A village in northern India in the fifth century CE has a goddess of smallpox. Her name is Śītalā (literally "the Cool One" — the polite-fearful name, because addressing her by her actual function would invite her to come). When smallpox arrives in the village, the people understand it as Śītalā's visit. Her possession of a body produces the pustules and the fever. She is fierce. She is dangerous. She is also the only one who can stop the epidemic. The village makes offerings — songs, food, animal sacrifice, sometimes things a respectable Brahmanical temple would not accept. The offerings are not for Śītalā's love. They are propitiation. They are to make her leave.
This is one of seven or nine sister-goddesses in the local pantheon — disease-demonesses, each with her own specialty (cholera, smallpox, the wasting illnesses of small children, the fevers, the seasonal sicknesses). Some of these goddesses have achieved independent fame: Śītalā for smallpox, Mariamman in South India for cholera and parallel fevers. Most of them remain at the village level, names known locally, propitiated through specific rituals.
By the eighth or ninth century, some of these goddesses have undergone transformation. They have been absorbed into the developing Hindu Tantric pantheon. They have new names: Kālī, Cāmuṇḍā, Yoginī, Ḍākinī. They are no longer just disease-goddesses — they are now wild Tantric goddesses, worshipped in cremation-ground rituals, partnered with male Tantric practitioners, eventually elevated as supreme deities of major non-dual Śaiva traditions. The smallpox-goddess has become the Goddess.
Samuel's argument in Chapter 10 is that this transformation is the historical mechanism by which Tantric goddess-cult emerged.1 The wild Tantric goddesses are not a new invention of the medieval period. They are the upgraded forms of disease-demonesses who had been operating at the village level for centuries. The Tantric tradition incorporated and elevated them.
Samuel asks the question directly: "where did these wild, dangerous goddesses come from?" His answer is that the standard goddess-iconography of the early centuries CE shows no real sign of a cult of wild or fierce goddesses. Early goddess imagery is relatively gentle. Even the early Kuṣāṇa-period "Warrior Goddess" images (the Durgā-killing-the-buffalo-demon scene, the Mahiṣamardini theme) are "quite mild" and "this continues throughout much of the first millennium CE."2 Later Durgā imagery becomes "a little more forceful, but it is still a long way away from the violence and horror of some of the imagery of the new deities."
The textual record shows more wildness. There are female demonesses in the Buddhist sūtras and jātakas (Sutherland 1991). Hāritī in the Pali Canon is a demoness who kills and eats small children before being converted by the Buddha. The Mahābhārata Book Three discusses graha ("graspers") who can be either male or female, emanated by Skanda, sources of illness.3 The medical literature (Suśruta-Saṃhitā, Aṣṭāṅgahṛdaya) contains discussion of female disease-demons responsible for specific illnesses of children and women.
What links these scattered references is the seven-sisters pattern. Kosambi (1960) traced references in North Indian texts to mother-goddesses occurring as a set, "described as mothers or sisters, and five, nine or, most often, seven in number."4 These are the operational pre-Tantric goddesses: female deities of illness, fertility, and disease, attached to specific local communities and propitiated through specific local rituals.
The Brahmanical tradition incorporated this goddess-complex through the Saptamāṭṛkā ("Seven Little Mothers") figures. The Saptamāṭṛkā are documented from the early fifth century CE in early Gupta sculpture. They are female consorts of major Brahmanical deities (Brahmāṇī, Vaiṣṇavī, Māheśvarī, etc.) along with the fierce goddess Cāmuṇḍā. They are usually depicted as a set of seven, often accompanied by the war-god Skanda.5
Katherine Harper has argued that the Saptamāṭṛkā first rose to importance "in the context of state sorcery, more specifically the performance of military rituals to aid the Gupta kings in their campaign to expel foreign rulers from India" (Harper 2002).6 The argument rests on the Udayagiri site, where Candragupta II built a shrine to Śiva that incorporated a Saptamāṭṛkā relief holding what Harper identifies as war-banners. If Harper is right, the Saptamāṭṛkā are very early evidence for the aggressive use of Tantric-style ritual at the level of Gupta state-religion.
