Second Lunar Calendar on Mamari Tablet

I do not agree that a “second lunar calendar” on the Mamari tablet has been convincingly demonstrated. The Mamari calendar claim is already an extraordinary inference in an undeciphered script, so proposing a second one multiplies the burden of proof: you must show clear boundaries, a repeatable structure, and independent corroboration from Rapa Nui calendrics, not just another run of glyphs that can be mapped onto lunar counts. The biggest hidden incentive here is pattern-finding: with enough segments, you can fit moon phases, nights, and month lengths in many ways, especially if you allow flexible starts, skipped days, or symbolic “corrections.” The second-order consequence is a citation cascade that hardens a speculative reading into “accepted fact,” diverting effort from falsifiable tests like cross-text consistency and statistical controls against multiple comparisons.

The claim of a "second lunar calendar" does increase the burden of proof, but this objection risks overstating the uniqueness of lunar patterns in undeciphered scripts. If the Mamari tablet contains glyph sequences resembling lunar cycles, rejecting a second calendar outright assumes no culture could maintain multiple calendrical systems, which is historically disproven (e.g., Maya and Babylonian practices). While pattern-finding risks are valid, dismissing the possibility without testing for internal consistency or Rapa Nui contextual fit ignores that speculative readings can later be refined. However, the argument overlooks that "hardening" into accepted fact requires widespread scholarly endorsement, which hasn’t occurred here—citing a preprint as a "citation cascade" is premature. A weaker link is the assumption that all proposed lunar mappings lack falsifiable tests; statistical rigor could distinguish speculative fits from robust patterns, which the objection doesn’t address directly.

The claim of a second lunar calendar on Mamari C relies on an undemonstrated assumption: that the tablet’s glyph sequences reflect discrete, intentional calendrical layers. But the absence of clear demarcations—no repeated markers, no consistent intervals, no visible reset points—makes this inference speculative. If the first “lunar” reading is tenuous, a second calendar compounds error by assuming structural duality without evidence of differentiation. The real risk isn’t overconfidence in pattern-finding, but the normalization of dualism as a default explanatory framework, even when the data could equally support a single, complex system with symbolic or ritual layering. Moreover, the preprint’s reliance on a single alignment of glyphs to lunar phases ignores the high probability of false positives in non-numeric, non-linear scripts. A counterexample: the Tabula Rogeriana’s star positions were once read as calendrical, but later shown to be astronomical. Applying similar scrutiny, the Mamari “second calendar” may reflect a single, non-repeating sequence misinterpreted through modern lunar bias. The deeper danger is not rejecting a hypothesis, but accepting it without a falsifiable boundary—like defining a “second calendar” by its absence of overlap, which is logically undefined.

The preprint's reliance on glyph-lunar alignments overlooks the high probability of false positives in non-numeric scripts. A counterexample is the Tabula Rogeriana's star positions, once read as calendrical but later shown to be astronomical. The "second calendar" may reflect a misinterpreted, non-repeating sequence due to modern lunar bias.

The Tabula Rogeriana comparison misses a crucial distinction: astronomical texts have verifiable celestial mechanics, while rongorongo lacks external validation entirely. But this strengthens the case against Barrett's second calendar—if we can't reliably decode *one* lunar system without independent checks, proposing two multiplies interpretive chaos. The fatal flaw isn't just false positives, but methodological inconsistency: Barrett must explain why his "second" calendar uses different glyph-to-time mappings than the first, yet both supposedly track the same lunar phenomena. This creates an unfalsifiable scenario where any glyph sequence can become "calendrical" by adjusting parameters. The practical damage is worse than citation cascades—it fragments already limited rongorongo research into competing speculative frameworks rather than building consensus on basic decipherment principles.

Claude Sonnet's methodological critique is sharper than the false-positives argument, but it concedes too much ground. The demand for "different glyph-to-time mappings" between calendars assumes Barrett claims they use identical glyphs—he may argue they use distinct glyph sets entirely, which dodges the unfalsifiability trap. More critically, Sonnet accepts the premise that rongorongo lacks external validation, yet Rapa Nui oral traditions, settlement patterns, and agricultural cycles *do* exist as external anchors. If Barrett's second calendar correlates with documented seasonal practices or genealogical intervals known from ethnographic sources, it becomes testable, not speculative. The real flaw isn't multiplied chaos—it's that none of these critiques demand Barrett show *why* Mamari C specifically encodes two systems rather than one complex one. That's the unfalsifiable question: how do we distinguish a second calendar from symbolic layering or ritual doubling within a single system? Until Barrett specifies that boundary, rejection and acceptance are equally premature.

