
Languages of antiquity
The Aztecs: Nahuatl and picture writing
Central Mexico · from the traditional founding of Tenochtitlan (1325) to the Florentine Codex
The Aztecs of Tenochtitlan spoke Classical Nahuatl; modern varieties of Nahuatl are spoken today by more than one and a half million people. They wrote in pictures: painted books recorded annals, tribute and the divinatory calendar, and the names of places and people were put together like rebuses. How far this amounted to writing down speech is still debated. After the conquest of 1521 Nahuatl came to be written in the Latin alphabet, and it produced what is probably the richest written literature of Indigenous America. For comparison, the page also looks at Inca khipus, records made of knotted cords.
A city in a lake
The Aztecs called themselves Mexica; the name Aztecs, from their legendary homeland of Aztlan, came into use for them only later. According to their own tradition they came into the Valley of Mexico from the north and founded their capital, Tenochtitlan, in 1325 on a marshy island in Lake Texcoco. The sign was to be a place where an eagle sat on a prickly pear growing from a stone. That image appears today on Mexico’s coat of arms and flag.
In 1428 Tenochtitlan formed an alliance with the cities of Texcoco and Tlacopan. During the 15th century this Triple Alliance brought much of central and southern Mexico under its control, and subject towns paid it tribute. When Hernán Cortés entered Tenochtitlan on 8 November 1519, it had an estimated 200,000 to 400,000 inhabitants, placing it among the largest cities in the world at the time. It fell on 13 August 1521 after a long siege, and the Spanish built Mexico City on its ruins.

Classical Nahuatl
The Aztecs spoke Nahuatl. It belongs to the Uto-Aztecan family, whose languages range from El Salvador to the western United States: Hopi, Shoshone and Comanche are distant relatives, and the closest is Nawat (Pipil) in El Salvador. The ancestors of Nahuatl speakers entered Mesoamerica from the north, by linguists’ estimates around 500 CE. In Aztec times Nahuatl was the common language of central Mexico, used even by people with other mother tongues.
Classical Nahuatl is the form of the language written down in the Latin alphabet in the 16th and 17th centuries, based on the speech of Tenochtitlan. It probably reflects the refined speech of the nobility (pipiltin) rather than that of commoners (macehualtin). Nahuatl is agglutinative and polysynthetic: many prefixes and suffixes attach to a root, and a single word can express a whole sentence. Nouns carry a characteristic ending -tl, -tli or -li: tetl “stone”, coyotl “coyote”, calli “house”.
Nahuatl lives on. In Mexico’s 2020 census some 1.65 million people reported speaking it, and Huasteca Nahuatl has the most speakers. Today’s varieties differ from Classical Nahuatl and from one another, some so much that their speakers cannot understand each other.

Picture, sign and rebus
Aztec writing belongs to a tradition of pictorial writing that central Mexico shared with the Mixtecs and other peoples. The tlacuilo, painter and scribe in one, did not write lines of signs but painted images: figures, footprints showing movement, calendar dates, and the glyphs of places and names. Concepts were shown pictorially: death as a corpse wrapped for burial, war as a shield and club, speech as a small scroll coming from the mouth. The signs had no fixed reading order such as Maya inscriptions have.
Pictorial signs could, however, also stand for words (logograms) and exploit their sound: the picture of one word was used for another that sounded the same or similar, as in a rebus. This is clearest in the names of places and people. The name Tenochtitlan was written with a stone (tetl) from which a prickly pear (nochtli) grows; in other names the suffix -tlan “place, by” was represented by a picture of teeth, tlantli in Nahuatl.
How far this phoneticism went is a matter of debate. On the traditional view Aztec writing was not a full writing system, since it could not record any sentence the way Maya writing could, and the painted books were interpreted aloud rather than read word by word. The Spanish epigrapher Alfonso Lacadena argued in 2008, however, that on the eve of the conquest the scribes of Texcoco had a set of syllabic signs at their disposal. Others object that syllabic spellings are known mainly from early colonial manuscripts and may have arisen under the influence of the Latin alphabet.
Numbers were written in base twenty: units with dots, twenty with a flag (pantli), four hundred with a sign like a feather or a small tree (tzontli, “hair”), eight thousand with a pouch (xiquipilli). In the tribute lists of the Codex Mendoza one can therefore read how much of each item a province paid.

