Bengal's goddess tradition: history and development

The tradition of goddess worship in Bengal predates the period which is the direct focus of this paper by many centuries. The forms in which the goddess is best known now, and the attributes most commonly associated with her, particularly in Bengal, appear to have primarily taken shape during the fifth and sixth centuries CE and the fifteenth and sixteenth centuries CE. There is a large body of literature which has studied the evolution of goddess traditions in general as well as for different regions within India, including Bengal.Footnote ⁵ The main stages in the evolution of the concepts, mythologies and worship of the goddess are also well-known. What follows therefore is a brief overview of the various factors and processes associated with it.

It is generally believed that there was an ancient tradition of goddess worship in the Indian sub-continent which predated the Vedic-Brahmanical tradition. This tradition was particularly long-lived and strongly rooted in particular regions within the sub-continent—for example in parts of Eastern India, such as Bengal and Assam. Some of the oldest textual references to the goddess (these occur in the Mahabharata and the Harivamsa) describe her as a virgin, and associate her with forests, wildernesses and craggy, mountainous places. She is, in this conceptualization, a fierce deity who is represented as a hunter and a warrior whose companions are ghosts and wild animals and whose battles are typically against demons. The worshippers of the goddess were supposed to be the people and communities who lived on the margins of settled society. The early stages of the Brahmanical tradition (roughly, pre-third century BCE) also proposed the existence of a great goddess who was linked to impersonal concepts such as Prakriti/materiality, Shakti/energy and Maya/Brahman's ‘creative yet delusive power’.Footnote ⁶ But, a fuller, more comprehensive and systematic notion of a great goddess, such as Devi or Mahadevi, emerged during the fifth and sixth centuries CE as a result of the interweaving of various conceptual and mythological threads.Footnote ⁷

The body of texts that best embodied the crystallization of the goddess tradition were the Puranas, a corpus embodying the ‘the confluence of various streams of thought already present in diverse conceptual and narrative environments’.Footnote ⁸ In Bengal, this produced a composite deity whose attributes and mythologies were derived from the Brahmanical, and various non-Brahmanical traditions including the Tantric and the Buddhist.Footnote ⁹ The development of the Puranic goddess was most likely tied to issues of Brahmanical hegemony. In order to maintain its status and acceptance in and by society at large, the Brahmanical tradition had to incorporate elements from the popular traditions. Kunal Chakravarty's excellent study explores the processes by which the non-Brahmanical goddess was Brahmanized—albeit imperfectly—via the Puranas.Footnote ¹⁰ The goddess’ most important connection to the gods of the Brahmanical-Puranic pantheon occurred through her performance of the vital service of protecting the cosmos against the forces of evil, represented by various demons (asuras).Footnote ¹¹ The most-celebrated text which describes the goddess’ mythological exploits is the Devi-Mahatmya segment of the Markandeya Purana (sixth century CE). This text extols the goddess as the source of all creation; since she has manifold forms, she is also given very many epithets such as Durga, Chandika, Chamundi, Kali, Vaishnavi, Narayani, Maheshwari, Shakti, Gauri, and Mahamaya.Footnote ¹² But by far the most important dimension of the goddess to emerge with clarity in the Devi-Mahatmya is of the goddess as the demon-slayer, or, as David Kinsley puts it, as ‘a great battle queen’.Footnote ¹³ Amongst the demons she kills are Madhu and Kaitabha, Mahisa or Mahisasura, and Shumbha and Nishumbha and their generals Chanda, Munda and Raktavija. Of the three myths that the Devi Mahatmya associates with the goddess, her encounter with Mahisasura is of central importance in her current identity. According to this well-known story, Mahisasura, the buffalo demon, prayed to Brahma, the father of all creation and was granted the boon that he could only die by the hand of a woman. Mahisasura then set off to conquer the universe. He conquered the earth, defeated the gods in a long drawn out battle and expelled the latter from heaven. In a desperate bid to save both heaven and earth, the gods combined their energies, producing thus:

Bedecked with jewels by the gods and armed with replicas of their weapons, the goddess went forth to meet the asura and his army in battle. The demon first assumed the form of a mahisa, and upon the goddess severing the head of this animal, he tried to emerge out of its body.Footnote ¹⁵ Thereupon, the goddess killed him by piercing his chest with a trident. ‘This is the moment’, as Sudeshna Banerjee writes, ‘that is frozen in the images [of the goddess] worshipped every autumn’.Footnote ¹⁶

The name most commonly given to this form of the goddess in Bengal is ‘Durga’, although many other epithets of the goddess are well known—some more than others. The term ‘Durga’ encapsulates a range of meanings amongst which the most basic meaning is ‘hard-to-go-by’, whereas its derivatives include, ‘inaccessible, unassailable, adversity, difficult passage, mountain fortress’. The epic-Puranic literature also provides additional meanings and significations of the term Durga: it denotes one who ends all miseries (‘durgatinashini’), the slayer of a demon named Durgama (meaning ‘difficult to access’) and the protector of a fortress (fortress = durga).Footnote ¹⁷ Durga is also regarded as the goddess of vegetation (shakambari)—an association that survives in particular rituals still performed during the annual Durga puja.Footnote ¹⁸ The deity's function as the giver of food, nourishment and abundance is captured in her forms as Annada, Annapurna and Jagaddhatri.Footnote ¹⁹ Another form of the goddess which is particularly current in Bengal is that of Kali, who is most often described as possessing a fearsome appearance. In the Agni and Garuda Puranas for example, she is mentioned in invocations that aim at success in war and against one's own enemies. She is depicted as a deity who ‘laughs loudly, dances madly, wears a garland of corpses. . .and lives in the cremation ground. She is asked to crush, trample, break and burn the enemy’. In the Devi-Mahatmya, during the goddess’ battle with the demons Shumbha and Nishumbha, the goddess Kali springs from the forehead of the latter, ferocious and battle-ready. In this text, she is, as Kinsley comments, ‘Durga's. . .embodied fury’.Footnote ²⁰

One of the most important developments in the evolution of the goddess tradition, particularly in Bengal was constituted by what Rachel Fell McDermott describes as the ‘Vaishnavization’ of the goddess and the permeation of the goddess tradition with the powerful and potent influence of bhakti or devotionalism. The most important result of the Vaishnavization of the goddess was evident in the much greater emphasis on her attributes as wife, but probably, even more important, as mother and daughter. McDermott's study sketches out the emergence of these phenomena which were centred on the goddesses Kali and Uma (wife of Shiva, also regarded as synonymous with Durga) and were manifest in the body of literature called Shyama Sangeet and Uma Sangeet or Agamani and Vijaya songs and poetry as well as in various rituals performed by women during the annual Durga puja. As the body of Bengali devotional Shakta poetry indicates, an intensification of the process of ‘sweetening’ the goddess (particularly, the goddess as Kali) from a fierce, frightening figure associated with blood and battle, was most apparent in the eighteenth and nineteenth centuries.Footnote ²¹ However, in the case of the goddess as Durga—the main focus of this paper—the process of transformation whereby she retained her demon-slaying image, but at the same time acquired the attributes of a benign, nurturing and protective mother, had begun in Bengal many centuries earlier. The best indicator of this is the practice of worshipping the goddess as a mother surrounded by the four deities who are represented as her children. Worship of the goddess in this particular form may be traced back to the late sixteenth century and probably earlier. Kabikankan Mukunda Chakrabarty's Chandimangalkavya deemed to have been composed around this time, depicts the goddess appearing before the hunter Kalketu in this form:

The family-oriented form of the Mahisamardini goddess is the one that has been worshipped every autumn for the past several centuries in Bengal with great pomp and festivities. As Tithi Bhattacharya points out, this particular iconography of the goddess possessed very important social affects which were related to the domestication of the deity who was conceptualized in extremely intimate familial terms such as mother and daughter.Footnote ²³ The transformation of the goddess from fierce warrior to benign mother and daughter was, as McDermott explains, symptomatic of a process whereby the deity was ‘softened, elevated, humanized and popularized’Footnote ²⁴ via the solvent of Vaishnava devotionalism (a particularly strong force in Bengal since the time of Sri Chaitanya) which offered important resources to the adherents of Bengal's Shakta tradition. The Bengali Mangalkavya genre also functioned as an important catalyst in the process of the ‘softening’ of the goddess. This corpus of narrative and performative poetry which held such sway in Bengal from about the fifteenth century until the early nineteenth century, emphasized the involvement of various gods and goddesses in the lives of human beings and thereby made them more familiar and accessible. As McDermott says, the emotional flavour of the Mangalkavyas do not qualify for description as a fully developed mode of bhakti. But, by absorbing gods and goddesses into the realm of domestic sentiments, they contributed to the domestication of the goddess in Bengal.Footnote ²⁵

In Bengal, an allegedly older practice of worshipping the goddess in the spring (Basanti puja) though rare, still endures. The worship of the goddess in the autumn for several days is a much more widespread practice and has been so for many centuries. The Devi-Mahatmya, so important in the evolving theology of the goddess also refers to the goddess asking her devotees to worship her in the autumn. According to the Bengali Ramayana of Krittivasa (though not the Sanskrit Ramayana attributed to Valmiki), the epic hero Rama, seeking the deity's blessings in his upcoming confrontation with Ravana, had worshipped the goddess in the autumn, thereby creating a precedent for celebrating Durga puja during this season instead of in the spring.Footnote ²⁶ According to Jogesh Chandra Roy Bidyanidhi, the autumnal Durga puja is basically a Sharadotsava—a festival which is celebrated after the season of rains is over and the harvest is brought in. The worship of the goddess was transposed on to a pre-existing festival celebrating the abundance of autumn.Footnote ²⁷

