Pramāṇa · Search · About · API

Early Tamil Epigraphy: From the Earliest Times to the Sixth Century A.D. c20.s2: 1.2 Earliest discoveries (1882-1903)

Early Tamil Epigraphy: From the Earliest Times to the Sixth Century A.D. c20.s2

Language
English

Contents

1.2 Earliest discoveries (1882-1903)

1.2.1 Mangulam: Pāṇḍiya inscriptions of the Caṅkam Age

c20.s2.t1.1

The cave inscriptions at Mangulam near Madurai were first noticed in 1882 by Robert Sewell who recorded the following observation about the site:

c20.s2.t1.2

Mauncolum [Mangulam]: 12 miles north-east of Madura [Madurai] and 6 miles west of Melur. A mile east of the village is a small hill called Karugumalai [Kalugumalai] in which is a cave with inscriptions. There is also a rock which the natives call *Pañchapāṇḍavar paḍukkai* or the beds of the Five Pāṇḍavas.¹

c20.s2.t1.3

¹ R. Sewell 1882: vol. I, p. 294.

c20.s2.t1.4

The stone beds on the Kalugumalai hill at Mangulam were also noticed by Francis² in 1906. The cave inscriptions at this site were rediscovered by Subrahmanya Aiyer³ in the same year.

c20.s2.t1.5

² W. Francis 1906: vol. I, p. 278.

c20.s2.t1.6

³ *ARE* 1906-07: paragraph 3. K. V. Subrahmanya Aiyer 1924: p. 276.

c20.s2.t1.7

(i) Confusion over the name of the site

c20.s2.t1.8

While Sewell and Francis had recorded that the Kalugumalai hill is in the Mangulam village, the next report in *ARE* 1906-07 stated that the hill belonged "partly to Arittapatti and partly to Mangulam". However, according to Krishna Sastri, the cave inscriptions are "nearly 3 to 4 miles away from the [Arittapatti] village and belong properly to a smaller village named Mangulam".⁴ In spite of such clear documentation, the site was inexplicably called 'Arittapatti' in *ARE* 1906-07 and the subsequent Annual Reports and this name was adopted by Krishna Sastri⁵ himself and by Subrahmanya Aiyer¹ and others² following their lead. It was only in 1966 that the original name Mangulam was restored to the site.³ By this time however, the eastern hamlet of Mangulam called Meenakshipuram had become a separate revenue village and the name of the site was again changed adding to the confusion.⁴ Considering the long association of Mangulam with the now famous cave inscriptions at this site, the name has been retained in this study.⁵ Incidentally, another cave with a Tamil-Brāhmī inscription has been discovered recently at Arittapatti itself.⁶

c20.s2.t1.9

⁴ H. Krishna Sastri 1919: p. 336.

c20.s2.t1.10

⁵ Ibid.

c20.s2.t1.11

² E.g., T.V. Mahalingam 1967: p. 201.

c20.s2.t1.12

³ I. Mahadevan, *Corpus of the Tamil-Brāhmī Inscriptions 1966* (1968): Nos. 1-6.

c20.s2.t1.13

⁴ The new name Meenakshipuram for the site was first proposed in R. Nagaswamy 1972d: p. 49.

c20.s2.t1.14

⁵ In ancient times, the Kalugumalai hill was most probably included in the village of Veḷḷaṟai mentioned in two of the cave inscriptions on this hill (*ARE 462* and *463-464 /1906* = Nos. *6* and *3* respectively. Numbers in bold italics refer to the inscriptions in the Corpus in this vol.) The place may be identified as the village of Veḷḷari-p-paṭṭi to the south of the hill (V. Vedachalam, personal communication).

c20.s2.t1.15

⁶ K.V. Raman and Y. Subbarayalu 1971: pp. 229-232. The inscription is included as No. 7 in the present Corpus.

c20.s2.t1.16

(ii) Mix-up of estampages

c20.s2.t1.17

There are four caves on the Kalugumalai hill (Fig. 1.¹), three of them at a higher level spread from south to north, and the fourth one at a lower level. The inscriptions in the upper caves are engraved on the brow of the boulders, and the one in the lower cave on its rear wall. The upper southern cave called the 'kitchen' has the longest inscription (*ARE 465/06*) of the site. The upper middle cave called the *paḷḷi-k-kūṭam* 'school' has two short inscriptions (461 & 462). The upper northern cave has a long inscription engraved in two segments (463-464) with only a small gap in between. The lower cave has another long inscription (460).⁷

c20.s2.t1.18

¹ K.V. Subrahmanya Aiyer 1924: p. 289.

