigned to the sixth century A.D.

c15.53

We have referred to earlier the traces of obliterated writing on some of the pillows in the cavern. One of them which is fairly well-preserved reads *namostu*.²⁴ The Grantha characters in which this label is written appear old and the script has been assigned to the fifth century A.D.

c15.54

²⁴ No. 132 of 1937–38.

c15.55

Thus we have the following records after the time of the Brāhmī inscriptions discussed above before we come to the inscriptions in the Tamil country which become available from the beginning of the seventh century A.D.

c15.56

| *Grantha script* : | One label of the fifth century A.D. and four labels of the sixth century A.D. all at Tiruchirappalli. |
| *Tamil script* : | Two labels, of which one is repeated in four other places, all at Tiruchirappalli and assigned to the sixth century A.D. |
| *Vaṭṭeḷuttu* : | One epitaph of the fourth century A.D. at Tirunātharkuṉru and the other (seventh century) at Tiruchirappalli written twice. |

Cites (6)