Despite the eminence for many centuries of the motherly yet martial Durga image with ties to the Brahmanical pantheon, vigorous traditions of worship surrounding scores of local or village goddesses as well as goddesses who serve as the presiding deities of specific families have remained current throughout Bengal. These local goddesses represent varying degrees of residual non-Brahmanical and in some cases, even anti-Brahmanical traditions. Known by various names, these deities are associated with a range of attributes and functions of which the most important is the protection of the village, locality and/or the family which worshipped them. However, as Kunal Chakrabarti writes, their functions also extended from the guardianship of a village and a lineage to things such as ‘. . .perhaps a cure for colic or prickly heat’.Footnote ²⁸ In Bankura district, for instance, the village of Chhandor was associated with the goddess known as Jangalasini Devi, Lakhershole village with Kamakhya Devi, Naricha village with the deity Sarvamangala Devi, Ajodhya village with the deity known as Kaluburi and Raipur village with Ambika Devi.Footnote ²⁹ The forms of these local village goddesses also differed considerably from what was to become the increasingly standardized Mahisamardini Durga image by probably the fifteenth and sixteenth centuries (more on this point below). Thus, in Bankura district and elsewhere, the goddess is worshipped in some places in her Kokamukhi or animal-faced form;Footnote ³⁰ Jangalasini Devi of Chandor village (also in Bankura district) is represented as a stone which reposes under a tree, believed to be the sacred site of the goddess.Footnote ³¹ The non-Brahmanical antecedents of many of these devis is underscored by the fact that the hereditary priests/priestesses or diyasis/diyasinis are usually drawn from among the low-status Bagdis, Bauris, Doms, and Majhis who inhabited these regions.Footnote ³² It is difficult to provide a tidy answer about the relationship of various local goddesses with the more ‘universal’ Chandi or Durga whose exploits were described in the Devi Mahatmya.Footnote ³³ In some cases, family deities are offered special worship during the autumnal Durga puja.Footnote ³⁴ Yet in other cases, the ritual worship of the demon-slaying Durga image in autumn is considered taboo in localities where presiding local goddesses are particularly potent. Thus, the worship of deities such as Durga, Kali, Jagaddhatri, and Basanti are forbidden within a certain area surrounding the famous temple of Bargabhima Devi of Tamluk in the Midnapur district. This feature is also noticeable in many parts of the forested South-Western margins of Bengal.Footnote ³⁵ But, despite the sensitive relationship between village goddesses and the more ‘universal’ Durga, as the discussion below shows, the annual autumn festival of the demon-slaying deity became identified as a quintessentially regional tradition in Bengal. One aspect of this regional tradition which does not usually receive the attention it ought to is the association of the goddess Durga or Chandi with political authority. The next section discusses this facet of Bengal's goddess tradition.

The goddess and kingly/political power

Despite the goddess’ predominant image in the past several centuries as ‘. . .a golden-coloured, ten-armed goddess with a gentle expression of countenance. . .’,Footnote ³⁶ the older conceptualization of her as a martial figure endures. Intertwined with this feature was the association of the devi with the acquisition of political/kingly power—particularly in the form of a kingdom and its subsequent preservation and prosperity. This angle, among the myriad myths associated with the goddess, can be traced to the Ramayana (and its commemoration over the centuries via the Dussera festival) as well as to the Markandeya Purana.Footnote ³⁷ In Bengal, connections between kingly/political power and the goddess can be traced back several centuries. Raja Ganesh (1400–1421) who usurped effective power from the sultan of Bengal, as well as his son, who is said to have converted to Islam and ruled as Sultan Jalaluddin Mahumud (1415–1432) proclaimed their association with the goddess by performing ritual worship and by issuing coins bearing symbols associated with this deity.Footnote ³⁸ It is also significant that the only nativist rebellion mounted against the Bengal sultanate was waged in the early fifteenth century in the name of Chandi by Danuja Mardana Deva and his son Mahendra Deva. Thus, as Richard Eaton points out, ambitious rajas who rebelled against the authority of the sultans chose as their inspiration, the goddess who had come to symbolize ‘a deeply rooted, local and nativist Bengali tradition and whose worship was obviously widespread’.Footnote ³⁹

Moving to the period of the late Sultanate and the inception of Mughal rule over Bengal, we find the continuing association of the goddess with a range of aspiring and successful rajas and landed magnates, particularly in the forested South-western tracts of Bengal. The origin accounts of very many rajas who founded kingdoms in these areas attribute their political success to the blessings of the goddess. These landed chieftains continued to profess their allegiance to the goddess thereafter by carefully cultivating an association between the devi and their own families. The examples of the rajas of Tamluk, Karnagarh, Chitua-Barda, Dhalbhum, Brahmanbhum and several other places in the Midnapur district, as well as elsewhere, bear testimony to this. Mukunda Chakrabarty, composer of the best-known Bengali Chandimangalkavya composed his landmark eulogy to the goddess Chandi under the patronage of Raja Bankura Roy who ruled over the Brahmanbhum area of Midnapur and was a devotee of the goddess known locally as Joychandi Thakurani.Footnote ⁴⁰

A great many of these devis who functioned as the presiding deities of families of landed chieftains also functioned as the protectors of fortresses or military strongholds of these rajas. The temple of Joychandi Thakurani mentioned above, was located within the fortress of Aradha Brahmanbhum. The fortress of Abhaygarh, controlled by the rajas of the Bhanjabhum-Baripada area of Midnapur, was named after the deity Abhaya who had been installed within the stronghold.Footnote ⁴¹ Chittaranjan Dasgupta characterizes Mrinmoyee, a presiding deity of the Malla kings of Bishnupur as essentially a fortress-protectress, ‘matrika’ goddess. According to the earliest layer of origin narratives of this line of kings, an ancestor of the Malla rajas was hunting in a deep forest when he encountered the goddess who commanded him to build a stronghold on that spot which came to be known as Bishnupur. Dasgupta points out that there were strongholds all over the Malla kingdom and all of them had protective devi images installed within them.Footnote ⁴² This phenomenon, though particularly evident for the south-Western Rarh region of Bengal was also evident in other areas and for periods of time very much earlier than the examples from the sixteenth and seventeenth centuries given here.Footnote ⁴³ The goddess’ connection with fortresses also linked her to the function of guarding and protecting specific territorial units such as a village or an entire kingdom.

In all these cases, what is striking is the particular association of the goddess with the establishment of forest kingdoms by political adventurers, who were almost always of low-caste or adivasi origin. It underscores yet again, the antecedents of the goddess Durga or Chandi as a forest deity or vanadevi who was worshipped by communities of low-caste, low-status people who inhabited these areas. During the fifteenth and sixteenth centuries, the phenomenon discussed above may well have provided the foundations for the emergence of a long-lived literary trope in the Kalketu Upakhyana, one of the principal story themes of the Chandimangalkavya. The genre of Chandimangala narratives which makes up a large part of Bengal's Mangalkavya tradition and constitutes a form of ‘vernacularization’ (in the sense as used by Sheldon PollockFootnote ⁴⁴) of the Sanskrit Puranic corpus, showcased the ability of the goddess Chandi to elevate humble, but deserving persons to kingship, to provide counsel for good and dharmic governance and finally to protect the king and his kingdom from enemies. The Kalketu upakhyana recounts the story of a poor hunter who, at the blessings of the goddess Chandi, acquired the wealth and the good fortune necessary to carve out a kingdom for himself within the forest. Chandi Devi also gave him detailed instructions about how to clear the forest of Gujarat and to found a kingdom.Footnote ⁴⁵ Chittaranjan Dasgupta remarks that Mukunda Chakrabarty's life, travels and experiences were confined by and large to the forested areas of the western Rarh region and could thus have well mirrored the real political processes occurring there during the later sixteenth and seventeenth centuries.Footnote ⁴⁶

The basic theme of the Kalketu story—the king or his representatives being empowered by the goddess to successfully destroy enemies and re-establish legitimate/dharmic authority—can be found also in later Chandi-oriented Mangalkavyas such as Bharatchandra Roy's famous Annadamangalkavya of the mid-eighteenth century, Gangarama's Marashtapurana of the same period and Raja Pritvi Chandra's Gaurimangala composed in the first decade of the nineteenth century. These later Chandimangalas feature much more high-status and sophisticated protagonists than the simple and rustic hunter Kalketu, and may well reflect the growing dissemination of the Durga/Chandi cult among the political elite and people at the upper levels of society. Sukumar Sen in fact commented on the fact that in his view, Chandimangala narratives slowly attained the status of sabha sahitya or courtly literature patronized by the courts of landed aristocrats.Footnote ⁴⁷ In fact, the impulse of the goddess tradition to draw into its ambit the wealthy and the powerful is evident in both foundation stories of the genre—the Kalketu story and the story of Dhanapati Sadagar and his wife Khullana. It is not that the goddess was indifferent to ordinary, humble people who offered her their worship; but she realized the importance of winning over kings and wealthy merchants as her devotees. It would however not be accurate to exaggerate the elevation of the goddess tradition to more elite levels of society. As the discussion above, on village/local goddesses in particular shows, the salience of the goddess cult remained just as powerful and influential amongst ordinary people.

In any case, with its adherents distributed at all levels of society, Bengal's goddess tradition continued its association with political power well into the eighteenth century. The Maratha Bargis who aspired to military success in Bengal are known to have worshipped the goddess in the autumn. In 1757, Nabakrishna Deb, who had attached his own fortunes to the rising star of the English Company hosted a Durga puja for the first time to celebrate the victory of the English against Nawab Sirajuddaula of Bengal. Lord Robert Clive participated in this thanksgiving to the goddess by personally sending over a number of goats to be sacrificed to the devi. Even today, rituals followed during the annual Durga puja ceremony, particularly at old pujas performed by families of erstwhile rajas and zamindars of Bengal, reinforce the intrinsic connection between kingly power and authority and the regional tradition of the goddess Durga.Footnote ⁴⁸

The Mughals as monsters: Mongol daitya (demon)

Bengal's Mangalkavya corpus provides clues to some perceptions of the Mughals in Bengal via their relationship to Durga or Chandi, the deity who had come to embody the region and its culture. One of the earlier of such depictions occurs in the Chandimangalkavya composed by Dwija Madhava, or the Madhavananda, in the late sixteenth century. In this kavya, the formulaic foundation stories of the Chandimangala genre—the Kalketu upakhyana and the Dhanapati Sadagar-Khullana story—are preceded by an account of the goddess slaying a demon known as Mangal (or Mongol) daitya.Footnote ⁴⁹ There is no asura of this name in the demon-slaying career of the goddess Chandi as given in the Puranas. Mongol daitya was therefore ‘invented’ specifically for this late sixteenth-century narrative and, according to Ashutosh Bhattacharya, appeared in only a few other Chandimangalas composed during the next century.Footnote ⁵⁰ It needs to be pointed out though that one of the most celebrated Mangalkavyas of all time—the composition of Mukunda Chakrabarty—which also belonged to this same period does not contain any reference to this asura known as Mongoldaitya. Nevertheless the appearance of a demon named Mongol in a Chandimangal which was composed around the time the Mughal military machine was engaged in trying to secure control over Bengal is something that merits closer attention and provides clues to perceptions of the advancing Mughal regime in Bengal during this initial phase.

