Sanskrit and the labour of gender in early modern South India

Abstract

In the early eighteenth century, a husband and co-wife trio undertook a household project in Maratha-ruled Tanjavur. These migrants from the Western Deccan jointly authored a set of Sanskrit commentaries invested in the idea of ‘Maharashtrianness’. The unusual authoring of a Sanskrit commentary by these women alongside their husband exemplifies broader changes that were taking place in Sanskrit intellectual circles in early modern South India. Tracing new formulations of regional identity, changing ideologies of gender, and shifts in the very labour of Sanskrit intellectual production, I demonstrate how new avenues of access to Sanskrit emerged for women in early modern South India. These new avenues of access were facilitated by the growing importance of the household as a site of cultural production and the rise of new regional courts in the Karnatak and Coromandel Coast regions.

Introduction

In 1730, Tukkoji Bhonsle I ascended the throne of Maratha-ruled Tanjavur as its sole ruler.¹ Maratha-ruled Tanjavur was located on the opposite coast of the subcontinent from Śivāji Bhonsle's Maratha polity sprawling out from the Pune area.² Despite both states sharing in the Bhonsle family lineage, Maratha-ruled Tanjavur displayed a political ambivalence towards Śivāji's Maratha state in the Western Deccan for the first 36 years of its existence.³ Looking to their Nayaka predecessors and the neighbouring courts of the Karnatak and Coromandel Coast regions, the Maratha-ruled court of Tanjavur fostered a highly multilingual cultural ecumene and offered a resurgence of centralized patronage for Sanskrit intellectuals.⁴

Although the multilingual Maratha-ruled court of Tanjavur remained politically independent, with the accession of Śarabhoji (r. 1712–1730), co-ruling with his brother Tukkoji (r. 1730–1735), Tanjavur started to memorialize and reaffirm its ties to the Maratha line of rulers in the Western Deccan.⁵ Elite families migrated to Tanjavur, just as they had migrated from the sixteenth century onwards to other Deccani courts and intellectual hubs such as Banaras. At crucial junctures these migrants chose to memorialize their ties to various state entities.⁶

Writing from Tanjavur, in the midst of this significant shift was the prolific Sanskrit author and commentator, Ghanaśyāma.⁷ A Brahman from Maharashtra and minister to Tukkoji, Ghanaśyāma called himself ‘The diamond-in-the-crown among the fine company of Maharashtra Brahmans'.⁸ As a scholar, his unique Sanskrit commentarial style mobilized vernacular language, particularly Marathi, as a critical philological tool for either establishing or undermining an intellectual position. This style epitomized both the multilingual vigour of the Tanjavur intellectual world and a growing orientation towards the Maratha-state in the distant Western Deccan.

More anomalous, however, is this: Ghanaśyāma, who regularly paid tribute to his two wives Sundarī and Kamalā in his writings, also undertook a joint commentarial project with them on the play, the Viddhaśālabhañjikā, written by the eminent author Rājaśekhara (active circa 885–950 ce).⁹ Their commentaries, which make the case for recovering evidence of the author's ‘Maharashtrianness’, also subtly subvert the social and aesthetic gender norms of Sanskrit intellectual production to which women rarely had access.

The commentaries of Ghanaśyāma, Sundarī, and Kamalā offer a unique opportunity to examine early modern trends in cultural production and social relations among migrant Maharashtrian Brahman circles and at the royal courts of the Karnatak and Coromandel Coast. The commentaries present a rich archival record of Brahman migration out of the Western Deccan, new eighteenth-century considerations around regional identity, the centrality of familial relations in Sanskrit intellectual production, and the abstraction and representation of these changes in processes of gendering alongside new avenues of access to Sanskrit for women. Buried in the intricacies of philological conventions, this comes to light most effectively through attending to what I term the labour of gender in the household project of Ghanaśyāma and his co-wives.

The problem of gender in Sanskrit

Despite the almost universal dictate that Sanskrit was the language of men, a certain subset of women could, and did, have access to Sanskrit learning.¹⁰ The authorial voices of women have been recorded as far back as the Vedic corpus, and can be found scattered across inscriptional records and anthologized verses in premodernity.¹¹ Nevertheless, this was uncommon. Between the fifteenth and seventeenth centuries, however, an unprecedented cluster of women authoring full-length Sanskrit texts emerged in the self-consciously multilingual and newly established Vijayanagara successor-states of the South.¹² Still, even against this backdrop, the project of Ghanaśyāma and his co-wives Sundarī and Kamalā was unorthodox.

For one thing, the project entailed the entry of these two women into an even more rarefied segment of the Sanskrit world—commentarial production.¹³ This marks a departure from other premodern and early modern women writing in Sanskrit. Secondly, the project entailed a division of labour between husband and co-wives, with Sundarī and Kamalā co-authoring their portion of the project together.¹⁴ Ghanaśyāma authored one commentary entitled the Prāṇapratiṣṭhā, and the co-wives authored a second accompanying commentary entitled the Camatkārataraṅgiṇī.¹⁵

Although joint authorship was common among early modern Sanskrit intellectuals, the way in which the trio divided their labour was unusual. The husband attended primarily to the Prakrit portions of the text (associated with the female voice) and the wives attended primarily to the Sanskrit portions of the text (associated with the male voice). This move emphatically breaks with the conventional gendering of the languages with which their project engaged. Through their commentaries, the trio set out to demonstrate the Maharashtrian identity of the author Rājaśekhara, in the process chipping away at the highly regulated cosmopolitan and gendered boundaries of Sanskrit linguistic usage.

