A study has revealed that the “W glyph” in Zapotec inscriptions tracks the visible days of the moon, enabling for the first time the precise dating of seven calendar records to over 2,200 years ago in absolute terms.
This breakthrough allowed the authors, John Justeson and Justin Patrick Lowery, to establish absolute chronological intervals between seven full dates marked by the W glyph, placing them in real time. A single sequence that aligns with the Zapotec calendar, as documented in the 17th century, spans the period from 496 to 221 BCE.
In their paper, the authors highlight that Monte Alban hosts the oldest corpus of hieroglyphic texts in Mesoamerica. This work demonstrates that the W glyph indicates the number of evenings the moon was visible since the new moon on a given date, establishes absolute intervals between seven fully preserved dates corresponding to W glyph records, and determines all possible positions of these dates in absolute time.
The W hieroglyph was identified by Mexican archaeologist Alfonso Caso in 1928, but its meaning remained enigmatic for decades. Three main hypotheses were proposed: that it signified a day within a 20-day month, that it represented the number of a trecena (a 13-day period), or that it counted the months of the year. The authors, using specific data, argue that all these interpretations are incorrect.
Instead, by analyzing the intervals between dates on two key monuments, they concluded that the number accompanying the W glyph increases and decreases in a cycle of approximately 29.53 days, precisely matching the length of a lunar cycle (the time between two new moons). In essence, the Zapotecs were recording the moon’s age at the time of each event.
The research is based on seven well-preserved inscriptions containing three primary components: a date in the 260-day divinatory calendar (called a piye), the year name within a 52-year cycle, and the W glyph with its numerical coefficient.
To grasp this discovery, it’s essential to recall that Mesoamericans employed two parallel calendars. On one hand, there was the 260-day divinatory calendar, constructed from the combination of 13 numbers and 20 day names (such as “Deer,” “Rain,” or “House”). On the other, there was an approximate 365-day year, similar to our own but lacking leap years, thus shifting with the seasons.
The Zapotecs named their years according to the date in the divinatory calendar that fell on the first day of the year. For example, a year might be called “Earthquake 6.” By the 17th century, northern Zapotecs still utilized this system, and it’s known that the first day of the year was “Earthquake 11.”
The W symbol consistently appears immediately after the divinatory date, followed by a number written with bars (valued at 5) and dots (valued at 1). The range of values is from 2 to 20. The researchers confirmed that if this number encompassed days of the month or trecenas, the predictions would not align with the data. However, interpreting it as the count of days since the first sighting of the waxing moon aligns perfectly.
The key to the decipherment lies in two sets of inscriptions. The first involves two monoliths (designated M-21 and D-142), which record the same divinatory date—2 Flint—twice within the same year (Year 12 Earthquake). One appears with W glyph-18, and the other with W glyph-14. The interval between these two events is exactly 260 days (one cycle of the divinatory calendar). If the W glyph counted lunations, a 260-day interval should have caused a change in the W number approximately equal to 260 divided by the length of a lunation. Indeed: 260 / 29.53 = 8.8. The difference between 18 and 14 is 4 (not 8.8), but this is because the W count resets after each new moon, much like our first day of the month repeats every 29 or 30 days.
The second dataset comprises four dates on an orthostat (stone slab) known as J-14, spanning two consecutive years. Using these dates, the authors calculated all possible values for the W glyph’s cycle duration, testing 7,401 candidates ranging from 18 to 92 days in increments of one-hundredth of a day. The results clustered around 29.53 days with minimal deviations.
An interval of approximately 29.525 days carries significant weight, write Justeson and Lowery: their estimates fall on either side of the average lunar cycle length of 29.5306 days. Therefore, we interpret W-1 as the first day of the lunar cycle, commencing with the first sighting of the waxing crescent after the new moon.
Once it was established that the W glyph tracked lunar cycles, the researchers calculated the absolute intervals between the seven well-preserved dates—how many actual days separated each event from the first in the series. They then searched for what absolute dates (BCE years) these events could have occurred, considering that the first sighting of the moon’s sliver at Monte Alban was only possible on specific days between 650 and 50 BCE.
They employed an astronomical model developed by Caldwell and Leynii (2001) and adjusted by Lowery to determine day by day when the moon’s sliver would have been visible from the Monte Alban plaza. Using this data, they filtered out all possible combinations of dates.
The result was the identification of seven potential sequences that adhered to archaeological constraints: the earliest monoliths (Danibán Period, 500–300 BCE) had to date before 360 BCE, and later tablets (Peñitas Period, 300–100 BCE) after that date. However, only one of these seven sequences aligned with other independent data: the Zapotec calendar recorded by the Spanish in the 17th century.
According to colonial records, in 1695, the Zapotec Year 11 Earthquake began on February 23rd. Working backward through 52-year cycles, the researchers calculated when the first year of a cycle in the inscription era must have commenced.
A sequence of dates spanning from 496 to 221 BCE aligns with these calculations, assuming the Zapotecs made 26 adjustments of 20 days over 36 52-year cycles to correct for seasonal drift.
The authors contend that only one of the seven candidates fits the established parameters of Mesoamerican calendrical history. The next closest contender would have required 440 20-day adjustments across 36 52-year cycles, which is unfeasible. This assures that the candidate series from 496 to 221 BCE belongs to the era of the W glyph records.
One of the most intriguing findings of the study is that the data allow for the determination of the time of day when the Zapotec divinatory calendar began. In the 16th century, Dominican friar Juan de Córdova documented that the day commenced at noon. Justeson and Lowery’s calculations corroborate that this same rule was in effect as early as 222 BCE.
The reason is that the crescent moon is first visible shortly after sunset, around 6:44 PM local time at Monte Alban. If the divinatory calendar changes at noon, an evening date corresponds to the following morning’s date. By comparing expected moon positions with those recorded in the W glyph, the authors found one-day discrepancies, perfectly explained by some events occurring in the afternoon and others in the morning.
Specifically, the return of the year-bearer (the 261st day of the year) was celebrated in the afternoon, and this celebration is recorded on tablet J-14 with the date 5 Soap and W glyph-10. Calculations show this ceremony was already taking place in the afternoon around 222 BCE.
The data indicates that dates in the Zapotec divinatory calendar likely began their count around midday, as in the 16th century, and that the return of the year-bearer on the 261st day was celebrated in the afternoon around 222 BCE, the paper states.
The significance of this discovery extends beyond Monte Alban. The authors note that the tracking of lunar days in Zapotec inscriptions predates the earliest known example in Maya writing by 857 years. The so-called “Maya Classic Lunar Series” emerged in the 4th century CE on stelae at Naachtun (Guatemala), linked to a military incursion from Teotihuacan.
Justeson and Lowery conclude that the Monte Alban texts attest to the antiquity of lunar daily counting in Mesoamerica, 857 years earlier than previously understood. The closest parallel to the Zapotec lunar daily count is that preserved in hundreds of Maya hieroglyphic texts.
The study also clarifies a point of confusion: three instances of the W glyph are depicted inverted (upside down) and without numbers. The authors propose that this inverted form was used when the moon was no longer visible (during the new moon) or when the count of visible days exceeded 20, which occurs about three times in every eleven lunar cycles. Statistical data align perfectly with the three documented instances of the inverted depiction.
About the Author