The Chandimangal of Dwija Madhava, described variously by its author as ‘Saradamangal’ or ‘Devir Mahatmya’Footnote ⁵¹ was composed most likely in 1579 CE. This author then was probably a contemporary of Mukunda Chakrabarty. But the image of the goddess that occurs in Dwija Madhava's work, seems more archaic and much more reminiscent of the martial goddess Chandi than the goddess image portrayed by Mukunda Chakrabarty. In the autobiographical atma-parichaya segment which had become formulaic in all Mangalkavyas, Dwija Madhava established the political parameters of his life by explaining that the supreme king who ruled over the area where he lived was the emperor Akbar. In his assessment of Akbar, Dwija Madhava wrote:

Clearly, Dwija Madhava's work was composed at a time when the Mughal armies had begun to enter Bengal and to engage local forces in battles. As is well-known, it was not until the early seventeenth century that Mughal authority was established in any meaningful manner over Bengal, and even then the Eastern and North-Eastern extremities of the region were yet to be completely subdued. However, to a contemporary such as Dwija Madhava, there were sufficient indications that the former political order in Bengal was in transition and that a distant political overlord had secured his rule, nominal or not, over Bengal.

According to the narrative of Dwija Madhava, the demon called Mongol, in the tradition of other Puranic monsters, was an ardent devotee of Shiva. Empowered by a boon he had received from Shiva, Mongol daitya conquered heaven and began a plundering spree there. Like the story in the Devi Mahatmya, here too the gods resorted to the goddess Chandi and begged her to destroy the monster.Footnote ⁵³ A fearsome encounter then took place between the goddess and the monster in which the former finally beheaded the demon with her chakra (disc). On the destruction of the monster, the gods, and other residents of heaven eulogized the goddess. ‘On annihilating Mongol daitya’, wrote Dwija Madhava, ‘the mother [the goddess] acquired the name of Mangalchandi’.Footnote ⁵⁴ In the seventeenth century, Bhavanishankar Das, in his Mangalchandir Panchalika, also attributed the epithet Mangalchandi, held by the goddess to her victory over this dreaded monster.Footnote ⁵⁵ The episode ended with: ‘the devi saying, O gods!, listen to me/remember me, whenever you are in danger’.Footnote ⁵⁶

As the work of Richard Eaton and others reminds us, one needs to be careful in positing a fully developed religious/theological identity denoted by terms such as ‘Muslim’ or ‘Hindu’ in medieval and early modern Bengal. However, vernacular literature provides evidence of a range of terminology derived from cultural practices and ethnic backgrounds that were used as markers to identify and recognize individuals and communities as Muslims. By the fourteenth and fifteenth centuries, and probably well before that, a considerable Muslim population had developed in Bengal. Contemporary literature indicates that different kinds of categories based on occupation, levels of wealth and ethnic origins were in vogue to classify or describe this population. Sanskrit inscriptions from the eighth and ninth centuries till about the twelfth and thirteenth centuries and beyond, used terms such as ‘Turushka’ and ‘Tajika’ to denote Muslims as well as the much more generic term of ‘Yavana’.Footnote ⁵⁷ Bengal's Mangalkavya corpus also records the use of terms such as ‘Turuk’ and ‘Yavana’. Mukunda Chakrabarty's Chandimangala holds up the picture of a large cross-section of ordinary Muslims in Bengal during the late sixteenth century in which occupational and artisanal affiliations formed the basis for sub-groupings within them. The ordering of the elite Muslim society in medieval Bengal however tended to derive from different criteria. But, in general even then, as later, an origin outside India was deemed the hallmark of aristocratic Muslim status. Thus, the elite Muslim society of medieval Bengal was believed to be made up of Sayyids, Mughals, Pathans, Arabs and Persians.Footnote ⁵⁸ Large numbers of such ‘immigrant’ Muslim groups were concentrated in and around centres of political-administrative power and port cities. The Manasa Vijaya of Bipradas Pipilai (1479 CE), refers to different types of Muslims visible in the port city of Saptagrama in the following terms:

Thus ‘Mongols’ were identified as a distinct ethnic group of non-Indian origin among the mix of ‘foreign’ elite Muslims in Bengal a full century or so prior to the Mughal conquest of Bengal. When Saptagrama was supplanted by the rise of Hugli as a seaport during the sixteenth and seventeenth centuries, Mughal merchants continued to comprise an important group amongst Persians, Armenian and other ‘foreign’ merchants concentrated there.Footnote ⁶⁰ The term ‘Mughal’, was a Persianized and Indianized version of the word Mongol and referred to first, the ruling Mughal dynasty of India as well as to ‘people of the Central Asian regions’, broadly defined.Footnote ⁶¹ Abdul Karim, and M. R. Tarafdar—both well-known scholars of the history and culture of Eastern Bengal/Bangladesh in the medieval and early modern periods—suggest that the term ‘Mangal’ or ‘Mongol’ was used to refer to the Mughals. Abdul Karim, however, held the view that this term gained currency from the sixteenth century following the Mughal conquest of Northern India and subsequently of Bengal. Tarafdar, on the other hand, argued that it was valid to hold that this term was deployed in a Bengali vernacular narrative such as Bipradas Pipilai's Manasamangala to denote the ‘Mughals’—not as the rulers of an empire, but rather as a people of Central Asian origin—well before the Mughal conquest of Bengal.Footnote ⁶² I find Tarafdar's argument in this regard to be much more persuasive given the fact that post sixteenth century Bengali narratives generally also used the term ‘Mughal’ to denote the political overlords of Bengal and large parts of the Indian sub-continent. This suggests that the usage ‘Mongol’, as used by the late fifteenth century poet Bipradasa Pipilai was most likely of pre-sixteenth century vintage and slowly ceded ground to the much more ubiquitous term ‘Mughal’.Footnote ⁶³ Given a longer pre-history of the use of the term ‘Mongol’ in Bengal, it is not difficult to suggest a connection between this term and its application to the imperial Mughal army as it entered Bengal as a conquering and occupying force in the late sixteenth century. The violence and terror which inevitably accompanied a military operation may well have provided the basis for the imagination of a monster named Mongol whom the region's protectress and presiding deity Chandi or Durga had to vanquish in order to provide relief to its inhabitants. The use of puns and the deployment of similar sounding words with two meanings (slesha) was a known feature of Sanskrit literature and its influence can be seen in some instances in the Mangalkavya corpus. This literary strategy may well have reached its peak in the famous exchange between the goddess (Ishwari) and the boatman (patni) in the mid eighteenth century Mangalkavya of Bharatchandra Roy.Footnote ⁶⁴ Against this literary context, one could suggest that the use of the term Mangal/Mongol in Dwija Madhava's late sixteenth century narrative needs to be located vis-à-vis this convention. The deliberate play on the word Mangal/Mongol allowed the sixteenth century composer to first situate his narrative within the Mangalkavya corpus, secondly, it enabled him to proffer an explanation for the epithet Mangalchandi used for the goddess, and finally, this made it possible for the poet to identify the Mughals/Mongols with a monster of the same name. Sukumar Sen, the doyen among literary scholars of Bengal noticed this feature and remarked: ‘It seems to me that the excitement/anxiety (uttejana) generated by the might (pratapa) of the Mughal badshahs. . .lay behind the imagination of Mongol daitya’.Footnote ⁶⁵

As noted above, Dwija Madhava offered profuse praise for the Mughal emperor Akbar and compared the latter to conventional mythological heroes such as Rama and Arjuna. However, there is little to suggest that Dwija Madhava had had any direct contact with the person of the Mughal emperor or, even with immediate representatives of the emperor. Such an encomium probably represents the convention of heaping praise on the monarch—particularly when he was physically far removed and not directly associated with acts of oppression and violence. Similar praises for the distant Mughal emperors are found in other works within the Mangalkavya genre and these are not without significance. Those belonging to the seventeenth century and later may indicate a growing acceptance of Mughal rule and the characterization of it in far from negative terms.Footnote ⁶⁶ However, Dwija Madhava's work composed during the initial phase of Mughal military penetration contains both a strongly positive image of the emperor Akbar, as well as a seriously negative image of a demon also called Mongol daitya. This seeming contradiction can be resolved by acknowledging that whilst this poet was prepared to shower praise on a distant monarch who was assumed to be just and compassionate, the immediate, contemporary reality in Bengal comprised a period of almost continuous military campaigns by the Mughal forces who were depicted in his narrative as a ferocious, evil and destructive monster who had to be vanquished by the goddess Chandi.

The conventional scholarly view of the Mughal conquest of Bengal does in fact emphasize its oppressive, violent and prolonged nature.Footnote ⁶⁷ Secondly, the general impression seems to be that Bengal's landed aristocracy, best typified by the celebrated barabhuiyans made it virtually impossible for the Mughals to really consolidate their power over this region. As discussed elsewhere, this scenario, repeated for decades and usually unexamined, needs to be subjected to much more careful scrutiny and modification. The Mughal military penetration of Bengal was indeed a prolonged affair; however, the much more chronic insurrection of the rajas and bhuiyans of Eastern Bengal was not necessarily paralleled by similar resistance in other areas of Bengal. Secondly, although the late sixteenth century—the period of Dwija Madhava's composition—was indeed one of more or less relentless battles waged by the Mughals against local rebels, Afghans and other anti-Mughal forces, by the early seventeenth century, the Mughal policy of diplomacy and selective clemency vis-à-vis many of the bhuiyans yielded much more useful results. Many of Bengal's landed chiefs accepted Mughal sovereignty, became their collaborators and provided assistance to the Mughal military drive into northern Bengal and into Assam. Besides, many new zamindars were created in Bengal by the incoming Mughal regime and a Persianized, courtly culture, associated with the Mughals was regarded as attractive and imitation-worthy by the aristocracy and gentry of this newly acquired Mughal subah.Footnote ⁶⁸ There is thus a scholarly need to modify the conventional, received picture about the Mughal conquest of Bengal.