To write in Sanskrit, whether as a woman or a man, was to write in a uniform, transportable, male, high-caste, and upper-class cosmopolitan language devoid of regional linguistic markers.¹⁶ This cosmopolitanism, identified by Sheldon Pollock, is what enabled Sanskrit to become a language of power for an ‘aesthetic state’.¹⁷ Pollock explains:

Constituted by no imperial state or church and consisting to a large degree in the communicative system itself and its political aesthetic, this order was characterized by a transregional consensus about the presuppositions, nature, and practices of a common culture, as well as a shared set of assumptions about the elements of power—or at least about the ways in which power is reproduced at the level of representation in language.¹⁸

Underlying this cosmopolitan order, however, is an absolute gendering of the Sanskrit voice as male—an issue that that generally falls outside the scope of Pollock's discussion of Sanskrit cosmopolitanism.¹⁹

Sundarī and Kamalā, and the labour of gender

Scholars, confronted with Sundarī and Kamalā's authorial production, have expressed remarkable anxiety over its attribution. Did they or did they not write it? Does it express a female voice or is it an artificial female voice?²⁰ In line with the cosmopolitan expectations of the language, women writing in Sanskrit did not always choose to represent themselves according to their ascribed gender, and even when they did this, they did not necessarily present a discernible female subjectivity.²¹ A gendered subjectivity was often determined by choice of language, not the other way around.

Anxiety over authorial attribution is not unique to Sundarī and Kamalā but routinely crops up in scholarship on other women who wrote in Sanskrit.²² We also see related discussions about women bhakti poets and female voiced-rekhti poetry.²³ Sanskrit scholarship, however, stands out for its overall resistance to gender studies.²⁴ This is further compounded by a lack of historical scholarship that can locate the labour of women authoring Sanskrit texts in relevant social institutions and contexts.

In what follows, I often refer to Sundarī and Kamalā collectively as ‘the co-wives’. Sundarī and Kamalā always refer to themselves in the dual and use the designations ‘co-wives (sapatnī)’ and ‘daughters-in-law (snuṣā)’ in their prefatorial remarks. The commentary is seamlessly co-authored with no distinctions in who wrote what. In line with their own self-designation, I use the term ‘the co-wives’ to signal their collective labour.

At times I also refer to Sundarī and Kamalā as ‘the co-wives’ in order to distance the question of authorial identity, as well as to highlight the significance of their familial relationships as a point of access to Sanskrit learning. Although there are good historical grounds for thinking that Sundarī and Kamalā did, in fact, author their commentary, I reject the drive to make a positivist determination as this closes off more productive avenues of enquiry. Instead, I focus on the labour of gender. There are two kinds of labour: 1) the labour of Sanskrit intellectuals, that is, who produces it, who has access to it, and what institutional structures support this; and 2) the labour that the act of gendering performs in the construction of new regional identities within the Sanskrit sphere; this entails thinking about processes of gendering and how ideas of gender are deployed. That some early modern women had access to Sanskrit or were associated with Sanskrit production does not necessarily indicate that this was a historical trajectory specifically propelled by women, for women. This article also examines the value attributed to gender and how it was leveraged in varying political, social, and cultural arenas.

Redefining regional identity

Ghanaśyāma, Sundarī, and Kamalā were part of a broader trend of early modern Brahmans who had migrated out of the Western Deccan to other Sanskrit intellectual centres. Negotiating their regional identity from the multilingual Maratha-ruled court of Tanjavur, the commentarial trio makes the case that Rājaśekhara, the author of their source text, was Maharashtrian and wrote in a Maharashtrian-inflected Sanskrit and Prakrit. They drew Marathi vocabulary into their Sanskrit commentary and urged flexibility in Sanskrit conventions in order to accommodate expressions of regional identity. In the way they divide their labour—whether real or imagined—the husband and co-wife trio reposition and re-gender these languages.

The tenth-century poet Rājaśekhara wrote from Kannauj under the patronage of Mahendrapāla (circa 885–910) and Mahīpāla (circa 912–940) of the Gurjara-Pratihara dynasty, and then later under Yuvarājadeva I of the Kalachuris of Tripuri/Cedi (circa 910–950).²⁵ Although these regions were all decidedly beyond the pale of what we might call Maharashtra, Rājaśekhara referred to his great-grandfather as ‘the crown jewel of Maharashtra’.²⁶ Ghanaśyāma and his co-wives, living in an eighteenth-century world where regional identity had taken on new connotations, made this a defining element of their commentary.

Expanding on this family-identifier, the trio repeatedly asserted Rājaśekhara's ‘Maharashtrianness’, producing an intra-textual Marathi glossary along the way.²⁷ Sundarī and Kamalā, for example, say, ‘By the phrase “mahaty eva prabhāte (in the early dawn),” the poet indicates his own Maharashtrianness. The reason being, Maharashtrians say, “moṭhyāca pahāṭe (in the early morning).”’²⁸ They also flag Maharashtrian sayings and idioms in the text, attending to references that might not register for readers outside the Marathi socio-linguistic sphere.²⁹ Quoting a line from Rājaśekhara's play—‘“What will I do, friend?” This durdolī of love is difficult break!’³⁰ — Ghanaśyāma explains that the word durdolī does not really mean ‘a tightly tied knot’. Instead, he asserts that it has a specialized meaning among certain caste groups in the Western Deccan. He explains ‘Koṅkaṇa and Deśastha Brahmans use the word “durdolī” to mean “abundance.”’³¹

By the eighteenth century, it seems that Rājaśekhara's Viddhaśālabhañjikā had been abandoned to dusty shelves and practically forgotten.³² Sundarī and Kamalā lament:

Some collected it, others copied it, some never touched it, and others hadn't even heard of it. Teachers are equal to students when it comes to its meaning. Ghanaśyāma, our husband, the one to revive this composition, the Viddhaśāla, descended to the earth in the land of Jambudvīpa (India) at the wish of sincere readers.³³

The text, they argue, had been spurned due to a propensity for local usage in both the Sanskrit and Prakrit portions of the play:

In Maharashtra, because of this commentary, this play of Maharashtrian origin which is but an agitation to people of other origins, will become fun.

