Nan Madol

Nan Madol is a group of artificial islands built on a shallow coral reef next to Temwen Island, on the eastern side of Pohnpei, one of the high volcanic islands of the Caroline Islands in the western Pacific. The site core is enclosed by stone walls and measures about 1.5 kilometres long and half a kilometre wide. Inside are 92 islets, each a platform of stacked stone filled with coral rubble, separated by tidal channels. Archaeologists count the wider archaeological district, including nearby shores and outlying islets, at more than 18 square kilometres.

The Pohnpeian name Nan Madol is usually translated as “within the intervals” or “spaces between,” a reference to the canals that criss-cross the complex. An older name, Soun Nan-leng, “Reef of Heaven,” is also recorded. Visitors have called it the “Venice of the Pacific,” but that comparison can mislead: Nan Madol was not a busy trading city full of ordinary households. It was an exclusive centre for rulers, priests and the dead.

Nan Madol was the seat of the Saudeleur, a line of rulers who, according to Pohnpeian oral history, were the first to unite the whole island under one authority. Archaeology shows people were already active on this reef in the early centuries AD, that islet building was under way by around AD 900, and that the characteristic walls of basalt columns began to rise around AD 1180–1200. Pohnpeian tradition places the fall of the Saudeleur, and the beginning of the system of chiefs that still exists on Pohnpei today, at the hands of the hero Isokelekel.

The easiest way to understand Nan Madol is to imagine it in two halves. The northeastern half is called Madol Powe. It contains 58 islets, many of which held the houses of priests, as well as the most impressive tombs. The southwestern half is called Madol Pah, and it was the area of the chiefs and those who served them, including the ruler’s residence on the islet of Pahnkedira. Early twentieth-century German researchers, working with Pohnpeian informants, divided the site into three sectors—the tomb wall, the royal area and the priests’ area—and later archaeological work has broadly supported this plan.

The most famous islet is Nandauwas, the royal mortuary islet. Walls of basalt columns between about 5.5 and 7.5 metres high surround a courtyard, and inside is a central tomb enclosure. Tradition says it was built for the first Saudeleur. Walking into Nandauwas you pass through an outer wall, then an inner wall, then reach the tomb: the architecture itself tells you that you are moving from the ordinary world towards the most restricted and sacred space.

Other islets had particular jobs. Pohnpeian tradition and archaeology connect Dapahu with canoe building and Peinering with making coconut oil. On the islet of Ideht, priests prepared sacred turtles as offerings, and a small enclosure held a sacred eel that acted as a messenger between people and a god. Food brought to the ruler was received at a special islet—first Peiniot, later the closer Usennamw.

Look closely at the walls and you will see the building method. Long basalt columns are laid in alternating layers: one layer runs along the wall, the next runs into it, like a log cabin. This “header and stretcher” pattern locks the stones together. The spaces behind are filled with coral and smaller stones.

Pohnpei is volcanic. In some places, thick lava cooled slowly and cracked into long, many-sided pillars called columnar basalt. There are no sources of columnar basalt right next to Nan Madol, so every column had to be brought from the main island. Geochemical tests on building stone, published in 2016, identified basalt from a volcanic plug on the opposite side of Pohnpei, showing that at least some of the stone travelled a long way.

Many of the columns weigh between one and five tonnes, and some boulders in the seawall behind Nandauwas are estimated at up to fifty tonnes. The builders had no draft animals, metal tools or wheels. Early twentieth-century Pohnpeian informants told the German ethnographer Paul Hambruch that heavy stones were moved up inclined tree trunks, especially coconut palms, using levers and ropes of hibiscus fibre. Most archaeologists think the stones were floated to the reef on rafts, perhaps at high tide.

Pohnpeian legend gives a different explanation. It says the stones were flown into place by magic. Many peoples tell such stories about their greatest monuments; for Pohnpeians the legend expresses the sacred power of the founders. Archaeologists, for their part, emphasise the enormous organisation of human labour: planning, quarrying, transport and construction carried out over several centuries by people from every part of the island.

Building did not happen all at once. Radiocarbon dates suggest that low platforms were being built on the reef from around AD 900, and that true megalithic building—the huge basalt walls—began around AD 1180–1200 and continued in phases for centuries. What we see today is the result of many generations of work, repair and addition.

