Computational analysis of “moon nights” reveals two ancient cultural branches and offers new insights into the peopling of Pacific islands.

Forty-eight traditional calendars, each built around the names given to the different nights of the lunar month, have preserved for centuries an unexpected trace of the dispersal of populations across the Pacific. By comparing them using quantitative methods, a group of researchers has identified two large families of calendars in eastern Polynesia: one connects the Marquesas Islands, Mangareva and Rapa Nui, while the other includes most of the other archipelagos in the region.

The separation takes on particular importance because it closely coincides with a classification of the Eastern Polynesian languages Obtained through independent evidence. Calendars and languages ​​thus appear to have preserved, through different mechanisms, part of the memory of ancestral communities who crossed thousands of kilometers of ocean and colonized some of the most isolated lands on the planet. The result does not reconstruct the entire migratory itinerary by itself, but introduces a new category of evidence to be compared with archaeology, genetics, and linguistics.

The research, published on PLOS One, was conducted by Fabio Tamburini e Michael Corazza of the University of Bologna together with Miguel Valerio of the University of Murcia. The work does not focus on material findings or DNA sequences, but on symbolic systems transmitted mainly orally: the lists of the so-called “nights of the moon”.

Each night had a name and indicated community activities

In pre-colonial Polynesia, the month followed the cycle of the Moon, and its nights were practical references for organizing daily life. The synodic cycle lasts approximately 29,53 days And, in the traditions of Eastern Polynesia, individual nights could be identified by specific names. Some invoked deities, others described characteristics of the moon, or occupied a recognizable position within particular sequences.

It was therefore not just a matter of measuring the passage of time. The calendar could help identify the periods considered most suitable for the fishing and agriculture, integrating astronomical observation, environmental experience, and social organization. In many communities, this information was preserved orally for generations, evolving as populations lived separated on islands hundreds or thousands of kilometers apart.

It is precisely this ability to change without completely losing their original structure that makes calendars interesting for historical reconstruction. A name could disappear, be replaced, or shift relative to others within the month. In some cases, entire series changed. Accumulating over time, these differences produced distinguishable and comparable local traditions.

“Like languages, calendars are also symbolic systems passed down from generation to generation,”

explains Miguel Valerio.

“As communities became increasingly isolated across the different Pacific Island groups, these documents evolved over time, while still retaining shared traces of the original calendar.”

The arrival of Europeans profoundly altered this system. With the introduction of the Western calendar and new religious and administrative references, traditional forms of timekeeping gradually fell into disuse. Some knowledge survived because indigenous experts and outside observers transcribed the lists, transforming predominantly oral traditions into documents that can be used today for comparative analysis.

“After the introduction of the European calendar, this traditional system of measuring time gradually fell into disuse in most of Polynesia,”

observes Fabio Tamburini.

“It has survived into contemporary times largely because it was written down by both indigenous experts and non-indigenous observers.”

A computational method measures names, sequences and movements

The methodological innovation lies in having transformed this cultural heritage into comparable data, without reducing it to the simple presence or absence of the same words. The researchers collected calendars from a vast territory that includes Hawaii, Aotearoa/New Zealand, Tahiti, Cook Islands, Marquesas, Mangareva, and Rapa Nui, as well as other island groups in eastern Polynesia.

For each tradition, both the correspondences between the names of the nights and their relative placement in the monthly cycle were considered. Two seemingly similar elements can tell different stories if, over the generations, one has remained in the same position while another has been moved, replaced, or integrated into a different sequence.

“We compared not only the names of the lunar nights, but also their relative position within the lunar month,”

explains Michael Corazza.

“Measuring the relative position of the lunar nights meant quantifying how much the corresponding nights had shifted within the monthly cycle, which we did using a quantitative computational method.”

This step is important because it applies tools normally associated with quantitative analysis and the construction of phylogenetic relationships to a set of cultural evidence. Accumulated differences can be converted into distance measures between calendars and subsequently organized into a branched structure, similar in function to the trees used to represent relationships between languages ​​or populations.

The result is the first “family tree” of the lunar calendars of Eastern PolynesiaRepresentation does not imply that a cultural tradition evolves exactly like a biological species: ideas, words, and practices can also be transferred laterally between communities. However, it offers a tool for identifying groups of documents that share a greater number of characteristics and for verifying whether this distribution is compatible with other historical reconstructions.