Because of the following and more: 1) the beauty but also the trickiness in the arrangement of deśī (non-Sanskrit-derived) Prakrit words 2) the use of sounds that seem like those in a manuscript with scribal mistakes but are enjoyable for learned people from Mahārāṣṭra, our husband, Ghanaśyāma, rescued the poet Rājaśekhara's if-not-already-elevated play just as the best of boars raised up the sunken earth.³⁴

The question of Maharashtrianness in eighteenth-century Tanjavur was conditioned by the unique position of Tanjavur in a broader history of Brahman migration out of the Western Deccan.

Marathi, Maharashtrianness, and Deccan Brahmans on the move

Brahman intellectuals on the move (1500–1650)

Like Ghanaśyāma, Sundarī, and Kamalā, many early modern Sanskrit intellectual households were on the move. Between the fifteenth and seventeenth centuries, a steady trickle of Brahmans migrated out of the Konkan and the Western Deccan, often cycling through various Deccani sultanate courts en route to Banaras.³⁵ By the early sixteenth century it is clear that familial connections between the Western Deccan and Banaras had thickened. Scholars have suggested that these Brahman intellectual households were increasingly attracted to Banaras in part because of new economic opportunities, extended circles of elite patrons, and an expansion in the pilgrimage infrastructure facilitated by the Mughal empire.³⁶ As Rosalind O'Hanlon has suggested, the cultural policies of Akbar promoting regional and religious exchange in the 1570s added further momentum to the intellectual milieu of Banaras.³⁷ By the beginning of the seventeenth century, Banaras had become firmly established as a flourishing centre of ‘new’ Sanskrit intellectual thought.³⁸

These intellectuals in Banaras operated not through a courtly patronage system but via diversified sources of economic subsistence.³⁹ Once in Banaras, many of the Brahmans who had migrated from the Western Deccan distinguished themselves as experts in legal matters, alongside their Sanskrit intellectual and literary activities. At various junctures, over the course of a century—in 1583, 1630/31, 1657, 1664, and 1683—these Brahmans in Banaras brought a translocal clout as they weighed in on disputes about the caste standing of communities in the Western Deccan and the Konkan.⁴⁰ Although these rulings were made through Pandit assemblies that included a good number Brahmans from these regions, overall, the assemblies were composed of regionally diverse assembly members.⁴¹

Among these migrants from the Western Deccan and Konkan, some Brahmans marked themselves as belonging to a community of ‘Maharashtra’ Brahmans, while in the same documents, other members identified themselves as belonging to specific Brahman subcastes of the Western Deccan, while Brahmans from the Konkan often identified themselves separately and as members of a distinct Konkan region.⁴² Certainly, these Brahman families from the Western Deccan and Konkan accumulated material resources along regionally defined caste lines, through marriages, and through active participation in the affairs of relatives and land disputes tied to their family networks there.⁴³ These Sanskrit intellectuals, however, still placed a premium on the translocal power of Sanskrit intellectual lineages.⁴⁴ One of the primary identifiers they used was the broader regional category of ‘Southern’ pandits.⁴⁵

Although a collective identity defined by the Marathi language had long been variously promulgated by the Vārkarī saints as early as the late thirteenth century, and since the sixteenth century among Maratha troops and service elites, Sanskrit intellectuals had not sought to define themselves in such terms until the close of the seventeenth century.⁴⁶ Instead, Sanskrit intellectuals saw themselves as part of a transregional Sanskrit intellectual community, even as their activities were increasingly defined by household ventures, which were inherently regionally bounded through kinship relations. This commitment to translocal prestige, well into the seventeenth century, is poignantly illustrated in a commemorative volume, Kavīndracandrodaya (circa 1650). Written in honour of the Banaras resident and Maharashtrian Brahman Kavīndrācārya, the volume compiled the praise of more than 70 intellectuals from across North India, pointing to an intellectual community that exceeded the boundaries of regional identity.⁴⁷

Marathi in the Sanskrit philological toolbox

In the second half of the seventeenth century, however, we see two parallel trajectories: Śivāji consolidated a Maratha state and Sanskrit intellectuals in Banaras began ruminating on their connections to a Marathi language. These ruminations, however, were not yet a fully fledged articulation of ‘Maharashtrianness’, which we will see in the work of Ghanaśyāma, and Sundarī and Kamalā down south in eighteenth-century Maratha-ruled Tanjavur. Although these two trajectories may not have developed in direct relation to one another, over the course of the late-seventeenth century, points of connection may have further catalysed reflections on regional-linguistic affiliation among Sanskrit intellectuals with family origins in the Deccan.

From the 1650s onwards Śivāji had begun expanding his territory beyond the bounds of his father's jagir.⁴⁸ By 1660 he had firmly carved out a polity which included the Deccan Desh and the Konkan and he had begun recruiting Brahmans into administrative positions in large numbers.⁴⁹ Śivāji's rule also saw a reintroduction of Sanskrit as a prestige language of central administration.⁵⁰ He began using a Sanskrit seal instead of a Persian one and commissioned a lexicon that provided alternative Sanskrit administrative terms to replace the Persian ones.⁵¹ Seeking to shore up his political base, in 1674 Śivāji recruited a prominent Maharashtrian Brahman from Banaras to curate a coronation ceremony that would put to rest debates about his caste identity.⁵² This Maharashtrian Brahman was a well-networked intellectual from a family of Sanskrit powerhouses.⁵³ Whether through connections of family networks in the Western Deccan, an orientation towards new potential patrons, or other tangential reasons, at this time, grammatically oriented Sanskrit intellectuals in Banaras were becoming increasingly preoccupied about the place of the vernacular in Sanskrit discourse.