c20.s2.t1.19

⁷ The following is the correspondence between the serial numbers of the inscriptions of 1906 in *ARE 1906-07* and those in the present Corpus: 460 = No. *2*; 461 = No. *5*; 462 = No. *6*; 463-464 = No. *3*; 465 = No. *1*. Estampages of only three inscriptions (460, 462 and the first segment of 463-464) have been published in *ARE 1917-18*: Pl. facing p. 6. Another inscription has been discovered more recently in the upper southern cave (*ARE B.242 / 1963-64* = No. *4*).

c20.s2.t1.20

Krishna Sastri mixed up the estampages from two different caves in his readings, interposing the two short inscriptions (461 & 462) in between the two segments of the longer inscription (463-464).⁸ Subrahmanya Aiyer also followed this scrambled order in his readings.⁹ Not surprisingly, neither reading makes much sense. Other epigraphists, relying on the earlier readings without visiting the caves themselves, have made the same mistake.¹⁰ It was only in 1966 that the correct order of reading was restored on the basis of my study of the inscriptions *in situ*.¹¹

c20.s2.t1.21

⁸ H. Krishna Sastri 1919: No. IV B to E at p. 338.

c20.s2.t1.22

⁹ K.V. Subrahmanya Aiyer 1924: No. IV A to C at p. 292.

c20.s2.t1.23

¹⁰ E.g., T.V. Mahalingam 1967: pp. 201-211.

c20.s2.t1.24

¹¹ *Corpus of the Tamil-Brāhmī Inscriptions 1966* (1968): Nos. 3, 5 and 6 corresponding to the same serial numbers in the present Corpus.

c20.s2.t1.25

Figure 1.1. Mangulam: general view of the Kalugumalai hill.

c20.s2.t1.26

Figure 1.2. Mangulam: cave with the 'forgotten' inscription of Neṭuñceḻiyaṉ.

c20.s2.t1.27

(iii) The'forgotten' inscription of Mangulam

c20.s2.t1.28

The longest inscription at Mangulam,¹ which also happens to be the earliest and historically one of the most significant of the Tamil-Brāhmī inscriptions, was totally forgotten after its rediscovery in 1906 and was not heard of again for the next sixty years and its estampage has never been published.²

c20.s2.t1.29

¹ *ARE 465/1906 = No. 1.*

c20.s2.t1.30

I visited the Mangulam caves for the first time in November 1965. The bright winter sun had lit up the facade of the upper southern cave (Fig. 1.²) and the long one-line inscription on the brow of the boulder caught my eye. It was weather-worn and looked more like bruising on the rock than engraving, but was otherwise undamaged. The bold and legible letters were clearly visible to the naked eye. As I spelled out *ne-tu-ñ-ca-li-ya-ṉ* (*neṭuñceḻiyaṉ*) and *va-lu-t-ti-y* (*vaḻuti*), I realised with astonishment that I was indeed looking at an inscription of a Pāṇṭiya king of the Caṅkam Age not reported earlier. However, I found out later that I had only rediscovered the long-forgotten inscription of Mangulam, which had till then existed only as a catalogue number (*ARE 465/1906*).³

c20.s2.t1.31

² See Fig. 8.1 (direct tracing) and Pl. 1 (computer-enhanced print from direct colour photographs) of the inscription. See also sections 2.11 and 2.17 on copying of the cave inscriptions.

c20.s2.t1.32

³ In May 1966, I called on the veteran epigraphist, K.V. Subrahmanya Aiyer, to apprise him of my findings at Mangulam and Pugalur and seek his blessings. In the course of our discussion that day, I asked him about the 'forgotten' inscription at Mangulam. He recalled that when he visited the site in August 1906, there was a large beehive on the brow of the cave very near the inscription, which made it hazardous to take a closer look or make copies. As the estampage made available later was too illegible, the study was not pursued further.