An earlier version of the work, made available as a preprint, describes a slightly larger corpus of 49 lists; the communication relating to the published study instead takes into consideration 48 lists. The essence of the result remains the particularly clear separation between the two main groupings of Eastern Polynesia.

Marquesas, Mangareva and Rapa Nui form a distinct branch

The first branch identified by the analysis brings together calendars of the Marquesas, Mangareva and Rapa Nui IslandsThe second group includes traditions traceable to Aotearoa/New Zealand, Hawaii, the Cook Islands, the Austral Islands, Tahiti, and the Tuamotu. Geography alone isn't enough to fully explain the configuration: some enormously distant islands share characteristics that point to a common cultural history.

The same division presents a significant correspondence with a recent proposal for linguistic classification which distinguishes an oriental group defined “Distal”, formed by marchesiano, mangarevano and rapanui, from a group “Proximal” including the other major languages ​​of Eastern Polynesia, including Hawaiian.

“This correspondence is unlikely to be coincidental,”

comments Fabio Tamburini.

“On the contrary, it suggests that calendars and languages ​​diverged in parallel, as ancestral Polynesian communities spread across the Pacific and gradually transformed into distinct island societies.”

The parallel is interesting precisely because the two systems have been analyzed separately. A language and a calendar are not the same type of cultural object: the former permeates everyday communication, the latter codifies astronomical knowledge, economic activities, and social conventions. If both maintain a similar division, the possibility arises that this structure reflects a separation that occurred in the populations that transmitted them.

The authors therefore interpret the results as consistent with an ancient bifurcation: one route would have led to the Marquesas, Mangareva, and Rapa Nui, while another would have been associated with the populations of the other eastern archipelagos. This interpretation enters into the scientific debate on the sequence in which the region was populated and calls into question models in which the Rapa Nui linguistic community was the first to autonomously separate from all the others.

The picture is necessarily more complex than a tree with perfectly separate branches. Ocean voyages did not cease after the first settlement, and subsequent interactions between different communities are documented. Even the calendars themselves may therefore have incorporated borrowings, reconstructions, and influences from other islands. Research considers this possibility especially in the central areas of eastern Polynesia, where contact was more frequent.

From oral tradition comes a new source of historical data

The value of work goes beyond the specific case of calendars. It shows how computational tools can be applied to cultural heritage not originally conceived as data, preserving lexical and structural differences that only gain meaning when analyzed on a large scale. Ethnographic documents, transcripts, and local memories thus become a corpus capable of generating hypotheses that can be tested by other disciplines.

Polynesia offers a particularly significant test bed. The colonization of its islands represented the last great human expansion into previously uninhabited territories and required ocean navigation skills over exceptional distances. Reconstructing its times and itineraries means bringing together archaeological dating, linguistic similarities, genetic data and cultural evidence, each characterized by different limits and time scales.

Genomic studies have already reconstructed a complex expansion across the Pacific and placed the settlement of Rapa Nui around the 13th century, within a network of connections that included other societies in eastern Polynesia. Calendars do not replace this evidence: they add an independent line of research based on the transformation of shared knowledge during the isolation of communities.

Rapa Nui also provides a special case. A sequence interpreted as a lunar calendar has been identified on the so-called Mamari tablet, belonging to the Rongorongo corpus, the island's glyph system, which still remains undeciphered. The link does not allow for automatic reading of the script, but it demonstrates how deeply rooted the computation of lunar phases was in the local culture and offers further opportunities for comparison between oral, ethnographic, and physical documentation.

The researchers' next goal is to go even further back. By comparing the elements conserved in the different branches, it will be possible to attempt a reconstruction of the ancestral lunar calendar from which the documented lists on the various islands may have derived. The corresponding family tree has been made available according to the principles of Open Science, so that the data and methodology can be examined and used for further comparisons.

This is the most scientifically interesting step: not to transform a tradition into a historical certainty, but to obtain from it a series of measurable relationships to be tested along with the other traces left by the population of the Pacific. The "moon nights," born to mark fishing, cultivation, and social life, thus become a cultural archive of human mobility, capable of preserving in the order of words a history that began many centuries before those words were put into writing.

Here are three insights that might interest you:

Te Henua Ènata, the Marquesas as a global cornerstone of conservation
Rapa Nui, an open-air laboratory for sustainable innovation
Rapa Iti, a remote outpost of adaptation and innovation