In the Sanskrit sphere, experimentation with vernacularity in Sanskrit cultural production was not new. Beginning at the turn of the millennium, scholars have identified a growing regional and vernacular self-awareness in literary texts.⁵⁴ The entrance and development of vernacular languages as a philological tool in big debates in Sanskrit was, however, a new phenomenon.⁵⁵

In the Sanskrit intellectual world, grammar and linguistic correctness could be a powerful tool to undo or defend against philosophical detractors.⁵⁶ For example, in early seventeenth-century Banaras, the influential grammarian Bhaṭṭoji Dīkṣita (active circa 1550–1630) levelled an attack against the Mādhva philosophical school. Bhaṭṭoji accused Madhva and his followers of using ungrammatical words (apaśabda). The accusation stuck and the commentators who followed Bhaṭṭoji spent much of their time justifying these ungrammatical words (apaśabda).⁵⁷ Folded into this category of ‘ungrammatical words’ (apaśabda) is the implication that such words are, at best, vernacular and, at worst, foreign (mleccha).⁵⁸

Within a generation the transition from vernacular language as a potential pitfall in philosophical argumentation to productive philological tool had started to shift. By the mid-seventeenth century, for one reason or another, the authorization of the grammatical efficacy of vernacular languages had begun to penetrate Sanskrit discourses.⁵⁹ Around 1650, the grammarian Kauṇḍa Bhaṭṭa,⁶⁰ the nephew of the Bhaṭṭoji Dīkṣita, offered a vigorous defence of the ‘signifying capacity’ of vernaculars, using Marathi as his prime example.⁶¹ If Kauṇḍa Bhaṭṭa was not himself from Maharashtra or a Marathi language-speaker, he must have had in mind interlocuters who could be compelled by such an example.⁶²

Regional identification, which had crept in as a tool in ‘big debates’ on grammar and philosophy, started to emerge more strongly at the same time that the regionally and linguistically defined Maratha state was on the rise.⁶³ Take, for example, Ananta Deva II, whose family had migrated from somewhere near the Godavari to Banaras.⁶⁴ Two generations earlier, his grandfather (fl. 1580) had expressed considerable anxiety about the place of the vernacular in Sanskrit philosophical traditions.⁶⁵ At around the time that Gāgā Bhaṭṭa oversaw Śivāji's coronation, Ananta Deva II, despite being a fully committed Sanskrit intellectual, memorialized his family ties to Eknāth, the bhakti-poet known for championing Marathi over Sanskrit.⁶⁶ The memorialization of this family tie points to the growing role of regional vernacular languages in the toolbox of Sanskrit intellectuals.

It is at this time that we also see the incorporation of vernacular language into Sanskrit commentary as a productive philological tool. Nārāyaṇa Lakṣmīdhara Ārḍe (circa 1650–1727) and Nīlakaṇṭha Caturdhara (fl. 1680–1693),⁶⁷ also migrants from the Western Deccan, both used Marathi glosses in their Sanskrit commentaries.⁶⁸ In his Mahābhārata commentary, Nīlakaṇṭha Caturdhara further offered glosses in Hindavi, Dravidian languages, Persian, Arabic, and possibly Bhojpuri.⁶⁹

This new orientation towards the vernacular, and Marathi in particular, came just as the Sanskrit intellectual centre of Banaras, and the infrastructure that supported it, was collapsing. Rosalind O'Hanlon has suggested that the execution of Dara Shukoh in 1659 precipitated the dismantling of support networks for the Pandit intellectual community in Banaras.⁷⁰ With the death of Aurangzeb in 1707, North India experienced a recalibration of courtly centres. Sanskrit intellectuals were drawn away to emerging regional courts such as Tanjavur.

The integration of vernacular Marathi glosses, seen in the late seventeenth-century commentaries of Maharashtrian Sanskrit intellectuals of Banaras, exploded in the commentarial style of Ghanaśyāma and his co-wives Sundarī and Kamalā down south in Maratha-ruled Tanjavur in the 1730s.⁷¹ These three commentators turned their Marathi glosses and regional expertise into the central focal point of their commentary on Rājaśekhara's Viddhaśālabhañjikā.

Language, lexicons, and Marathi in the broader Deccan

While the Banaras intellectuals from the Western Deccan started to gloss words in Marathi, Ghanaśyāma, Sundarī, and Kamalā incorporated this into a broader vision of commentarial practice that displaced grammatical explanations with lexical concerns. Redefining commentarial praxis, Sundarī and Kamalā argued that there was no need to chew-over-what-has-already-been-chewed by rehashing grammatical principles:

Prior⁷² grammarians have explained every word in multitudes of dictionaries. Thus, the rules concerning the formation of words have a long-winded format that is unbearable to sincere readers. It is chewing-over-what-has-already-been-chewed. Dictionaries and usage are both approved by the learned. As the sayings go: ‘grammarians take refuge in usage,’ and ‘definitely an ācārya (learned teacher) is someone in possession of dictionaries’.⁷³

They define dictionaries as final authoritative sources and cast grammatical explanations as unnecessary and even distasteful. Likewise, Ghanaśyāma silences his objectors by defending his decision to rely on inflected forms in his exposition rather than grammatical derivations:

Rivers of inflected verbs and nouns have found their way into the oceans of verbal glossaries, and dictionaries.⁷⁴ Here, [in the Prāṇapratiṣṭhā] there are glosses from those in plenty. I have explained all of the words there⁷⁵ [in the Viddhaśālabhañjikā].⁷⁶

Such a lexically driven approach is not just stated in the commentarial prefaces but is abundantly clear throughout the Prāṇapratiṣṭhā and the Camatkārataraṅgiṇī. The commentators routinely include Marathi glosses, as well as occasional glosses in Tamil (Drāviḍa) and Kannada, and hash out concerns about undesirable secondary meanings in vernacular languages.⁷⁷

