Phase 5: Individual Tablet Analyses and Comparative Glyph Patterns
Phase 5.1: Mamari Tablet (Text C) β Content and Pattern Analysis
The Mamari tablet is one of the best-understood rongorongo texts, primarily because it contains a lunar calendar sequence that has been partially deciphered. Thomas Barthel first identified a series of ~30 glyph clusters on Mamari corresponding to the nights of the Rapa Nui lunar month. This calendrical text spans the end of line 6 of side A through lines 7 and 8 (and perhaps the start of line 9) of the tablet.
Later researchers confirmed this interpretation by correlating the glyph sequence with recorded Rapa Nui moon-night names: William Thomson's 1886 list of moon nights aligns one-for-one with the Mamari glyph sequence in order, with no additions or omissions, strongly validating the calendar reading. In other words, each glyph (or glyph group) in that part of Mamari corresponds to a night such as Hiro, Kokore, Taκu and so on, in the same sequence local informants gave those names.
Calendrical Pattern Discovery: Within the calendar sequence, distinctive patterns emerge. There are 28 main night glyphs plus two special glyphs (making ~30 total, matching a lunar month) arranged symmetrically with the full moon at the center. Certain glyphs repeat in a regular cycle: for example, a particular "fish" glyph appears eight times β four times with its head oriented upward, and four times inverted (head downward).
Researchers interpret this as a clever device to mark the moon's waxing vs. waning phases: the fish glyph is upright during the waxing half of the month and upside-down during the waning half. This orientation pattern, along with other recurring sequences, shows the scribes encoded astronomical information in a structured way, reinforcing that this section is a functional lunisolar calendar.
Indeed, Jacques Guy demonstrated that the sequence of glyphs on Mamari correlates with the traditional Polynesian metaphor of the moon as a growing and shrinking fruit (glyph 74 hua "fruit" appears for the first quarter moon night). The full moon night sits at the midpoint of the sequence, and the nights on either side mirror each other in glyph composition, consistent with the waxing/waning symmetry.
Beyond the Calendar
Outside of the famous calendar portion, the Mamari tablet's content is more enigmatic and appears varied. There are indications of other structured passages on Mamari that parallel sequences on different tablets, suggesting some shared formulaic texts. For instance, lines Cv2βCv4 of Mamari contain a repeated glyph sequence involving glyph 380.1 that also occurs in other inscriptions.
This implies that Mamari includes at least one common phrase or liturgical sequence found elsewhere in the corpus, beyond its unique calendar. The rest of Mamari's lines have not been deciphered, but some may also deal with calendrical or ritual content. It's noteworthy that an early 20th-century oral tradition claimed the Mamari tablet (known as Kohau o te Ranga) was a war tablet listing the names of enemy prisoners, used magically to ensure victory.
However, the actual inscriptions that have been decoded β like the lunar calendar β do not obviously match that description. In sum, Mamari stands out as the only rongorongo text with a broadly agreed-upon meaning for a large portion of its content, thanks to the lunar month sequence. This deciphered segment provides a crucial key for researchers, as it demonstrates how repeated glyph clusters and contextual clues (moon phases, known night names) can reveal the text's subject matter without forcing interpretation.
Phase 5.2: Aruku Kurenga Tablet (Text B) β Repeated Sequences and Structure
The Aruku Kurenga tablet is another substantial text in the corpus, comprising around 1,290 glyphs on 22 lines (recto and verso). It is especially intriguing for its internal repetition patterns. Butinov and Knorozov (1956) discovered that Aruku Kurenga contains three large sequences of glyphs that are nearly identical β in their words, "evidently, this is one and the same text, given in three variants".
Repetition Discovery: A lengthy passage of glyphs is written, then later in the text a very similar passage is repeated, and yet again a third time. This is a rare phenomenon among the rongorongo tablets. The repeating sequences suggest that the tablet might be a compiled collation of texts or verses.
One interpretation is that Aruku Kurenga could be recording a ritual chant or narrative in a refrain-like format, where the same invocation or story is told multiple times (perhaps to preserve different versions or for recitation by multiple performers). Steven Fischer indeed proposed that Aruku might not be a single continuous narrative but rather a composite of several texts or iterations, possibly assembled by a scribe.