Nevertheless, that this enterprise involved the use of considerable violence and turmoil is undoubted. For one thing, the Mughal regime was particularly motivated to intimidate the inhabitants of the area by projecting an image of overwhelming force and might. Mirza Nathan, a military officer who was involved in the imperial campaigns in Bengal during the seventeenth century wrote in the following terms about the effects of the deployment of Mughal artillery in Bihar and Bengal:

As Richard Eaton has shown, the use of artillery may not have been a critical advantage enjoyed by the Mughal imperial forces in Bengal.Footnote ⁷⁰ Indeed, by the late sixteenth century, there had occurred a not inconsiderable diffusion of cannon and other gun-powder weapons throughout Bengal. By the mid-sixteenth century—the period of the rule of Afghan sultans in Bengal—cannons were being manufactured in the Gauda-Malda area of Northern Bengal. Rising territorial barons, such as Isa Khan of Bhati, also possessed cannons which they may have seized from the Bengal sultans or from an external army sent from Delhi, or, they may have been locally manufactured for them.Footnote ⁷¹ Local chiefs such as Raja Prapaditya of Jessore also secured cannons and other gunpowder weapons from the Portuguese and used Portuguese ‘renegades’ as gunners in their armies.Footnote ⁷² In some parts of Bengal, exposure to Mughal military tactics and the later establishment of alliances with the Mughals, enabled landed chiefs to effect improvements in the quality of artillery manufactured within their kingdoms.Footnote ⁷³ But, as Mirza Nathan's account shows, even if allowance is made for a degree of exaggeration, the imperial Mughal forces periodically fired cannons—often purely for demonstration effect—in order to frighten and terrify ordinary people. There are other references in the Baharistan-i-Ghaibi to artillery pieces being moved from Rohtas to Bengal in preparation for military encounters against insurgents in the latter area.Footnote ⁷⁴ There are many other examples of the sufferings endured by ordinary people due to the activities of Mughal armies. Murad, the brother of Mirza Nathan for example, also campaigned in Bengal on behalf of the Mughals.Footnote ⁷⁵ He once brought back as captives, a large group of women who had been stripped of their clothes. Thus, whereas the older picture of Mughal efforts to establish itself in Bengal needs to be altered, we need to acknowledge that it was neither painless nor a pleasant experience for scores of ordinary people as well as the better-off landed magnates. As the detailed research of R. D. Banerji shows, the forested South-Western borderlands between Bengal and Orissa, became killing fields in encounters between the Mughals and their local allies on the one hand, and the Afghans, entrenched in Orissa, on the other.Footnote ⁷⁶ When the dust had settled after these battles, the memory of this phase of warfare was commemorated in these areas in a board game (which is still played), called ‘Mughal-Pathan khela’ in which two contending sides in the game took on the identities of Mughals and Pathans/Afghans respectively.Footnote ⁷⁷ Dwija Madhava's narrative composed during the first phase of Mughal military efforts in Bengal may thus have reflected the fear, panic and dislocation produced by this period of warfare. The evil entity of the Mughal/Mongol army was represented in Bengal's regional idiom—one that was particularly evocative within the goddess tradition—as a monster.Footnote ⁷⁸

It is difficult to make a definite claim but it may also be possible that there emerged a more direct connection between the warrior goddess Chandi who protected her devotees and the Mughal forces operating in Eastern India, particularly in the borderlands of Bengal's Rarh region and Orissa. R. D. Banerji, following Blochmann, associates the place called Tukaroi—the site of one of the best-known battles between the Mughal army and Daud Khan Karrani in 1575 CE—with Mughalmari, located on the road from Midnapur to Jalesar in Orissa. The same area, described by Banerji as ‘Tukaroi or Mughalmari’ was also witness to another battle in 1592 between Raja Man Singh and the Afghans.Footnote ⁷⁹ The name ‘Mughalmari’ may have arisen from the occurrence there of battles involving the Mughals as well as a great deal of slaughter and killing. Chittaranjan Dasgupta refers to several villages and fields also named ‘Mughalmari’/‘Mughalmarir Maath’ in the Midnapur district and in the border areas of Bankura and Burdwan districts. Even more intriguingly, Dasgupta maps out the location of Chandi shrines at or near the places and fields called Mughalmari.Footnote ⁸⁰ This could point to the invocation of these deities by villagers for protection against the Mughals. Dasgupta does not doubt the Mughal connection with such place names, but believes that the term ‘Mughalmari’ referred to areas through which the Mughals travelled and their comings and goings stimulated the development of settlement clusters named after them.Footnote ⁸¹

The goddess and the Mughals

Bengal's literary tradition also indicates that the image of the Mughals discussed above may have undergone modifications due to changing circumstances. By the seventeenth and eighteenth centuries, Mughal rule was neither new nor untried in Bengal and there had emerged a segment of gentry and aristocracy who had benefitted considerably from Mughal rule. The goddess motif, symbolizing Bengal's regional traditions was deployed to depict the Mughal relationship with the region somewhat differently than the figure of the asura developed during the late sixteenth and early seventeenth centuries. This point is explored here with reference to two well-known Mangalkabya narratives from the mid eighteenth century—the Annadamagala and the Maharashtapurana.Footnote ⁸² These dense, multi-layered texts yield rich insights into different facets of Bengal's political and cultural history.Footnote ⁸³ Space constraints do not permit a discussion of these here; the following section concentrates therefore on the use in these narratives of the figure of the goddess vis-à-vis the Mughals.

The Annadamangalkavya (1752) composed by Bharatchandra Roy, the court poet of the Raja of Nadia, for instance, represents the Mughals and their agents as initially misguided perhaps, but ultimately embracing the worship of the goddess.Footnote ⁸⁴ The principal human protagonist in this narrative is Bhavananda Majumdar, the founder of the fortunes of the Nadia rajas and an ancestor of Raja Krishnachandra, at whose command, this kavya had been written. One of Bhavananda's principal earthly missions was to propagate the worship of the goddess who had singled him out especially for her blessings.Footnote ⁸⁵ An important opportunity in this direction occurred when according to the Annadamangala, the Mughal mansabdar, Raja Man Singh, was sent to Bengal to suppress Raja Pratapaditya of Jessore who had displayed blatant defiance towards Mughal authority. The goddess decided that she wanted Man Singh to become her devotee. To this end, and at her machination, the raja and his soldiers faced seriously inclement weather and an acute shortage of food. Due to the goddess’ favour, Bhavananda, whose storehouse was overflowing with abundant provisions was able to come to the rescue of the Mughal forces by offering these to the latter. Convinced that without the goddess’ blessings, Bhavananda would not have succeeded in alleviating the crisis for the Mughal forces, Raja Man Singh too, became one of Annada's devotees and proceeded to offer her ceremonial worship. Fortified by the goddess’ protective power, Man Singh and his ally Bhavananda Majumdar defeated Raja Pratapaditya in battle. The real reason for Pratapaditya's defeat, according to this kavya, was the fact that the devi Abhaya-Jessoreshwari, the presiding deity of the former's kingdom had withdrawn her protection of him on account of his unethical deeds.Footnote ⁸⁶ Man Singh invited Bhavananda to accompany him back to the imperial capital at Delhi and the latter accepted the invitation.

At the Mughal darbar, Man Singh, informed the emperor Jahangir of the crucial services rendered by Bhavananda to the Mughal army in Bengal—mainly due to the blessing of the goddess—and requested the emperor to offer him a handsome reward. The Mughal padshah however did not react well to this request. He proceeded to belittle the power of the goddess and described her as a ‘ghost’. This produced a protracted argument with Bhavananda Majumdar and culminated in the emperor ordering the former to be thrown into prison. The goddess, angered at the emperor's denunciation of herself and his ill-treatment of her protégé, Bhavananda, proceeded to take terrible revenge on the city of Delhi. Accompanied by her entourage of spirits and fearsome creatures (dakini and yogini), she began a ferocious assault (utpaat) on the city: she and her hordes terrorized the citizens and created a serious shortage of food there. Large numbers of people died of starvation. The emperor, confronted by a crisis in his own capital city, ordered Bhavananda to be released from prison and offered the latter an apology. He issued a farman (order) bestowing the title of raja on Bhavananda as well as the right to wear honorific robes (khelat), to fly flags and to have kettle drums played.Footnote ⁸⁷ He also acknowledged the greatness of the devi and agreed to worship her. The goddess, convinced of the Mughal emperor's genuine repentance, revealed herself to him in her full divine splendour. Jahangir, now converted from his former views, ordered that all citizens of Delhi should worship the goddess in their homes. The grand finale of this segment of the poem comprised a scene in which the Mughal darbar became the venue for Annada/Durga puja:

The goddess, symbolic of Bengal's regional tradition thus succeeded in converting the Mughal emperor, his courtiers and all the citizens of Delhi, into her devotees and worshippers.

The Maharashtapurana also composed in the mid-eighteenth century by a certain Gangarama depicts an interesting relationship between the goddess, in her role as protector on the one hand and the Nawab of Bengal on the other.Footnote ⁸⁹ Here, the recurring raids of the Maratha Bargis which convulsed Bengal during the mid eighteenth century provide the central motif of this narrative. According to this text, the goddess had instructed the Bargis to invade Bengal as a punishment for the sinful and unethical activities being committed there, both by its ruler as well as by ordinary people. The nawab referred to in this poetic work was Nawab Ali Vardi Khan (1740–1756 CE) and his misdeeds included political disloyalty, breach of political obligations and the anti-dharmic act of attacking the city of Bhuvaneshwar, recognized to be the abode of Shiva. However, the indiscriminate plunder and brutality practiced by the Bargis, and their total disregard for ethical scruples, caused the goddess to withdraw her sponsorship of them and instead to transfer it to the Muslim nawab of Bengal. Armed with the blessing of the goddess, the nawab of Bengal succeeded in killing the Bargi general Bhaskar Pandit and saved his kingdom and subjects from much suffering. As Gangarama, wrote:

The representation of the relationship between Bengal's goddess and the Mughals—whether the emperor, or his agent in Bengal—in these two Mangalkavyas is significant. In Bharatchandra's work, the victory of the goddess in first transforming Raja Man Singh and then, the emperor Jahangir, his courtiers and the residents of Delhi into her worshippers is significant in many ways. First, it successfully upholds the ultimate power and greatness of the goddess Annada/Durga vis-à-vis the most powerful person within the Mughal empire—the emperor himself. In fact, Annada's interest in including within her circle of devotees increasingly powerful and high-status human beings—first, Raja Man Singh and then, the Mughal emperor and his court—underscores the point made above that the cult of the goddess slowly rose from the lower levels of society to become a tradition adhered to by the political and social elite (the historical, as opposed to the merely literary association of Raja Man Singh with goddess worship is discussed further below). Secondly, the depiction of Bhavananda as a recipient of an imperial farman in the Annadamangala points to the pride taken by Bengal's landed aristocracy in tracing their status and privileges to direct grants from the Mughals, the paramount political overlords of much of the sub-continent. However, by far the most important function performed by Bharatchandra Roy's magnum opus, may have been implicitly to make the Mughal regime much more acceptable to Bengal's society through depicting them as capitulating to the great goddess associated with this region. Thus, implicitly, there may have existed also the suggestion that the Mughal administration was not an alien regime which imposed its ways and customs over Bengal—this regime was also willing to respect the much revered and very popular goddess tradition of this region. This kavya may have been attempting also to make a case for the cultural investments of the Mughal empire in Bengal, a point which relates to the much broader question of the nature of the Mughal polity as well as the role of cultural factors in the consolidation of Mughal power.Footnote ⁹¹