The focus on lexical form over grammar mirrors a broader trend seen in Tanjavur's cultural production. The politics of cultural production in Tanjavur were influenced by the movement of Marathi-speaking Maratha troops and service elites throughout the broader Deccan, including the Karnatak and Coromandel Coast, as well as by the social, cultural, and political practices of the Tanjavur Nayaka court prior to Maratha rule. By the early seventeenth century, Marathi-speaking Maratha troops were being employed in the tens of thousands in regional states whose military affairs saw them settle throughout the broader Deccan.⁷⁸ They brought with them a new language of administration and literary expression. The establishment of a new Maratha-ruled court on the far-flung Coromandel Coast, however, forged a new milieu for such Marathi-speaking elites and Brahman intellectuals alike.

In 1676, after becoming involved in the succession struggles of the Tanjavur Nayaka court, Śivāji's half-brother Ekoji swooped in and installed himself on the throne of Tanjavur.⁷⁹ What followed was tense military confrontation and negotiation over spheres of influence between Śivāji and Ekoji. Though a formal agreement was eventually reached, this new Maratha court in Tanjavur set itself apart in other ways as well.⁸⁰

Looking to their Nayaka predecessors who had refined the practice of bilingual literary production, the Maratha-ruled court of Tanjavur further expanded on this and experimented with new literary forms.⁸¹ These cultural politics extended beyond just literary works, registering, for example, in the production of special Marathi lexicons in Tanjavur. These lexicons included guidance on the gender and conjugation of Marathi words, adhered to earlier Marathi lexical forms, focused on Marathi ‘vocabulary and its corruption’, and offered glosses in Kannada and Tamil.⁸² As Sumit Guha has suggested, they were aimed at a regionally and culturally distanced community of Marathi speakers in a new dynamic polylingual court.⁸³

The labour of gender in vernacularizing Sanskrit

The expansion of new regional polities in the Deccan, and the migration of Maratha military elites and Sanskrit scholar households in early modernity, spurred new ways of articulating regional identity. One of the ways in which this manifested was in new philological tools that repositioned the vernacular in relation to Sanskrit. This is also displayed in diverse cultural forms of production where abstract regional identity was concretized by identifying regionality with the female body.

Such impulses are seen in the new artistic styles of Maratha painters who combined the cosmopolitan Persianate culture of the Deccani courts with expressions of Maharashtrian regionality. New genres of paintings emerged depicting women in their local attire, namely the nine-yard sari, while men were depicted in the standard cosmopolitan Islamicate dress of ‘Hindu’ ruling elites.⁸⁴ These new Maratha paintings also drew on Marathi literary tropes of the mobile warrior juxtaposed to his wives, rooted at home in their regional homeland of the Western Deccan.⁸⁵ The female body, a crucial site of contention in debates about regionally bounded social relationships such as caste and kinship, was positioned as the pre-eminent ‘matter’ in which regionality was signified.⁸⁶ Through identification with the female body, regionality, in turn, was articulated as both a cultural code and a bodily substance.⁸⁷

Across diverse forms of cultural production, abstract regional identity was thus concretized by identifying regionality with the female body. It is unsurprising, then, that Sanskrit intellectual production also formed new associations with female bodies as it reconstituted its own relationship to regional identity. In the case of the husband and co-wife trio discussed here, regionality was articulated through a careful deployment of the co-wives’ gendered bodies in their joint commentarial project.

The commentaries on Rājaśekhara's Viddhaśālabhañjikā were undoubtedly envisioned as a joint project, albeit with divided labour. The co-wives wrote:

May this trio of works (the Śālabhañjikā of Rājaśekhara, the Prāṇapratiṣṭhā of Ghanaśyāma, and the Camatkārataraṅgiṇī of Sundarī and Kamalā), interconnected like the tristhalī cities (of Kashi, Prayag, and Gaya) prosper. If even one were missing, those seeking merit would not accomplish their aim.

If the words of Ghanaśyāma, Sundarī, and Kamalā are attached to the Śālabhañjī, it can circulate among learned people and not hide on a shelf like the statue-of-a-lady (śālabhañjī) carved into a wall.⁸⁸

Alluding to the rising popularity among Maharashtrians of performing the tristhalī-yātrā from the sixteenth century onwards, the co-wives suggest that a complete understanding of, and wide audience for, Rājaśekhara's play (nāṭikā) depended on the circulation of both commentaries together.⁸⁹

The basic form of Sanskrit plays—applicable to almost all genres and sub-genres, including nāṭikā—is in fact multilingual. The dialogue is a mix of Sanskrit and various types of Prakrit, which is to say, Middle Indo-Aryan languages with putatively regional characteristics. These languages essentially stand in for vernacular speech.⁹⁰ On the whole, Prakrit is cast as the language of women, although it is also spoken by lower-status male characters who are arguably feminized through their lack of social power. Sanskrit is the language of men, though it is occasionally spoken by female characters who transgress socially preferred female roles.⁹¹ Whether due to Prakrit's historically smaller share in literary production or a further marginalization of the language in the second millennium, Sanskrit plays are often circulated with a translation (chāyā) of the Prakrit dialogue into Sanskrit.⁹²