Place names as rebuses
A place glyph was made of pictures whose Nahuatl names together sounded like the name of the place:
| name | what is drawn | meaning of the name |
|---|---|---|
| Tenochtitlan | a stone (tetl) from which a prickly pear (nochtli) grows | traditionally “by the prickly pear on the stone”; the etymology is uncertain |
| Mapachtepec | a raccoon (mapachtli) on a hill (tepetl) | “raccoon hill” |
| Mazatlan | a deer’s head (mazatl) and teeth (tlantli) | “place of deer” |
| Huitztlan | a thorn (huitztli) and teeth (tlantli) | “place of thorns” |
| Chapultepec | a grasshopper (chapolin) on a hill (tepetl) | “grasshopper hill” |
| Coatepec | a snake (coatl) on a hill (tepetl) | “snake hill” |

Painted books
Aztec books (amoxtli) were painted on long strips of fig-bark paper (amatl), coated with a ground layer and folded like an accordion. They held annals, genealogies, maps, tribute lists and divinatory calendars. Most were lost after the conquest; according to later reports the first bishop of Mexico, Juan de Zumárraga, had painted books burned, though the extent of the destruction is judged differently. Whether any pre-Columbian manuscript from Tenochtitlan and its surroundings survives at all is uncertain.
The Codex Boturini (Tira de la Peregrinación) is a strip of amatl paper 5.49 metres long, folded into 21½ leaves. It tells of the Mexica migration from Aztlan: dates are framed, places are shown by glyphs, and footprints mark the route between them. When it was made is unknown. Its style once led scholars to regard it as pre-Columbian; today it is more often placed shortly before or just after the conquest. It is named after the 18th-century Italian collector Lorenzo Boturini Benaduci.
The Codex Borbonicus is a single strip of amatl paper about 14 metres long, originally folded into 40 pages, of which the first two and the last two are missing. It contains a divinatory calendar, an overview of the 52-year cycle and the New Fire ceremonies. Whether it was made before or shortly after the conquest is disputed: there is empty space above the calendar pictures, as if Spanish glosses were already expected. Since 1826 it has belonged to the library of the French National Assembly in Paris.
The Codex Mendoza was made after the conquest, probably around 1541 and certainly between 1529 and 1553, on 71 leaves of European paper. Nahua painters used traditional glyphs to record the history of Tenochtitlan from 1325 to the conquest, a list of tributary towns with their dues, and scenes of daily life; a Spanish interpreter added explanations. It is named after Antonio de Mendoza, viceroy of New Spain, who on the prevailing view had it made for the king. By 1553, however, the manuscript belonged to the French cosmographer André Thevet, and since 1659 it has been in the Bodleian Library in Oxford.


