Coral houses dot the Pacific, and researchers now know when they were built.

Source: French to English Tester   Published on: 2026-03-29

Source: The Conversation – in French– By James L. Flexner, Senior Lecturer in Historical Archaeology and Heritage, University of Sydney

In French Polynesia, dozens of houses built with coral testify to a profound upheaval caused by the arrival of missionaries in the 19th century. Thanks to a very precise dating method, archaeologists are now able to trace for the first time the timeline of their construction.

---

The Mangareva Islands(Maʻareva)are located approximately 1,600 kilometers southeast of Tahiti, in French Polynesia. They derive their name — which means “floating mountains” — from the effect produced by the spray breaking on the surrounding coral atolls, or motu: the ancient volcanic peaks then give the impression of floating above the waves.

Today, the islands are home to about 2,000 inhabitants, many of whom work in pearl farms established in the turquoise lagoon. Across the islands also remain the remains of dozens of remarkable buildings: houses built from coral.

As part of a broader project dedicated to the transformations of daily life in Mangareva in the 19th century, my research team in archaeology has documented dozens of these coral houses, notably on the islands of Aukena, Akamaru, Mangareva, and Taravai.

Today, in a newarticlepublished in the journal Antiquity, we established the first precise chronology of the construction of these coral houses.

These results highlight new dynamics in the way Pacific societies transformed their built environment after contact with Europeans—and show how this colonial legacy continues to shape local life today.

Colonization has transformed the lives of communities in the Pacific

French Catholic missionaries established a post in Mangareva starting in 1834. In addition to learning prayers, attending religious services, and reading the Bible, the inhabitants of Mangareva profoundly changed their daily lives through their contact. Among the many changes was a complete transformation of domestic spaces.

In a few decades, traditional wooden and thatched houses were replaced by a new type of small stone houses. Missionaries often recorded precise dates for their constructions, particularly for the Rikitea cathedral, the churches spread throughout the archipelago, and the main Catholic schools.

On the other hand, for the most numerous category of buildings from this period — houses — we generally have no information about their construction date, the people who built them, or the individuals who lived in them.

A precise dating method

During fieldwork in October 2024, I noticed that one of the coral blocks fallen from the wall of the ruined house we were excavating contained branching corals that looked very fresh, almost as if they had just been cut from the living reef.

We used an advanced technique called uranium-thorium (U-Th) dating to determine the age of these branching corals — and the structures built from them.

Unlike radiocarbon dating, which is more well-known and has margins of error measured in decades, U-Th dating is extremely precise: it allows determination of the date when corals died, to within a few years, leaving behind only their hard exoskeleton.

Unlike radiocarbon dating, which is not very reliable for materials less than about 400 years old, U-Th dating works for much more recent periods, even almost up to the present day.

We took a “control” sample from a building whose date is known — the boys’ school of Aukena built in the 1850s — as well as samples from eight other houses and a coral watchtower.

We also collected a branching coral from a pit layer located in the same house where I had first noticed these coral branches with a “fresh” appearance in the blocks.

At the time, we thought that this pit contained the remains of a feast organized just before the construction of the house. The overlapping dates in our U-Th results confirmed this hypothesis.

The mysteries of the “old coral”

Afteranalysis of samples, we were surprised to find that several dates did not match what we had planned.

Some corals seemed to have died before the 1830s, when the missionaries arrived. Some even dated back to a period before contact with Europeans in the 1790s.

A similar problem is well known in radiocarbon dating: the issue of the “old wood“where the date of death of an organism can precede by several decades, or even centuries, the event that the archaeologist is trying to date. Are we facing here a problem of “old coral”?

Two explanations are possible.

An archaeologist who visited Mangareva in the 1930s hasreportedheaps of coral debris that he thought were the remains of marae, sacred structures once toppled during the missionary period. This opens up the possibility that this ancient coral was reused for new constructions.

Another possible explanation for this type of coral, belonging to the scientific genusAcropora, is that some branches die at a distance from the active growth zone of the reef over the years or decades, while retaining their “fresh” appearance.

This scenario is perhaps the most likely, as our “too old” dates only precede the expected chronology by a few years or decades, and not by several centuries. But we cannot completely rule out the hypothesis of the marae either.

We still have much to learn about how populations used coral to build their structures in the past — and perhaps also about how coral reefs regenerated, or did not, after decades of human exploitation. This last point could be important to consider more carefully our own relationships with coral reefs today.

James L. Flexner is an associate professor of historical archaeology and heritage at the University of Sydney. This research is funded by the Australian Research Council (FT210100244).

–ref. Coral houses dot the Pacific and researchers now know when they were built. –https://theconversation.com/coral-homes-dot-the-pacific-and-researchers-now-know-when-they-were-built-279467