c20.s2.t1.33

During the same visit, I discovered that the name *neṭiñcaḻiyaṉ* (*neṭuñceḻiyaṉ*) also occurs in the long inscription⁴ in the lower cave at the site. It is strange but true that the incorrect reading of a single letter (*li*) as *ṭṭhi* by Krishna Sastri⁵ and as *ṟi* by Subrahmanya Aiyer⁶ resulted in their failure to recognise the famous name and the historical significance of the record. Both inscriptions, illustrated with tracings directly made from the stone, were published by me in 1966.⁷

c20.s2.t1.34

⁴ *ARE 460/1906 = No. 2*. See also Fig. 8.2 (tracing) and Pl. 2 (computer-enhanced print from an estampage).

c20.s2.t1.35

⁵ H. Krishna Sastri 1919: pp. 337-338.

c20.s2.t1.36

⁶ K.V. Subrahmanya Aiyer 1924: pp. 289-292.

c20.s2.t1.37

⁷ *Corpus of the Tamil-Brāhmī Inscriptions 1966* (1968): Nos. 1 & 2 = *Nos. 1 & 2*.

c20.s2.t1.38

The two cave inscriptions of Neṭuñceḻiyaṉ at Mangulam are the oldest historical records in Tamil discovered so far.⁸ Their archaic linguistic and palaeographic features indicate a date around the 2nd century B.C. The inscriptions record the gift of a monastery to Nanta-siri Kuvaṉ, a senior Jaina monk. These are the earliest known Jaina inscriptions of South India and among the oldest of the kind in the country.

c20.s2.t1.39

⁸ In the course of three lectures in the Seminar on Archaeology at Madurai University (I. Mahadevan 1970a: pp. 12-13 ), I referred to these two inscriptions. T.P. Meenakshisundaran, the Vice-Chancellor of the University, proposed that an expedition be organised the very next day to see the inscriptions at Mangulam, hardly an hour's drive from the city. The expedition was led by R. Nagaswamy, Director of Archaeology, Tamil Nadu, and included, besides myself, many archaeologists, epigraphists and linguists who were participating in the Seminar. The expedition proved to be a success as both the inscriptions were clearly visible and could be read without difficulty.

1.2.2 Edakal: Cēra inscriptions of the Caṅkam Age

c20.s2.t2.1

Edakal hill, about 6 km. south-west of Sultan's Battery (Sultan Batheri) in Wynad District, Kerala, is near the junction of the present states of Karnataka, Kerala and Tamil Nadu. The hill is about 1500 ft. above the local ground level and 4000 ft. above the mean sea level. On the western slope of the hill near the summit is a large cave whose walls are covered with prehistoric carvings, among which a few inscriptions of later times are found (Fig. 1.³).

c20.s2.t2.2

The cave was discovered in 1894 by Fawcett, the Superintendent of Police, Malabar District. He visited the cave several times more in the next two years and made careful drawings and took photographs of the rock carvings and inscriptions. He transmitted the photographs to Hultzsch, the Government Epigraphist, for his comments on the inscriptions. Hultzsch made immediate arrangements for taking estampages of the inscriptions and published a brief note (without illustrations) on the discovery.¹ He also forwarded the estampages with his detailed comments on the inscriptions to Fawcett who incorporated them along with his own drawings and photographs in the paper published by him in 1901.² Hultzsch identified four inscriptions at Edakal as in "cave characters" (the older name for the Southern Brāhmī script).³ He read two of them, one each in Sanskrit⁴ and Tamil,⁵ and assigned them to about the end of the 5th century A.D. He declared that the other two inscriptions⁶ were "unintelligible" to him, but "decidedly archaic".

c20.s2.t2.3

¹ Madras Government Orders Nos. 1062 & 1063, Public, dated 10th August 1897, paragraph 14; Nos. 120-123/ 1897 (in *ARE 1896-97*). Cf. one-volume reprint of *ARE 1887-1905* (1986).

c20.s2.t2.4

² F. Fawcett 1901(reprint 1985): pp. 409-421. See also Rajan Gurukkal and Raghava Varier 1999: pp. 85-95, 191-197.

c20.s2.t2.5

³ *ARE 120-123/1897* corresponding to estampages 1-4 in Fawcett 1901 (Pl. facing p. 412). No. 120 is a Sanskrit inscription not included in the present study. Nos. 121 & 122 corresponding to Nos. ***118 & 119*** are Tamil inscriptions in the Early Vaṭṭeḻuttu script. No. 123 is actually a pair of short inscriptions engraved in one line, but separated by a natural cleft in the rock. They are included as Nos. ***81 & 82***.

c20.s2.t2.6

⁴ *ARE 120/1897*.

c20.s2.t2.7

⁵ *ARE 121/1897*.

