In memory of Paul Dundas (1952–2023)

Earlier in the pages of this journal, I addressed some of the problems connected with the city of Dhār, king Bhoja, and the legendary image of the goddess of learning connected with him and the Paramāra kings of central India more generally.Footnote ¹ An important piece of evidence for the cult of Sarasvatī in the eleventh century comes from an inscription of Bhoja’s time that is engraved on the plinth of a sculpture that has been in the collection of the British Museum since the nineteenth century.Footnote ² Although some early authors associated the sculpture with the campus of the Kamāl Maula masjid, it was actually found on the site of the city palace in 1875, as reported by C. B. Lele in the 1940s.Footnote ³ A recent study of the historical geography of Dhār has resulted in a new map of the old city, shown here in Figure 1. This shows that the city palace and the Kamāl Maula campus are almost one kilometre apart, the latter standing on the edge of the settled urban area.

Figure 1.

Dhār (Madhya Pradesh). Topographical study of the old city prepared in 2025.

Figure 1 Long description

The map illustrates the topographical layout of Dhār, Madhya Pradesh. It includes features such as the medieval earthen rampart, grid footprint of the inner city and bodies of water. Key locations are marked, including the Kamāl Maula campus and the city palace. Water bodies surround much of the city, with labels naming the tanks. The fort, in a commanding position on a hill to the north-east, was rebuilt in the fourteenth century.

Source: Courtesy of Saarthak Singh and the author.

The earlier reading of the inscription published in 2012 was adequate as far as it goes but was frustrated by the abraded surface and the diagonal fractures across the inscribed part of the stone. In recent years, however, new tools have emerged, in particular Harry Falk’s Indoskript, an online tool that has transformed the study of Indian palaeography.Footnote ⁴ Coupled with a specially prepared photograph, we are in a position to re-evaluate this record (Figure 2).

Figure 2.

Dhār (Madhya Pradesh). Ambikā sculpture found on the site of the city palace in 1875, detail of the inscription on the plinth. (British Museum, no. 1909,1224.1).

Source: Courtesy of the Trustees of the British Museum.

In the first place, my re-evaluation was prompted by a small votive figure of a kneeling woman on one of the receding faces of the plinth (Figure 3). In accord with conventions prevailing in medieval India, this can be understood as the donor. The image is labelled, the reading being sāna. With this as a starting point, I went back to the inscription. Here I present the results of my work in two parts: firstly, a diplomatic reading, giving the text as it appears to be written, line-by-line without corrections and, secondly, a critical reading, with the text corrected and displayed as a verse with a prose colophon.

Figure 3.

Dhār (Madhya Pradesh). Ambikā sculpture found on the site of the city palace in 1875, detail showing kneeling votive figure on a receding face of the plinth (British Museum, no. 1909,1224.1).

Source: Courtesy of the author.

DIPLOMATIC EDITION:

CRITICAL EDITION (the metre is Śārdūlavikrīḍita):

Colophon (Prose):

The key change from earlier published editions is the addition of the donor’s name: sosā. That the letters here are sosā seems reasonably certain when comparisons are made to inscriptions of the period; for example, the Paramāra inscription from Udaypur (Figure 4). The name Sosa appears in the Jaina Prosopography Database, so we are on firm ground historically even if the donor named in our inscription is not documented otherwise.Footnote ⁵ However, on the face of it, the label (sāna) and the inscription (sosā) do not agree. My understanding of the label is that it is a mason’s abbreviation of the main text, sosā nāma being abbreviated as sanā. Sosā itself is a vernacular name, perhaps coming from Soṣā (su+uṣā).

Figure 4.

Palaeographic chart showing the letters sosā nāma, drawing on the examples provided by Harry Falk and Oliver Hellwig, Indoskript 2.0 (Berlin, 2017). Key to letters shown: so (left to right): Udaypur stone inscription, ID: 476; Bilhāri stone inscription, ID: 704; Udaypur stone inscription, ID: 476; sā (left to right): Udaypur stone inscription, ID: 476; Māndhātā plates, ID: 582; Bilhāri stone inscription, ID: 704; nā (left to right): Khairha plates, ID 55; Māndhātā plates, ID: 582; Udaypur stone inscription, ID: 476; ma (left to right): Nadol plate, ID 53; Arthūṇā stone inscription, ID 726; Udaypur stone inscription, ID: 476.

Source: Courtesy of the author and Indoskript.