The mechanism Samuel proposes is therefore: the disease-demonesses operating at the village level get Brahmanically adapted into the Saptamāṭṛkā for state-sorcery purposes (controlling and unleashing disease and misfortune on the kings' enemies). Then this adapted set, with its operational structure of male-deity-or-male-female-couple-at-centre-surrounded-by-female-deities-in-eight-directions, becomes the template that the developed Tantric goddess-maṇḍala will use. The eight Māṭṛkā in the maṇḍala of the Hevajra Tantra (Gaurī, Śavarī, Caurī, Caṇḍālī, Vetālī, Ḍombī, Ghasmarī, Pukkasī)7 are the operative architectural form. The disease-demonesses have made their way through Saptamāṭṛkā into developed Buddhist Tantra.
Samuel makes a specific suggestion about the ḍākinī figure. If the wild Tantric goddesses descend in part from disease-demonesses operating at the village level, then the ḍākinī may also have a specific social origin. Samuel proposes: "if the human ḍākinīs were folk healers from low-status castes such as the modern birth attendants or dāī, constantly suspected also of being sorceresses, then their ambivalent role as partners for the male Tantric practitioners, and their elevation to high status in Tibet where this whole complex of ideas was more or less irrelevant, might make sense."8
The hypothesis is operational. The ḍākinī in the early Tantric material is a real category of woman — low-caste folk-healer, midwife, herbalist, also frequently suspected of sorcery. The Tantric practitioner who partners with a ḍākinī is partnering with a specific kind of woman who occupied a specific socio-economic role at the village level. The Tantric tradition incorporated this role into its initiated practice, elevated the ḍākinī into a religious-symbolic category, and gradually (especially in Tibet, where the social-historical referent did not exist) transformed her into the khandroma — the high-status spiritual category that ḍākinī eventually became.
The argument matters for how the vault should engage Tantric texts. The figures who appear in early Tantric texts as ḍākinīs, ḍombīs, caṇḍālīs are not purely symbolic. They are partly literal — referring to actual women in actual socio-economic positions — and partly symbolic-coded. Reading the texts requires holding both registers.
Connects to Kāpālika-Style Ascetics — the male-practitioner counterpart to the female-deity-and-partner complex. Connects to Auspiciousness vs Purity — the value-system inside which the female-deity material operates. Connects to Two-Worlds Thesis: Kuru-Pañcāla vs Central Gangetic — the cultural-historical frame that gives the female-deity material its strength on the Central-Gangetic side. Connects to Śaiva Tantra Hierarchical Succession — the institutional architecture that integrated the goddess-cults. Connects to Trika Non-Dual Śaivism — the elaborated philosophical form the goddess-cult eventually produced.
The Devīmāhātmya ("Glorification of the Goddess") is one of the central early-medieval Goddess-cult texts. It is part of the Mārkaṇḍeya Purāṇa, generally dated to the sixth or seventh century CE. The text describes the Goddess (Devī) as the supreme cosmic-creative power who manifests to defeat various demonic enemies of the gods: Madhu-Kaiṭabha, Mahiṣa (the buffalo-demon), and Śumbha-Niśumbha.
In the text, Kālī appears as an emanation of Durgā during the Śumbha-Niśumbha battle. Cāmuṇḍā appears as another emanation. The wild-goddess figures are operating as manifestations of the central Goddess rather than as independent deities. The text is doing important consolidation work: it is taking the previously scattered disease-demonesses and folk-goddesses and organising them under a single supreme Goddess who manifests in different forms for different cosmic purposes.