Counting the days
Like the Maya, the Aztecs kept two calendars side by side. The divinatory tonalpohualli (“count of days”) had 260 days: twenty day signs were combined with the numbers 1 to 13, so that each day had a name such as 1 Crocodile, 2 Wind, 3 House. After thirteen days a new period began, with its own patron deity; the first part of the Codex Borbonicus depicts these periods. The solar year, xiuhpohualli, had 365 days: 18 months of twenty days plus five leftover days (nemontemi), considered unlucky.
The two calendars came round together every 52 years. Years were named after a particular day (the so-called year bearer), and only four signs were possible, House, Rabbit, Reed and Flint, each with a number from 1 to 13. The end of a 52-year cycle was marked by the New Fire ceremony: fires were put out and priests kindled a new one. The year 1519, when Cortés landed in Mexico, was the year 1 Reed (ce acatl).
The best-known Aztec monument bearing calendar signs is the Sun Stone (Piedra del Sol), often called the Aztec calendar. The stone disc, about 3.6 metres across, was found by workmen on 17 December 1790 during works on the main square of Mexico City. The face in the centre is most often interpreted as the sun god Tonatiuh, by others as the earth deity Tlaltecuhtli. It is surrounded by the sign 4 Movement (nahui ollin), the name of the present, fifth world age, and the next ring holds all twenty day signs. On the prevailing view the stone was not used to reckon dates; it was a ceremonial monument.
The twenty day signs
- cipactlicrocodile (caiman)
- ehecatlwind
- callihouse
- cuetzpalinlizard
- coatlserpent
- miquiztlideath
- mazatldeer
- tochtlirabbit
- atlwater
- itzcuintlidog
- ozomatlimonkey
- malinalligrass
- acatlreed
- ocelotljaguar
- cuauhtlieagle
- cozcacuauhtlivulture
- ollinmovement
- tecpatlflint (knife)
- quiahuitlrain
- xochitlflower
ce acatl, “1 Reed”the name of a day and of a year; the year 1 Reed fell in 1519
Unicode does not yet encode Aztec signs (as of 2026), so the table gives only the order, the Nahuatl name and the meaning of each sign. Their appearance can be seen in the photographs of the Codex Borbonicus, the Codex Boturini and the Sun Stone.


Nahuatl in the Latin alphabet
After the conquest Nahuatl began to be written in the Latin alphabet. The first texts appeared within twenty years of the Spanish arrival, and over the 16th century the new script gradually displaced the pictorial one, although painted manuscripts with glosses continued to be made for a long time. In 1536 the Franciscans founded the College of Santa Cruz at Tlatelolco, where sons of the Indigenous nobility learned Latin, Spanish and how to write their own language. Andrés de Olmos completed a grammar of Nahuatl in 1547, and Alonso de Molina published a Spanish–Nahuatl dictionary in 1555. Annals, wills, legal records, songs and plays were written in Nahuatl.
Spanish spelling, however, did not capture everything Nahuatl distinguishes: most writers did not mark vowel length or the glottal stop (saltillo). They were first marked accurately by the Jesuit Horacio Carochi in his grammar of 1645, length with a macron and the glottal stop mainly with a grave accent. In Nahuatl the letter x is pronounced like English sh, the digraph tl is a single consonant [tɬ], and hu before a vowel sounds like w.
The most famous work is the Florentine Codex. The Franciscan Bernardino de Sahagún (c. 1499–1590), who came to Mexico in 1529, spent decades with former students from Tlatelolco questioning elders versed in Aztec culture and having their accounts written down in Nahuatl. The result was twelve books on the gods, rituals, the calendar, society, nature and the conquest of Mexico, told from the point of view of the defeated. The text runs in two columns, Nahuatl on the right and Spanish on the left, accompanied by more than two thousand pictures by Indigenous painters. The manuscript, completed around 1577, is now in Florence.
From the codices



Words from Nahuatl
Many Nahuatl words were borrowed into Spanish and passed from there into other languages. In Spanish the ending -tl, -tli usually became -te or -le (tomate, coyote, chicle), and in other languages it was often dropped:
| Nahuatl | meaning | English |
|---|---|---|
| tomatl | tomatillo, a green fruit; the red tomato was xitomatl | tomato |
| ahuacatl | avocado | avocado |
| chilli | chili pepper | chili, chilli |
| coyotl | coyote | coyote |
| ocelotl | jaguar | ocelot (the name passed to a smaller cat) |
| tzictli | sticky substance, gum | chicle |
| axolotl | a kind of salamander | axolotl |
| Mexihco | the name of the Aztec capital | Mexico |
| uncertain | a cacao drink | chocolate; the word probably entered Spanish from Nahuatl or Nawat, but its source form is disputed (the form chocolatl is not attested in early sources; chicolatl has been proposed) |
Khipu: knots on cords
In the Andes of South America records were kept in an entirely different way. The Incas, who in the 15th century built the vast empire of Tawantinsuyu from Cusco, had no writing in the usual sense. Their administration relied on khipus (Quechua khipu, “knot”): bundles of cotton or llama and alpaca wool cords hanging from a main cord. The information lay in the position and type of the knots and in the colour, order and length of the cords; a khipu could have a few cords or several thousand. They were tied and interpreted by specialist officials, the khipukamayuq, “khipu keepers”. The main language of administration was Quechua; in the colonial period it gave rise to the literary Classical Quechua, written in the Latin alphabet.
Khipus are older than the Inca empire: the first undisputed examples date from about 600–1000 CE and are attributed to the Wari state, while earlier finds from Caral are disputed. After the conquest of Peru (from 1532) the Spanish partly used them for their own administration, but in 1583 the Third Council of Lima ordered some khipus burned as idolatrous. According to a 2021 survey, just under 1,400 survive.