The Maharashtapurana not only portrayed a positive relationship between the rulers of the Mughal successor state in Bengal and the goddess, but also posited stronger associations of both the Nawab and the goddess with Bengal in comparison to the Annadamangala. The nawab's identification with the land he ruled is manifest first, when both he as well as his subjects stray away occasionally from the dharmic path and secondly, in his fierce defense of his kingdom and his subjects vis-à-vis the external invaders—the Bargis. There is not a single reference in the Maharashtapurana to the nawab either worshipping the goddess or calling on her for help. Yet, in the end, the goddess chose to support the nawab because he upheld the ethical values expected of kings—protection of their subjects from oppression. Thus, she too overlooked the nawab's earlier moral lapses and chose to end the suffering of both the land and the people of Bengal. There is a strand of scholarship which has been devoted to emphasizing the oppressive, unpopular and exploitative nature of the Mughal connection with Bengal to the virtual disregard for other aspects of this phenomenon. The representation of the goddess electing to bless the Muslim nawab is extremely significant. It underlines the strong possibility, borne out by other aspects of Bengals’ social and cultural history during the early modern period, that to a segment of Bengal's population, the nawabs of Murshidabad were not alien oppressors, but rather, their immediate ruler and someone who was doing his best to expel the external invaders. Yet, this important development depicted in the Maharashtapurana has been overlooked and its implications disregarded by those interested in painting a dark, oppressive picture of Mughal and Nawabi rule over Bengal. Also, this narrative, whilst it acknowledged the theoretical relationship between the Mughal overlord at Delhi and the Nawab of Murshidabad, did not describe the Nawab of Bengal as a ‘Mughal’. The term ‘Mughal’ was applied mainly to the distant overlord in Delhi who had actually acquiesced in the despatch of the marauding Bargis to Bengal. Here, the regional ruler, the Nawab of Murshidabad is identified much more closely with the fate of Bengal and its people.

Mughal rule, the public performance of religion and Bengal's sharadiya Durga puja

There exists a strong consensus in scholarly literature as well as in other types of popular traditions that allegedly the ‘first’ Durga puja was celebrated in Bengal during the late sixteenth and early seventeenth centuries. Three candidates are usually named as the hosts of this alleged ‘first’ Durga puja: Raja Kangshanarayan of Tahirpur in Northern Bengal (currently in Bangladesh), Bhavananda Majumdar, the Raja of Nadia in Western Bengal (mentioned above in connection with the discussion on the Annadamangalkavya) and Lakshmikanta Majumdar of the famous Sabarna Chowdhury family who controlled large parts of the area around what later grew into the metropolis of Calcutta.Footnote ⁹²

Interestingly enough each of the three alleged hosts of the much publicized ‘first’ puja, were closely allied to the Mughal regime in Bengal. There are other striking similarities in the nature of their relationships with the Mughal regime. All three individuals were military entrepreneurs who were also equipped with managerial, clerical, and bureaucratic skills. The genealogies of all three individuals aver that their protagonists offered their collaboration and military services to the incoming Mughal regime in Bengal during the late sixteenth century and were rewarded in exchange with the grant of revenue-collecting rights and titles such as ‘Mazumdar’, ‘Roy’ and ‘Roy Chowdhury’.Footnote ⁹³ Each of these individuals had existed in conditions which were fairly obscure or, at any rate, unremarkable, prior to their alliances with the Mughals. Each of them used their newly acquired political and material power via the Mughals to establish themselves as leaders and arbiters of Brahmanical samajs in the immediate areas which they controlled. All three of these resourceful adventurers patronized Brahmin scholars, and Brahmanical institutions and established their credentials as adherents of various Brahmanical deities, of whom the goddess, in her various manifestations (Kali, Jagaddhatri, etc.) was certainly one of the most important. Raja Krishnachandra Roy of Nadia, a descendant of Bhavananda Majumdar was particularly notable for his energy and zeal in establishing goddess temples and popularizing the worship of different forms of the goddess (see Appendix I below). These three individuals, Lakshmikanta Majumdar, Bhavananda Mazumdar and Kangshanarayan, Raja of Tahirpur, are also given the honour of having staged the so-called very ‘first’ Durga puja in Bengal. By some accounts, The Raja of Tahirpur performed his annual worship of the goddess Durga around 1580 CE, Bhavananda Majumdar in 1606 and Laksmikanta Majumdar in 1610.Footnote ⁹⁴ Thus, for these three rajas of the late sixteenth and early seventeenth century Bengal, the performance of Durga puja in the autumn represented the demonstration of wealth, political clout, social prestige and cultural authority.

The tradition of Durga Puja has very old roots in Bengal. Sculpture and other forms of art from the Pala-Sena periods of Bengal's history depict images of the goddess in her Mahisamardini form. The Ramacharitam of Sandhyakara Nandy (twelfth century CE) refers to the celebration of the worship of the goddess Uma in the Barendri region of Bengal;Footnote ⁹⁵ Bhavadeva Bhatta, probably a high official of a king named Harivarmadeva of the eleventh century also referred to the worship of Durga, as do the Chandimangalkavyas of the later sixteenth century. A significant body of Smriti literature produced in Bengal and its adjoining areas, for example, Mithila, from the eleventh to sixteenth centuries and thereafter, codified the proper rituals to be used to perform the worship of Durga or Chandi. More than anything else, the smriti nibandhas indicate that the worship of the goddess Durga for several days in the autumn may have been crystallizing into a definite tradition and thus created the need for a body of literature which could clearly and systematically delineate the rituals needed for this performance.Footnote ⁹⁶ However, the strength, homogeneity and persistence of the tradition which identifies either Raja Kangshanarayan of Tahirpur, Lakshmikanta Majumdar of the Sabarna Chowdhury family, or Bhavananda Majumdar of Nadia as the host of the very ‘first’ Durga Puja, is difficult to dismiss without due consideration. The position adopted here is that either one of these rajas, or perhaps all of them, had performed the ‘first’ puja, not in its literal sense but rather to indicate that the late sixteenth and early seventeenth centuries represented some kind of landmark ‘moment’ in the evolution of this ceremony in Bengal. Secondly, it is suggested that the pujas performed by all three potentates had showcased some of the ‘new’ features and forms which gained much greater currency from this time onwards. They were thus early examples—though perhaps not literally the very ‘first’—of some features that have now come to be regarded as indispensable characteristics of the sharadiya event.

Amongst these features are the use of large-scale clay images of the goddess, together with her family which were commissioned especially for the several days long puja in the autumn. Many of the smriti treatises of Bengal recommended the use of clay images, or paintings (pat) or a ghat (water pot) for the worship of the goddess.Footnote ⁹⁷ The worship of the goddess represented by a water pot or a painted scroll (pat) still endures.Footnote ⁹⁸ Reference to the use of clay images of the goddess in smriti literature indicates that this practice was probably known prior to the late fifteenth and early sixteenth centuries. However, it may have acquired much greater currency, and ultimately have become a characteristic feature of the autumnal Durga Puja from about this time. Existing scholarship on the history and evolution of Durga Puja, as well as the myriad popular memories and traditions about it, are in agreement on this point: whoever was the host of one of the fabled pujas during the late sixteenth and the early seventeenth centuries (whether Kangshanarayan, Lakshmikanta Majumdar or Bhavananda Majumdar—there is no clear consensus about this) had used a large scale clay image constructed especially for the several days long sharadiya puja.Footnote ⁹⁹ As is well known, the manufacture of clay images, dolls and other artefacts is a particularly well-developed artisan craft in certain parts of Bengal, notably, Kumortuli in Calcutta, and Krishnanagar, the principal town of the Nadia zamindari. It is difficult to ascertain chronological accuracy on the basis of memories and popular traditions held by communities of clay artisans. However, it is interesting that Kumortuli and Krishnanagar artisans trace the history of their craft back to the making of large-sized Durga images on order from rajas and zamindars in the eighteenth and early nineteenth centuries. Some groups of Krishnanagar artisans believe that they were taken to North Bengal by the Rajas of Tahirpur and asked to create a Durga image for Raja Kangshanarayan's Durga image.Footnote ¹⁰⁰ It is also commonly believed that the practice of worshipping the goddess together with her family and, moreover, positioning them within the same semi-circular ‘chaal’ or backdrop, either originated or gained popularity from about the late sixteenth century. The reference in the first section of this paper to the goddess revealing herself en famille to Kalketu in the Mangalkavya of Mukunda Chakrabarty, points to this. The smirti nibandhas referred to above had not contained any mention of the goddess being surrounded by her family.Footnote ¹⁰¹ Thus the coincidence between the literary allusion and popular notions handed down over the centuries point to the strong possibility that the practice of using especially made clay images of the goddess Durga and her family for the annual puja may have become more widespread from about this time.

These ‘old’ pujas, associated with aristocratic families who rose to power during the early phase of Mughal rule in Bengal, also exhibit some other features not commonly seen in present Durga pujas as well as in the recent past. The goddess image worshipped by the Nadia rajas, for example, is dressed in trousers and warrior attire replete with armour and a bejewelled breast-plate. This is in sharp contrast to the extremely feminine, motherly appearance of typical Durga images. The Nadia raj Durga image is said to be modelled on the Durga image used by the rajas of Tahirpur whose celebration of the autumnal puja had taken place a little before theirs.Footnote ¹⁰² Clearly, a trouser-clad warrior goddess Durga indicates a stronger emphasis on the older association of the deity as a martial heroine. The Tahirpur, Nadia and the Sabarna Chowdhurys’ images of Durga also depict as her vahana (mount) not a lion, as is typical now, but instead a mythical animal described as ‘ghora-simha’—a combination of a horse and a lion.Footnote ¹⁰³

