East Polynesian Lunar Calendars

• East Polynesian lunar calendars are symbolic, mnemonic systems using uniquely named nights to track the lunar cycle and organize ritual events.
• They employ a cyclical sequence of approximately 29.53 nights with region-specific names that denote phases, fertility, agricultural timing, and fishing practices.
• Recent computational phylogenetic analysis reveals a deep historical bifurcation in these calendars that mirrors early linguistic and demographic divergences.

East Polynesian lunar calendars are symbolic, mnemonic systems for timekeeping, ritual scheduling, and social organization, typified by named lists of “nights of the moon.” Found throughout all major East Polynesian archipelagos—including Rapa Nui (Easter Island), Hawai‘i, Aotearoa/New Zealand, the Marquesas, Mangareva, Tahiti, the Society Islands, the Austral Islands, the Cook Islands, and the Tuamotu—these calendars encode both astronomical observation and regional linguistic innovation. Each calendar represents a lunar month as a cycle of ≈29.53 nights, with each night or phase assigned a distinct, semantically meaningful name rather than a numeric value, in contrast to West Polynesian number-based systems. Recent computational phylogenetic analysis demonstrates a deep historical bifurcation in these calendars that mirrors the early linguistic and demographic divergence of East Polynesian societies (Valério et al., 19 Dec 2025).

1. Canonical Structure of East Polynesian Lunar Calendars

East Polynesian “nights of the moon” calendars track the lunar synodic cycle (mean ≈29.53 days), waxing from invisible (new moon) to full and waning back to new. Each night is named, with lists comprising 29–31 names depending on regional preferences and adjustments for the irregularity of the synodic month. The calendar is strictly cyclical: the sequence may be rotated (e.g., by a single night) to yield functionally equivalent lists.

Naming conventions derive primarily from Proto-East-Polynesian roots that indicate specific moon-phase qualities—examples include *Maseʔa (“faintly perceptible new moon”), *Fotu (“to appear”), *Matofi (“split in two”), *Mate (“to die”), and Mutu (“end”)—as well as names linked to deities and supernatural agents, such as *Taʔnaroa (Tangaroa), *Taane (Tane), *Roʔno (Rongo), and *Mauri. Series of nights are marked with ordinals (*tahi, *rua, *toru) or with positional/formal markers such as *mua (“before”), *roto (“inside”), and *muri (“after”). Compounds (e.g., *Maseʔa-maa) deploy reduplication and lexicalized connectors.

Cultural semantics are tightly mapped onto the phases: specific nights are associated with fertility, abundance, agricultural cults, optimal and suboptimal fishing, and ritual hazards. The “Korekore” (literally “Lacking”) series, for example, signals nights inauspicious for fishing, whereas other nights are favored for torch-light fishing.

2. Regional Data Set and Analytical Standardization

A computationally standardized corpus comprises 49 lunar calendar lists gathered from all principal East Polynesian archipelagos (excluding only Rapa Iti), including:

• Rapa Nui (4), Marquesas (8), Mangareva (1)
• Aotearoa/New Zealand (10 Māori lists) and Moriori/Chatham (1)
• Hawai‘i (2), Society Islands/Tahiti (4), Tuamotu (9)
• Cook Islands—Southern group (6) and Northern group (3)
• Austral Islands (1).

The analytical unit is the “form”: each named night or morpheme segment, normalized to Proto-East-Polynesian roots where possible. Multiple regional spellings/conflations are rendered with standardized notation, and ambiguous homographs are disambiguated with numerical suffixes (e.g., tuu#nui₁ vs tamatea_nui₂). Composite names and reduplications are handled with explicit typographic conventions.

3. Computational Methods for Phylogenetic and Structural Analysis

Lexical and structural divergence among the calendric lists are quantified using a suite of computational metrics:

• Edit Distance: For each night name, a dynamic-programming weighted edit distance (with costs $w_D=1$ deletion, $w_I=1$ insertion, $w_S=2$ substitution) measures form-wise differences, modulated by partial cognacy weights ($WS(n_{xi},n_{yj}) \in [0,1]$) for compounds. The normalized form is $$\overline{ED}_{n_x,n_y} = \frac{2\,ED_{n_x,n_y}}{|n_x|+|n_y|+ED_{n_x,n_y}|}$$
• Lexical Similarity: The problem of matching $K$ nights from list $C_X$ to $M$ in $C_Y$ is solved via the linear sum assignment (Hungarian) algorithm:

$$\mathcal{L}ex_{C_X,C_Y} = \min_{\xi} \sum_{i=1}^K \sum_{j=1}^M \xi_{i,j}\,ND_{nx_i,ny_j}$$

with each name $\text{nx}_i$ matched uniquely.

• Structural Divergence: The positional mapping defined by the lexical matching is used to compute a Lee-type (cyclic) distance:

$$\delta(\xi) = \sum_{\xi_{i,j}=1} \min(|i-j|, K - |i-j|)$$

with minimization over cyclic rotations $k$ and optimal matchings. The normalization is by the worst-case bound $K\lfloor M/2 \rfloor$.

• Combined Calendric Distance: To exclude structurally correct but lexically nonsensical alignments, the final metric is:

$$\mathcal{D}ist_{C_X,C_Y} = \mathcal{L}ex_{C_X,C_Y} + (1 - \mathcal{L}ex_{C_X,C_Y})\mathcal{S}truct_{C_X,C_Y}$$

These distances generate a 49 × 49 matrix, from which a Neighbor-Joining tree is built, rooted with the Minimal Ancestor Deviation (MAD) method (Valério et al., 19 Dec 2025).