c20.s2.t2.8

⁶ *ARE 122 & 123/1897*.

c20.s2.t2.9

(i) *Inscriptions neglected and lost

c20.s2.t2.10

One would have expected that the announcement by Hultzsch of the discovery of inscriptions in 'cave characters' not known till then to occur in the Tamil country (including the Kerala region) would have aroused the curiosity of other investigators to examine the unread inscriptions. What happened thereafter is an incredible story of long neglect and the eventual loss of the two earliest inscriptions at Edakal.⁷ During the next century, that is, between 1897 when Hultzsch had the estampages taken and 1995-96 when our team visited Edakal, there is no record of any attempt to read the inscriptions. In course of time, the inscriptions fell into total oblivion as there is virtually no mention of them in any later epigraphical publications.⁸

c20.s2.t2.11

⁷ *ARE 123 / 1897*.

c20.s2.t2.12

⁸ The only exceptions are: *SII*. VI (1928): Nos. 74-77 = *ARE 120-123/1897* without the texts. The Sanskrit inscription at Edakal as read by Hultzsch was republished in the *Corpus of the Kadamba Inscriptions*. No. 50.

c20.s2.t2.13

The rediscovery of the Edakal cave inscriptions came about accidentally. I was attending a seminar on epigraphy held in March 1995 in the University of Kerala at Thiruvananthapuram. There I saw exhibited enlarged photographs of Hultzsch's 1897 estampages of the Edakal inscriptions (taken from the 1985 reprint of *Indian Antiquary*, vol. XXX). None of the epigraphists from Karnataka, Kerala and Tamil Nadu who were participating in the seminar had heard of these inscriptions. The photographs clearly showed that two of the inscriptions were in the Tamil-Brāhmī script of ca. 3rd century a.d. Both commence with the Tamil word *kō* 'king, chieftain' and one of them ends with the alveolar *ṉ* available only in Tamil-Brāhmī.¹

c20.s2.t2.14

¹ *ARE* 123 / 1897 = estampage No. 4 in Fawcett 1901= Nos. *81* & *82*.

c20.s2.t2.15

The unexpected discovery of the occurrence of Tamil-Brāhmī inscriptions in Kerala led to organising an expedition to the Edakal cave in October 1995 and again in February 1996.² During our first visit to the cave, we copied the Sanskrit inscription and the two Early Vaṭṭeḻuttu inscriptions, all deeply engraved on the south rock wall of the cave and still in an excellent state of preservation. But in spite of intensive search we could not locate the pair of Tamil-Brāhmī inscriptions reported to be engraved on the north rock wall of the cave. We had to discontinue the search as it was getting late and we had to reach the plains before darkness fell. A second visit was organised by the team in February 1996 and this time, we were equipped with exact measurements based on Fawcett's photographs of 1894-96. We were able to locate the spot ³ where the inscriptions ought to have been, but found, to our dismay, that they had been totally obliterated by the mindless vandalism of tourists who had incised their names over them on the soft rock.⁴ At present, the only source for these two inscriptions is the excellent reproduction of Hultzsch's 1897 estampage in Fawcett 1901.⁵

c20.s2.t2.16

² The members of the team which visited the cave were, besides myself, Rajan Gurukkal and M.R. Raghava Varier from the Mahatma Gandhi University, Kottayam, S. Swaminathan from the Directorate of Epigraphy, ASI, Mysore, and A. Seetharaman, Thanjavur.

c20.s2.t2.17

³ See Fig. 1.3 in the present volume reproducing the photograph in Pl. VI: Fig. 1 in Fawcett 1901. The Tamil-Brāhmī inscriptions are visible at top left. They are no longer extant.

c20.s2.t2.18

⁴ The vandalism had started right from the discovery of the cave. Hultzsch noted, "To the right of the second line, [photograph] No. 1. shows the signature of a certain C. Kannan in modern Malayalam characters". Fawcett added drily, "In reference to the above remarks, I may note that the C. Kannan was the work of one of my own men" (Fawcett 1901: p. 412). Mercifully, the said C. Kannan did not engrave his name over the inscriptions, a fate which later befell *ARE* 123 / 1897 (estampage No. 4 in Fawcett 1901).

c20.s2.t2.19

⁵ Reprinted in Pl. 37 in this volume.