The fame of Raja Kangshanarayan's Durga puja, was also due in part to the lavish expenditure associated with it. Kangshanarayan is said to have spent rupees nine lakhs on his legendary puja. The Durga puja of the Nadia rajas too is associated with much pomp, extravagance and splendour. In the time of Raja Krishnachandra, their Durga image would be decked out in precious jewellery made of pearls and diamonds.Footnote ¹⁰⁴ The performance of Durga puja, which is several days long, involves elaborate rituals and is also associated with large-scale feeding of people for several days, the staging of popular entertainments and other festivities by the host inevitably entailed significant expenditure and could therefore be celebrated in this manner only by the wealthy. Much has been written about the unimaginable—bordering on the vulgar—ostentatiousness of Durga pujas held by the ‘new’ rich of Calcutta, mainly a comprador class of entrepreneurs who had become significantly enriched via their collaboration with the English East India Company during the nineteenth century. Kaliprasanna Simha, author of the famous lampoon on Calcutta's babu culture entitled Hutom Pechar Naksha, however, pointed out that the wealthy babus of late eighteenth and early nineteenth century Calcutta, were essentially modelling their Durga pujas—particularly in terms of ostentation and pomp—on the precedent set by the Maharaja Krishnachandra Roy of Nadia.Footnote ¹⁰⁵ Thus, the role of Bengal's aristocracy as hosts of elaborate and expensive Durga pujas certainly predated the rise of Calcutta and the concentration of a segment of colonial compradors there. The traditions that pinpoint the late sixteenth and early seventeenth centuries as a sort of landmark in the evolution of Bengal's Durga puja, also suggest that this period may have witnessed the greater involvement of rajas and zamindars as hosts of these ceremonies. Indeed, the growing association of goddess worship in Bengal with wealthy and high-status people suggests that this particular feature predated Mughal rule. This phenomenon may have been further strengthened during the early phase of Mughal rule and then continued throughout the late seventeenth and eighteenth centuries to become an established feature. Bimala Chandra Datta's extensive survey of Durga pujas all over West Bengal is admittedly general and occasionally impressionistic. Nevertheless, it is a remarkable coincidence that a significant number of families with aristocratic and affluent antecedents—particularly in rural West Bengal—anecdotally trace the origins of their Durga pujas back by about 300–250 years. There are also other concrete examples of families other than the Tahirpur and Nadia rajas and the Sabarna Chowdhurys who made handsome professional and material gains under the Mughal regime and then sought to proclaim their prosperity and status, by staging Durga pujas, amongst other things.Footnote ¹⁰⁶ Questions as to why the landed aristocracy of Bengal may have been especially encouraged, first, to play more prominent roles as hosts of the annual Durga puja and secondly, to make lavish displays of wealth during the pujas in the period marked by the establishment of Mughal rule over Bengal, are addressed below.

The popular traditions surrounding Durga puja also contain the persistent suggestion that the hosts of the apocryphal ‘first’ Durga puja in the late sixteenth and early seventeenth centuries—particularly Raja Kangshanarayan of Tahirpur—may have made his puja much more accessible to ordinary people than hitherto.Footnote ¹⁰⁷ In fact, the elements associated with the most celebrated Durga pujas, whether staged in pre-Mughal Bengal, Mughal Bengal or in colonial and contemporary times, are associated typically with the attendance of large numbers of people. The large expenditure and logistical planning for the many days of the puja are necessitated precisely by the fact that large numbers of people attend the ceremonies. In the case of Durga pujas hosted by rajas, zamindars and the affluent gentry, the viewers and attendees at these events were their subjects. As many nineteenth-century descriptions of Durga puja celebrations in Calcutta reiterate, the city palaces and mansions were accessible up to a point for ordinary residents of the city.Footnote ¹⁰⁸ It has been said that baroari pujas (the practice of ordinary people combining their resources to stage a neighbourhood puja, which is the most typical form of the Durga puja now) began precisely because rich and powerful hosts could and often did impose restrictions on the free access of ordinary people to certain events associated with the many days long event. The very first baroari puja, is commonly dated to the late eighteenth century, but this phenomenon may actually have begun earlier.Footnote ¹⁰⁹ It certainly gave the collective group which hosted the celebration a much greater say and proprietorship in all matters connected to the puja and also for a more direct participation in all stages of it. However, as the discussion below underscores, seriously restricting or barring the access of ordinary people to the puja by wealthy and powerful hosts, would in a sense defeat the very logic and rationale of staging these spectacles. In terms of the accessibility of the Durga puja for ordinary people and their participation in it, it also seems that this feature certainly did not originate with the pujas of the Mughal allies, for example, the Lakshmikanta Majumdar, Bhavananda Majumdar and Kangshanarayan of Tahirpur. The Ramacharitam of the twelfth century stated that on the occasion of the worship (archana) of the Goddess Uma, general festivities occurred; in his treatise entitled the Kalaviveka, the smarta pandit Jimutavahana (twelfth–thirteenth centuries) echoed a similar sentiment; Phullara, in Mukunda Chakrabarty's Chandimangala, also suggested that the worship of the goddess Durga or Chandi was an occasion of joy and excitement for humble people as well.Footnote ¹¹⁰ Phullara in fact said that women took pains to put on their nicer clothes during this annual celebration.Footnote ¹¹¹ This feature may have become more pronounced during the late sixteenth and early seventeenth centuries.

The hypothesis regarding the larger scale and greater openness of pujas organized by the landed aristocracy from about the late sixteenth and early seventeenth centuries is further reinforced by the greater currency of a particular architectural feature during this period: the chandimandapa or the thakurdalan as an element in the architecture of aristocratic and gentry residences. These structures, typically located within the courtyard of the mansions or rajbaris of the wealthy were in fact separate annexes or buildings which were used for big, annual festivals such as the Durga puja. Thakurdalans were flat-roofed structures with pillared arches.Footnote ¹¹² They provided a covered area for the performance of the puja and other events; they were open on most sides and thus gave unrestricted access to the large numbers of people who came to view the image and to participate in the ceremonies in various ways. Although not unknown earlier, annexes or buildings may have become a more common feature of aristocratic residential mansions from about the late seventeenth and eighteenth centuries. According to Nazimuddin Ahmad, chandimandapas and thakurdalans became a much more common feature in the mansions of zamindars from about the late seventeenth century.Footnote ¹¹³ According to Bimala Chandra Datta, most extant durgamandapas, chandimandapas and thakurdalans (as well as rasmanchas), in western Bengal at least, were built during the eighteenth and nineteenth centuries.Footnote ¹¹⁴ The royal residence of the rajas of Nadia is supposed to have had one of the most ornately decorated chandimanadapas in Bengal. It was built either in the late seventeenth century or during the eighteenth century.Footnote ¹¹⁵

Many nineteenth century observers were sharply critical of the excessive revelry, ribaldry and drunkenness that accompanied the pujas of the indigenous urban elite of Calcutta. Some lamented the fact that the solemnity of this religious ceremony had been lost. This study concurs with scholars such as Acharya Jogesh Chandra Roy Bidyanidhi and Shashibhushan Dasgupta in suggesting that for many centuries, the annual Durga puja (as opposed perhaps to the daily worship of family deities or of goddess images in temples) had in fact become synonymous with fun, merry-making and a carnival-like atmosphere.Footnote ¹¹⁶ As Jogesh Chandra Roy Bidyanidhi writes:

Dasgupta too remarks that ‘the fun-festivity aspect of the sharadiya Durga puja dwarfs the [purely] ritual aspect of it’.Footnote ¹¹⁸ Thus the sharadiya Durga puja may never have been only a solemn religious ceremony. This is more than attested to by evidence stretching back to the twelfth and thirteenth centuries. According to the smriti nibandhakara, Jimutavahana, the four days of Durga puja were associated with a great deal of dancing, singing and festivities.Footnote ¹¹⁹ Literary texts from the sixteenth century echo this scenario. In the Chaitanyacharitamrita, Krishnadasa Kaviraja offered a vivid description of Shri Chaitanya engaging in a dramatic performance of a segment of the Ramayana on the day of Vijaya Dasami (the final day of the Durga puja) together with his devotees. Chaitanya's stellar performance as Hanumana, the monkey-hero of the epic, apparently drew loud accolades from the audience.Footnote ¹²⁰ These examples in fact provide an important insight into some aspects of Durga puja which have continued into present times, made evident through the association of Durga puja with popular entertainment and performance. Indeed, contemporary as well as nineteenth century accounts of Durga puja, whether baroari (communal), or hosted by an individual family, whether in a rural area or in an urban centre such as Calcutta, Durga puja was always a much-anticipated event, not just because of the religious ceremonies surrounding the worship of the most-beloved of Bengal's deities, but equally perhaps because it was also associated with dramatic performances, music, dancing, and acts by clowns and acrobats. Fun and merry-making were thus essential to the event. The increased involvement of the landed elite with the staging of pujas which became more extravagant, the greater currency of building open-sided, pillared halls (chandimandapas) where elaborate religious rituals as well as musical and dramatic performances could be held, all point towards the continuation, and maybe the expansion of the element of popular entertainment as an indispensable aspect of the annual Durga puja. Datta mentions many Durgamelas founded by local zamindars. Rajas and zamindars had deep purses which allowed them to host the entertainment and spectacles so enjoyed by ordinary people. In many areas, fairs and Durgamelas which further heightened the carnival-like aspect of Durga puja were founded and supported by the landed elite.Footnote ¹²¹ The rajas of Nadia were particularly known for the grand entertainments they arranged at the time of their puja. Bharatchandra Roy referred to them in his Annadamangalkavya; other accounts mention the direct involvement of the raja and his immediate family in some types of audience-pleasing activities.Footnote ¹²²

Much has been written about the association of the goddess in her various forms as the embodiment of forms of community—ranging from an Indian nation (Bharat mata) to a linguistically grounded Tamil nation (Tamilttay)—together with the gendered and broader cultural implications of such conceptualizations.Footnote ¹²³ Tithi Bhattacharya quite rightly emphasizes the ‘secular’ use of religious symbols such as the goddess Durga and her autumnal puja for community formation in late nineteenth and twentieth century Bengal.Footnote ¹²⁴ Moreover, the greater openness of Durga puja celebrations raises the question about the nature of the ‘public’ realm or arena (terms used by Sandria FreitagFootnote ¹²⁵) and its implications for the social and cultural history of early modern/Mughal Bengal. The classic Habermasian model of a print-based public sphere which was distinct from the stateFootnote ¹²⁶ has been critiqued as much for its assumption of neat distinctions between civil society and state, its privileging of print culture as an essential attribute, its Eurocentrism and ahistorical character amongst other things,Footnote ¹²⁷ and there is no question of attempting to locate exact equivalences between the public realm in India between the sixteenth to eighteenth centuries and the kind of public sphere envisaged by European political philosophy. However, the public realm, as a sphere that was conceptually distinct (notwithstanding the possibility of connections between the two) from the private one associated with household and family is useful and necessary for the political and cultural history of early modern South Asia, if its distinctions from the type of civil society sphere described by Habermas are kept in mind. As Freitag's work showed, public arena activity and collective ceremony such as processions, festivals and the like, which revolved around ‘symbolic enactments of events and rituals that simultaneously delineated common values and drew on shared historical moments and locally significant cultural referents’,Footnote ¹²⁸ functioned as means through which community identity was both constructed and expressed. In the case of Mughal Bengal, the access (albeit with some restrictions and some differential treatment), of ordinary people (including Muslims and low-caste people) to the pujas hosted by rajas and zamindars at the mansions of the latter, represented a degree of collective participation in festivities associated with this event. In the context of Mughal and nawabi Bengal, lavish, annual celebrations of Durga puja served to tie the local raja and zamindar to his prajas (subjects) through creating a venue where the latter could enjoy and participate in this spectacle. Moreover, the many artisans, craftsmen and service providers who worked all year to prepare for the annual Durgotsava were often maintained by the local raja as well. Durga pujas thus, showcased the wealth, prestige and generosity of a particular zamindar; these events created a specific sense of community among the subjects of a particular landlord. Needless to say, such communities co-existed with other overlapping and intersecting communities created by various sources of commonalty and affiliation current at the time.Footnote ¹²⁹ Shashibhushan Dasgupta points out that the inseparable association of the annual Durga puja with popular enjoyment and entertainment probably explains why Durga, by far the most beloved and popular of Bengal's deities, may have rendered this deity less important in Bengal's Shakta spiritual traditions. The form of the goddess which acquired greater significance, and in fact became central to Shakta spiritual/metaphysical endeavours, was the goddess Kali and various other deities associated with the Dasamahavidyas.Footnote ¹³⁰ But the goddess Durga compensated for this by functioning as a broad, overarching symbol of community—both local and regional. The Mughals, as the ultimate political masters of the local rajas and zamindars may have reaped indirect benefits as enablers of such public performances of religion.