These repeated clusters emerged "naturally" from the text and were not forced by any modern analysis β they are plainly visible upon close comparison of the lines, highlighting a pattern of deliberate duplication.
Historical Context and Analysis Challenges
Historically, Tablet B (Aruku Kurenga) played a prominent role in early attempts to understand rongorongo. It was among the tablets given to Bishop Florentin Jaussen, and his Rapanui informant Metoro Tau'a Ure was asked to "read" it aloud. Metoro did recite something while pointing to the glyphs (starting correctly at the bottom of the recto, per the rongorongo reading order), and Jaussen transcribed those chants.
However, the result β known as the "Jaussen List" β proved incoherent and of no use in deciphering the script. The transcribed chant did not correspond to the glyph sequences in any straightforward way. Modern analysts believe Metoro's recitation was likely a series of memorized chants or prayers triggered by seeing familiar symbols, rather than a direct reading of the text.
This episode underscores the difficulty in interpreting Aruku Kurenga's repeated sequences: even a 19th-century islander versed in oral lore gave interpretations that diverged from the text's actual content, indicating we must analyze the patterns without imposing assumed meanings.
Structural Connections
From a structural standpoint, Aruku Kurenga shares some affinities with other large tablets. Researchers have noted that certain prominent glyph pairings and sequences in Aruku also appear in the so-called "Great Tradition" texts (a proposed grouping of texts with mythic or ritual content). In fact, comparative analysis finds overlapping segments between Aruku (Text B) and tablets like Large Santiago (Text H) and Large St. Petersburg (Text P).
This hints that Aruku's content might belong to a broader narrative or genre circulated in multiple copies. The presence of glyph 32 as a delimiter in Aruku is one example: a structured sequence in Aruku is delimited by a specific glyph (Barthel #32), and a similar construction is observed in other texts, suggesting a standard phrase boundary or refrain marker.
Phase 5.3: Santiago Staff (Text I) β Genealogical Patterns and Unique Features
Text I, the Santiago Staff, differs from the flat wooden tablets in form β it is a 126 cm long wooden staff incised with glyphs β but it carries the longest rongorongo inscription (about 2,920 glyphs). Analysis of the Santiago Staff reveals a markedly formulaic and repetitive structure consistent with genealogies or lineage chants.
Statistical studies have shown that Text I's content is distinct from most other inscriptions, sharing almost nothing in common with non-genealogical texts. In fact, Pozdniakov (1996) noted that the Staff shares short repeating phrases only with the Small Santiago tablet (G) and the Honolulu tablet (T) β and not with the rest of the corpus.
Genre Separation: This means the staff, G, and T form a cluster of texts with similar patterns (presumed genealogical lists or king chronicles), whereas the Staff has virtually no overlap in phrases or sequences with tablets like Mamari, Aruku, or the large "Great Tradition" tablets. This clean separation is a strong hint that the Staff's content belongs to a different genre.
The Genealogical Glyph 76
One striking feature of the Santiago Staff is the abundant repetition of a particular glyph, Barthel sign 76, which depicts a stylized phallic or copulating figure. Glyph 76 occurs with unusually high frequency in Text I and in the few texts allied with it (G and T). Conversely, this sign is rare or absent in the other rongorongo texts.
Scholars widely believe that glyph 76 functioned as a relational or generative marker β essentially a verb meaning "to beget" or "to copulate/produce offspring" in a genealogical or creation context. Steven Fischer's attempted decipherment of the Staff hinged on this idea: he interpreted repeated sequences like "Person A β γ76γ β Person B" as "A begat B" (or A copulated [with B] to produceβ¦).
Together, the frequent pairing of adult figures engaged in "procreative" action (glyph 76) and references to offspring (glyph 400) strongly suggest the Staff records sequences of parentβchild relationships β essentially genealogies or mythical lineages. This pattern is not forced by analysts; it emerges from the repetition frequency and context of the signs.
Unique Structural Features
In addition to its content patterns, Text I is unique in showing potential punctuation or segmentation marks. It has been called "the only [rongorongo text] which appears to have punctuation" by virtue of a recurring glyph or separator that breaks the text into clauses. This might be a visual cue separating generation entries or narrative verses β a feature not clearly seen in other tablets.