Pohnpeians divide their island’s history into periods: an age of people (or of building), the age of the Saudeleur, the age of the Nahnmwarki, and the age of foreigners. The name Saudeleur is usually explained as “Lord of Deleur,” Deleur being an old name for Pohnpei or part of it. The Saudeleur period is remembered as the first time a single ruler governed all of Pohnpei, whose population has been estimated at around 25,000 at its height.

In the founding story, two brothers with sacred powers, Olisihpa and Olosohpa, arrived by canoe from a place called Western Katau, looking for a site to build an altar to the god of agriculture. After several failed attempts elsewhere, they succeeded on the reef off Temwen Island. When Olisihpa died, Olosohpa became the first Saudeleur. Tradition says he married a local woman and that a line of rulers followed, though accounts differ about how many there were—historian David Hanlon noted counts ranging from eight to seventeen.

According to oral history, the Saudeleur owned the land and everything on it. Chiefs in each district organised tribute, and ordinary people brought breadfruit in the season of plenty and yams, taro and preserved breadfruit in the lean season, as well as fish and other seafood. The island was divided into three districts, called wehi, a word that also means “turtle”: Madolenihmw in the east, Kitti in the west and Sokehs in the north. At Nan Madol, officials held titles tied to particular tasks, such as chief adviser, food preparer and guards of entrances and dwellings.

Religion was at the heart of Saudeleur power. The main cult at Nan Madol honoured the thunder god Nahn Sapwe, a Pohnpeian deity from whom the rulers claimed legitimacy. Priests led offerings, divination and funerals, and a ritual of turtle sacrifice and a sacred eel linked the people with the gods. Some traditions say the Saudeleur also tried, without much success, to impose worship of their own god on Pohnpeians.

Nan Madol has no fresh water and no farmland. Every basket of food and every container of water had to be brought by canoe from the main island. This is one of the strongest clues to the site’s purpose: it was deliberately set apart from the ordinary world, and it could only function if the rest of Pohnpei supported it.

Archaeologists estimate that no more than about 1,000 people lived there at any one time, and perhaps fewer than half that. Some were chiefs and priests, but many were commoners who served them. One widely discussed interpretation holds that requiring rival chiefs to live at Nan Madol, rather than in their home districts, allowed the Saudeleur to keep a close watch on potential challengers—an arrangement often compared to the way some later rulers elsewhere in the world kept nobles at court.

Archaeologists Joyce Bath and J. Stephen Athens argued in 1990 that words like “city” or “town” can give the wrong impression. Their excavations found no evidence that Nan Madol was a centre of craft production or a place where tribute was stored and redistributed. Instead, they described it as a place where the elite were physically separated from the people they ruled: a political, religious and above all mortuary centre.

Evidence on the islets supports what oral history says. Utilitarian tools such as adzes and grinding stones are concentrated in the priests’ sector, while large stones used for pounding sakau (kava) for ceremonies are found mainly on the tomb islets and on Pahnkedira. On Ideht, a mound of burned coral cooking stones more than two metres high, containing turtle bone, matches traditions of the turtle ceremony, and a small basalt enclosure matches the pool of the sacred eel.

Pohnpeian oral history remembers the later Saudeleur as increasingly harsh. Stories describe rulers who taxed people until they starved, and others notorious for greed and cruelty; there are accounts of at least two assassinations. Scholars read these stories as moral history—as ways of explaining why centralised rule was rejected and why the later system was better—rather than as literal records of each ruler’s deeds.

The fall of the dynasty is told through the story of Isokelekel. In most versions, the last Saudeleur offended the thunder god Nahn Sapwe, who fled to Katau, often identified with Kosrae, an island to the east. There Nahn Sapwe fathered Isokelekel, who grew up knowing his destiny. He sailed to Pohnpei with 333 companions. More than a dozen published versions tell the story of the war differently: in some, a quarrel between children sparks the fighting; in others, Isokelekel wins over oppressed locals, or is helped by a woman of the ruling family. In the end, the Saudeleur retreats inland, where, tradition says, he turned into a fish in a stream.

Isokelekel took the title Nahnmwarki and at first lived at Nan Madol. He divided the island into autonomous chiefdoms, each with its own Nahnmwarki, and a parallel line of chiefs, the Nahnken, was established. This dual-chief system, in which each chiefdom has two ranked lines of titles, remains an important part of Pohnpeian society today; Pohnpei now has five traditional chiefdoms.