Mughal rule, the public performance of religion and Durga puja

Despite the abrogation of several elements of it during the reign of Aurangzeb (1659–1707), the dynastic ideology elaborated and articulated during the reign of the emperor Akbar (1556–1605) became the core of the Mughal state's long-term legacy.Footnote ¹³¹ The conceptual underpinnings of this ideology synthesized ‘elements from the vast landscape of evolving political practices, conscious and unconscious social ethos, a mosaic of secular and religious streams and their strands [and] the positing of an alternative reconstitution of history. . .’.Footnote ¹³² The centrality given in this ideology to the ultimate goal of achieving ‘universal peace’ (sulh kul) or harmony, became instrumental in the fashioning of an imperial policy in which discrimination based on religious-sectarian differences, ethnicity, etc., was set aside as was the primacy of orthodox, Sunni Islam as the foundation of the state. There emerged instead an emperor-centric ideology in which the emperor was held to be the embodiment of a divine light, sometimes (as in the case of Akbar) regarded almost as a prophet and a spiritual preceptor as well as (in the case of Jahangir, for example) to be endowed with mystical, healing powers. The will of this almost super-human, divine ruler took precedence over religious law—including the Sharia. Together with compassionate, but just and firm paternalism, Mughal emperors aspired to bind the nobility and high officials to the cause and service of the empire through a political culture which was manifest in elaborate court rituals and ceremonies and, as discussed insightfully by J. F. Richards, a mystical relationship of ‘discipleship’, associated with the utmost loyalty and devotion to the emperor.Footnote ¹³³ The displacement of formal religion—any religion—as the primary foundation of the state, its institutions and its public culture was paralleled by concrete measures to ensure much wider participation in the polity by different groups who comprised the multi-religious, multi-sectarian, multi-ethnic society of Mughal India. Acts such as marriages to Hindu princesses as well as the patronage of and interactions with scholars, poets, musicians, theologians, ascetics of various religious-cultural backgrounds, further reinforced the primary ideological imperatives of the empire,Footnote ¹³⁴ as did the demonstration of respect for the sensibilities of regionally entrenched groups, such as the Jains of Gujarat at the time of Akbar.Footnote ¹³⁵

A cluster of formal policy measures represented the practical implementation of this ideology under Akbar and despite some erosions from the reign of Shah Jahan, these were continued overall through the reigns of his two immediate successors. In the early years of his reign, Akbar abolished discriminatory taxes against Hindus (i.e. the tax levied on Hindus travelling to pilgrimage and attending festivals and the jiziya tax, a graduated property tax) and permitted both the construction of new temples as well as the repair of existing ones. Under the sultans of Delhi, official policy had banned Hindus from doing both. As Sri Ram Sharma admits, it is impossible to ascertain to what extent this law was enforced. There are also examples of the Sultans of Delhi protecting and repairing temples within their jurisdictions. Secondly, the Sultans of Delhi are not known to have attempted to regulate the private practice of religion by Hindus and other non-Muslim groups.Footnote ¹³⁶ This was in significant contrast to many parts of early modern Europe for instance, where there were prohibitions on the private practice of religion when it was different to that of the ruler. However, from the time of the emperor Akbar, the clear signal given by the monarchy that the public performance of all religions was to be an accepted feature of its public life (as long as they were not perceived to be political challenges to the regime, threats to law and order and/or offensive to ‘mainstream’ cultural norms) may well have strengthened and reinforced such activities. If the construction and use of temples by Hindus provides an important example of the public profession and display of religious-cultural affiliation, then, it is not insiginificant that Akbar's reign witnessed the commencement of monumental temple architecture at prominent sites of Hindu pilgrimage such as at Benaras and Brindavan-Mathura. Contemporary observers testify to gatherings of multitudes of Hindu devotees at Hardwar and Prayaga. At the imperial court, Mughal emperors participated in the celebration of festivals such as Deepavali and rakhi.Footnote ¹³⁷

The role of Mughal nobles and mansabdars in disseminating imperial culture in various parts of the empire and their assimilation of regional/local cultural traditions is an important topic which deserves far greater attention than it has merited so far. For Bengal, we know of Mughal mansabdars extending support and patronage for at least certain selective types of regional traditions and practices. During his appointment as subahdar of Bengal during 1626–1627, Muqarrab Khan began to celebrate the locally popular bera bhashan festival in honour of Khizr pir. Raja Man Singh, who had a very long professional connection with Eastern India (Bihar and Orissa) as well as with Bengal itself, also played a critically important role in these regions through his non-military activities. During his tenure in these regions he built mosques and palaces in places such as Chunar and Rajmahal, repaired existing temples, founded new ones and made land grants for the support of various shrines as well as religious orders such as Vaishnavas.Footnote ¹³⁸ Various local traditions in Bengal associate the raja with a predilection to associate himself with Hindu sanyasis and yogis. His marriages to princesses from the families of local rajas whom he subdued on behalf of the empire, his seeming partiality to elevate as zamindars, high-status Brahmin lineages who combined scholarship and piety with military entrepreneurship,Footnote ¹³⁹ all contributed to the creation of an almost legendary image for Man Singh, particularly amongst the Bengali Hindu gentry and the landed elite of Bengal. The family traditions of all three of the rajas supposed to have hosted the so-called ‘first’ Durga puja (Bhavananda Majumdar, Lakshmikanta Majumdar and Kangshanarayan of Tahirpur) fit this mould. A persistent tradition also developed in Bengal which represented Raja Man Singh as the Mughal commander who had successfully vanquished the rebel Raja Pratapaditya of Jessore. Literary productions such as the Annadamangalkavya (discussed above) and the Raja Pratapaditya Charitra of Ramram Basu (1801) also reflect this. Yet, at the time of the military showdown between Pratapaditya and the Mughal forces, Raja Man Singh was not present in Bengal. This campaign was led by Islam Khan Chishti, the subahdar of Bengal in the early seventeenth century. Man Singh's cultural activities in Eastern India and in Bengal, therefore, served to present him to the Bengali gentry and rajas as a role model whom they should emulate. The raja embodied the example of a grandee who was a high-profile Mughal mansabdar, adept in Mughal courtly culture and, at the same time, a Hindu ruler who cultivated well-accepted Brahmanical practices and norms.Footnote ¹⁴⁰

Raja Man Singh's attentions to Bengal's favourite deity, the goddess, are particularly relevant here. A temple to the devi, known locally as Sarvamangala, was constructed in the Western part of Midnapur by a Mughal official in 1604 CE. Man Singh endowed lands for the seva of the deity.Footnote ¹⁴¹ Local memories also associate Raja Man Singh with the deity known as Karunamayi Kali in the village of Amdanga (Twenty Four Parganas, near Barasat) and further reinforce the constructed tradition regarding the raja's participation in the battle with Pratapaditya by positing that he had established the deity here on his way back from that encounter.Footnote ¹⁴² The strongest instance of Raja Man Singh's connection with Bengal's goddess tradition is of course comprised of his action in carrying back to Amber the deity known as Shila Devi which had once been worshipped by Kedar Roy of Sripur. This deity, ensconced within the Amber fortress, is still worshipped by Bengali priests whose ancestors were apparently brought here by the raja.Footnote ¹⁴³ As the work of Richard Davis, Richard Eaton, Romila Thapar, Norbert Peabody, and others, has emphasized, the worship of images housed in temples had become the dominant form of public religious practice for various denominations of Hindus since medieval times.Footnote ¹⁴⁴ Since images and the temples in which they were situated were closely tied to the political order, it was customary for conquerors to seize temple idols ‘as a statement of conquest’.Footnote ¹⁴⁵ Kedar Roy, one of the Barabhuiyans, had been defeated by the raja; as was the practice of victors, he carried the deity away with him. If the goddess had become synonymous with Bengal's regional tradition, then, Man Singh, as an agent or representative of the Mughal polity and its religious-cultural and political ideology, in a sense, succumbed to the former by placing the deity in a position of prominence within his fortress. In the imagination of the poet Bharatchandra, the goddess had converted the Mughal emperor, his court, and the citizens of Delhi into her devotees; Raja Man Singh's actual action vis-à-vis Shila Devi, represented the incorporation of Bengal's goddess indirectly into the cultural ethos of the empire through her installation in the watan of one of the polity's pre-eminent Hindu mansabdars.