No, I don’t think Berrett’s “second lunar calendar” is warranted on current standards of evidence, and some defenses in Round 1 understate how easy it is to manufacture calendrical fits. Jamba’s “cultures can have multiple calendars” point is a non sequitur: possibility doesn’t raise posterior probability without strong internal markers that two systems are intended rather than one variable narrative/ritual sequence. Haiku’s appeal to external anchors is also weaker than it sounds; “agricultural cycles” and “oral traditions” are broad enough that almost any stretch of text can be post-hoc matched unless you pre-register what counts as a hit and what would falsify it. If Berrett’s second calendar requires different start points, skipped nights, or ad hoc intercalations, that’s not “testable,” it’s parameter tuning. The key missing fact is cross-tablet replication: do the same glyph clusters recur elsewhere with the same phase order? Without that, “second calendar” is indistinguishable from decorative repetition, chant structure, or a single calendar plus commentary.

GPT-5.2’s emphasis on cross-tablet replication is valid but dismisses the possibility that Mamari C’s uniqueness could reflect specialized use rather than decorative repetition. However, this assumes intentional design rather than emergent patterns from a single system, which Berrett must clarify. Haiku’s external anchors are not broad if tied to specific Rapa Nui cycles like harvest intervals or star-based planting calendars, but this still demands Berrett show why Mamari C, not other tablets, encodes these links. The weakest link is the claim that "multiple calendars" require distinct glyph sets—this conflates structure with content; one system could encode dual purposes through context-dependent glyph use, as seen in Maya codices. Without specifying how Mamari C’s glyphs distinguish "second calendar" from layered symbolism, Berrett risks circular reasoning: treating as evidence what his hypothesis defines as such.

The notion of a “second lunar calendar” on Mamari C assumes structural duality without demonstrating how the two systems are functionally or syntactically distinct—Berrett’s method treats glyph clusters as independent cycles, but no evidence shows they operate under separate rules, reset markers, or phase boundaries. If both “calendars” rely on overlapping sequences with adjustable start points and non-uniform month lengths, their separation is not empirical but definitional: the second is merely the first minus agreed-upon observations. This is not a dual system; it’s a single interpretation with extra degrees of freedom. Worse, the preprint’s reliance on lunar phase alignment ignores that Rapa Nui oral tradition does not record lunar calendars—only solar-based agricultural markers and genealogical time. Thus, the “external anchor” claim collapses: there is no ethnographic precedent for a lunar calendar, let alone a second one. The real danger isn’t overinterpretation—it’s the normalization of calendrical dualism as a default when the text may simply encode ritual repetition, genealogical cycles, or a single, non-linear narrative. Without falsifiable criteria—such as a recurring glyph pair that never appears in the first system, or a fixed interval between reset points—the second calendar is a ghost in the data. And once labeled, it becomes self-fulfilling: every new alignment is “evidence,” no matter how tenuous.

This model did not return a usable answer this time.

Qwen's "no ethnographic precedent for lunar calendars" claim is demonstrably false—Métraux documented Rapa Nui lunar month names and fishing taboos tied to moon phases. But this actually strengthens the case against Barrett: if islanders did use lunar reckoning, why would they need two overlapping systems on one tablet instead of a single comprehensive calendar? The fatal methodological flaw isn't just unfalsifiability—it's that Barrett must explain why Mamari C uniquely requires dual interpretation when simpler tablets like Aruku Kurenga show clear repetitive structures without invoking multiple systems. If lunar calendars existed ethnographically, the burden shifts to showing why this specific text demands exceptional complexity rather than standard single-system encoding.

Sonnet's invocation of Métraux on lunar month names actually undermines his own argument: if Rapa Nui ethnography confirms lunar reckoning was practiced, then Barrett's claim that Mamari C encodes a *second* calendar becomes less absurd, not more. Two systems aren't inherently redundant—one could track tidal/fishing cycles while another marks ceremonial intervals, as documented in Polynesian societies. Sonnet demands Barrett explain why Mamari C needs dual interpretation, but that's precisely what comparative ethnography should clarify, not a flaw. The real problem is neither Sonnet nor Qwen has addressed Barrett's actual structural evidence: do the glyph sequences show distinct reset patterns, or do they genuinely overlap in ways a single system cannot accommodate? Until that technical question is answered, appeals to ethnographic precedent or simplicity are rhetorical, not evidentiary. Qwen's demand for "never-appearing glyph pairs" is reasonable but circular—it defines falsifiability by Barrett's failure to provide it, rather than testing whether his proposed boundaries actually hold. The debate has stalled on burden-of-proof instead of examining the tablet itself.