However we choose to understand the name, this is the proper subject of the first half of the verse. The word following is much damaged, as can be seen in Figure 2, but the Śārdūlavikrīḍita metre and the thematic context restrict the possibilities. The word following sosā nāma is anyway an adjective describing the donor. Based on the parts of the letters visible, I suggest here suśāsanā (i.e. ‘she of good discipline’ or ‘she whose commands are good’). Another possibility is subhāṣaṇā (i.e. ‘well-spoken’). After this, the compound sukhaprasthāpanāyāpsarāḥ can be construed two ways, either as sukhaprasthāpanā yāpsarāḥ (i.e. ‘who is an apsara sending forth/distributing happiness’) or as sukhaprasthāpanāya+apsarāḥ (i.e. ‘an apsara for sending forth/distributing happiness’).Footnote ⁶ The second is the most likely.

Describing our donor Sosā as an apsara might seem out of place and unusually sensuous for a Jain patron. In Jainism, however, these are semi-divine beings who have the celebration of the Jinas as their primary function. Such figures appear often in the image surrounds of Jain sculptures. By way of example, I illustrate here an Ambikā from Sehore, Madhya Pradesh, a town in the district neighbouring Dhār. In iconographic terms, this image is close to that under study in this article.Footnote ⁷ Although damaged, a small standing female figure can be seen in the upper part, as well as a broken celestial couple carrying garlands (Figure 5).

Figure 5.

Sehore (Madhya Pradesh). Standing figure of Ambikā, eleventh century. As documented in the Collector’s Office, Sehore.

Source: Courtesy of the author.

Our donor Sosā is qualified as dharmmadhīḥ. The dha at the beginning has been omitted by the scribe but is required by the metre. This can, of course, be both masculine and feminine. In my earlier publication, I took it as masculine and construed it with vararuciḥ in the second half of the verse. We will come back to that shortly, but my guess that vararuci was a masculine name forced me to understand dharmmadhīḥ as masculine and thus as a technical term, some sort of religious supervisor who had his mind fixed on the dharma. This is possible but contrived. It is more plausible and natural to take the obvious meaning in the feminine—‘she whose mind is fixed on the dharma’.

With these observations providing a route into understanding the first half of the verse, we come to the long compound at the start: śrīmad-bhoja-nareṃdra-caṃdra-nagarī-vidyādharī. Thanks to an observation made by Paul Dundas in a seminar at an early stage of my research on this inscription, vidyādharī can be taken as referring to a Jain lineage, the Vidyādharī śākhā. In his characteristic way, Paul presented this as a modest aside, but it was transformative and, in a stroke, allowed me to understand the gist of the inscription. The Vidyādharī is an old school with documented roots in central India and with further inscriptions from the Akota bronze hoard showing it was active in medieval times.Footnote ⁸ The name is given as vidyādharī and vidyādhara in the sources, the latter appearing to be somewhat later in time. For example, an inscription of the reign of Kumāragupta in the fifth century records the donation of an image at Mathurā by a woman and lay follower named Śāmāḍhyā at the instance of Dattilācāryya, a preceptor belonging to the Vidyādharī śākhā. Footnote ⁹

For the rest, there are various ways of construing śrīmad-bhoja-nareṃdra-caṃdra-nagarī. While there are lineages with the name Candra, Candranagarī is not found. So nagarī has to be taken independently, its ending in ī explained by the fact that the word is referring to Dhārā, the capital of Bhoja and, as we know, the find spot of the sculpture itself. We are left, then, with śrīmad-bhoja-nareṃdra-caṃdra, which can be glossed as ‘the illustrious Bhoja, that moon among kings’, a fairly standard turn of phrase in inscriptions.

The second part of the verse is not problematic with regard to its core meaning: the donor Sosā made the image of Ambā (i.e. the sculpture on which the inscription is incised (sosā … ambāṃ nirmame)). The only problem is the word vararuciḥ. As mentioned above, I took this earlier as a masculine name in the nominative singular. With the subject appearing in the first part of the verse as Sosā, this now needs to be understood adjectivally, describing the donor as ‘she of excellent taste’. The word order is unusual, however, and the description of the donor in this way is not compelling. In terms of metre, the syllable needs to be heavy, and both vararuciḥ and vararuciṃ meet the requirements. Considering that the first half of the verse is given over to the donor and the second half is given over to the image, and given also the number of scribal errors and the need to restore all the anusvara marks, the best critical decision seems to be an emendation to vararucīṃ.Footnote ¹⁰ The image is thus ‘of excellent beauty’. This is more natural and logical.

Colophon does not present difficulties. The sūtradhāra is the mason, while a vijñānin is a craftsman, usually a term describing those responsible for engraving inscriptions.Footnote ¹¹ The date is Vikrama saṃvat 1091 current, as with most Paramāra records. The equivalent is 1034–1035 CE. This places the inscription in the middle of king Bhoja’s reign.

This is now the place to pull together the observations in the foregoing paragraphs and offer a new translation.