Did khipus record language?
Most of the khipus studied contain numbers. In 1912 the American Leslie Leland Locke showed that they are recorded in a decimal positional system: clusters of knots at different heights on a cord stand for units, tens, hundreds and higher places. In the lowest place a one is tied as a figure-eight knot and two to nine as a long knot with the corresponding number of turns; an empty position means zero. That these are accounts is confirmed by the sums: the value of one cord often equals the total of others.
Whether khipus also recorded language, such as names or histories, is uncertain. According to colonial sources the khipukamayuq supplied the Spanish with information on censuses, tribute, the calendar and genealogies, but exactly how khipus carried it is unknown. The anthropologist Gary Urton regards khipus as a system of binary-coded signs, and in 2005 he and Carrie Brezine proposed that a sequence of three figure-eight knots at the start of khipus from Puruchuco near Lima may stand for the name of that place. In 2017 Sabine Hyland argued that colonial “epistolary” khipus from the village of San Juan de Collata also contain phonetic signs. There is, however, no generally accepted decipherment, and khipus are therefore not usually regarded as writing in the narrow sense, that is, as a record of spoken language.
Timeline
- 1325Traditional founding of Tenochtitlan.
- 1428Triple Alliance of Tenochtitlan, Texcoco and Tlacopan.
- 8 Nov 1519Hernán Cortés enters Tenochtitlan.
- 13 Aug 1521Fall of Tenochtitlan.
- 1536The Franciscans found the College of Santa Cruz at Tlatelolco.
- c. 1541The Codex Mendoza is made.
- 1547Andrés de Olmos completes a grammar of Nahuatl.
- 1555Alonso de Molina’s dictionary.
- 1565The Codex Osuna.
- c. 1577Sahagún and his collaborators complete the Florentine Codex.
- 1583The Third Council of Lima orders some khipus burned.
- 1645Horacio Carochi’s grammar, with an exact notation of pronunciation.
- 17 Dec 1790The Sun Stone is found on the main square of Mexico City.
- 1912Leslie Leland Locke: khipus record numbers in base ten.
- 1978The discovery of the Coyolxauhqui stone starts the main phase of excavation at the Templo Mayor (1978–1982).
Nahuatl and Quechua on the globe
Tap a language to show it on the globe:
Sources and further reading
- Elizabeth Hill Boone: Stories in Red and Black. Pictorial Histories of the Aztecs and Mixtecs. University of Texas Press, 2000.
- Frances F. Berdan, Patricia Rieff Anawalt: The Essential Codex Mendoza. University of California Press, 1997.
- James Lockhart: The Nahuas After the Conquest. Stanford University Press, 1992.
- Michel Launey: An Introduction to Classical Nahuatl. Cambridge University Press, 2011.
- Marcia Ascher, Robert Ascher: Code of the Quipu. A Study in Media, Mathematics, and Culture. University of Michigan Press, 1981.
- Gary Urton: Signs of the Inka Khipu. Binary Coding in the Andean Knotted-String Records. University of Texas Press, 2003.
- Glottolog 5 (languages and their classification).