archaeology

Locating Domains of Devotion, Patronage and Power in the Eleventh Century

by MICHAEL WILLIS

The large hill known as Udayagiri, not far from ancient Vidisa in central India, is a crucial site for the history of the Gupta kings and for the artistic and cultural dispensation that was forged during their rule. A number of well-known inscriptions, incised in or beside the hill’s cave-shrines, mention Candragupta II and members of his court.1 Equally well-known are the monumental relief sculptures of Visnu’s incarnations and a number of other deities.2 The amount of writing on this material over the last hundred years might give the impression that little of substance remains to be tackled, but such an impression is unwarranted. In the first place, Udayagiri has a long but unacknowledged history which goes back to at least the second century BCE. This is  shown by archaeological traces, not yet properly investigated, as well as inscriptions and sculptural remains which pre-date the Gupta period. We may note, as an example, a large lion capital of the late second century BCE now in the collection of the Archaeological Museum at Gwalior.3 This and related material shows that the Gupta presence at Udayagiri represents a reworking of an ancient site. The intrusive character of this reworking is demonstrated by the way in which fifth-century caves and images cut directly through older shell-inscriptions.4 The scale of change in Gupta times is further indicated by the massive sculpture of Varaha in Cave 5, the single-most important image at Udayagiri. Although the modern road has altered  the ground level to some extent, it is not difficult to see that the adjacent pond once came up to the edge of the hill and that water washed across the base of the image during the rainy season. In other words, the great Varaha once rose out of the water, exactly as the incarnation is visualised in Matsya Purana, chapter 248.5 This demonstrates that the changes made in Gupta times were simply a random series of additions by pious donors. Rather, the new caves, sculptures and environmental features involved a significant modification to the immediate landscape and the dedication of the hill to gods who were being reconceived in dynamic new ways. Extant scholarly writing ignores this and related aspects of the geographical setting. The myopia extends to the sculptures, which have been discussed as single items out of context, and to the cave-shrines, which have been described individually rather than as parts of a larger architectural ensemble.6

1. Udayagiri, Cave 19. Entrance door, fifth century. Courtesy of the Department of Oriental Antiquities, British Museum.

Turning to Udayagiri after the Guptas, there is no history at all, at least to the extent that historians have not written it. The aim of the present article, therefore, is to publish five inscriptions which provide information about the hill in the centuries following the Gupta dynasty. As will be seen in the concluding remarks, these inscriptions provide a number of insights into the Inscriptions from Udayagiri: locating domains of devotion, patronage and power in the eleventh century MICHAEL WILLIS 1. Udayagiri, Cave 19. Entrance door, fifth century. Courtesy of the Department of Oriental Antiquities, British Museum. character of religious patronage in the time of the Paramara kings. Placing the inscriptions in a regional context, also taken up in the concluding remarks, further shows that Udayagiri was a cult centre of considerable importance from at least the ninth century, a conclusion that is not immediately obvious from remains presently on site.

2. Udayagiri, Cave 19. fifth century with eleventh-century inscriptions. Courtesy of the Department of Oriental Antiquities, British Museum.

All the records discussed in this article are engraved on the pillars inside Cave 19 (Fig. 2). This cave, at the northern end of the hill, appears to have been excavated out of the hillside in the second quarter of the fifth century, probably in the time of Kuma -ragupta. The door of the cave (Fig. 1) is elaborately carved and there was a free-standing hall before the entrance.7 The records incised on the pillars record the renovation of the shrine and donations of land in its favour. One is dated Vikramayear 1093; the others probably belong to the same period.

1. INSCRIPTION DATED VIKRAMA YEAR 1093

This inscription is on the north face of the pillar on the left as one enters Cave 19.8 It is written in eight lines of nagari and carries the date Vikrama year 1093 (circa CE 1036-37). The language is corrupt Sanskrit. The inscription was first noted by Cunningham who erroneously took it to be a pilgrim’s record.9 It was subsequently published by Fleet in 1884 and later mentioned in the Annual Report on Indian Epigraphy.10 It does not seem to have drawn any further attention and is illustrated here for the first time (Fig. 3).

3. Udayagiri, Cave 19. Inscription number 1. Drawing by A. Searight. Courtesy of the Department of Oriental Antiquities, British Museum.