Given the highly regulated gender conventions of Sanskrit, the husband and co-wife trio divided their labour in a surprising way along the Sanskrit-Prakrit axis. Ghanaśyāma's commentary, the Prāṇapratiṣṭhā, acts as more of an annotated translation, focusing on Prakrit issues (female voice), with only occasional quick notes on the Sanskrit. By contrast, Sundarī and Kamalā's commentary, the Camatkārataraṅgiṇī, provides extensive comments focused primarily on the Sanskrit dialogue (male voice) and Sanskrit issues. While Ghanaśyāma's concern with Prakrit might be representative of a renewed interest in, and prestige associated with, knowing Prakrit grammar which emerged between the fifteenth and eighteenth centuries, the underlying gender identifications of the languages remained intact.⁹³ This striking division of labour between husband and co-wives, then, might be seen to disrupt the binary nexus of associations: Sanskrit/male versus Prakrit/female. While the relationship between Sanskrit and Prakrit does not exactly replicate the relationship between Sanskrit and the vernacular (that is, Marathi), Prakrit sets a fundamental precedent for imagined intersections between gender and place.⁹⁴

Prakrit occupies a precarious in-betweenness. It is imagined as a pre-eminent cosmopolitan literary language (even if ‘not-quite-so high’ as Sanskrit) and the metonymized female voice; it is also imagined as a low literary register of the common people, which simulates a regionality, namely Mahārāṣṭrī.⁹⁵ In Ghanaśyāma's imagination, Prakrit is intimately associated with a highly vernacular register.⁹⁶ He signals this through his reading of Rājaśekhara's Prakrit, pointing out the many places where Rājaśekhara uses deśī (regional) words.⁹⁷ While the husband and co-wife trio cast both Rājaśekhara's Sanskrit and Prakrit as representative of a regional identity, the reassociation of these languages with gendered bodies through the division of labour between husband and co-wives points to the force of new early modern ideologies of gender and regionality.

The inversion of the literary languages’ encoded genders, combined with a renegotiation of the boundaries between Sanskrit and the vernacular was not an accidental aberration. It mirrors a transaction between gender, the cosmopolitan, and the vernacular that was already taking place in Tanjavur at the court of Raghunātha Nāyaka (r. circa 1597–1626) just prior to the arrival of the Marathas.⁹⁸

At least two women at Raghunātha's court were involved in Sanskrit literary production.⁹⁹ Rāmabhadrāmbā, who was trained in eight languages, composed an important Sanskrit biography of Raghunātha;¹⁰⁰ and Madhuravāṇī translated Raghunātha's Telugu Rāmāyaṇa into Sanskrit. Madhuravāṇī recounts the conditions under which she produced the translation, explaining that Raghunātha had specifically sought out a woman to translate his vernacular Telugu composition into Sanskrit.¹⁰¹ The gendered bodies of Madhuravāṇī, Sundarī, and Kamalā are positioned as mediating the changing relationship between Sanskrit and the vernaculars while also concretizing regional identity. Though Sanskrit was never impervious to social, cultural, and political changes, it would have required significant and novel forces of influence for these changes to register in Sanskrit production.¹⁰²

Gender and the labour of intellectual production

Labour in Sanskrit intellectual households

Brahman articulations of Maharashtrianness were driven by highly localized and historically contingent factors. There were, however, also broader trends emerging throughout the subcontinent that influenced Ghanaśyāma, Sundarī, and Kamalā's household project. The household, for example, was emerging as a crucial site of economic, political, and cultural activity.¹⁰³ In the arena of cultural production, the household facilitated access to the cultural capital of specialized training and family style.¹⁰⁴

The rise of Sanskrit scholar households between the sixteenth and seventeenth centuries has recently been elucidated by scholars tracking the migration of Maharashtrian Brahman households to the ‘hyper-competitive’ city of Banaras.¹⁰⁵ These households sought to consolidate and maximize their scholarly resources through marital alliances, building up manuscript collections, and participation in local assemblies.¹⁰⁶ The absence of centralized state patronage meant that they had to diversity their sources of income at a time when their respective Brahman communities in the Western Deccan were also accruing wealth and ‘nested rights' associated with the rise of early capitalism.¹⁰⁷

Brahman intellectuals in Tanjavur did not appear to follow the same economic trajectory, operating instead through centralized court patronage.¹⁰⁸ The Maratha-ruled court of Tanjavur aggressively endowed tax-free agrahāras (villages) to Brahmans. Nevertheless, the household was an important means of consolidating ‘intellectual capital’.¹⁰⁹ Like their counterparts in Banaras this was also facilitated through marital alliances or quasi-familial intellectual lineages.¹¹⁰ Intellectual capital in turn manifested in material capital. For example, in 1693 King Śāhaji endowed a particularly large agrahāra in Thiruvisanallur (also called Śāhajirājapuram).¹¹¹ The 45 Brahmans who received land grants there received shares according to their intellectual stature.¹¹²

Across the board, it was through the Sanskrit intellectual households that budding intellectuals gained access to the social prestige of intellectual lineages. This ranged from cross-pollination of ideas between family members, to the joint authorship of texts, to the joint appearance of fathers and sons, brothers, and scholar-brothers in debates.¹¹³ As O'Hanlon has pointed out in the case of Banaras households:

As sons matured, they might write alongside their fathers and in some cases complete their works. Gagabhatta completed his father's digest the Dinakaroddyota. Samkarabhatta edited and added to the work of his father Nilakanthabhatta.… Krsna Sesa very likely completed the Govindārṇava of his father Narasimha.¹¹⁴

Such joint production blurred the lines of intellectual ownership.¹¹⁵ This is similarly evident in the commentary of Sundarī and Kamalā. They deploy what we might think of as Ghanaśyāma's signature commentarial remarks: first, the regular citation of regional, particularly Marathi, words (iti mahārāṣṭrāḥ);¹¹⁶ second, flagging dhvani when no dhvani is readily apparent (iti dhvaniḥ);¹¹⁷ third, the problematizing of literary usages (iti kaveḥ pramādaḥ/acāturyam);¹¹⁸ and fourth, the negotiation of linguistic flexibility (‘x’ kṣantavyaḥ(m)/dayanīyaḥ(m)).¹¹⁹ Based on the commonalities of commentarial vocabulary across the works of Ghanaśyāma and his co-wives, scholars have either hastily attributed this to stylistic imitation or else raised doubts about the attribution of the Camatkārataraṅgiṇī to Sundarī and Kamalā.¹²⁰