Why was Nan Madol eventually abandoned? One explanation is practical: the Nahnmwarki at Nan Madol no longer received tribute from the whole island and had to fetch their own water and grow their own food, so they moved back to their districts. Another points to a sharp decline in population. One Pohnpeian account says a typhoon and famine led to its abandonment in the time of the fifth Nahnmwarki of Madolenihmw. By the time Western visitors arrived in the nineteenth century, Nan Madol was no longer a residential centre, though some ceremonies continued there into the 1800s.

Europeans first described Pohnpei in some detail in 1828, when members of the Russian expedition of Fyodor Litke visited; F. H. von Kittlitz wrote about the offshore ruins. In 1836 a ship’s surgeon named Campbell, visiting on the British cutter Lambton, compared the ruins’ antiquity to that of the Pyramids but concluded that Pohnpeians knew nothing of their origin. As historian David Hanlon points out, what Campbell met was not ignorance but language barriers and a culturally prescribed silence about sacred knowledge. The Polish naturalist John Stanislaw Kubary made the first detailed description of Nan Madol in 1874, the British anthropologist F. W. Christian investigated it in 1895, and in the early twentieth century the German ethnographer Paul Hambruch mapped the complex and recorded what Pohnpeian informants told him.

Hanlon shows that nineteenth-century visitors repeatedly exaggerated the site’s size and read it through the racial theories of their day, and that many refused to believe that Pacific Islanders could have built something so large. Nan Madol became the setting for A. Merritt’s adventure novel The Moon Pool (1918), and it later inspired stories in the Cthulhu Mythos. In 1926 the writer James Churchward claimed it as a remnant of an imaginary sunken continent called Mu. There is no scientific basis for these claims.

Such ideas continue today. When a 2022 television series presented false claims about Nan Madol’s age, experts in Pacific archaeology and geography criticised it as insulting to the Pohnpeian ancestors who built the site and connected it to nineteenth-century racist ideas. Underwater surveys have also tested rumours of drowned cities offshore: teams in 1988–89 and in 2013 found that supposed submerged “pillars” were natural coral growths or fallen building stones, and that a nearby blue hole is a natural sinkhole in the reef.

The lesson is important. Nan Madol is not a mystery left by vanished strangers. It is the achievement of Pohnpeian people, whose descendants still live on the island, still hold the chiefly titles that grew out of Nan Madol’s history, and still tell its stories.

Nan Madol was listed on the U.S. National Register of Historic Places in 1974 and designated a National Historic Landmark in 1985, during the period when Pohnpei was part of the U.S.-administered Trust Territory of the Pacific Islands. Objects from the site were displayed at the Lidorkini Museum until it closed in 2012.

In 2016 UNESCO inscribed “Nan Madol: Ceremonial Centre of Eastern Micronesia” on the World Heritage List under criteria (i), (iii), (iv) and (vi)—recognising it as a masterpiece of human creativity, a testimony to a vanished cultural tradition, an outstanding type of architecture, and a place tied to living traditions and beliefs. In the same year it was placed on the List of World Heritage in Danger.

The threat cited for the Danger listing is the continuing siltation of the waterways, which encourages vegetation to overgrow the site and undermines the existing structures. In other words, the canals that once defined Nan Madol are slowly filling in, and plant growth is pushing into walls that were built without mortar. Conservation therefore involves not only repairing stonework but also managing water, sediment and vegetation across the whole complex.

Establishing Nan Madol’s chronology has been one of the central tasks of Pohnpeian archaeology since the University of Oregon’s Ponape Archaeology Survey, directed by William S. Ayres, began work in 1977. Radiocarbon samples from islet fill indicate occupation of beaches, sand bars or perhaps stilt houses over the reef from around AD 1, with human activity dated to about AD 80–200. Low platform construction began by about AD 900. Ayres’s excavations on Pahnkedira identified two major building phases—around AD 900–1100, perhaps without true megalithic construction, and around AD 1300–1500, with it.

In 1990 Bath and Athens argued that if megalithic construction and the turtle ceremony on Ideht define the Saudeleur era, it began around AD 1200. Smithsonian excavations in 1965 had dated the Ideht oven mound to about AD 1200–1400; their own dates confirmed a start near 1200 but suggested an end in the 1500–1600 range. A decisive refinement came in 2016, when Mark McCoy and colleagues used high-precision uranium–thorium (230Th/U) dating on coral from the tomb of the first island-wide chief on Nandauwas, together with geochemical sourcing of the basalt. They placed the beginning of monument building at Nan Madol at AD 1180–1200, earlier than monumental construction on most other Pacific islands, which followed over AD 1300–1600.