Text 11
(1) namo s´ri jirnnodhari
(2) kanha pran amati (3) visnupadau nityam
(4) sam vat 12 1093
(5) candraguptena ki
(6) rtanamkirtitamh
(7) pas´cat vikra
(8) madityarajyamh

Translation

Obeisance! Kanha, the glorious restorer of that which has decayed, bows forever to the feet of Visnu. The year 1093 after the reign of Vikramaditya.13 The temple was made by Candragupta.

4. Gold coin of Candragupta II. Department of Coins and Medals, British Museum (H. Nelson Wright Collection 1910. 4-3. 36).

5. Intaglio with devotees worshipping Vis.n. upada or Buddhapada, c. fourth century. Provenance unrecorded, probably ancient Gandhara. Department of Oriental Antiquities, British Museum (1880. 3787).

This inscription is of interest because it shows that six centuries after Candragupta there was a living tradition which associated him with Udayagiri and Vikramaditya.14 Candragupta’s use of the epithet Vikramaditya is documented by gold coins, one of which has an extended legend reading: ‘Having conquered the earth with good conduct, Vikramaditya conquered heaven’ (Fig. 4).15 The epithet Vikramaditya indicates that Candragupta was drawing an analogy between his own acts as king and Visnu’s trivikrama, the heroic three strides by which Visnu redeemed the world from evil. The requisite link between vikrama and aditya is provided by the association of Visnu’s three steps with the position of the sun at dawn, midday and sunset. This use of Vais.n. ava imagery as political allegory was much favoured by the Gupta monarchs, the most elaborate instance being the Udayagiri relief of Varaha to which reference has already been made.16 The present inscription becomes significant on this count for it specifies that Kanha ‘bows forever to the feet of Visnu (Visnupada)’. That the god’s feet were a common icon of worship from the early centuries of the current era is indicated by an intaglio now in the collection of the British Museum (Fig. 5). The most celebrated Visnupada temple is at Gaya in eastern India, but the present inscription shows that the Visnupada were also worshipped at Udayagiri and that special emphasis was placed on this worship by Candragupta. The importance and antiquity of the cult in the vicinity is shown by the Heliodorous pillar inscription which mentions ‘three steps to immortality’ (triniamutapadani).17 Not far from the pillar is Caranatirtha, a pilgrimage spot and bathing place on an island in the River Betwa (Fig. 6).18 The temples at Caranatirtha date to the nineteenth century, but incorporated into the platform above the bathing steps there are pillar fragments of about the sixth century. There are additionally a number of carved Visnupada, the oldest of which belonging to the Paramara period.

2. UNDATED LAND RECORD OF SODHA

This inscription, like the previous example, is on a pillar inside Cave 19. It is was noted by the Archaeological Survey of India in 1913-14 and subsequently listed in the Annual Report on Indian Epigraphy for 1957-58.19 It seems to be otherwise unpublished and is illustrated here for the first time in Fig. 7. The inscription is in three lines of nagari and the language is Sanskrit. The record may be assigned tentatively to the eleventh century in view of the palaeographic similarities to the inscription of 1093 just mentioned. Its purpose is to record the gift of some land to the temple, probably at the time of its renovation. I have not been able to trace the village of Maniyaraka. 

6. Udayagiri and key ancient sites in the neighbourhood. Drawing by A. Searight. Courtesy of the Department of Oriental Antiquities, British Museum.

7. Udayagiri, Cave 19. Inscription number 2 (bottom) and 3 (top). Drawing by A. Searight. Courtesy of the Department of Oriental Antiquities, British Museum.

8. Udayagiri, Cave 19. Inscription number 4 (bottom) and 5 (top). Drawing by A. Searight. Courtesy of the Department of Oriental Antiquities, British Museum.

Text
(1) maniyarakagrame
(2) rajaputrasodhena
(3) bhumi nivartanah.20

Translation

A plot (nivartana21) in the village of Maniyaraka [was given] by rajaputra Sodha.

3. UNDATED LAND RECORD OF SOMAPA - LA

This inscription is written directly above the land record just discussed. The history of its publication is the same.22 It is illustrated here for the first time in Fig. 7. The inscription is in three lines of nagari, more casually written than the land record of Sodha, and the language is corrupt Sanskrit. The meaning of the inscription is not clear by itself, but taken in conjunction with the previous inscription it can be understood as recording the gift of an additional plot of land to the temple, probably at the time of its renovation in the eleventh century.

