The Burned Library of Leningrad

Yuri Knorozov and the Phonetic Key to Maya Hieroglyphs

Walking past a shelled library on the outskirts of Berlin in May 1945, Red Army artillery spotter Yuri Valentinovich Knorozov pulled two boxes from the ash. Inside lay two rare books that had survived unburned. The first was an 1864 Paris edition of Bishop Diego de Landa's sixteenth-century manuscript Relación de las cosas de Yucatán, edited by Charles Étienne Brasseur de Bourbourg. The second was Paul Schellhas's 1930 Berlin reproduction of the three surviving pre-Columbian Maya bark-paper books: the Dresden, Madrid, and Paris Codices. Knorozov carried both volumes back to Moscow.

Two years later, Knorozov asked his ethnography professor at Moscow State University, Sergei Tokarev, whether anyone could decipher the Maya script. Tokarev pointed to a 1945 article by Paul Schellhas arguing that the decipherment of Maya hieroglyphs was an insoluble problem. In the West, Maya epigraphy was ruled by British archaeologist J. Eric S. Thompson of the Carnegie Institution. Scholars since Ernst Förstemann in the 1880s could read the bar-and-dot numbers and Long Count calendar dates on Maya stelae. Yet the remaining eighty percent of the glyphs remained mute. Thompson enforced a strict orthodoxy. He maintained that the Classic Maya were peaceful astronomer-priests who never recorded wars, biographies, or spoken language on stone. According to Thompson, Maya glyphs were esoteric ideograms representing abstract concepts without spelling phonetic words.

Moving in 1949 to a tiny room inside the Museum of Ethnography in Leningrad, where he slept on a cot beside his desk, Knorozov ignored Thompson's ban. He analyzed the three Maya codices using positional linguistics and sign-frequency statistics. Counting every distinct graphic element across the Dresden, Madrid, and Paris Codices while excluding compound ligatures and minor variants, he found that the Maya texts employed roughly 355 unique active signs.

That single number demolished Thompson's dogma. Every known writing system on Earth falls into one of three regimes based on its active sign inventory:

• - The Typological Sign-Count Law: Purely alphabetic scripts (such as Latin or Cyrillic) represent phonemes and require 20 to 35 signs. Purely syllabic scripts (such as Linear B or Japanese kana) represent open syllables and require 50 to 100 signs. Mixed logosyllabic scripts (such as Egyptian hieroglyphs or cuneiform) combine logograms with phonetic syllable signs and employ several hundred to a few thousand core signs. With 355 active signs, Maya writing had to be a mixed logosyllabic script capable of spelling spoken Maya words phonetically.
• - Reinterpreting De Landa's Alphabet as Syllables: In 1566, after burning Maya books at Maní, Bishop Diego de Landa forced a literate Maya noble, Gaspar Antonio Chi, to write down the glyphs matching the Spanish alphabet (A, B, C...). When nineteenth-century scholars plugged de Landa's twenty-seven glyphs into the codices as single letters, the results produced gibberish. Knorozov reconstructed the scene. When the bishop pronounced the Spanish letter names aloud, the scribe drew the Maya syllabic glyphs matching those spoken syllables (a, be, se, ku).
• - The Law of Synharmony ($CV$-$CV$ Echo Spelling): Armed with de Landa's syllable signs, Knorozov turned to the Dresden Codex. Above a drawing of a turkey stood a two-glyph compound. In colonial Yucatec dictionaries, a turkey was called kutz, a closed consonant-vowel-consonant ($CVC$) syllable. The first glyph was de Landa's sign for ku (the letter name Q). Because Maya phonetic signs represented open consonant-vowel ($CV$) syllables, Knorozov deduced synharmony. Maya scribes wrote a closed $C_1 V_1 C_2$ root using two open $CV$ signs ($C_1 V_1$-$C_2 V_1$), where the second vowel echoed the first. Since turkey was kutz, the second sign had to be tzu (ku-tzu). Checking the glyph compound above a dog (tzul in Yucatec), Knorozov found that its first sign was the exact same tzu glyph, followed by de Landa's sign for lu, spelling tzu-lu (tzul, dog).

When Knorozov published his report in Sovetskaya Etnografiya in 1952 and defended his dissertation in March 1955, the committee awarded him an immediate doctorate. Thompson launched a twenty-year campaign to discredit him. Only after Thompson's death in 1975 did Western epigraphers led by Floyd Lounsbury, Linda Schele, Peter Mathews, and David Stuart test Knorozov's grid on the monuments of Palenque and Copán. Every glyph block obeyed Knorozov's logosyllabic rules. In 1990, Knorozov finally stood atop the temples of Tikal and Palenque, reading aloud the dynastic names of Classic Maya kings.