Teotihuacan

The city began around 100 BC and ceased to be a political force in the second half of the sixth century AD, although people went on living among its ruins for another thousand years. In that span it produced a third great temple platform, more than 2,000 walled apartment compounds, an extraordinary body of mural painting, and an obsidian industry that reached from the Basin of Mexico to the Maya lowlands. It produced no named king. There is no royal portrait, no dynastic stela, no tomb inscribed with a ruler's name, no king list, no scene of a captive being humiliated before a throne. Set beside the boastfulness of Maya monuments or the later Mexica sculptures, the silence is not modest. It looks like a political choice.

Teotihuacan's reach is easier to trace abroad than at home. In AD 378 an expedition linked to the city arrived at Tikal in the Petén and replaced its king with a child who wore highland costume. Green obsidian from the Pachuca source, north-east of the city, travelled in the same centuries to Tikal and beyond. Its architecture — the sloping talud with the projecting tablero above it — was copied from the Guatemalan highlands to the Gulf coast. Long after the city was empty, Maya dynasts still claimed a western origin, and the Mexica who built Tenochtitlan gave the ruins the name by which we know them.

At a glance

How did a valley of springs become the largest city in the Americas?

The Basin of Mexico is a high closed basin ringed by volcanoes, with a chain of shallow lakes along its floor at more than 2,000 m above sea level. Teotihuacan sits not on the lakeshore but in a side valley to the north-east, where springs fed the San Juan river and volcanic soils lay deep on the slopes. The setting had one great advantage and one great weakness: water rose from the ground in quantity and could be led into fields, but the valley is high enough for frost and dry enough for harvests to fail. Irrigation was not an optional improvement there. It was the condition of living at scale.

People had farmed the valley since at least 300 BC, in scattered hamlets of perhaps a few thousand souls in all. Around 100 BC those hamlets began to coalesce into a town with a planned ceremonial core, and within a century or two the town had become a city of a kind Mesoamerica had not seen. The old explanation for the surge was catastrophe next door. Cuicuilco, on the southern shore of Lake Texcoco, was the largest centre in the basin in the late first millennium BC, and the eruption of the Xitle volcano buried its fields in lava. Displaced Cuicuilco families, the argument ran, moved north and built Teotihuacan. The city's three-temple complexes, moreover, reproduce a domestic arrangement known from Tetimpa, a village destroyed by Popocatépetl, which makes the migration story more than a demographic guess.

The dating of that eruption has since moved, and with it the neatness of the story. Radiocarbon work summarised in a 2014 study of Teotihuacan's political organisation places the Xitle flow in the third century AD, some 300 years too late to explain the city's first growth. What fits the chronology better is an earlier eruption of Popocatépetl, in the middle of the first century AD, which one line of argument credits with displacing some 50,000 people northwards into the Teotihuacan valley; a smaller eruption about fifty years later would have pushed a second wave. Volcanism, migration and the pull of a growing market probably worked together, and the sequence of eruptions fits the two pulses of growth better than any single catastrophe does.

What is not in doubt is the speed and scale of the result. By about AD 150 the population had reached a conservative 80,000–100,000, and much of the basin's population had either moved into the city or was supplying it. Most of the growth was over by about AD 200; after that the built-up area changed little while the population levelled off. That sequence matters for everything that follows: the pyramids, the grid and the first apartment compounds were all laid out before the city was full, by a generation or two of people who could see what they were building.

Why is the whole city on one grid?

The spine of Teotihuacan is the Avenue of the Dead, called Miccaotli by the Mexica, who saw the mounds along it and assumed they were tombs. It runs north–south for about 2.4 km from the Ciudadela to the Plaza of the Moon, and with its northern continuation towards Cerro Gordo it is often described as 5 km long. It is 40 m wide and lined with close to 80 platforms and structures, most of which once carried temples. The terrain rises along its length; the builders dealt with the slope by inserting sunken plazas at intervals, so that the route moves in steps rather than a ramp.