The presence of such regular dividers reinforces the impression of a list or verse structure (like lineage stanzas). The small Santiago tablet G (which may record a genealogy on one side) and the fluted Honolulu tablet T (a shorter text) follow similar repetitive patterns, though on a smaller scale. But the Staff, being the longest, provides the most extensive example of this repetitive formulaic genre.
Phase 5.4: Cross-Tablet Glyph Pattern Analysis and Comparative Insights
Examining all the tablets together, researchers have identified clear pattern groupings and cross-textual relationships through both single-glyph frequency analysis and multi-glyph cluster comparisons. One of the most salient findings is that the rongorongo corpus seems to fall into at least two major content groups based on their internal structures and favored glyph combinations:
1. Genealogical/List Texts
This group includes the Santiago Staff (I), Small Santiago (G), and the fragmentary Honolulu (T). They are characterized by extremely frequent use of glyph 76 (the "procreative/genealogy connector") and often contain short repetitive phrases that recur among themselves. These texts share phrases with each other β for example, Text I and G have specific sequences in common β but significantly, they share almost nothing with the rest of the corpus.
The heavy use of a "begat" glyph and possibly a child/offspring glyph in these texts suggests they record genealogies or similar repetitive lists. Butinov and Knorozov's early observation that line Gv6 on the Small Santiago tablet reads like a genealogy (a hypothesis still considered plausible decades later) fits this pattern.
2. Narrative/Ritual Texts (Great Tradition)
Most of the other tablets fall into a different cluster often termed the "Great Tradition" by researchers. These include large tablets like Large Santiago (H), Large and Small St. Petersburg (P and Q), Aruku Kurenga (B), Tahua (A), EchancrΓ©e (D), Large Washington (S), and even fragmentary texts like Berlin (O) and Small Washington (R).
They are unified by shared glyph sequences and the relative scarcity of glyph 76 (the genealogical marker) in their texts. Instead of parent-offspring repetitions, these tablets show narrative or ritually repetitive sequences.
Textual Duplication Discovery: Large Santiago (H) has been found to "nearly duplicate" the texts on the Large St. Petersburg (P) and Small St. Petersburg (Q) tablets β indicating that multiple physical copies of the same text were made. (H, P, and Q share so many sequences that they appear to be versions of one composition, perhaps a standard chant or mythic account.)
Cross-Reference Pattern Matching
Similarly, the Berlin tablet (O), though heavily eroded, has been shown via digital enhancement to contain several glyph sequences that appear in Aruku (B), Large Santiago (H), and the St. Petersburg tablets. These repeated clusters across independent artifacts suggest a common corpus of chants or narratives that multiple scribes reproduced β a kind of canonical text.
For instance, analysts have identified a specific glyph pair or phrase (for example, a pair of identical glyphs followed by a bird glyph) that recurs in both the Berlin text and the Large St. Petersburg text, among others. Such findings imply that these tablets likely encode mythological, cosmological, or ritual sequences that were widely known and copied.
Chronological and Environmental Markers
Beyond broad groupings, cross-comparison of individual glyphs across tablets offers additional insights. Certain glyphs are widespread and appear in many texts, while others are extremely rare (hapaxes) or confined to one or two texts. For instance, glyph 67 (often identified as the shape of an Easter Island palm tree) is notable for its rarity and potential chronological significance.
This glyph likely depicts the native palm (Paschalococos disperta) that went extinct on Rapa Nui by around 1650 CE. Glyph 67 appears only in a limited subset of tablets (scholars have noted its presence in texts such as Large St. Petersburg Q). The implication is that any text containing glyph 67 must have been composed while the palm tree was still part of the culture's lived environment, suggesting those inscriptions (or the prototype texts they copy) date to at least the mid-17th century or earlier.
This kind of single-glyph distribution analysis helps cross-date the tablets relative to each other and to known historical events. Another example is glyph 302 (often described as a figure with a distinctive headdress) β if it appears predominantly in one group of tablets but not the other, it might indicate a deity or concept specific to the genre of that group.