Text
(1) mahasamamtaso
(2) mapalabhumi
(3) nivartana

Translation
A plot (nivartana) [of] mahasamanta Somapala

4. UNDATED LAND RECORD

This is another inscription in Cave 19 recording a gift of land. It does not seem to have been previously noted and is illustrated here in Fig. 8. It is written in two lines of casual nagari, palaeographically similar to the inscriptions just discussed. The language is corrupt Sanskrit. The term pali would seem to refer to part of the land belonging to a hamlet; more common forms are palli and pallika.23

Text
(1) rajaputravahilava
(2) hadena bhumipali

Translation
Village land (pali) [was given] by rajaputra Vahilavahada

5. RECORD OF DA- MODARAJAYADEVA

This inscription, like the previous examples, is on a pillar in Cave 19. The publication details are the same as inscription number 2 above. It is illustrated here for the first time in Fig. 8. The inscription is written in two lines of nagarii, and the language is Sanskrit. This has the appearance of a pilgrim record, but given its similarity to the neighbouring inscriptions could well be a cryptic or unfinished donative record. Palaeographically this inscription does not differ much from those just mentioned, but a more direct clue to the possible date is given by a record which mentions the name Damodara in connection with the construction of a memorial. That inscription is dated [Vikrama] year 1236.24

Text
(1) rajaputradamodara
(2) jayadevena

Translation
[Given?] by rajaputra Damodara Jayadeva

CONCLUDING REMARKS

The Cave 19 inscriptions record donations of land by minor lords (rajaputra, mahasamanta) and for this reason are comparable to the Kalvan copper-plates which provide an account of the land (also measured in nivarttana) given by samanta S´ri Amma to a Jaina Tirthankara.25 Unlike the Kalvan plates, however, the Udayagiri inscriptions make no mention of the ruling king. This restricted focus is remarkable considering that the activities date to the time of Bhoja, the mostcelebrated monarch of the Paramara dynasty.26 This circumstance prompts a number of questions. Perhaps the most compelling concerns how Cave 19 and its donors are to be situated in the cultural and political landscape of the eleventh century. This landscape needs to be considered both literally and metaphorically, that is to say, in terms of religious topography and dynastic history. While the evidence from Cave 19 is rather meagre in itself, other data of an archaeological and epigraphic nature provides considerable information about the period in question. As will be seen in what follows, this information has a direct bearing on Udayagiri.

As our concern is to situate a shrine and its donors in the broadest sense, it is not out of place to begin with a brief consideration of Vidisa. The river Vidisa (now called the Bes) is one source from which we can trace the town’s various old names: Vidisa, Vaidisa, Vedisa, Vedisanagara.27 All these refer to the ancient city that flourished up to about the sixth century and which occupied a large triangular area of land between the rivers Betwa and Bes (Fig. 6). The names ‘Bes’ and ‘Besnagar’ are less convincingly derived from the ancient name of the river because we would anticipate the preservation of the letter ‘d’ in later forms. A satisfactory solution seems to be found in Pali the vessa. The term Vessanagara (‘Town of Merchants’) goes back to at least the third century BCE when Asoka formed a liaison with Devi, the daughter of a leading merchant who lived there. That Devi was a vaisya (Palivessa) is documented by literary references.28 Devi continued to live in her native place after Asoka became king and the name Vessanagara could have enjoyed currency from that time.29 From Vessanagara, there is, of course, no difficulty getting to Besnagar. As to the actual history of the city, architectural fragments indicate that it was still a reasonably important place in the early fifth century. In later times the site was never completely abandoned, but after circa 600 most of the population seems to have shifted to the location of the present town. The reason for this move has not been explored but a clear indication of the change is provided by a massive stone lintel incorporated into one of the town’s gates. This lintel formed the over-door of a substantial temple dating to the seventh century. A step-well in the Bijamandal complex also belongs to this period.30 Epigraphic evidence demonstrates that in post-Gupta times the new town was not called Vidisa but rather Bhaillasvamipura or Bhailasvamidevapura, the ‘town of Bhaillasvami.31 From this comes Bhilsa, the name in the venacular from at least the eleventh century (it is recorded by alBiruni) and used until the Sanskrit form Vidisa was revived after Independence. What emerges from this evidence for Udayagiri during the eleventh century is the following: the hill and its cave-shrines stood near to a city-site that was in significant decline; all but a portion of the population appears to have shifted several kilometres to Bhaillasva mipura, the settlement popularly known as Bhilsa until fairly recent times.