Seven years, one empire.

The Brihadeeswara temple was begun in 1003 CE, in the eighteenth year of Raja Raja Chola's reign, and consecrated in 1010, in the twenty-fifth. The dates are not approximate. They are recorded on the plinth of the temple itself, alongside the names of the donors, the dancers, the accountants, the goldsmiths and the priests. The building is its own document.

Seven years is, by Indian temple standards, fast. The Khajuraho group took roughly a century. The Hoysala temples at Belur and Halebid took decades each. Brihadeeswara, larger than any of them, was finished inside a single regnal lifetime — and the documentary record of how it was financed has survived almost intact.

The commission of 1003.

The dedicatory inscription names the temple Rajarajeswaram — “the temple of Rajaraja's lord”, a possessive formulation that fuses the king and the deity in a way that is not coincidental. The commission almost certainly post-dates the conquest of Sri Lanka, completed in 993, and the consolidation of the Chera and Pandya territories by 1000. The temple is, in chronological terms, the architectural ratification of an empire-builder's mid-career.

The political logic is straightforward enough. By 1003 the Chola court was wealthy, the empire stable, the dynasty's religious affiliations clear: Raja Raja was a Saiva, of the orthodox vaidika tradition, with a personal devotion to Shiva that the inscriptions document with unusual specificity. The temple was a state project, financed from the treasury, staffed from the imperial rolls — but it was also, plainly, a personal one.

Quarry and transport.

The Thanjavur plain has no native granite. The bedrock lies tens of metres below the alluvium of the Kaveri. The stone for the temple — by modern estimate some 130,000 tonnes of granite — had to be brought from quarries in the rocky country to the north-west, somewhere in the catchment of the Kollidam and Vellar rivers. The most commonly cited site is Sittannavasal, roughly 80 km away.

Transport almost certainly used the river system for the long leg and elephant-drawn rollers for the final approach. The blocks at Brihadeeswara are large — the largest in the plinth course are over five metres long and weigh upwards of forty tonnes — which rules out animal-only transport over land. Floating barges down the Kollidam to the Kaveri and into a dedicated canal at Thanjavur is the only feasible reconstruction.

The river-sand foundation.

The foundation system used at Brihadeeswara is, structurally, more impressive than the tower above it. There is no excavated footing into bedrock. The Chola engineers laid a broad, shallow raft of interlocking granite slabs directly onto compacted river sand, and built the temple on the raft. The whole assembly, vimana included, behaves like an inverted ship's hull — distributing the load across a wide area and settling uniformly with the floodplain.

This is now the textbook explanation for why the temple has not failed structurally in a thousand years. The raft settles; it does not crack. Pichard's 1995 survey measured a uniform 60 cm subsidence across the footprint, with no differential settlement above four centimetres. For a 43,000-tonne structure on sand, that is an extraordinary result.

Raising the vimana.

The 66-metre tower was raised in dry-jointed granite — no mortar — with each course keyed into the one below by interlocking shoulders cut into the stone. The walls thin as the tower rises: 2.5 m at the base, around 1.2 m at the eighth tier, perhaps 70 cm in the upper storeys. The hollow interior is partially filled with structural rubble; the upper chambers carry only their own dead weight and the kalasam above them.

For lifting, the consensus reconstruction is an earthen ramp. The village of Sarapallam, six kilometres north-west of the temple, is locally said to be the borrow pit from which the ramp's fill was taken; the name itself derives from sarapali, “scaffolding hollow”. A 1:60 grade from Sarapallam would arrive at the right height for the kalasam, which is approximately the structural maximum a clay-bound earthen ramp can support without internal collapse.

“Of all the medieval Indian temple projects we can document, only Brihadeeswara gives us the schedule, the budget, the personnel and the consecration date with a precision that approaches the modern.”— Pierre Pichard, Tanjavur Brhadisvara: An Architectural Study, 1995

The consecration of 1010.

The consecration is dated to the twenty-fifth regnal year of Raja Raja, a date the inscriptions fix to early 1010 CE — most commonly given as 25 January 1010, though scholars differ on the precise day. The ceremony, an Agamic kumbhabhishekam, would have lasted several days and required the participation of hundreds of priests drawn from across the empire.

On the day of consecration the temple already had a fully constituted establishment: 400 named devaradiyars (temple dancers), 212 musicians, scores of priests, an administrative staff and a treasury endowed in perpetuity. The inscriptions also list substantial donations of gold and silver liturgical equipment from the king, his consorts and his courtiers. It was, in modern terms, a fully institutionally formed enterprise on its opening day.

What the inscriptions say.

The 107 surviving inscriptions on the plinth are, collectively, the most detailed documentary record of any medieval Indian building. They list, by name and home village: the dancers, the musicians, the singers, the temple servants, the accountants, the priests, the supplier-villages, the endowed lands and the precise quantities of paddy, oil and ghee due each day to support the worship.

The total endowed paddy — counted in kalams, a Chola dry-measure of about 8.5 kg — is in the order of 70,000 kalams a year, or roughly 600 tonnes. The grain came in from a long list of named villages across what is now Tamil Nadu and Sri Lanka. The documentary footprint is national.

After the consecration.

Raja Raja died four years after the consecration, in 1014. His son Rajendra Chola I succeeded him, having been co-regent since 1012, and within a generation had built his own answer — Gangaikonda Cholapuram, dedicated 1035, deliberately the same scale and deliberately slightly different. The Brihadeeswara establishment continued, fully endowed, through the entire Chola period and was carried over without interruption by every successor regime.

It is, in fact, the only medieval Indian temple that has been continuously functioning since its consecration, with the same six daily poojas, the same Saiva Agamic ritual order and the same lingam in the sanctum, for over a thousand years.

Further reading

• Pierre Pichard, Tanjavur Brhadisvara: An Architectural Study, IFP/EFEO, 1995 — the definitive architectural survey.
• K. A. Nilakanta Sastri, The Cōḷas, University of Madras, 1955 — the surviving standard reference.
• Y. Subbarayalu, South India under the Cholas, Oxford, 2012 — the inscription-led economic reading.
• South Indian Inscriptions, Volume II — the published corpus of Brihadeeswara plinth inscriptions.