Anxieties about stylistic similarity are predicated on a Eurocentric model of authorship that privileges the idea of individuality. This historically maligned paradigm fails to take account of the centrality of family enterprises in early modern South Asia. In this context, collective and even composite authorship were forms of cultural capital and prestige across multiple contexts of cultural production. This has been particularly well-theorized by B. N. Goswamy and Eberhard Fischer who have identified the early modern development of a coherent ‘family style’ with stylistic trademarks in family painting workshops.¹²¹ That the co-wives’ commentary would resemble Ghanaśyāma's is neither unusual nor evidence of rote imitation.

Joint authorship and family style

In Ghanaśyāma's family, several other household members also took to the family enterprise. Sundarī and Kamalā, who frequently cite Ghanaśyāma's commentaries, also refer to Ghanaśyāma's elder brother Cidambara, who appears to have been a commentator.¹²² Likewise, Ghanaśyāma's sons, alongside Sundarī and Kamalā, also became commentators. One son, Candraśekhara, wrote a commentary on his father's original composition, the Ḍamaruka. His other son, Govardhana, authored a commentary on the Ghaṭakarparakāvya, and a text called the Rukmiṇīcampū.¹²³ These texts require further study. However, in a small sample of preamble verses from the sons we already see indicators of a shared stylistic vocabulary.

Ghanaśyāma's commentarial style is often quite terse. He, as well as Sundarī and Kamalā, quickly move from lexical notes to authorizing examples culled from literary sources without much explanation. For Ghanaśyāma, the adoption of a concise style is a point of great pride. He asserts:

Moreover, those who act like Raṅgarāja send noble people wandering in the wilderness of poor wording.¹²⁴ O sober-minded people! Since I am a virtual Mallinātha, with concise and essential explanations, this is a blessing to all you wise people.¹²⁵ For idiots, what's even the point of ten million words? A single lamp is enough in the dark for the god Indra, who wields a lightning-bolt. For the blind king Dhṛtarāṣṭra, what's even the point of ten thousand candles?¹²⁶

Drawing from his father's laconic commentarial style, Govardhana carries on the tradition, valuing brevity and efficient productivity:

I praise my own father, Ghanaśyāma, who is partial to the word ‘śeṣa’ (the commentarial expression ‘this is what is left-out’) who knows the śāstras with nothing left-out (aśeṣa). The pain of my blindness became easily bearable through the sight of the scriptures which he bestowed upon me.

Even though I am slow-witted, by the grace of my father, my learning is flawless. I, the poet named Govardhana, just now quickly produced a concise commentary on the Ghaṭakarpara. ¹²⁷

Candraśekhara, on the other hand, dutifully employs the idiosyncratic Marathi maxim of jāmātṛśodham (a son-in-law's wrongheaded revisions) used by his father, warning critics to hold their tongue.¹²⁸ He says:

I bow to the illustrious lord Śiva as Kāmeśvara and the goddess Śakti as Kāmeśvarī,¹²⁹ and to my father who remains only in name, to fulfil my wish that I, Candraśekhara, should compose a commentary on the Ḍamaruka, the work of Ghanaśyāma who holds the title of ‘most excellent of poets’.

Those who don't make a son-in-law's wrongheaded revisions shall appreciate the entire array of verbal roots and so forth, like a palette of mineral-paints, in the gallery of dazzling pictures that is the Ḍamaruka, composed by my venerable father.¹³⁰

The reference to the misguided, critical son-in-law draws directly on the introductory comments of Ghanaśyāma, who states:

I have eschewed chewing-over-what-has-already-been-chewed. Wise scholars, drink up the wine of my words, and then without a son-in-law's wrongheaded revisions, belch silence, the incarnation of Śiva-Dakṣiṇāmūrti.¹³¹

Although the warning that critics should hold their tongues is directed at a wide audience of potential critics, this signature phrase betrays a trace of anxiety about the potential departure of a younger family member from the family style by casting the son-in-law as the exemplary critic. As scholars of Indian painting have pointed out, although families of painters developed stylistic trademarks, new generations made their own revisions, an act that might register as a form of critique.¹³²

The regional, political, and economic factors in the development of ‘family style’ in Indian painting may have been different from that of Sanskrit intellectual production. Nonetheless, the rising significance of the household in early modernity was a far-reaching structural change. The cultural sphere of early modern Indian painting offers a productive point of reference for how the household facilitated access to cultural capital. In painter families, for example, it was common to pass down master sketches, underdrawings, and plans for paintings from which the next generation could work.¹³³ Similarly, Ghanaśyāma's annotated translation (chāyā) provided the commentarial outline upon which Sundarī and Kamalā built.

