Mayan Glyphs: How The Maya Script Was Read Again

Maya writing is logosyllabic. It combines logograms, signs standing for whole words or morphemes, with syllabograms, signs standing for consonant-vowel syllables, along with semantic determinatives and phonetic complements. Our research file puts the total inventory at roughly 800 distinct signs, of which roughly 500 are logograms and roughly 200 syllabograms. It is the only Mesoamerican writing system known to represent spoken language fully, which is the fact everything else on this page depends on: these are not symbols standing for ideas, they are a script for a language.

Syllabograms follow a consistent consonant-vowel pattern: ba, ka, la, ma and so on. To write a word ending in a consonant, two syllabograms are used and the final vowel of the last one is not pronounced. So ba-la-ma writes balam, jaguar. The corpus calls this the rule of synharmony. A logogram is also often accompanied by a syllabogram that confirms the reading, a phonetic complement: the jaguar logogram with a -ma sign attached, telling the reader the word ends in an /m/. The system is a belt and braces arrangement, and that redundancy is part of why it could eventually be recovered.

Maya texts were written in paired columns, read left to right and top to bottom in paired glyph blocks. Each block typically carries a word or short phrase, with the individual signs fitted together inside the block in visually complex arrangements rather than strung out in a line. This is one of the reasons the script resisted reading for so long: a European eye looking for a row of letters finds instead a grid of dense, self-contained pictures, each of which is doing several jobs at once.

The script was in use among the ancient Maya of what is now southeastern Mexico, Guatemala, Belize and Honduras from at least the third century BCE through the sixteenth century CE. That is a span of roughly eighteen centuries of continuous literate practice, which ends not because the script failed but because of what arrived in the sixteenth century.

Bishop Diego de Landa presided over a burning of Maya books at Mani in 1562, and other missionaries destroyed books elsewhere. An unknown number of codices carrying historical, astronomical, medical and ritual knowledge were destroyed. You will often see a specific figure for Mani, a precise count of books on a precise day. This page does not print one. That number descends from Landa's own account through four centuries of retelling and was not verified for this article, and the total destroyed across the whole campaign is simply unknown. What can be said is the shape of it: a literate civilization's library, reduced to four books.

The same man is on both sides of this story. Landa's Relacion de las cosas de Yucatan, written around 1566, documented his understanding of Maya writing, calendar and culture, and became the key source for later decipherment. Our own research file puts it as a paradox, and the word is right, but the irony has to be stated accurately or it turns into a redemption arc it does not deserve. Landa did not save the script. He recorded data he had misunderstood, and the misunderstanding is what made the data usable three hundred years later.

Yuri Valentinovich Knorozov (1922 to 1999), a Soviet epigrapher, published the foundational paper in 1952: Drevnyaya pis'mennost' Tsentral'noy Ameriki, Ancient Writing of Central America, in Sovetskaya Etnografiya, issue 3, pages 100 to 118. He demonstrated that Maya signs carried phonetic-syllabic values, taking Landa's sixteenth-century plate as a starting point, applying its assignments to known texts, and producing coherent readings. That is the breakthrough, and it is exactly as large as it sounds: he established that the marks were sound.

There is a famous story in which Knorozov, a Red Army soldier in Berlin in 1945, rescues a volume of Maya facsimiles from a burning library, and that book becomes the seed of the decipherment. This page does not tell it, because it is not established. Accounts by and about Knorozov place his wartime service away from the capture of Berlin, and he is reported to have denied the fire story himself. The true version is narrower and better: he did the decisive work in Moscow, from published facsimiles and printed sources, without having stood at a Maya site. Nor is the isolation permanent, which the legend usually implies. He travelled to Guatemala in 1990 and to Mexico from the early 1990s, and the Mexican government awarded him the Order of the Aztec Eagle in 1994.

One excuse can be removed immediately. Knorozov's argument was available in English, in the flagship American archaeology journal, within six years: The Problem of the Study of the Maya Hieroglyphic Writing, American Antiquity volume 23, issue 3, pages 284 to 291, in 1958. Western Mayanists were not unable to read him. They read him and did not accept him.