Statistical Patterns and Frequency Analysis
Indeed, analysts create frequency tables of each glyph per text: these show that some glyphs, presumably "function words" or grammatical markers, occur hundreds of times in a text (for example, glyph 6 or 8 might appear very commonly as part of composite signs), whereas others appear once. Such distributions echo patterns in real writing systems (where, say, a comma or article is frequent, but a specific noun is rare).
In rongorongo, glyph 76 serves as one such high-frequency "connector" in genealogical texts, analogous to a repeated phrase like "son of" or "begat". Meanwhile, complex ligatured glyphs (where two or more base glyphs are fused) often appear as set pieces in multiple tablets, hinting at fixed compound meanings.
Cross-table cluster analysis has identified several of these common compound segments. For example, a compound of a crouching figure plus a circlet glyph appears in at least three different tablets in the Great Tradition set, possibly indicating a recurring concept or name.
Calendrical Cross-Reference: It is also instructive to compare the calendrical content identified in Mamari with other tablets for any similar structures. Recent research by Wieczorek (2011) posits that Tablet Keiti (Text E) contains sequences analogous to Mamari's lunar calendar. By examining Keiti's recto, scholars noted a pattern of crescent-shaped glyphs and interval markers that suggest an astronomical or calendar text.
In fact, Keiti may encode lunar observations or instructions and appears "similar in content to the only other text [Mamari] whose function is known". If confirmed, this would mean Mamari is not entirely unique β at least one other tablet likely records calendrical or astronomical information, reinforcing that some glyph clusters (like those for nights of the moon, crescents, etc.) recur across tablets.
Researchers have pointed out specific delimiters in Keiti (such as pairs of crescent moon glyphs) that could mark month divisions or lunar events, matching structures in Mamari's known calendar. This cross-table pattern mapping (Mamari vs. Keiti) was achieved by structural comparison of glyph clusters, again without assuming meaning beyond the reasonable hypothesis that repeated crescent signs relate to moons.
In summary, the cross-table analysis β both at the level of individual glyph frequencies and larger repeating sequences β reveals a naturally organized diversity in the rongorongo corpus. Some texts share so much material that they likely transmit the same composition (e.g. H ~ P ~ Q), pointing to scribal copying. Some glyph sequences are nearly universal, possibly common phrases or chants, while others are exclusive to one genre of text.
This approach has allowed scholars to make progress in understanding the script's structure without "forcing" a translation, instead relying on the emergence of patterns: duplication of passages, common motifs, and statistical outliers all guide us to classify texts and even guess their function. It is through this careful comparative method that we can say, for instance, that Mamari's calendar is unique but Keiti might join it as an astronomical text, or that the Staff and Small Santiago likely record genealogies, while tablets like Aruku and Large Santiago encode mythic or ritual recitations.
Sources:
- Barthel, Thomas S. (1958). Grundlagen zur Entzifferung der Osterinselschrift. (Identified the lunar calendar on Mamari)
- Guy, J. (1990). "On the Lunar Calendar of Tablet Mamari." JSO 91(2):135β149. (Correlated Mamari glyphs with Rapa Nui moon-night names)
- Wikipedia: Rongorongo text C (Mamari) (General description of Mamari and its deciphered calendar)
- Wieczorek, R. (2011). "Astronomical Content in Rongorongo Tablet Keiti." JSO 132:5β16. (Hypothesis that Keiti contains lunar calendar akin to Mamari)
- Fischer, Steven R. (1997). Rongorongo: The Easter Island Script. (Attempted decipherment, especially of the Staff)
- Pozdniakov, Konstantin (1996). "Les bases du dΓ©chiffrementβ¦" JSO 103(2):289β303. (Noted shared phrases among I, G, T and absence in others)
- Butinov, N., & Knorozov, Y. (1956). "Preliminary Report on the Study of the Easter Island Writing." (Noted repeating sequences in Aruku Kurenga)
- Horley, P. (2007 & 2010). Structural analyses comparing rongorongo texts and Rapanui chants (Identified content-based sub-groups of texts)
- Thomson, William J. (1891). Te Pito te Henua, or Easter Island. (Recorded Rapa Nui lunar calendar in 1886, used to verify Mamari's sequence)