The avenue and every building around it are set about 15.5° east of astronomical north. That single decision, repeated for centuries, is the strongest evidence that someone was in charge of the whole enterprise. It is also the most argued-over fact about the city. Alignments with the sun on particular days, with Venus, with the zenith passage, and with the sacred mountain behind the Pyramid of the Moon have all been proposed, and none has carried the field. What can be said is that the Pyramid of the Sun appears to have been laid out first and the rest of the city followed its axis, which suggests that the orientation came out of a specific cosmological observation at that building and then became canonical.

Two qualifications keep the grid from being a simple proof of central planning. The first is that most of the streets between compounds are short and discontinuous rather than long and straight, and the internal layout of each compound is unique, which is not what a state housing programme produces. The second is that the city's water system ran along the same axis: Mexico's national archaeological institute describes the avenue as a hydraulic as well as a ceremonial spine, with canals and channels feeding off it. A grid that carries both processions and irrigation is doing two kinds of work at once, and obedience is only one possible explanation for why everybody kept to it.

The alternative is that the grid was sacred rather than enforced. If the city was understood as a cosmogram, a model of the ordered world with the avenue as its axis, then building your compound on the correct alignment was a religious act, not a tax. Immigrants from a dozen regions, arriving over a century, could have reproduced the alignment voluntarily because the alignment was what made a house part of the city. That reading fits the archaeological detail better than the image of a planner with a ruler, and it is the reading that the most recent modelling work on Teotihuacan's government has adopted.

What were the three great pyramids for?

The Pyramid of the Sun is the largest structure at the site: roughly 222 by 225 m at the base and about 63–65 m high, built of compacted soil, stone and volcanic rubble faced with dressed stone and stucco, and oriented so that its main stair faces west. Its core was raised over a natural lava tube that was enlarged into a tunnel about 103 m long, ending in a four-lobed chamber. Doris Heyden read that cave as a place of emergence, the womb from which people came into the world in central Mexican thought, and the pyramid as a built mountain over it. The dedication of the pyramid itself is unknown. The name is Mexica, given six or seven centuries after the city fell.

Offerings were placed at the corners and stages of the Pyramid of the Sun as it rose, including the remains of children, and the building was enlarged more than once before it reached its final size. The Pyramid of the Moon, at the northern end of the avenue, is smaller — about 140 by 150 m at the base and 45 m high — but stands on higher ground and closes the view north with Cerro Gordo behind it, so that the mountain appears to continue the pyramid. It was rebuilt over at least seven phases. Its summit carried no durable temple, which has been read as evidence that it was a stage for rituals performed in the plaza below and watched by crowds rather than a house for priests.

The third great platform, the Temple of the Feathered Serpent, sits inside the Ciudadela, a sunken enclosure whose four platforms measure about 400 m to a side and which contains fifteen smaller temples. It has one public entrance, on the west. The temple itself had seven sculpted tiers, of which four survive; feathered serpents undulate through shells and conches, and faces with obsidian-inlaid eyes project from the façade. Before construction began, an estimated 200–260 people were killed and buried in and around the structure, 137 of them documented archaeologically, accompanied by a dedicatory deposit of around 100,000 objects. Many wore military dress, with slate discs on their backs and necklaces of imitation human jaws carved from shell. Isotope analysis of 41 of them showed that most had lived in the city for years, some from childhood, and others had come from several foreign regions. They were not, for the most part, captured strangers.

That deposit is the strongest single argument that early Teotihuacan had rulers of the familiar Mesoamerican kind, people who could command the labour of thousands and the lives of hundreds for their own consecration. It is also, read differently, an argument for the opposite: a state temple whose dedication was a collective act by a whole community, performed once and never repeated. There is no second Feathered Serpent temple at Teotihuacan, no series of later royal burials, and no ballcourt of the I- or T-shaped kind that elsewhere in Mesoamerica went with elite power and sacrifice.

Who lived in the apartment compounds?