Our research file states that the phonetic approach was rejected for decades by the dominant figure in the field, J. Eric S. Thompson, who insisted Maya writing was ideographic, recording concepts and calendar computations rather than spoken language, and that Thompson's authority delayed acceptance until after his death in 1975. That is the mechanism, and it is worth being precise that the opposition was published rather than merely social: Thompson set out his methodological case in American Antiquity in 1959, and his book-length statement of the ideographic position is Maya Hieroglyphic Writing: An Introduction, first published in 1950. This was an argument conducted in print, by a serious scholar, and lost.

Our file also says the Cold War context exacerbated the dispute, that Western scholars dismissed Soviet scholarship, and that Knorozov could not travel to Maya sites or Western institutions. Hold that at its real strength and no higher. The Cold War is genuine context, and it is not the mechanism. What made the phonetic reading unpublishable in practice, for years, was one man's standing inside a small field, not a clean East-versus-West split. The distinction matters because the two explanations point at different lessons. One says politics can poison science from outside. The other says a discipline small enough to have a single dominant figure can be wrong for as long as he is, and that is the harder problem.

Heinrich Berlin, in 1958, identified emblem glyphs: title compounds attached to rulers, combining a site-specific main sign with a holy prefix, k'uhul, and the ajaw or lord title. Recognising them made it possible to map Maya political geography for the first time. Emblem glyphs have since been identified for Tikal (Mutal), Palenque (Lakamha'), Copan, Calakmul (Kaan), Caracol and Yaxchilan, and what they reveal is a landscape of competing city-states rather than a unified empire. Knorozov gave the texts sound, Proskouriakoff gave them people, Berlin gave them places. All three are required.

At the first Mesa Redonda de Palenque in 1973, Linda Schele and Peter Mathews constructed the first full dynastic history of a Maya city, reading king lists, dates and events straight off the monument inscriptions at Palenque. This is the moment the method becomes ordinary work: not an argument that the texts could be read, but a room full of people reading them.

The language recorded in Classic Period (250 to 900 CE) inscriptions has been identified as Classic Ch'olti'an, an ancestor of the modern Ch'ol and Ch'orti' languages rather than of Yucatecan. Ch'olan morphological patterns, ergative-absolutive alignment, verbal suffixes and possessive constructions, match the patterns observed in the glyphs. Our own research file files this as settled and then, in its own Counter-Arguments section, says the identification remains debated by scholars. Both halves are carried here, because printing only the first would be quoting our file against itself.

Pre-Classic inscriptions may record a different Mayan language, or a more archaic stage of one, and Late Postclassic texts from Yucatan, including the Madrid and Paris codices, show Yucatecan features. So the script was used to write more than one Mayan language across its life, which is what you would expect of a working writing system used for eighteen centuries over a large region, and which also means that reading a passage aloud correctly does not settle which language you are hearing.

You will see a figure of about 85 percent, and this page will not print it without saying what it is a percentage of, because the two available meanings differ sharply. Our research file says roughly 85 percent of KNOWN SIGNS can be read with confidence, with the remaining share covering rare or damaged signs and signs whose values are debated. The figure as it usually circulates in the wider literature means something else: roughly 85 percent of RUNNING TEXT, which is a much easier number to reach, because the commonest signs are the best understood and a small number of them carry most of any given inscription. A script can be 85 percent readable by the second measure while a large minority of its sign catalogue still has no secure reading. Both things are true at once and only one of them sounds like an ending.

The earliest known Maya writing includes the San Bartolo murals at roughly 300 BCE, the Kaminaljuyu and El Porton stelae at roughly 400 to 200 BCE, and, more contentiously, the Cascajal Block. San Bartolo has a named primary publication, Saturno, Stuart and Beltran in Science in 2006, and the site produced a further result our research file has not caught up with: a 2022 paper in Science Advances by Stuart, Hurst, Beltran and Saturno reporting an early Maya calendar record from the same site.

