We Finally Know What The Maya Calendar Was For

A seventeen-minute popular-history video from Michael Button’s channel, built toward one finding about the Maya 819-day count. It starts with what was lost. Diego de Landa burned the books of Maní in July 1562, the last Maya kingdom fell at Nojpetén in 1697, and four codices survive. What the Maya knew about the sky has to be read from stone, from building alignments and from those four books.

The middle of the video is a tour of naked-eye astronomy. Buildings at Uaxactún and elsewhere were lined up on sunrise points, and an archaeoastronomer’s survey found the dates they mark fall at intervals of 13 or 20 days, units of the Maya calendar. The Dresden Codex Venus table rounds the planet’s 583.92-day cycle to 584 and carries corrections that, on one modern reading, keep the error to about two hours over roughly five centuries. Its eclipse table restarts at intervals that match what astronomers now call the saros and inex. Button is careful in places: the Chichén Itzá equinox serpent can be seen for about a week, the “Star Wars” link between Venus and Maya warfare does not hold up against the dates, and the Venus accuracy rests on how one page is read.

Then the count itself. Carved at Palenque and Yaxchilán, it was read in 1961 by Heinrich Berlin and David Kelley as four stations tied to directions and colours, which fits Mercury (7 × 117 days) and nothing else, so for decades it was treated as ritual. In 2023 John Linden and Victoria Bricker ran it to 20 stations (16,380 days, about 45 years) and found that the synodic periods of all five naked-eye planets line up with its station points. The video’s ending irony: Landa, who burned the books, also wrote down the notes on the script that later helped decipher it.

The details check out. Landa’s own tally was 27 books and about 5,000 images; the Grolier codex, long called a forgery, was authenticated in 2016 and is now the Maya Codex of Mexico; the 9.5 to 16 million population figure is a 2025 lidar estimate for the central lowlands; the eclipse-table claim is Justeson and Lowry’s 2025 paper (55 of its 69 entries read as solar-eclipse warnings). The headline claims more than the research does. Linden and Bricker show that the count fits, which is not the same as reading what the Maya said it was for, and the video concedes that the glyphs never say. The fits use rounded whole-day periods (117 days for Mercury, whose cycle is about 115.9), and Venus fits only roughly.