Why Village Goddesses Were Never Absorbed Into the Brahminic Pantheon

Every village in Tamil Nadu knows Mariamman before it knows Lakshmi. Every lane in rural Bengal has a stone smeared with sindoor that no priest installed. These presences, Mariamman, Yellamma, Sitala, Bonbibi, Meldi, Khodiyar, Pochamma, sit at the edge of settlements, under old trees, inside termite mounds, on the boundary stones between one field and the next. They were never carried into the great temples. They were never given Sanskrit genealogies, celestial husbands, or chapters in the Puranas. Scholars of religion have spent decades asking why. The more useful question is what that refusal actually protects.

The process by which local cults were drawn into the Brahminic mainstream is well documented. Shiva absorbed Rudra. Vishnu absorbed dozens of regional heroes and tribal deities through the avatar framework. Durga was assembled, partly, from goddess traditions that had no Sanskritic origin. The mechanism was dasification, giving the local deity a role in a cosmic hierarchy, a Sanskrit name, a Puranic backstory, a priestly ritual calendar. Once absorbed, the deity became portable. You could build a temple anywhere, install the image, perform the agama rites, and the presence would travel with the form.

The village goddesses did not travel. Mariamman belongs to the red soil of Tamil Nadu. Bonbibi belongs to the Sundarbans. Khodiyar belongs to the Saurashtra coast. When their devotees moved to cities, they built small shrines in the new neighbourhood and re-consecrated the goddess there, but the original stone at the original boundary did not become less real. The goddess was not transferred. She was duplicated, because she is not a form that can be carried. She is a presence that belongs to a specific ground.

The theologian Alf Hiltebeitel, writing on South Indian goddess traditions, observed that the boundary goddesses resist Sanskritisation not because they were overlooked but because absorption would have destroyed the thing they actually are. A Puranic goddess has a biography. A village goddess has a jurisdiction.

Mariamman governs smallpox, chickenpox, heat rash, and the fevers that come with the hot season. Sitala, her North Indian counterpart, whose name means the cool one, governs the same diseases from the opposite direction: she is propitiated to cool the body the fever has set on fire. Pochamma in Telangana governs cholera and epidemic disease. These are not metaphors. For most of South Asian history, epidemic disease was the primary cause of death in rural settlements, and no Puranic deity had jurisdiction over it. Brahma, Vishnu, Shiva, Lakshmi, Saraswati, their domains were cosmic. Creation, preservation, dissolution, wealth, learning. The boundary goddess governed the one thing that actually killed you before the monsoon broke.

Her jurisdiction also covers the threshold between the village and what lies beyond it. The forest, the uncultivated land, the road that goes somewhere else. Bonbibi in the Sundarbans is invoked by honey collectors and wood cutters before they enter the tiger mangrove. She is not a goddess of nature in any romantic sense. She is a goddess of negotiated entry into a space that does not belong to human beings. Her protection is conditional on the terms of the negotiation: take only what you need, do not be greedy, come back before dark. The tiger, Dakshin Ray, is her counterpart in the forest. The devotee stands between them.

The honest answer is that the Brahminic synthesis left the village goddesses alone partly because it could not absorb them without losing the thing that made them useful. A goddess of epidemic disease who accepts blood sacrifice, who is propitiated with rice beer and goat meat, who is served by non-Brahmin priests, sometimes by women, sometimes by members of communities outside the varna system entirely, cannot be brought inside the agama framework without stripping away the practices that made her efficacious to the people who actually needed her.

There was also a matter of caste geography. The village goddesses were primarily the property of agricultural and artisan communities, of fishing communities, of forest-dwelling communities. Their priesthoods were hereditary within those communities and outside Brahminic control. To absorb the goddess would have meant absorbing the priesthood, or displacing it, and displacing a local hereditary priesthood is a political act, not a theological one. The synthesis chose, again and again, to leave the boundary intact.

What resulted was a two-tier religious landscape that has persisted for at least fifteen hundred years. The great temples drew the merchant class, the administrative class, the literate class. The boundary stone drew the farmer, the fisherwoman, the person whose child had a fever at the start of the hot season. Both landscapes were real. Neither was a degraded version of the other.

The village goddess guards the specific. She guards this field, this child, this body, this threshold. The Puranic goddess guards the principle, wealth as such, learning as such, the preservation of cosmic order. The specificity of the village goddess is not a limitation. It is the source of her power for the people who approach her.

She also guards a form of religious authority that does not require Sanskrit literacy, Brahminic initiation, or access to a temple complex. Her worship is available to anyone in the village. The ritual knowledge belongs to the community, not to a specialist class. The offerings are things the community produces: rice, coconut, turmeric, a goat from the flock. The goddess does not require what the community cannot give.

This is why the attempts to Sanskritise the village goddesses, to give Mariamman a Shaiva genealogy, to make Sitala a form of Parvati, have always been partial and contested. The communities that depend on her resist the absorption, because they understand, without necessarily articulating it, that absorption is also a transfer of authority. Once the goddess has a Sanskrit biography and a Brahminic priesthood, she no longer belongs to the lane, the field, the boundary stone. She belongs to the institution.

Mariamman festivals in Tamil Nadu still involve firewalking, still involve possession states, still involve offerings that no agama temple would permit inside its walls. The Bonbibi puja in the Sundarbans is performed in Bengali by a hereditary singer, from a text that blends Islamic and Hindu narrative without anxiety, because Bonbibi herself belongs to both communities and the Sundarbans does not observe the boundary between them. Yellamma's festival at Saundatti in Karnataka draws hundreds of thousands of devotees, most of them from communities that the mainstream pilgrimage economy has historically ignored.

These are not survivals of a dying tradition. They are living religious systems that have maintained their integrity precisely because they were never fully absorbed. The boundary goddess stayed at the boundary. That is where the boundary needed her.

The Puranic pantheon answers the question of how the cosmos is ordered. The village goddess answers the question of whether the child will survive the fever, whether the harvest will come in, whether the person walking into the forest will come back out. These are not smaller questions. They are the questions that most people, across most of history, actually needed answered.