The Devīmāhātmya is, structurally, the Brahmanical consolidation of the goddess-complex into a coherent theological architecture. The disease-goddesses and folk-goddesses have been re-positioned as forms of one supreme Devī. The Saptamāṭṛkā emerge in the text as her forces in battle. The wild figures are still wild, but they are now wild under a unifying theological frame.
This is precisely the move Samuel describes for the broader political-cosmological shift (see Polytheism-to-Monism Shift). The plurality of village goddesses is being consolidated under a single supreme deity. The mechanism that consolidates the male-deity-pantheon (Viṣṇu and Śiva as supreme forms) is the same mechanism that consolidates the female-deity pantheon. The Goddess as supreme cosmic principle is the parallel theological move to the supreme-Viṣṇu and supreme-Śiva moves of the early-medieval period.
What the Devīmāhātmya does not yet do is fully integrate the goddess-cult with the male-deity Tantric architecture. That integration happens in the seventh-and-eighth-century Tantric texts that follow. By the time we reach Abhinavagupta in the late tenth and early eleventh centuries, the integration is fully developed and the Goddess is theologically articulated as the active creative power (śakti) of Śiva himself.
You read a Tantric text featuring wild female deities. You ask three questions to surface the substrate.
First, what is the deity's function? Does she protect against specific illnesses, ward off specific dangers, control specific malevolent forces? If yes, you are looking at a probable disease-demoness substrate. The function survives the transformation into elevated Tantric form.
Second, what are her epithets and offerings? Does she receive blood-offerings, animal-sacrifices, polluting substances? Does the text mention cremation-ground locations, midnight-rituals, low-caste partners? The substrate persists in the ritual specifics.
Third, what is her iconographic ancestry? Cāmuṇḍā retains the skeletal-emaciated form of the cremation-ground goddess. Kālī carries the garland of skulls. The wild iconography preserves the village-level origin even after the theological elevation.
The text will rarely tell you directly: "this goddess was originally a smallpox-deity at village X." But the function, the ritual specifics, and the iconography preserve the trail. Reading the text for the substrate makes the historical trajectory visible.
You read Kālī, Cāmuṇḍā, the Yoginīs, and the Ḍākinīs as originating in the developed Tantric tradition, missing the disease-demoness and folk-goddess substrate that they emerged from.
You assume the wild iconography of Tantric goddesses is symbolic — that the cremation-ground, the skull-garland, the blood-offerings are metaphors for inner states — and miss that the iconography preserves a literal earlier ritual-historical reality.
You read the ḍākinī as a purely symbolic-spiritual category and miss the low-caste-folk-healer-and-suspected-sorceress social origin.
You collapse the Devīmāhātmya's consolidated Goddess-architecture with the village-level disease-demonesses she has absorbed, losing the developmental trajectory that makes the historical picture clear.
You assume the elevation of these goddesses to supreme-deity status is a theological innovation of the medieval period, missing that it is the parallel move to what was happening with Viṣṇu and Śiva at the same time.
Strongest evidence: the persistence of the disease-goddess cults at the village level into modern times provides a comparative case. The Saptamāṭṛkā iconographic record from the early Gupta period documents the Brahmanical adaptation step. The Devīmāhātmya provides the theological-consolidation step. The developed Tantric maṇḍalas (Hevajra etc.) document the final integration.
Tension on chronology: Samuel suggests the major emergence of the wild-goddess Tantric form is "between the fifth to eighth centuries CE." Other scholars place specific elements earlier or later. The Devīmāhātmya's dating is contested (some place it in the sixth century, some in the seventh, some later).
Tension with practitioner-lineage warrants: vault practitioner-lineage corpora that engage Goddess-cult material often treat the Goddess as eternal-cosmic-principle rather than as historically-developed-form. Hold both vantages.
Open question: how much of the contemporary popular devotional cult of Kālī (in modern Bengal) and of Durgā (in pan-Indian popular Hinduism) preserves the disease-goddess substrate operationally vs. how much has it been theologically domesticated away? S. Gupta (2003), McDermott (1996), and McDaniel (2004) have documented the contemporary practice. The vault could investigate.
