One damaged stone panel from the Mexican state of Chiapas preserves the earliest known date in the system the whole world knows as the “Maya calendar”—and yet it has almost nothing to do with the Classic Maya.
On the western bank of the Grijalva, in the Central Depression of Chiapas, lies a place known in archaeological literature as Chiapa de Corzo. People settled here early—around 1200 BCE, in the Early Preclassic period, when Mesoamerica was only beginning to feel its way toward its great forms. By 700–500 BCE, the settlement had grown to roughly 70 hectares and controlled the trade routes leading to coastal Soconusco—one of the richest regions on the Pacific coast. It was a hub through which cacao, obsidian, and ideas all passed.
In 2008, at the base of Mound 11, archaeologists discovered an Olmec cache dated to around 700 BCE, as well as a so-called E-Group astronomical complex—an architectural configuration that made it possible to observe solstices and equinoxes. Then, in April 2010, that same mound yielded a find that took Mesoamericanists’ breath away: the tomb of a dignitary some 2700 years old, the earliest pyramidal tomb in all of Mesoamerica. It predated the famous royal burials of Tikal and Kaminaljuyu by six centuries. The grave goods bore unmistakably Olmec traits—the very iconography that, in the Middle Preclassic period, linked elites from the Gulf Coast to the Pacific.
Chiapa de Corzo’s cultural and linguistic affiliation is associated with speakers of Mixe-Zoquean languages—the linguistic family many researchers consider the language of Olmec civilization. By 700 BCE, a Zoque civilization was already taking shape here, with its own ritual practices, architecture, and, apparently, the beginnings of writing. When the Guanacaste phase began around 100 BCE, the city became a regional capital: hieroglyphic inscriptions appeared on seals, vessels, stelae, and panels, while Maya ceramic types began to make their way into the pottery assemblage. One pottery sherd bears a probable example of Epi-Olmec writing, dated to around 300 BCE. More than 250 burials from the Formative period have been excavated, and Mound 5 contains an early Mesoamerican palace complex from the 1st century CE.
This was a crossroads city: Olmec heritage, Zoque speech, Maya pottery on the shelves, trade contacts with the coast. And it was in a city like this—not Tikal, not Palenque—that the earliest date in the system we habitually call Maya was found.
Stela 2 from Chiapa de Corzo is, strictly speaking, a wall panel rather than a freestanding monolith. It is badly damaged. Most of its surface is gone, and only fragments of the long inscription remain: a day sign and numerals, enough to read a date in the Long Count system—7.16.3.2.13. Under the standard GMT (Goodman–Martínez–Thompson) correlation, accepted by most modern epigraphers, that date corresponds to 7 December 36 BCE.
What fascinates me is how fragile this evidence is. The entire case for the earliest Long Count date rests on a handful of surviving glyphs on a stone panel found in a city whose history spans more than three thousand years of continuous habitation. The stone could have broken differently, slipped into the river, been reused in the masonry of a later building—and then the laurels would belong to another monument, and the debate over the origins of the Long Count would have taken on rather different contours.
But the stone survived—just enough for the date to be legible. And that date turned out to be a few years older than the next contender.
To understand why the date 7.16.3.2.13 means anything at all, we need to take apart the mechanism behind it. The Long Count is a positional system for recording days counted from a fixed starting point. In the GMT correlation, that point is 11 August 3114 BCE in the proleptic Gregorian calendar. The date carried no historical meaning: it was a mythological “zero mark,” the moment when the gods completed the current creation of the world.
The system’s unit is the kin, one day. Twenty kin make a uinal (20 days). From there, the logic of the vigesimal system makes one exception: the next unit, the tun, contains not 20 but 18 uinals, or 360 days—a number conveniently close to the solar year. Twenty tuns make a katun (7200 days, roughly 19.7 years). Twenty katuns make a baktun (144 000 days, around 394 years). A date was written as a sequence of five numbers: baktuns, katuns, tuns, uinals, kins.
Basically, it is an odometer. It clicks through kins the way a car’s counter clicks through miles, and it can rack up enormous distances in time. The date 7.16.3.2.13 means that 7 baktuns, 16 katuns, 3 tuns, 2 uinals, and 13 kins had passed since the starting point. Multiply and add it up, and you get 1 100 533 days—a little over three thousand years. The system could record dates thousands of years into the future and the past; inscriptions from the Classic period contain calculations reaching millions of years back. The thirteenth baktun, whose completion fell on 21 December 2012, was merely the rollover of one digit on the counter—like the moment an odometer flips from 99 999 to 100 000. No ancient text describes it as the end of the world.
The panic of 2012 sprang from two misunderstandings: the idea that 13 baktuns was the system’s “maximum” (it was not; higher units exist—the piktun, kalabtun, kinchiltun) and the romantic notion of “cyclical time,” according to which the end of a cycle means the end of everything. Ancient Maya texts that mention the date 13.0.0.0.0 treat it as a celebration, not a catastrophe. At Tortuguero, the only monument that speaks directly of events on that date, the text is damaged—and even the surviving fragments point more toward a ritual act by a deity than cosmic annihilation.
Chiapa de Corzo, with its 36 BCE, is not the only early witness. The second-oldest surviving Long Count date is carved on Stela C from Tres Zapotes—a monument of the Olmec and Epi-Olmec tradition in the state of Veracruz, on Mexico’s other coast. This stela was discovered back in 1939 by an expedition led by Matthew Stirling, and its date is reconstructed as 7.16.6.16.18, corresponding to 31 BCE—just five years younger than the Chiapas inscription. Stela C from Tres Zapotes was long considered the oldest Long Count date, until the reading of the Chiapa de Corzo panel moved it into second place.
