Lost Worlds · Mahabalipuram (Mamallapuram), Tamil Nadu, India

Mahabalipuram's 'Seven Pagodas'

A surviving Pallava temple on the shore, a persistent legend of six lost siblings beneath the waves — and structures that the 2004 tsunami briefly laid bare.

Mainstream: c. AD 700–728 (Pallava dynasty; Shore Temple and monuments)Alternative: A far older drowned complex proposed by some, though no firm early date has been demonstrated offshore12.62°, 80.19°

Lost Worlds Myth Belief

At a glance

Mahabalipuram's 'Seven Pagodas' — Photo: Rangan Datta Wiki · CC BY-SA 4.0

Mahabalipuram, on the Coromandel Coast south of Chennai, is a UNESCO World Heritage site famous for its rock-cut caves, monolithic 'rathas', the giant relief known as Arjuna's Penance, and the sea-facing Shore Temple built under the Pallava kings around the early eighth century AD. For centuries European sailors called it the 'Land of the Seven Pagodas', and a durable local tradition holds that the Shore Temple is the last of seven such temples, the other six having been swallowed by the sea. In December 2004 the Indian Ocean tsunami briefly pulled the water back from the beach, exposing rows of stones, carvings and what looked like walls on the seabed — turning an old legend into a live archaeological question.

See it on the globe →

The mainstream view

What archaeology says

Archaeologists treat the visible monuments as securely Pallava, carved and built in the seventh and eighth centuries AD under rulers such as Narasimhavarman I 'Mamalla' and Rajasimha, with inscriptions, iconography and architectural style all consistent. The 'seven pagodas' tradition is generally read as a real memory of coastal erosion and shoreline change, not as a claim of great antiquity: this is an active, retreating coast, and structures once inland can end up in the surf zone within centuries.

Serious underwater work has been done here. In April 2002 a joint expedition by India's National Institute of Oceanography (NIO) and Britain's Scientific Exploration Society located masonry, walls and scattered dressed blocks under five to eight metres of water several hundred metres offshore, consistent with submerged temple-related structures. After the 2004 tsunami exposed carved stones and a long wall on the beach, the Archaeological Survey of India (ASI) and the Indian Navy conducted sonar and diving surveys in 2005 and identified remains of further structures within about 500 metres of the shore — including a wall roughly 70 metres long.

The mainstream reading is that these are genuine but broadly medieval remains: outlying shrines, gopurams or ancillary buildings of the Pallava-and-later temple town, lost to a combination of erosion, subsidence and storm action over the last thousand-odd years. In this view Mahabalipuram is a real drowned townscape, but one measured in centuries rather than millennia, and firmly within the historical Tamil past.

Key evidence cited

Visible monuments securely dated to the Pallava period by inscriptions and style (c. AD 700–728)
2002 NIO / Scientific Exploration Society survey locating submerged walls and dressed blocks offshore
2005 ASI and Indian Navy sonar/diving surveys identifying a ~70 m wall and further structures near shore
An actively eroding, subsiding coast that readily moves inland structures into the surf zone
Carved stones exposed by the 2004 tsunami that are stylistically Pallava or later, not archaic

The alternative view

What the skeptics propose

Alternative writers, above all Graham Hancock, have championed Mahabalipuram as evidence that flood legends preserve memories far older than orthodox chronology allows. Hancock featured the site in his book Underworld and in television work, arguing that the 'seven pagodas' story, the tsunami exposures and the NIO discoveries together hint at structures that could predate the Pallavas by a very long way — part of his broader thesis of a lost maritime culture drowned by the post-Ice-Age sea-level rise.

Proponents stress that the offshore remains are real and were partly confirmed by exactly the mainstream institutions (NIO, ASI, the Navy) that are usually sceptical of fringe claims, and that local fishermen had long described 'temples' beneath the water before divers looked. They point to the tsunami's dramatic unveiling of carved stone as vindication of the oral tradition, and argue that the coast should be surveyed far more aggressively and to greater depth before assuming everything is medieval.

The weaker part of the alternative case is dating: no offshore structure at Mahabalipuram has been shown by any dating method to be older than the historical Tamil period, and the exposed carvings that can be studied are stylistically Pallava or later. Hancock himself frames the site as suggestive rather than proven. The honest position is that Mahabalipuram offers a vivid, well-documented example of a real submerged townscape matching a real legend — while the leap to Ice Age antiquity remains, so far, unsupported by the evidence recovered.

Key evidence cited

The durable 'Seven Pagodas' tradition of six temples lost to the sea
Long-standing fishermen's reports of 'temples' on the seabed, later partly confirmed by divers
The 2004 tsunami dramatically exposing carved stone and a long wall on the seabed
Offshore remains confirmed by mainstream bodies (NIO, ASI, Navy), showing the legend has a real basis
Hancock's argument that the coast has not been surveyed deeply enough to rule out older structures

Genuinely open questions

How much of the offshore complex is medieval temple-town outbuildings, and could any of it be older?
How far out and how deep do the submerged structures extend beyond the surveyed 500-metre zone?
Can the 'seven pagodas' number be tied to specific drowned buildings, or is it a symbolic tradition?

Worth knowing

In the hours before the 2004 tsunami struck, the sea drew back hundreds of metres and briefly revealed carved stones and a long wall on the exposed seabed — witnesses effectively watched a piece of the 'Seven Pagodas' legend surface before the water rushed back in.