A study published by Cambridge University Press shows that coral used in 19th-century buildings can be dated directly, offering a clearer timeline for construction in parts of the Pacific where written records are limited.
The research, led by James Flexner, focuses on the Mangareva Islands. By applying uranium-thorium (U-Th) dating to coral blocks, the study tests whether building materials themselves can provide reliable dates for structures linked to early missionary activity.
Dating sites from the colonial period have often depended on documents or artefacts such as ceramics and coins. These can give rough timeframes, but they are not always precise. Objects might arrive years after they were made, especially in remote regions, and some sites have little or no associated material. Scientific dating methods have limits as well. Radiocarbon dating is less effective for more recent centuries, and timber suitable for tree-ring analysis is rarely preserved in tropical climates.
Coral offers a different option. The U-Th method measures radioactive decay to determine when a coral organism died. Because coral was widely used as a building material across Polynesia, this makes it possible to estimate when stone structures were put together, at least within a certain range.
Flexner’s team collected samples from nine buildings, including houses, a watchtower and a school. The school, which has known construction dates from historical records, was used as a reference point. Laboratory analysis produced dates that were generally close to the expected period, often within a few years.
Most of the buildings examined belong to a type of stone house known locally as ’are po’atu. These small structures became common after the arrival of French Catholic missionaries in the 1830s. Missionaries introduced new building techniques and trained local converts to construct in stone, changing settlement patterns across the islands.
Several samples returned dates from the 1830s and 1840s, matching the early phase of missionary construction. One case stood out: coral taken from a house and from a nearby pit feature produced similar mid-1840s dates. The pit contained food remains, glass and metal objects, suggesting it may be linked to a single event, possibly a gathering or feast, followed by the building of the house.
Not all results were as straightforward. Some coral samples gave dates that appeared too early, even before European contact with the islands. The study links this to how materials were collected and reused.
Coral blocks used in construction were not always freshly cut.
In some cases, builders may have taken material from reef areas where coral had already died years earlier. There is also evidence that stone from older sites, including traditional ceremonial platforms known as marae, may have been reused. This creates what archaeologists call “inbuilt age,” where the material predates the structure it was used in.
Because of this, the dates do not always match the exact year a building was constructed. Instead, they provide a point after which construction must have taken place. Researchers say this still makes the method useful, especially when combined with other forms of evidence.
The findings suggest that U-Th dating can be applied to relatively recent buildings, not just ancient sites. This is important in places like the Mangareva Islands, where many structures are tied to local communities rather than formal records.
The islands themselves have a long history. Polynesian settlement dates back to around the ninth century, while more permanent environmental changes appear by the 11th century. European contact began in the late 18th century, followed by sustained missionary activity from 1834. Construction during this period included churches, schools, workshops and housing, much of it built with coral and local stone.
The study is based on a limited number of samples, but it points to wider applications. Similar buildings exist across French Polynesia and in other tropical regions where coral limestone has been used. Expanding the dataset could help confirm how consistent the results are and how often the “inbuilt age” issue appears.
There is also interest in comparing results from different islands and regions. If patterns hold, U-Th dating could become a standard tool for studying colonial-era architecture in the tropics.
For now, the work adds detail to a period that is not always well documented. By focusing on building materials rather than artefacts alone, it offers another way to track how communities lived and built during a time of major change in the Mangareva Islands.
Sources : Antiquity – Flexner J. Direct dating of colonial-era coral building materials using the U-Th method in the Mangareva Islands, French Polynesia. Antiquity. Published online 2026:1-12. doi:10.15184/aqy.2026.10325