While the particularities of education for women in these early modern Brahman households have yet to be uncovered, in general family homes were a crucial locus of education. This has been observed by Rosalind O'Hanlon in the context of Maharashtrian Brahman families in Banaras. Although Banaras was full of great sectarian maṭhas, much of the Sanskrit education took place in small ‘private houses’. O'Hanlon has pointed out that these ‘private houses’ in turn facilitated the exchange of sons between families. A son might be sent to another household and then be married into that family. Thus, in effect, they received their training in their wife's family home.¹³⁴

J. B. Chaudhuri has speculated that Ghanaśyāma and his first wife were married as children, and that his second wife, whom he married after the age of 29, would have been substantially younger.¹³⁵ Given the structure of access to Sanskrit training, it is possible that Ghanaśyāma's wives were trained in Sanskrit, either alongside him or by him in the case of the younger second wife. There is certainly evidence of such arrangements in the nineteenth century.¹³⁶ Although the details of the Sanskrit education of women in the eighteenth century have yet to be uncovered, what we can say for sure is that women scholars who were not part of royal households gained access to Sanskrit through intellectual households.¹³⁷

Conclusion: Positioning gender in the royal courts of early modern South India

The emergence of the household as a significant political and economic unit was not the only early modern shift that impacted on the gendering of social and cultural spaces in the broader Deccan, Karnatak, and Coromandel Coast regions. From the Vijayanagara empire until its fragmentation, women were becoming increasingly visible in historical records. The early decades of the Vijayanagara empire saw the first of several royal-affiliated women composing full-length Sanskrit historical and literary works. Gaṅgādevī (active circa 1343–1379), who wrote the dynastic memoir the Madhurāvijaya, was either the wife or concubine of the Vijayanagara prince, Kamparāya.¹³⁸ New visibility and new roles for women seemed to have emerged with greater consistency, however, as Vijayanagara's power waned and new states were formed.

Over the course of the sixteenth century, power was recalibrated in the empire of Vijayanagara. As the state expanded territorially, it became increasingly reliant on military elites (nāyakas) to govern these annexed regions. These military governors could accumulate significant political and economic clout and eventually began establishing new regional states and royal households throughout the Karnatak and Coromandel Coast.¹³⁹ The conditions for such social mobility, particularly the rise of new royal households, were the same ones that made elite women visible and, at times, prominent actors.

Just as the Vijayanagara empire was beginning to fragment, another woman at court, Tirumalāmbā (active circa 1529–1542), distinguished herself in the Sanskrit arena with a royal biography of the Vijayanagara king, Acyutarāya. Though not the chief queen, she was an influential consort.¹⁴⁰ Her public persona is amplified by her association with the founding of a new state. When the low-born military commander Sevappa gained favour at the Vijayanagara court, his marriage to Tirumalāmbā's sister was arranged. As a dowry, he was granted the governorship of Tanjavur where he subsequently established an independent Nayaka court.¹⁴¹ The Nayaka court of Tanjavur, which had particularly uneasy successions, then saw two more women from the newly minted royal household enter the public sphere of Sanskrit literary production: Rāmabhadrāmbā (active circa 1597–1626) and Madhuravāṇī (active circa 1597–1626).¹⁴²

The Vijayanagara successor states, with newly established royal households, saw a relatively large number of royal successions that brought queens to the throne as paramount rulers: in Ikkeri, Cennamāji (r. 1673–1697) and Vīrammāji (r. 1757–1763); in Madurai, Maṅkammāḷ (r. 1691–1707) and Mīnākṣī (1732–1739); and in Maratha-ruled Tanjavur, Sujāna Bāī (1736–1738).¹⁴³ The dynamic politics of these courts were further punctuated by increased political and economic engagement with European actors. Dutch East India Company archival records indicate several instances of trade negotiations being overseen by court women. For almost a decade, the courtier ‘Wengama’ (active 1658–1666) liaised with the Dutch on economic matters as a representative of the Nayaka Tanjavur.¹⁴⁴ At the Nayaka court of Ikkeri, the courtier ‘Maribassuama’ (1681–1684) was a point person in the discussion of trade matters with the Dutch.¹⁴⁵

The fluid social situation through which women found new visibility and agency might also be seen as a moment of social uncertainty. Sometimes this uncertainty crystalized in royal or scholastic authority, as in the case of Ghanaśyāma and his co-wives Sundarī and Kamalā; in other instances, it inspired new projects of social regulation. In Maratha-ruled Tanjavur there was increased investment in new genres of texts treating strīdharma or ‘women's dharma’ (that is, the ideal conduct and social obligations of women). The powerful minister Tryambaka (active circa 1684–1730) systematized and indexed all such concerns in a single standalone treatise on strīdharma that catalogues the minutiae of daily activities prescribed for women.¹⁴⁶ King Śāhaji, under whom Tryambaka served as minister, composed several works on thematics of strīdharma, while the queen mother Dīpāmbā commissioned a Marathi text on the subject entitled Pativratādharma. ¹⁴⁷ Even a popular Marathi text on food and consumption composed under the patronage of Dīpāmbā included what Anand Venkatkrishnan refers to as a ‘legalistic section on women's duties (strīdharma) in a medicalised discussion of the female body and sexual performance’.¹⁴⁸ These strīdharma texts, which project the desire to regulate every detail of women's social activities, their bodies, and their sexuality, are also a record of a social fabric in flux, which made it possible for women to find new avenues of access to Sanskrit and public roles alike.

These social changes came at a time when Sanskrit was no longer the primary language of power or among the primary means for producing cosmopolitan subjects. Although discrete centres of Sanskrit intellectual production existed well into early modernity, this kind of Sanskrit thought usually struggled to circulate widely or elicit comment from a broader community of readers.¹⁴⁹ In many respects the concerns of these texts had become too hyperlocalized to register outside their region of production. Attending to the labour of gender makes these hyperlocal concerns—spurred by migration and changes to social and political structures—more legible and intelligible.

In the case of the household project of Ghanaśyāma, Sundarī, and Kamalā, several new insights emerge: first, the incorporation of vernacularity in the Sanskrit philological toolbox was facilitated through kinship networks even as modes of regional identification shifted according to very local contexts and politics; second, processes of gendering were constitutive forces in the articulation of more abstract concepts such as regional identity; and third, one of the many outcomes of the household's increasing prominence was the cultivation of a ‘family style’ in scholar households; this offered a new rubric for women's access to Sanskrit. The labour of gender is the thread connecting all of these. It is what turns obscure texts like the commentary of Sundarī and Kamalā into valuable early modern archives.