Almost everybody. More than 90 per cent of the city's inhabitants appear to have lived in walled apartment compounds, possibly the only premodern city of its size in which apartment living was the norm rather than the exception. A compound was a single-storey rectangle, typically around 60 m on a side, with high windowless walls facing the street, one or two entrances, and a series of patios around which rooms were grouped. Each patio and its surrounding rooms formed a dwelling for one household; a compound held three to five of them, and a household probably numbered five to twelve people. Drainage, plaster floors, shrines and small family tombs were built in. The richest compounds stood near the avenue and had painted walls and cut stone; the poorest were smaller and further out, and the poorest of all lived in adobe houses that have left little more than scatters of sherd.

The compounds were also workshops. Obsidian working is the best documented: surface survey found roughly 150 localities with heavy obsidian debris, and the densest clusters were specialised workshops turning out blades in quantities far beyond household need. Shell, mica, slate, cotton, feathers and pottery were worked in other neighbourhoods, and the murals that give the city its colour — the processions and deities of Tepantitla, Tetitla and Atetelco — were painted on compound walls, not on temples. Craft specialisation was organised by household and by neighbourhood rather than by a central workshop attached to a palace.

Neighbourhoods had their own centres. At least 22 three-temple complexes have been identified, each with a plaza and three shrines, distributed across the city; they are the best candidates for the administrative cells out of which Teotihuacan was built. Above them, the city seems to have been divided into four quarters, separated by the two great axes. Some neighbourhoods were markedly foreign: the Zapotec enclave at Tlailotlacán kept its own customs and took up the city's ways unevenly, and other compounds held people who had grown up on the Gulf coast or in the Maya region. Isotope studies of burials point the same way: many residents were born elsewhere. It was a city of immigrants that never became homogeneous.

The murals show what the residents thought worth painting: priests in feathered headdresses scattering seeds, processional figures with shields and darts, jaguars playing conch trumpets, a mountain of water and flowers with people swimming and diving in it. They do not show battles, defeated enemies or a king receiving tribute. The absence is not total — some processional figures carry weapons and wear the costume of warriors — but the genre scenes of Mesoamerican royal art are missing from the walls where ordinary people lived.

How many people lived at Teotihuacan?

The number usually quoted, 100,000–200,000, is a range that has been repeated so often that it reads like a measurement. It is not. It comes from René Millon, who directed the mapping project that produced the first complete plan of the city in the 1960s, and who estimated the population from the number and size of compounds. Millon assumed about 2,300 compounds, calculated occupants per compound from a few excavated examples, and produced figures of 75,000, 125,000 and 200,000 at different times, with 125,000 as his "probable" estimate; in 1976 he narrowed the range to 100,000–200,000. The top of the range has had the longest life in popular accounts, largely because it makes the best sentence.

Later work has moved the figure down. Cowgill, reviewing Millon's assumptions, thought 60–100 people per compound too high, preferred 30–60, and arrived at a range of 30,000 to 140,000 with a midpoint of 85,000. A 2019 study by Michael E. Smith and colleagues went back to the mapping project's own records, measured dwelling sizes in excavated compounds, and extrapolated to the whole city; it concluded that 100,000 is the most reasonable figure for the Xolalpan–Metepec period. The uncertainties are real and quantifiable: how many households shared a compound, how many people were in a household, and how many compounds stood empty at any one time. Change any one of those and the total moves by tens of thousands.

The map behind all these numbers is itself an achievement, and a fragile one. Millon's teams walked the whole built-up area in the 1960s, recorded more than 5,000 surface collections, and published a plan at a scale of 1:2,000 that still underlies every reconstruction of the city. Because the survey captured walls and platforms that have since been lost to ploughing, quarrying and the growth of the modern towns, that 1960s plan now functions as a primary source: much of what is known about Teotihuacan's layout is known from a document made by people who could still see it.

Even at the lower end, Teotihuacan was in a class of its own in the Americas. It was larger than any contemporary Maya capital, and its population was concentrated in a single built-up area rather than spread across a kingdom. Whether it was among the largest cities in the world in AD 400 depends on how one counts the declining cities of the Mediterranean and the great centres of India and China, but it was certainly in the conversation, which is remarkable for a city with no navigable river and no wheeled transport.

The decline was as steep as the rise. After about AD 600 the population fell markedly, and between 600 and 800 the city probably held 30,000 people or fewer, scattered among ruins that were no longer maintained. By 750 the ceremonial centre was empty of authority.