If the religious-cultural policy and attitudes of the Mughal polity, together with the actions in Bengal of Raja Man Singh in particular, fed into and strengthened the tradition of goddess worship, then the latter was reinforced further by a body of popular perceptions regarding the empathy of the emperor Akbar, in particular for Hindus and with Hindu practices and his overall resolve to associate deserving Hindus with the functioning of his regime. As Harbans Mukhia points out, such perceptions, derived from ‘folk tales and bazar gossip’ served a critically important function in first, legitimizing distant emperors in the eyes of ordinary people and secondly, minimized the psychological-cultural distance between the two by humanizing the rulers.Footnote ¹⁴⁶ Bengali narratives from the seventeenth and eighteenth centuries contain eulogies to the distant Mughal emperors who are compared to and equated with Hindu divinities and epic heroes. Akbar is compared to Arjuna and Brihaspati and even Aurangzeb is compared to Ramachandra. Resonances of this are also found in other parts of India.Footnote ¹⁴⁷ Local traditions current in the Northern parts of Bengal recount an incident in which a Brahman who had turned outlaw, in rebellion against the Mughal regime, was reconciled to the latter when he had a conversation with Bhan Singh, Man Singh's brother. Bhan Singh is said to have informed Beni Roy, the Brahman-turned-dacoit:

This story, had a very long life: it was recounted by European travellers in India during the seventeenth century as well as by the Brahman pandit Mrityunjoy Bidyalankar, author of the Rajabali, which had been commissioned by the English East India Company in the early nineteenth century.Footnote ¹⁴⁹ Clearly, this story had considerable appeal—it underscores the desire of at least some Hindu subjects of the empire to ‘own’ (a term used by Mukhia) Akbar by folding him into their ‘community’, if such a term can be used to refer to the vast diversity of Hinduism, via the latter's status, in a previous life, as a brahmachari. According to the story in question, Beni Roy, the ‘pandit-dacoit’, was apparently won over by the argument presented to him by Bhan Singh and became a peaceful and respectable subject who received a zamindari sanad from the Mughal emperor.Footnote ¹⁵⁰ Thus folk/popular perceptions regarding the Mughals—particularly the emperor Akbar, went a long way towards strengthening the impression amongst Hindu subjects that the imperial government was sympathetic to their cultural-religious practices, and secondly, accepted, maybe even encouraged, public displays of religion. As indicated by Alam and Subrahmanyam, Mukhia and others, there is an urgent need for historians of Mughal India to exploit the rich potential of such ‘folk’ perspectives in order to widen the parameters of this period of South Asian history.Footnote ¹⁵¹

As seen above, a marked trend towards the Brahmanization and gentrification of goddess worship was noticeable particularly from the sixteenth to the eighteenth centuries. One of the most important manifestations of this trend was the production of a large body of performative narratives during this period (See Appendix II below). These included Mangalkavyas, in honour of the goddess, described variously as Chandi, Durga, Gauri, Abhaya, Ambika, and Kalika and, from the eighteenth century, a large and impressive outpouring of Shakta devotional padavalis (See Appendix II below). This trend coincided with a spurt of temple-building in Bengal from about the sixteenth century into the early nineteenth century. These temples, mostly built of brick, and profusely embellished with terracotta ornamentation, represented the emergence of a unique vernacular mode of temple architecture which was indebted for inspiration, to mosques built in sultanate Bengal.Footnote ¹⁵² The latter, in turn were inspired by the building style of the humble peasant's thatched hut. What is also remarkable about this phase of temple building is that there are very few extant temples in Bengal which can be dated to the period covered by the region's sultanate. The few that exist from this period are generally concentrated in South-Western Bengal (corresponding to present day districts such as Midnapur, Bankura, Purulia, and parts of Burdwan), in areas that adjoined Orissa and were susceptible to Orissan culture and periodically, to Orissan political control as well. Architecturally, these temples resemble the Orissan rekha-style.Footnote ¹⁵³

As is well known, the sultans of Bengal, particularly from the fifteenth century onwards, were sympathetic to regional, vernacular culture. Yet, existing literature tells us practically nothing about the position taken by various sultans regarding the public performance of religion by non-Muslims—specifically, by Hindu subjects.Footnote ¹⁵⁴ The panchali literature of Bengal, of which the Mangalkavyas are a part, was performed in front of audiences. We also know of collective devotional singing (samkirtanas) and dancing sessions in which Sri Chaitanya and his associates participated as well as the samkirtana processions which they took down the streets of Navadipa. Even in the post-Chaitnaya period, large assemblies of Vaishnavas, such as the ones held by Nityananda, are common knowledge. But, Ramakanta Chakrabarty suggests that at least in the time of Sri Chaitanya, collective and probably public displays of religious/sectarian affiliation may not have been common, or acceptable, at least to orthodox Brahmans.Footnote ¹⁵⁵ Thus, even if some modes of public/collective displays of religious/sectarian-cultural affiliation were not officially banned by the Bengal sultans, Brahmanical society itself, particularly, its upper/elite levels may have had limited toleration for it.

Literary as well as art-historical evidence from the centuries immediately preceding the establishment of Islamic rule over Bengal (the Pala-Sena periods, corresponding to the period from the eighth–ninth centuries to the twelfth century) suggests that temples had in fact existed in Bengal at that time. Some of these temples were established and patronized directly by the Pala and Sena kings and were spectacular in terms of size and grandeur.Footnote ¹⁵⁶ Yet, there are very few extant temples in Bengal from the sultanate period. This study does not subscribe to the (often) unsubstantiated view that Hindu temples were destroyed en masse during the entirety of Sultanate and Mughal rule over Bengal. But, as Richard Eaton pointed out in his seminal article on temple destruction, Indo-Muslim powers are known to have attacked and destroyed temples symbolizing the sovereignty and legitimacy of a raja against whom they were at war, just as Hindu states did too, during times of military conflict.Footnote ¹⁵⁷ Thus, during the initial period of Turkish conquest in Bengal, certain temples, as symbols of the power and autonomy of pre-existing rajas were targeted and destroyed. Thus, to provide one example, the tomb of Zafar Khan Ghazi, at Tribeni (one of the earliest extant Muslim monuments of Bengal), bears testimony to the re-use of materials which had been salvaged in all probability from a pre-existing Hindu structure.Footnote ¹⁵⁸ There are few substantive evidences of temple destruction in Bengal during the period of the late Sultanate and the subsequent period of Mughal and nawabi rule.Footnote ¹⁵⁹ George Michell suggests that perhaps pre-Mughal temples in Bengal have not survived because they were built primarily of straw and thatch.Footnote ¹⁶⁰ But, in either case, the re-emergence of temple construction in Bengal from the sixteenth century, as well as the fact that these were mostly built of brick—a more durable material than straw and thatch—opens the door to the possibility that there had occurred a change in the political-cultural environment which rendered more attractive and feasible the construction of temples which would be more enduring.

Against these considerations, the proliferation of temple building from the sixteenth century onwards, particularly the later sixteenth century, is indeed remarkable. Temple-building in its most concentrated form is supposed to have occurred in particular during the late eighteenth and the early nineteenth centuries—the builders and patrons of these temples were mostly the zamindars and entrepreneurs who had successfully enriched themselves by collaborating with the English East India Company.Footnote ¹⁶¹ If we move back a bit in time, the temples that began to emerge during the late sixteenth, seventeenth and early eighteenth centuries, are typically associated with rajas, zamindars and affluent gentry who had collaborated with the Mughal regime and then, with its lineal descendant, the regime headed by the nawabs of Murshidabad. The dedicatory inscriptions of these temples attest to the social-professional backgrounds of those who financed and commissioned the building of temples.Footnote ¹⁶² The largest numbers of them were Vaishnava temples dedicated to Radha-Krishna, but the emergence of goddess temples during the seventeenth century is also a distinctly identifiable trend. The worship of the goddess as Kali, gained greater currency from about the middle of the eighteenth century onwards and is borne out by the construction of larger numbers of Kali temples during this period and also by the growing popularity of vernacular narratives entitled Kalikamangala, during the seventeenth and eighteenth centuries. Although not identical, the emergence of goddess temples (no matter what the form of the goddess) and the trend towards more public, ostentatious Durga pujas lasting several days in the autumn, are not unrelated either and reflect complementary religious-cultural processes.

The support of Mughal emperors such as Akbar, Jahangir and even Shah Jahan for various Vaishnava sects is well known, as is the close affiliation of many important mansabdars for Vaishnavism, in particular, Raja Man Singh and the Kachhvahas of Amber-Jaipur.Footnote ¹⁶³ The period of Mughal rule also coincided with the greater consolidation of Vaishnavism in Bengal in the form of a more Sanskritized and Brahmanized version which had been crafted in the Braj region by the Chaitanyite goswamis of Brindavan.Footnote ¹⁶⁴ This more gentrified mode of Vaishnavism is said to have been remarkably successful in securing as patrons and sponsors already wealthy rural elites and rajas, but even more particularly, those upwardly mobile lineages and communities who had material affluence, but lacked social-cultural status and sophistication.Footnote ¹⁶⁵ A good example of the latter are the Malla rajas of Bishnupur, who had low-status, maybe even tribal or adivasi antecedents—they went on to become one of the foremost patrons of the Brindavani mode of Vaishnavism in Bengal. They are also known for their generous support of the celebrated navaratna-style temples of which literally hundreds are once supposed to have graced their kingdom. The devotional culture of Vaishnavism which gave such centrality to collective and congregational activities such as samkirtana, and celebrations such as dol-yatra, ratha-yatra, and rasa utsava acquired greater elaboration and definition in and around temples built during this period. As George Michell points out, the architectural features of Vaishnava temples certainly, but also other temples, began to incorporate elements which would allow for such public, congregational activities in halls, porches and other structures such as rasa manchas, dol-manchas and jhulan manchas.Footnote ¹⁶⁶ This coincides with the emergence in the domestic architecture of aristocratic families who hosted the autumnal Durga puja, of the chandimandapas and thakurdalans, mentioned above. Thus, the devotional-congregational culture of Vaishnavism (also supported by the Mughals) may have, in Bengal, reinforced the trend towards making the greatest of all Shakta festivals, the Durga puja, more accessible to ordinary people and more ‘open’ than before. Scholars have reminded us that the great populist potential of Vaishnavism (which, at least in Bengal, was never completely denuded despite its greater Brahmanization in the seventeenth century) was instrumental in persuading adherents of the region's shakta culture of the need and wisdom of incorporating elements of the former.Footnote ¹⁶⁷ Thus, together with the experience of Mughal rule, the region's Vaishnava culture may have acted jointly to render the annual Durga puja grander, more open and more accessible. The involvement of the landed elite with these celebrations was designed to secure to them the loyalty and goodwill of their subjects who could participate in the festivities associated with the worship of Durga.

It has been suggested that the chaos, turbulence and insecurity that engulfed Bengal during the eighteenth century, explains the growing currency of goddess worship in the forms of Durga and Kali during this period.Footnote ¹⁶⁸ The middle to the end of the eighteenth century indeed witnessed a series of military, political, economic and climatic (the great famine of 1770) crises and there is something to be said for the expanding tendency to worship at such times, a deity, who could protect worshippers and dispel enemies. The worship of the goddess as Kali grew in particular during the later part of the eighteenth century; the mother-child relationship, such a dominant metaphor for the relationship between the goddess and her worshipper in the Shakta literature produced during this period, may indeed hold a clue to the former phenomenon. As this paper shows, ostentatious Durga pujas staged by rajas and zamindars may not have been a response to the ‘crisis’ of the eighteenth century alone. The connection between the worship of the goddess and kingly personages was of very much older vintage. The inception of Mughal rule reinforced this connection and encouraged the greater definition and development of certain features now considered to be almost indispensable to this sharadiya celebration.