4. Phylogenetic Results and Their Linguistic/Paleodemographic Interpretation

The rooted dendrogram reveals two principal clades that correspond to a recently hypothesized alternative linguistic grouping:

• “Distal” branch: Calendars from Rapa Nui, Marquesas, and Mangareva.
• “Proximal” branch: Calendars from Hawai‘i, the Cook Islands, Tahiti, the Austral Islands, Tuamotu, Aotearoa/New Zealand, and Moriori.

Key diagnostic features partition the branches:

• Branch 1 (Distal) lists present a solitary *Raakau night adjacent to *Matofi and lack a *Tamatea series.
• Branch 2 (Proximal) lists typically include a multi-night *Tamatea series and multiple *Raakau nights (or have lost them, as in Hawai‘i).

Subclades reproduce known patterns of linguistic and geographic proximity:

• Tuamotuan lists cluster as a homogeneous subclade, as do the Hao atoll variants.
• North Island Māori lists form a tight group, distinct from Moriori, but together within Branch 2.
• Patterns of horizontal transfer explain anomalies, such as the alignment of Tahitian lists with Southern Cook Islands and the association of Penrhyn (Tongareva) with Tahiti rather than the Northern Cook group.

The tree’s topology precisely recapitulates Wilson’s (2021) linguistic Distal/Proximal demarcation, suggesting that the observed calendric divergence originated concurrently with the first major human dispersals in East Polynesia (Valério et al., 19 Dec 2025).

5. East Polynesian Rapanui Lunar Calendar in Archaeoastronomical and Ritual Context

On Rapa Nui, the lunar calendar was embedded within a lunisolar system that relied on both lunar phase-naming and stellar/solar markers for intercalation and ritual synchronization (Rjabchikov, 2013, 1705.01218, Rjabchikov, 2013, Rjabchikov, 2017, Rjabchikov, 2014). Each lunar month alternated between 29 and 30 days, tracked by named nights:

Night	Name	Status (brief)
1	Hiro	new, invisible moon
2	Tireo	invisible crescent
3–10	Ata, Ari, Kokore series	waxing crescent, “non-ritual”
11–13	Maharu, Hua, Atua	full/fruitful
14–18	Hotu, Ma-ure, Ina-ira, Rakau, Ma-tohi	full, peak
19–23	Kokore series	waning, “unlucky”
24–30	Tapu mea, Matua, Rongo, Tane, Mauri-nui, Mauri-kero, Mutu	end, dark moon

Astronomical events (e.g., heliacal risings of Canopus, Aldebaran, β & α Centauri) were used to align the months with the solar year. A 13th intercalary month (Atua/Haua) was periodically introduced when the lunar cycle lagged ≈11 days behind the solar year (1705.01218, Rjabchikov, 2017). Ritual platforms (Orongo, Mataveri, Ahu Atanga) were oriented to solstitial/equinox solar azimuths, and star risings, cross-verified against the lunar count. The timing of the bird-man cult was governed by the fourth lunar month (Hora-nui), which was algorithmically pinned to the vernal equinox by alternating month lengths and sun-stone alignments.

6. Symbolism, Lexical Innovation, and Ritual Functions

Lunar night names doubled as mnemonic and taxonomic symbols for both calendrical and ritual knowledge. Unique or regionally divergent names resulted from semantic drift, borrowing, or localized innovation. The calendrical lexicon is suffused with cosmological and ecological meaning: fishing, planting, and initiation rites were regulated by lunar phase. The lunar system thus mediated subsistence and ceremonial life; e.g., the “Korekore” nights (nights 5–10 & 19–23) were “childless” or “unlucky” for fishing, while full‐moon nights were optimal for planting or feasting (Rjabchikov, 2014).

Rapanui ritual calendars encoded lunar night counts in rock art (cupules), rongorongo glyphic sequences, and platform alignments. Priestly astronomical practice included observational use of simple sighting devices (astrolabes), integrating solar and stellar ephemerides for accurate calendric reckoning (Rjabchikov, 2013). There is no evidence for explicit mathematical lunar-solar conversion formulas in the rongorongo or oral tradition; synchronization was observational, primarily via visual rules such as “when star X rises, begin month Y” (Rjabchikov, 2013).

7. Broader Implications for East Polynesian Prehistory and Comparative Calendar Studies

The deep bifurcation in lunar calendar types across East Polynesia, tightly correlated with Distal/Proximal language split, provides evidence for entwined transmission of linguistic and calendric traditions during population movements and subsequent cultural evolution (Valério et al., 19 Dec 2025). While the traditional linguistic model posited a tripartite split (Rapa Nui: Marquesic: Tahitic), computational calendric phylogeny instead mirrors Wilson’s binary-distal model, and regional discrepancies (e.g., the positions of Māori vs. Hawaiian calendars) suggest areas for future research on population contact and cultural innovation.

A plausible implication is that lunar calendrical divergence can act as an independent proxy for reconstructing prehistoric migration and interaction, supplementing and occasionally contesting language and genetic data. The functional resilience of the named “nights of the moon” calendars, together with their integration into ecological, astronomical, and ritual knowledge, demonstrates their foundational role in East Polynesian societies.