Who ruled a city with no kings?

The list of things Teotihuacan did not produce is long and consistent. No royal tomb has been found, despite a century of looking and some very well-funded looking. No portrait statue of a named individual survives. No stela records a ruler's accession, parentage or conquest. No inscription names a king, because there are almost no inscriptions at all: the city used a small repertoire of signs, some of them painted in murals, but never developed the kind of public text that Maya and Zapotec cities carved on stone. There is no ballcourt of the monumental kind associated with elite display elsewhere, and no scene of a captive being trampled or stripped.

Against that, the city was obviously unequal. Compounds near the avenue were larger and better built, with lime plaster, cut stone and murals; those on the periphery were small and plain. Burials differ in their grave goods. Murals show richly dressed lords with elaborate feather headdresses, some of them armed. The Feathered Serpent Pyramid required the labour of thousands and the deaths of hundreds. A society that could do that had rulers of some kind. The question is what kind.

The most widely used description is oligarchic: a city run by a council of leading families, compared by some to the republics of Tlaxcala and Cholula that Spanish observers later described. Linda Manzanilla has argued for a more specific arrangement — four co-rulers, one from each of the city's four quarters, an arrangement she links to the four figures on the Las Colinas vessel and to the four-fold division of the urban plan. On this reading, Teotihuacan was a federation of descent groups or ethnic corporations that met as equals, and the "state" was the pact between them.

A third line of argument, developed through computer modelling by Tom Froese, Carlos Gershenson and Manzanilla, treats the 22 three-temple complexes as the seats of 22 neighbourhood representatives and asks whether a network of that kind could hold together without a central authority. Their answer is that it could, provided the whole community periodically suspended normal life in shared ritual — which is exactly what the murals of Tepantitla seem to show. The model proves possibility, not fact, and its authors say so. Its value is that it removes the assumption that a city of 100,000 people must have had a king.

Iconography offers a fourth route. Zoltán Paulinyi has catalogued more than 35 images of lords wearing a distinctive Great Tasseled Headdress, against about ten images for each other elite headdress type, and argues that these are the rulers, depicted as a body rather than as individuals. He also reviews the candidates for a palace — the Ciudadela, the Avenue of the Dead Complex, the Xalla compound north of the Pyramid of the Sun, and the Plaza of the Columns — and concludes that the Plaza of the Columns is the best fit. Nobody has found the throne room.

What complicates all of this is time. The city lasted 600 years, and its constitution probably changed. George Cowgill's compromise was that powerful rulers did build the Feathered Serpent Pyramid in the second century, but that their existence cannot be projected onto the whole of the city's history; authority may well have become more collective later. The burning and desecration of that pyramid within a century or so of its completion may mark the moment when the ruler's platform was repudiated. If so, Teotihuacan was not a kingless city but a city that abolished its kings, which would explain both the early monuments and the later silence.

What did the city live on?

Obsidian first. The city sat within reach of the most important source in Mesoamerica: the green obsidian of Pachuca, in the mountains of Hidalgo to the north-east, which is visually distinctive and was prized everywhere. Teotihuacan's workshops turned it into blades, knives, scrapers and points on a scale no other centre matched. Michael Spence's survey of the industry counted about 150 localities with heavy obsidian debris, traced a steady rise in the proportion of green obsidian used for blades until it reached roughly 84 per cent of cores in one phase, and argued that the industry became increasingly specialised, with some workshops producing almost nothing but blades. The finest blades were exported; the Maya lowlands took them in quantity, and green obsidian is one of the clearest archaeological traces of Teotihuacan's long reach.

Food came from the valley and beyond. The springs and the San Juan river were managed with canals and channels, and the slopes around the city were terraced and irrigated. Maize, beans, amaranth and maguey were the staples; turkeys and dogs were kept. The valley's fields are unlikely to have fed 100,000 people on their own, so grain and produce came in from the basin and from the warmer valleys to the south. The resettlement of the basin's population into Teotihuacan was also a resettlement of its farmers into the city's hinterland, which is one reason the urban explosion was possible at all.