How Maya writing relates to earlier Mesoamerican scribal traditions, particularly Zapotec at Monte Alban from roughly 500 BCE and Epi-Olmec or Isthmian from roughly 300 BCE to 200 CE, is genuinely debated. The script may share common roots with these systems in a broader Mesoamerican scribal tradition, or may represent independent development inside a shared cultural framework. Olmec marks on monuments at La Venta and San Lorenzo consist of isolated symbols whose status as true writing remains uncertain.

The Cascajal Block, published in Science in 2006 as the oldest writing in the New World, is contested more deeply than our research file records. Our file flags only that a Maya attribution is debated. In fact the object's authenticity and its status as writing at all were both challenged in print, in the same journal, within months: Bruhns and Kelker's technical comment, Did the Olmec Know How to Write?, appeared in Science in 2007. An object whose authenticity is in question cannot carry weight in an origins argument, and this page does not let it.

The fourth book is worth following, because a decades-long authenticity dispute actually resolved. The manuscript surfaced in private hands in the 1960s, was first exhibited at the Grolier Club in New York, and was doubted for decades. Coe, Houston, Miller and Taube argued the case for authenticity in The Fourth Maya Codex in 2015, Mexico's Instituto Nacional de Antropologia e Historia then ran a multidisciplinary study, and the manuscript is now recognised as genuine and renamed the Maya Codex of Mexico. The lesson is not that doubters are always wrong. It is that the dispute was settled by material analysis rather than by argument, which is the difference between this case and the Cascajal Block.

Thompson's long-dominant insistence that Maya signs were ideographic or calendrical rather than phonetic is refuted, and this page does not present it both-sidedly. It was definitively overturned by the successful syllabic reading of thousands of texts; the script records full spoken language in syntactic detail. This is a settled refutation and not a live debate. It is also, note, the only claim on this page that was ever the mainstream position, which is a useful thing to remember about mainstream positions.

Claims that Maya inscriptions encode secret messages about alien contact, Atlantis or future predictions are refused, and the refutation here is unusually direct: we can read the texts. They record dynastic history, warfare, ritual dedication, astronomical observation and divinatory practice. The content is culturally rich and historically grounded, and it is about the people who wrote it.

The notion that Maya inscriptions predicted catastrophe or transformation on 21 December 2012, the 13.0.0.0.0 Long Count station, is a modern fabrication. The Maya texts that mention the date treat it as the completion of a calendrical cycle, a milestone comparable to a millennium rollover, not as destruction. Two corrections to how this is usually reported. Our own file says only two ancient inscriptions mention the date; a second monumental reference was published in 2012, Block V of Hieroglyphic Stairway 2 at La Corona, Guatemala, excavated that spring, alongside the long-known Tortuguero Monument 6. And the date itself is printed two ways in serious Mayanist writing, 21 December and 23 December 2012, a two-day discrepancy that traces to the unresolved debate over how Long Count dates map onto the Gregorian calendar.

Millions of Maya people are alive today across Mexico, Guatemala, Belize and Honduras, and roughly thirty Mayan languages are spoken. This is not a coda. Nothing on this page is an elegy for a disappeared civilization, and the register that treats Maya writing as the relic of a vanished people is both false and, once you notice it, strange: the descendants of the people who cut these inscriptions can be asked. Mayan language communities are the subject of active revitalisation work, much of it led by Maya linguists themselves.

Maya people are participants in the epigraphy rather than only its object of study. The workshops teaching hieroglyphic writing to Maya people are documented in the mainstream anthropological literature, in a chapter of Maya Cultural Activism in Guatemala published by the University of Texas Press in 1996. A script that was read again by outsiders is now being read by the descendants of the people who wrote it, which is a better ending than the one where a code gets cracked.

How widely Maya writing was read beyond the elite scribal class is unclear. Some scholars argue literacy was restricted to specialists, the aj tz'ib, he of the writing. Others note that public monuments were designed to be seen by large audiences, which suggests at least partial visual literacy among non-elites, and that painted ceramics carrying hieroglyphic captions were widely distributed, possibly extending some text exposure beyond palace contexts. This is genuinely open and the evidence available does not settle it.