The familiar end date of about 1628 needs careful handling. It is not a directly observed event. Athens calculated it in 1984 from an estimated average reign length applied to a presumably complete list of 22 Nahnmwarki of Madolenihmw. Oral accounts generally place the fall of the Saudeleur earlier, in the 1500s. The 1628 figure is therefore an informed estimate, and the transition may have been more gradual than a single conquest.

The dating also bears on the relationship between Nan Madol and Leluh (Lelu) on Kosrae, the other great basalt-walled complex in eastern Micronesia. Pohnpeian tradition connects Nan Madol’s builders with Kosrae, and the Isokelekel story links Kosrae to the fall of the Saudeleur. The radiocarbon evidence indicates that Nan Madol predates Leluh, which suggests that influence flowed—at least in part—from Pohnpei towards Kosrae rather than the other way.

Nan Madol is a classic case in the debate over how archaeologists should use oral tradition. Much of what is known about the Saudeleur comes from Pohnpeian sources, including the Book of Luelen, written over many years by the Pohnpeian Luelen Bernart and published in English translation in 1977, and an unpublished account by his grandson Masao Hadley, along with records made by ethnographers such as Hambruch and Saul Riesenberg. Bath and Athens noted that Pohnpeian custom discourages sharing esoteric knowledge with outsiders, which makes these detailed accounts especially valuable.

Bath and Athens tested nine archaeological implications of the oral accounts. Several were supported: the site’s location; a division between priestly, chiefly and mortuary sectors; turtle-cooking remains and an eel enclosure on Ideht; a U-shaped meeting house on Palakapw; and subsidiary centres built in the same columnar-basalt style inland at Sapwtakai in Kitti and in Madolenihmw. One was not: the tradition that Nan Madol was built within the lifetimes of the two founding brothers. Archaeology shows a long, gradual history of occupation that began well before the Saudeleur era.

Anthropologist Suzanne Falgout, who recorded oral traditions in the Wene area of Pohnpei, argued that the question is not simply whether oral traditions are “true.” For Pohnpeians, important knowledge is tied to persons and titles, is highly restricted, and is carefully managed and edited in transmission; slightly different versions are expected. Understanding who holds knowledge, and why it is shared, is part of evaluating it. Her fieldwork, guided by knowledgeable Pohnpeians, led to the recording of dozens of previously unknown sites.

Historian David Hanlon, in a 1990 essay on early Euro-American “visions” of Nan Madol, showed how nineteenth-century observers—ship’s surgeons, beachcombers, naturalists and anthropologists—produced inflated measurements, neat but inaccurate maps, and theories of vanished races that said more about their own times than about Pohnpei. He argued that archaeologists should recognise how the concepts they bring shape what they see. Together, these studies reflect a wider shift in Pacific archaeology: away from treating Nan Madol as an enigma and towards reading it through Pohnpeian history, language and social organisation, with Pohnpeians as authorities on their own past.

Pohnpei’s oral history describes something that social theorists find unusual: an island-wide centralised polity (the Saudeleur) that was replaced not by a larger state but by several smaller autonomous chiefdoms (the Nahnmwarki system). Bath and Athens called this the “Saudeleur to Nahnmwarki transformation” and asked whether it represented a genuine decline in social complexity or a reorganisation. They examined models by Jonathan Friedman, who linked the power of small theocracies to elite control of prestige goods, and by Ross Cordy, who related the number of social strata to population and territory.

Archaeologically, a decentralising society might be expected to show abandoned centres, the end of elite funerary monuments and changes in craft organisation. Nan Madol’s eventual abandonment, the construction of upland centres such as Sapwtakai (occupied roughly AD 1325–1700), and the development of meeting houses (nahs) in the Nahnmwarki period all bear on this question. Bath and Athens also warned that the Saudeleur state may not have been as tightly centralised in practice as later tradition suggests; the western district of Kitti, for example, had a reputation for defying eastern authority.

In a broader comparative frame, McCoy and colleagues used Nan Madol to pinpoint when monument building—usually taken as a sign of a shift in authority—first appeared in the remote Pacific. Its early date makes Nan Madol a key reference point for studying the rise of chiefly power across Oceania.

Scholars disagree about many details: how many Saudeleur reigned, how centralised their rule really was, whether Isokelekel’s conquest affected the whole island or mainly Madolenihmw (as the anthropologist Petersen proposed in 1990), and what finally ended occupation at Nan Madol. These open questions are part of what makes the site so important for understanding how Pacific societies built, and rebuilt, their political worlds.