Craft production supplied the rest. Mould-made figurines, incensarios, Thin Orange pottery imported from the Puebla region, shell ornaments, mica, slate mirrors, feathers, cacao and cotton all moved through the city. Where the exchange happened is uncertain. The Great Compound, a low platform opposite the Ciudadela, has been proposed as a forum or assembly place, and possibly a market, and the scale of craft output implies distribution beyond the household, but Teotihuacan had no coinage and no written accounts, so the mechanism — market, palace redistribution, or neighbourhood exchange — has to be inferred from debris patterns.

What the economy did not produce was a territorial empire. There is no evidence of provincial governors, garrisons or tax archives outside the valley. Influence travelled through prestige goods, through colonists and through the adoption of Teotihuacan styles by local elites, above all at Kaminaljuyú in the Guatemalan highlands and at Matacapan on the Gulf coast. The city could project power a thousand kilometres into the Maya lowlands, as AD 378 showed, but it did not administer what it reached.

What happened in AD 378?

The event is known almost entirely from Maya inscriptions, and it is the best-documented intervention by Teotihuacan anywhere. On 8.17.1.4.12 of the Long Count — 16 January AD 378 — a figure named Siyaj K'ak', "Fire is Born", arrived at Tikal. The same day, Tikal's king Chak Tok Ich'aak, known as Jaguar Paw, died. Within a year a new king was installed: Yax Nuun Ayiin, a boy, shown on Stela 31 in highland costume with a rectangular shield and an atlatl, the spear-thrower that was a central Mexican weapon. His father was not Jaguar Paw but a person recorded as Spearthrower Owl, who had acceded in AD 374 and who would die in AD 439, and who is the best candidate for a ruler of Teotihuacan itself.

The texts are careful about what happened and reticent about how. David Stuart's reading of the four surviving accounts — Stela 31 at Tikal, Stelae 5 and 22 at Uaxactún, and the Marcador at Tikal — established that the verb is hul-iy, "he arrived", not "he conquered", and that no war glyph appears in any of them. Siyaj K'ak' is named as overlord of the new king, and a stela at El Perú-Waka', west of Tikal, records his arrival there days earlier. The sequence looks like a column moving east along the rivers, removing one dynasty and installing another with a claim on a foreign throne. Whether it was sent by the Teotihuacan state, or was the venture of a faction acting on its own, cannot be settled from the Maya side alone.

The consequences were durable. Tikal's art changed: Teotihuacan costume, talud-tablero architecture and central Mexican imagery enter the monumental record, and the new dynasty lasted. Spearthrower Owl's name appears on Tikal monuments for generations, and in AD 695 a later king, Jasaw Chan K'awil, celebrated the anniversary of his death. Maya rulers elsewhere kept the connection alive as a source of legitimacy: the name of the city in Maya texts was puh, "place of cattails", the same metaphor as the Nahuatl Tollan, the ancestral city from which legitimate dynasties claimed to come. Dynasties such as Copán's would make stronger claims of the same kind.

What AD 378 did not produce was a Teotihuacan province. No garrison, no administrative building, no tax record and no Teotihuacan settlement has been found in the Petén. The intervention looks like regime change by a small, militarily effective group, followed by the absorption of its imagery into local politics — which is why the debate between "conquest" and "prestigious emulation" has run for fifty years without resolution. Both are partly true, at different moments and in different places.

How did Teotihuacan end?

The end came in two stages, and the first was violent. Sometime around AD 550–570 a fire swept the ceremonial core. It burned the temples and platforms along the Avenue of the Dead, and it burned Xalla, the palace compound north of the Pyramid of the Sun that Manzanilla has excavated since 2000. Sculptures were smashed, not stolen; walls were pulled down. The burning was concentrated on the buildings of the ruling elite, and it left no destruction of ordinary housing and no sign of a foreign army. The consensus now is that this was an uprising from within the city, perhaps by the neighbourhoods against the people in the palaces.

