Ancient India — what is known, and how confidently

Why the dates come as ranges

Two reasons, and they are not evasions.

The Vedic texts were composed and transmitted orally with extraordinary phonetic fidelity and no interest whatever in dates. They can be placed relative to one another by language, and placed absolutely only by inference.

And Indian archaeology, though extensive, rarely connects to named rulers before the Mauryas. The first firm chronological anchor is Ashoka, whose inscriptions name Hellenistic kings whose dates are known independently.

Anything more precise than a range, before roughly 300 BCE, is a guess with a confident voice.

Before the cities

Human occupation of the subcontinent runs far further back than any city in it. People were hunting, gathering, farming, settling and painting long before anything here was written down, and the record of that is archaeological throughout.

The rock shelters at Bhimbetka, in the Vindhyan foothills of Madhya Pradesh, are the clearest single case. People sheltered and painted there across an enormous stretch of time: the images do not belong to one period but accumulate through several, and they carry on into the centuries for which written history exists. What they show is mostly ordinary life — animals, hunting parties, groups of figures dancing. The site went onto the World Heritage list in 2003.

What that material cannot do is tell a continuous story. It shows presence, technique and subject. It does not give names, dates in years, or the meaning of what was painted.

The Indus cities

Between roughly 2600 and 1900 BCE, an urban civilisation extended across what is now Pakistan and north-western India — larger in area than contemporary Egypt or Mesopotamia, and far more widely distributed than the two famous cities suggest. Harappa and Mohenjo-daro are the names everyone knows; the settlements run from Rakhigarhi in Haryana to Lothal and Dholavira in Gujarat.

Dholavira is worth dwelling on, because it shows what these communities could do with a site that should not have supported a city at all. On an arid island in the Rann of Kutch, the excavated town is walled, with its burial ground outside it, and the work that kept it alive was hydraulic — reservoirs, channels and drains built to take a season's rain and hold it through the months with none. UNESCO, inscribing the site in 2021, puts occupation at roughly 3000–1500 BCE.

What the evidence shows, across the civilisation: planned cities on a grid, standardised fired brick, covered drainage, sophisticated water management, bead-making and metalwork at scale, standard weights, long-distance trade with Mesopotamia and the Gulf, and a writing system.

What it does not show: any structure unambiguously identifiable as a temple or a palace, royal burials, or unambiguous evidence of an army. That absence is genuinely striking and is argued about rather than explained.

The decline was gradual and its causes are debated — shifts in the monsoon, changes in river courses, and the failure of the trade networks are all proposed, and they are not exclusive.

The script, and what it costs us

Around four hundred signs survive, almost all of them on seals and other short inscriptions — usually fewer than five signs at a time. There is no bilingual text to work from and the underlying language is unknown. Some researchers have argued that the sequences may not encode language at all.

The consequence is easy to state and hard to absorb: no Harappan text can be read. Nothing survives in which these people describe their own politics, their own beliefs, or who they thought they were. Everything said about them is inferred from what they left behind.

The Vedic period

The Ṛgveda describes a pastoral, mobile, cattle-centred society organised in clans, with ritual at the centre of its literature. That society is materially unlike the Indus cities, and the texts contain no memory of urban life.

Two things are worth saying about the period before anything else, because both are commonly assumed and neither is supported. It was not a single uniform civilisation that suddenly covered India — other communities and material cultures continued across the subcontinent throughout, and Vedic culture itself changed considerably over the centuries it spans. And it is not simply the origin of modern Hindu practice. The Vedas became foundational to later Hindu traditions, but the traditions kept developing for two thousand years after them, through the Upanishads, the epics, the philosophical schools, devotional movements and temple worship.

Over the following centuries the material record changes: iron, settled agriculture in the Ganga plain, larger polities, and by the middle of the first millennium BCE the mahājanapadas — sixteen substantial states — and the cities that go with them.

The century of arguments

That same period is one of the most consequential in Indian intellectual history, and it is not well described as the beginning of two religions.

What the sources show is a crowded field of teachers and renunciants arguing in public about the same set of questions: what the self is, whether action carries consequence beyond death, what knowledge is worth having, and whether ritual does anything. The Upanishads belong to this world. So does the Buddha, whose teaching set aside the idea that liberation depended on Vedic sacrifice, and Mahavira, whose Jain tradition placed an extraordinary weight on ahiṃsā. So do several teachers whose schools did not survive and who are known mainly from the polemics of those who disagreed with them.

The pattern established here recurs for the rest of Indian history: traditions living beside one another, arguing with one another, borrowing from one another, and competing for royal and popular support.

Maurya, and then Gupta

The Mauryas, from roughly 322 BCE, built the first empire to cover most of the subcontinent — though not the far south, and Ashoka's own edicts name the Cholas, Pandyas and Keralaputras as neighbours beyond his borders. His inscriptions, on rock and pillar in several languages and scripts, are the earliest Indian writing that can be read with confidence, and they record a state apparatus and a ruler's public argument about ethics after a brutal war.

Whether anything earlier can be read is a live question: excavations at Keezhadi and Porunthal in Tamil Nadu have been argued to push Tamil-Brahmi before Ashoka, and the dating is disputed.

The Guptas, from roughly 320 CE, presided over the period conventionally called classical: Kālidāsa's poetry, Āryabhaṭa's astronomy, the maturing of temple architecture, and the decimal place-value system that reached the rest of the world through Arabic transmission.

None of this happened in isolation

Ancient India was connected outward for the whole of the period on this page.

Harappan materials and finished goods reached Mesopotamia and the Gulf in the third millennium BCE. In the early centuries CE, merchants crossed the Arabian Sea and the Bay of Bengal, overland routes ran to Central Asia, and Indian ports traded with the Roman Mediterranean — attested from both ends, in classical accounts and in Tamil poetry. Buddhism travelled out of India to Sri Lanka and across much of Asia. Sanskrit became a language of learning and political culture well beyond the subcontinent.

The traffic ran both ways, and it did not make either end a copy of the other. Connection without uniformity is the pattern, and it holds for the rest of Indian history.

For where this section's dates come from, how each kind of evidence is read, and what happened after 600 CE, see Indian history and civilization.