Both monuments lie outside what we habitually call “Maya territory.” Tres Zapotes is in southern Veracruz, a zone of Olmec heritage and Epi-Olmec culture. Chiapa de Corzo lies in the Grijalva basin, a Zoque setting. Early Long Count dates cluster on and around the Isthmus of Tehuantepec, in a region where the Olmec tradition gradually transformed into the Epi-Olmec one, while contacts with emerging Maya societies were close and constant.
Later epigraphic monuments of the Epi-Olmec tradition—such as La Mojarra Stela 1 (2nd century CE), with its long text in Epi-Olmec script—show that the Long Count continued to be used in non-Maya settings. But it was in the Maya cities of the Classic period—Tikal, Palenque, Copán, Calakmul—that the system reached its greatest complexity and frequency of use. The earliest secure Maya Long Count date, on Tikal Stela 29, belongs to 292 CE—more than three centuries later than the records from Chiapa de Corzo and Tres Zapotes.
The chronological gap between the early dates from the isthmus zone and the first Maya inscriptions is not an abyss, but more a zone of fog: monuments may not have survived, may have been written on wood or plaster, may still be waiting for a shovel. But the map we have stubbornly suggests that the Long Count was born somewhere at the intersection of the Olmec, Epi-Olmec, and Zoque worlds, then migrated eastward.
The temptation is simple: look at the map, point to Chiapa de Corzo or Tres Zapotes, and say, “There—it was invented by the Zoque,” or “It was invented by the Olmecs.” But an archaeological category is not an ethnic group. Olmec culture is a collection of stylistic, iconographic, and architectural traits identified by modern scholars. We do not know for certain what language was spoken by the people who carved the colossal heads at San Lorenzo. The hypothesis that Mixe-Zoquean was the language of the Olmecs is persuasive, but it remains a hypothesis: the Olmecs left no deciphered texts.
At the time Stela 2 was made, Chiapa de Corzo was a city with Olmec heritage and a Zoque cultural environment, one already penetrated by Maya elements. Who exactly carved the date 7.16.3.2.13—a Zoque-speaking priest, a bilingual trader, an invited master from Veracruz? The stone is silent. We know the system, but not the name or language of the person who first thought to count days from a mythical zero point and set them down in positional notation.
I admit, I like that uncertainty. It is more honest than any simple answer. To assign the Long Count to a single culture is to project onto pre-Columbian Mesoamerica the modern logic of intellectual property, where every invention has a patent holder. The Mesoamerican world worked differently: ideas, rituals, and iconographic motifs traveled along trade routes together with obsidian and cacao. Calendar systems—the 260-day ritual cycle, the 365-day Haab—were the common inheritance of all Mesoamerica, from the Zapotecs to the Maya, long before the Long Count appeared. The Long Count was, essentially, a superstructure built atop that shared foundation—an instrument that made it possible to anchor local cycles to an absolute chronology.
If the invention of the Long Count belongs to the misty zone of the isthmus, its flowering is unquestionably tied to Maya civilization in the Classic period (250–900 CE). Maya kings turned the positional count of days into a language of power. Every stela, every lintel, every hieroglyphic stairway began with a date—and that date was not merely a timestamp, but a declaration of the ruler’s place in the cosmic order.
In Palenque, priest-astronomers calculated dates reaching millions of years into the past, linking the birth of the dynasty to the deeds of the gods. In Copán, altars recorded exact intervals between astronomical events—eclipses, the synodic periods of Venus—with an accuracy that still astonishes today. In Calakmul and Tikal, dates on monuments marked military victories, accessions to the throne, dedications of buildings. The Long Count became the backbone of what epigraphers call “dynastic history”—a continuous chronological record with no equivalent in pre-Columbian America.
The Maya added further elements to the basic five-place date: the Lunar Series (information on the current lunar month), the Supplementary Series, and the so-called Number of the Lord of the Night—a cyclical coefficient with nine positions. The simple odometer acquired a dashboard. Every full Long Count date in a Classic Maya text is a complex construction containing information about the day’s position in several interwoven cycles at once. In no other Mesoamerican tradition was the Long Count used with such a degree of detail.
When Classic Maya civilization underwent the collapse of the 9th century, the Long Count gradually fell out of use. Postclassic Maya societies in Yucatán switched to the abbreviated Short Count system—a count of katuns—while the Long Count remained on abandoned stelae in the jungles of Petén, until European explorers began clearing and reading them.
The date on Stela 2 from Chiapa de Corzo—7 December 36 BCE—remains the oldest known surviving Long Count date. But “the oldest surviving” and “the first ever recorded” are fundamentally different things. Tomorrow, a fragment bearing an earlier date may turn up in the jungles of Petén, on Guatemala’s Pacific coast, or in the mountains of Oaxaca—and the map will shift. Archaeology is a science in which every field season can rewrite the textbook.
What this stela truly destroys is the phrase “Maya calendar,” spoken as if it referred to the patented invention of a single people. The Long Count emerged in a contact zone between several traditions: Olmec heritage, Zoque civilization, emerging Maya societies. Its early traces are scattered across the Isthmus of Tehuantepec—from Veracruz to Chiapas. The Maya made something magnificent from it, but the original impulse belonged to a broader world.
And certainly no one in that world built a prophecy of the end of the world into the system. An odometer does not predict that the road will end when the numbers roll over. It simply measures the distance. The damaged panel from Chiapa de Corzo—with its surviving glyphs, its stubborn numerals—reminds us of that better than any textbook. It also reminds us of something else: that in the Grijalva basin, six centuries before Tikal’s flowering, someone was already counting the days from the beginning of the world—and recording the result on a stone we are still trying to read to the end.