There had been warning signs. Around AD 540–550, large foundationless walls were thrown up around Xalla's main plaza, apparently to protect it — a fortification built by a regime that felt threatened. The compounds had been growing more autonomous for a century, with controlled entrances and their own trading partners. Manzanilla's model of the city as a pact between 22 neighbourhoods predicts exactly this: when the shared rituals that held the pact together lost their force, the parts went their separate ways.

The second stage was slow. The centre was abandoned, but the city did not empty at once. A poor late phase, Metepec, built mean structures in the ruins; the population fell to 30,000 or fewer. A long drought that affected central Mexico in the Epiclassic period made recovery harder. From about 640 to 820 the Coyotlatelco people, newcomers from the west, lived in and around the ruins, digging into the tunnels under the Pyramid of the Sun and leaving their pottery across the site. They were followed by Mazapa groups from about 940 to 1230, and much later by Mexica-period occupation in the fifteenth century. The archaeological record of Teotihuacan after the fire is not a record of emptiness but of a series of small societies camping in a very large ruin.

The political vacuum was filled elsewhere. Cacaxtla, Xochicalco and Cholula competed for the succession, producing the hybrid art of the Epiclassic — Cacaxtla's murals, with their Maya and Gulf coast elements, are the most vivid example — and none of them approached Teotihuacan's size. Cacaxtla, the best studied, held perhaps 15,000–20,000 people. The age of the giant central Mexican city was over until the Aztec Triple Alliance built another one.

What did later Mexico make of the ruins?

The name is the Mexica's. Teotihuacan in Nahuatl has been glossed as "place of the gods" or "place where gods were born", and it attached to the ruins because the ruins looked like the setting of a creation myth: the Mexica told of the gods gathering at a place called Teotihuacan to make the sun rise. Their identification of the two largest pyramids as the houses of the sun and the moon is the origin of the names we still use. None of it is a memory of the city itself. The inhabitants' own name for it is lost; the Maya called it "place of cattails", and the Nahuatl word Tollan, applied to several later cities, carried the same sense of the ancestral metropolis.

Colonial and early modern writers got the city wrong in an instructive way. Because the Mexica revered Tollan and because the sixteenth-century chronicles attached that name to the Toltecs, Teotihuacan was attributed to the Toltec — a people whose capital at Tula flourished from about the tenth century, two centuries after Teotihuacan emptied. The error survived into the twentieth century and is still visible in guidebooks. It took stratigraphy to break it: the Toltec ceramic assemblage sits above Teotihuacan's, not below.

The monument itself has been rebuilt as much as excavated. Carlos de Sigüenza y Góngora probed the Pyramid of the Moon in 1675. Leopoldo Batres dug and reconstructed at the Pyramid of the Sun in 1905, under Porfirio Díaz, and founded the first site museum; the pyramid's present profile, with its restored stair and cleared terraces, is partly his work. Large INAH projects followed in 1962–64, 1980–82 and 1992–94, and Millon's mapping project in the 1960s produced the plan on which every later discussion of the city's layout depends. The cave under the Pyramid of the Sun was found in 1971, and work on the sealed deposits beneath the Feathered Serpent temple has continued into this century. UNESCO inscribed the site in 1987; 652 acres are open to visitors.

Modern Mexico has made the city a national emblem, and the name now covers a whole civilisation as well as a place. That is a larger afterlife than most fallen capitals get. It is also a slightly ironic one: a society that built no statues of its rulers is remembered through two pyramids named after gods it may not have worshipped, by a people who arrived six centuries after it emptied.

Common myths about Teotihuacan

Timeline of Teotihuacan

What the map shows next

At AD 690 on the map the border data still draws Teotihuacan, though by that date the city had been a ruin for two generations and central Mexico had broken into competing centres. The vacuum it left is visible in the entries that follow. Classic Veracruz was rising on the Gulf coast, the Zapotec state at Monte Albán was still standing, and the Maya city-states of the Petén were entering their Classic prime at 425.

The contrast is the point. One giant city in a highland basin, planned in a single campaign and governed, apparently, without a king; and a lowland landscape of rival kingdoms that never became one. The map's border data stops Teotihuacan at 704, 46 years before the conventional end of the city, which is a reminder that the drawn range is a dataset convention and not an event.