Samuel, Harper, Kosambi, Sutherland, Coburn, Yokochi, Dimock converge on the basic descriptive pattern: the wild Tantric goddesses are historically-developed forms whose substrate lies in village-level disease-and-fertility goddess cults that were Brahmanically adapted through the Saptamāṭṛkā, theologically consolidated through the Devīmāhātmya, and finally integrated into Tantric architecture. The disagreements concern specific chronology and the extent of cross-domain (Buddhist / Hindu) borrowing.
The interesting tension with the vault's practitioner-lineage corpus is that lineages treating the Goddess as cosmic-eternal principle do not naturally engage the developmental-historical story. Samuel's reading is at the critical-historical vantage. The vault's practitioner-lineage Gauḍīya Vaiṣṇava material engages Goddess through the Rādhā-Lakṣmī axis; the vault's Kashmir Śaiva material engages Goddess through the Śakti-and-Śiva architecture; both engage from inside theological frames that do not require the historical-developmental story. The vault holds the vantages parallel.
A religious tradition that incorporates and elevates village-level disease-and-fertility cults into supreme-deity architecture has cross-cultural and cross-domain parallels.
The handshake to african-spirituality sits in the structurally cognate pattern documented in vault Odwirafo material. The Akan and Igbo cosmological systems incorporate village-level disease-and-fertility-and-ancestral spirits into broader theological architectures. The Igbo dibia tradition's negotiation with disease-and-misfortune spirits operates with structural similarity to the disease-demoness propitiation of pre-Tantric Indic village practice. The vault's African-spirituality corpus on Agwu (the Igbo cosmological-spirit-system that operates at multiple scales including disease and healing) shares structural form with the Indic case. The insight: the village-level cult of disease-and-misfortune deities is a recurring world-religious form that multiple major religious traditions have built upon, and the comparative work the vault could do would illuminate both the African and Indic cases by mutual reference.
The handshake to psychology sits in what the disease-demoness theology does for the practitioner relating to illness. A cosmology that personifies disease-as-deity gives the practitioner an interpersonal frame for engaging illness. The patient is not battling an abstract pathology — the patient is in relationship with a specific deity who has visited the body and who can be propitiated, negotiated with, eventually persuaded to leave. The vault's psychology corpus on illness-as-narrative and on healing-rituals-and-placebo can absorb this. The insight: ancient disease-deity cosmology may have provided psychotherapeutic infrastructure that secular biomedical-frame healing has lost. The patient who relates to illness as a being whose visit can be ritually managed has access to coping resources that the patient who relates to illness as random-biological-event does not.
The handshake to behavioral-mechanics sits in the village-level operational architecture for managing collective danger. The disease-demoness cult is a social-ritual technology for coordinated community-response to epidemic and individual response to illness. The village makes collective offerings; specific designated practitioners (often low-caste women, the dāī-tradition that becomes the ḍākinī-substrate) take on the operational work of negotiation with the deity. The vault's behavioral-mechanics corpus on collective-coordination-under-uncertainty can absorb this case. The insight: the disease-deity cult operates as coordination technology — it gives the village a structured set of behaviours to enact under conditions where the underlying causation is unknown but the danger is real. The community is held together through the ritual response. The ritual response works coordinatively whether or not its theological premises are accurate.
Sharpest implication: the entire developed Tantric Goddess-tradition — including the supreme-Devī tradition, the Kālī-tradition of Bengal, the Yoginī-Tantra of the Trika, the elaborated Tibetan Vajrayāna ḍākinī-tradition — has its operational substrate in the disease-demoness cults of pre-medieval Indic village religion. The substrate is not embarrassing; it is the source. Reading the developed tradition without recognising the substrate misses what gives the tradition its peculiar operational force.
Generative questions: