Veḷi

Made visible · deep time

Before the Indus, the graves already faced the sun.

Thousands of years before the first Indus city and tens of thousands before the Vedas, people in South Asia buried their dead in ordered cemeteries aligned to the sun, and set up shrines to a goddess whose living descendants still worship her. This is the oldest religion of the subcontinent — and no text records it, because the people who kept it did not write.

One of the deepest human landscapes on Earth.

the Son valley · ~80,000 years of occupation

The Middle Son valley in north-central India holds a record of continuous human presence spanning roughly eighty thousand years — through the Toba supereruption of ~74,000 years ago, with fine microlithic tools appearing by ~48,000 years ago. This is the corridor Homo sapiens moved through while dispersing out of Africa, and people never left. It is one of the longest unbroken human occupations documented anywhere in the world.

Clarkson et al. (2020), Nature Communications — Dhaba, ~80 ka occupation across Toba; Jones et al. (2009), J. Anthropological Archaeology. verified · consensus

The Baghor stone.

an Upper Palaeolithic shrine · and its living worshippers

At Baghor, in that same Son valley, archaeologists found a triangular stone with natural concentric colour-banding, set into a rubble platform — dated to the Upper Palaeolithic, deep in the Ice Age. They interpreted it as a possible shrine associated with a female principle — an interpretation, not a demonstrated fact. The interpretation is not a guess imposed from outside. They recognised it because the living hunter-gatherer communities of that same region — the Kol, the Baiga, and especially the Musahar — still set up and venerate exactly this kind of triangular stone today.

The goddess who never left

Among the Musahar the great power of the universe is Bansapatti — goddess of the forest, who makes the trees bear fruit and the bulbs grow, in whose name the medicine-person expels sickness, and to whom the childless pray for a child. Goddess of the forest, of healing, and of childbirth. The excavators proposed a date of roughly 9000–8000 BCE on contextual grounds; the living veneration is documented in the present.

Evidence panel — Baghor I

Observed: a triangular stone with natural concentric banding, placed within a constructed rubble platform.

Proposed dating: c. 9000–8000 BCE, argued by the excavators from context. No directly datable material was recovered at the feature itself.

Interpretation: possible ritual feature or shrine — proposed, not demonstrated.

Comparative observation: similar triangular stones receive ritual attention among Kol, Baiga and Musahar communities in the region today.

Not established: the identity or name of any deity; the language or ethnicity of the makers; any continuous doctrine linking the Palaeolithic feature to present-day practice.

Kenoyer, J. M., Clark, J. D., Pal, J. N. & Sharma, G. R. (1983). "An Upper Palaeolithic shrine in India?" Antiquity 57(220): 88–94. Primary excavation report. Dating proposed on contextual grounds. proposed · not conclusively established

The oldest shrine we can find in South Asia is a goddess of the forest and childbirth — and the people who built it never stopped praying to her.

Who claimed her later.

the absorption

This deep, indigenous veneration of a feminine principle runs the length of the subcontinent — and it was absorbed into the Sanskritic tradition much later, then presented as if it had always belonged there. The yoni-shrine of Kāmākhyā in Assam was originally the kirāta dharma, the religion of the Kirātas — the local, non-Brahmanical people — and was only afterwards folded into the Brahmanical fold and given a Sanskrit myth. The Indus cities, a thousand years before any Vedic text, kept aniconic ring-stones and mother-goddess figurines of the same aniconic cult.

So the pattern, across the whole region and across tens of thousands of years, is one thing: an indigenous goddess-veneration, deep and pre-Vedic, later claimed by the tradition that arrived last. Śakti was not given to these communities by the Vedas. It was taken from them, and sold back with a new name.

Rosati (2017, 2021), Religions of South Asia — Kāmākhyā, kirāta dharma; Banerjea; Schumann (2024), Indus ring-stones & pre-Śaiva goddess. verified · consensus

The graves that face the sun.

Damdama & Mahadaha · Mesolithic Ganges valley

Long before farming, hunter-gatherers in the Ganges valley buried their dead in formal cemeteries — and the graves were oriented to the sun. At Damdama and Mahadaha the grave alignments track solar variation, and the cemeteries mark something specific: a community's rights over land and water, held across generations through the ancestors buried in it. To bury your dead in an ordered, sun-facing ground was to say this river is ours, and the dead prove it.

This is a whole religion of place, ancestry, and the sky — with no Sanskrit anywhere in it, thousands of years before the Vedas claimed to define what religion in India was.

Being illustrated · flagged for the interactive build

The solar-aligned grave orientation at Damdama and Mahadaha will be shown as an animated figure — the graves turning to face the rising sun across the seasons. The exact absolute dates are being verified from the excavation record before the animation is fixed; the platform will not fill a date it has not confirmed.

Chattopadhyaya (1996), "Settlement pattern and the spatial organization of subsistence and mortuary practices in the Mesolithic Ganges Valley," World Archaeology (solar-aligned graves; Saxe-Goldstein corporate-rights model). verified · consensus

The oldest death rites, in one line of time.

a chronology · deepest first

~80,000–48,000 ya
Son valley, continuous occupation. Homo sapiens through Toba; microliths by ~48 ka. Clarkson 2020
Upper Palaeolithic
The Baghor goddess-shrine. Triangular banded stone; worship continuous to living Kol/Baiga/Musahar. Kenoyer 1983
>10,000 ya
Smoke-dried mummification, S China / SE Asia. Among the oldest artificial mummification on Earth — older than Egypt. Hung 2025 (PNAS)
Mesolithic
Damdama & Mahadaha. Formal HG cemeteries, graves aligned to the sun, marking rights to river & land. Chattopadhyaya 1996
~3000 BCE
South Indian ashmounds. Deliberately burnt cattle-dung mounds; sometimes founding rituals beneath villages. Fuller 2007
Neolithic → Iron Age
Megalithic burial, peninsular India & Sri Lanka. Predates the "Iron Age" label; still practised today in parts of India. Bauer 2015 · Dissanayake 2022

A single chronology of death-ritual, deepest first. The Indus urban phase (2600–1900 BCE) sits late in this sequence, not at its start. Sources per row. Global mummification included to show South/Southeast Asia was a centre of the world's earliest death-ritual, not a periphery.

What language did they speak? We don't know — and we say so.

the honest limit

It is tempting to call these ancient communities "Dravidian," because their region's later inhabitants speak Dravidian languages and there is real genetic continuity to modern South Indians. But we resist it, because it is the same error in reverse that this platform exists to correct.

The rule this page keeps

Calling a Mesolithic hunter-gatherer "Dravidian" is the same move as calling the Indus "Tamil": a modern label pressed backward onto people who left no words. The oldest layer — the Ancient Ancestral South Indian substrate, some forty-five thousand years deep — has no known language at all. The honest answer is the interesting one: we do not know, and here is exactly how far the evidence reaches before it stops.

Language history: Kolipakam et al. 2018 (Proto-Dravidian ~4,500 yrs); ancestry: Narasimhan et al. 2019 (AASI substrate). Substrate-language research beyond Tamil/Indo-Aryan (Vedda, isolates) is an open thread on this platform. doi:10.1126/science.aat7487 open · being researched

Not fossils. Living carriers.

the continuity, stated carefully

The communities who carry the oldest layer — Musahar, Kol, Baiga, Bhil, and the Munda-speaking peoples — are not living fossils, and this platform will never present them as unchanged relics of an ancient past. They are present-day peoples with their own histories, changes, and futures. But the memorial-stone death rites of the Bhils and the Munda communities are continuous with the ancient megalithic and ancestral-worship complex, and the Musahar goddess-veneration is continuous with the Baghor shrine. The honest frame is neither "they are the ancient past" nor "the past is gone" — it is: the oldest religious life of South Asia was carried by the communities the later tradition pushed to the very bottom, and they are still here, still keeping it.

Mendaly (2024), OMEGA — J. of Death and Dying — Bhil & Munda memorial-pillar death rites, continuity with the ancient complex. verified · consensus

What this page establishes.

the ledger

ClaimStatusSource
Son valley: ~80,000 yrs continuous human occupationCitedClarkson 2020
Baghor: Upper Palaeolithic goddess-shrine, living continuity to MusaharCitedKenoyer 1983
Mesolithic Ganges cemeteries, graves solar-alignedCitedChattopadhyaya 1996
Ashmounds ~3000 BCE; megaliths predate Iron AgeCitedFuller 2007; Bauer 2015
Goddess/yoni cult absorbed into brahmanism lateCitedRosati 2017
>10,000-yr mummification in S China/SE Asia (global context)CitedHung 2025 (PNAS)
Language of these deep-time populationsUnknownNot determinable in either direction
Exact Damdama/Mahadaha absolute datesOpenVerifying before the animation is fixed
vibhūti / sacred-ash continuity with ashmoundsOpenWill not assert without a source