c20.s2.t2.20

(ii) *New discoveries at Edakal

c20.s2.t2.21

The proverbial silver lining in the dark cloud appeared in the form of the discovery, as a result of our intensive search in February 1996, of two more Tamil-Brāhmī inscriptions ⁶ on the north rock wall of the cave, which had earlier escaped the attention of Fawcett. All the four short Tamil-Brāhmī inscriptions appear to be contemporary and may be assigned to ca. 3rd century a.d. on palaeographic grounds. The occurrence of the expression *kaṭummiputa cēra* in one of the newly discovered inscriptions¹ proves that these are records of the times of the Cēra dynasty of the Caṅkam Age. The Tamil-Brāhmī inscriptions from the Edakal cave have now been read and are included as Nos. **79-82** in the present *Corpus*.²

c20.s2.t2.22

⁶ Nos. *79* and *80*. The inscriptions were discovered by S. Swaminathan of our team.

c20.s2.t2.23

¹ In No. *80*.

c20.s2.t2.24

² Our expedition to Edakal also led to an improved interpretation of the Sanskrit inscription in the cave, a revised reading of the Early Vaṭṭeḻuttu inscription read by Hultzsch and the decipherment of the second Early Vaṭṭeḻuttu inscription in the cave regarded as "unintelligible" by him (the latter two in Nos. *118 & 119* respectively). See the detailed report on the expedition to the Edakal cave (I. Mahadevan 1999: pp. 1-19.)

c20.s2.t2.25

Figure 1.3. Edakal: cave with prehistoric rock carvings. Photograph taken in 1894-96. The Tamil-Brāhmī inscriptions (top left) are no longer extant.

c20.s2.t2.26

Figure 1.4. Kilavalavu: cave with Tamil-Brāhmī inscription. (The pillars are recent structures.)

1.2.3 Kilavalavu: the earliest reference to Tondi on the east coast

c20.s2.t3.1

In 1903, Venkoba Rao discovered an inscription in "archaic characters" in a cave on the rocky outcrop at Kilavalavu, a small village about 38 km. north-east of Madurai (Fig. 1.4).³ Even though the inscription is engraved in bold and legible characters and is reasonably well-preserved, it could not be read and its significance was not recognised until much later after many similar inscriptions had come to light. Neither Venkayya⁴ nor Krishna Sastri⁵ refers to it in the Annual Reports on the discovery of the cave inscriptions up to 1909. It was only in 1910, seven years after the discovery, that the Kilavalavu cave and its inscription find a mention in Krishna Sastri's report.⁶ Even thereafter, the estampage of the inscription was not included in the large number of photographs of the cave inscriptions published in the Annual Reports during 1911-18.⁷ The correct reading of this inscription has eluded the investigators until recently probably because of the unusual manner in which it is engraved.⁸ The present Corpus includes the most recent direct reading from the stone and the interpretation based on it.⁹ The inscription is notable for the earliest reference to the port of Tondi on the east coast.

c20.s2.t3.2

³ G. Venkoba Rao: *ARE* 135 /1903 = *SII.* VIII: No. 422 = No. *10*. He also discovered another inscription in similar characters in this cave. Krishna Sastri makes a passing reference to it as "much damaged" (*ARE* 1909-10: paragraph 4). The inscription is not included in *ARE* 1903 or *SII.* VIII, but I have seen the almost completely illegible estampage (133B / 1903) in the Directorate of Epigraphy, ASI, Mysore. During my visits to the cave, I have not been able to locate this inscription which seems to have been completely eroded as it was engraved above the drip ledge.

c20.s2.t3.3

⁴ *ARE 1906-07*: (paragraphs 1-5) and 1907-08 (paragraphs 1-7).

c20.s2.t3.4

⁵ *ARE 1908-09*: (paragraphs 1-9).

c20.s2.t3.5

⁶ *ARE 1909-10*: (paragraph 4).

c20.s2.t3.6

⁷ The estampage was first published in T.V. Mahalingam 1967: Pl. 7.

c20.s2.t3.7

⁸ The inscription is engraved from right to left and with most (but not all) of the characters turned upside down. See section 5.8 (Direction of writing) for discussion on this peculiar feature found at Kilavalavu and a few other caves.

c20.s2.t3.8

⁹ No. *10* is a revised version of my earlier attempt in the *Corpus of the Tamil-Brāhmī Inscriptions 1966* (1968): No. 9). See also Fig. 8.8 (tracing) and Pl. 8 (estampage).

Linked texts (2)