Maya city-states

The Maya matter out of proportion to their political scale because they wrote. Their inscriptions name kings and record their births, accessions, marriages, wars and deaths, often to the day. Combined with excavated palaces, chemical analysis of bones, lake and cave sediments that record rainfall year by year, and since 2016 airborne laser scanning that strips the forest from the ground, the Classic Maya are the best documented pre-Columbian society anywhere in the Americas. Few ancient worlds can be reconstructed at this resolution, and fewer still offer so clear a case of a sophisticated political order coming apart.

The vocabulary needs care. "Maya" is a language family and, later, an ethnic label; it was never a state. "City-state" is a modern convenience for polities that were neither fully urban nor organised like the Greek or Sumerian models, but it captures the essential fact: sovereignty sat in individual kingdoms, not in a nation. The Classic period itself is defined less by pottery than by a habit — the erection of monuments carrying Long Count dates — which is why its beginning and end are dated differently in different places, and why the abandonment of that habit is one of the clearest signs that something had broken.

At a glance

How did the lowlands come to be filled with cities?

The Maya lowlands are a limestone shelf, flat or gently rolling, covered in tropical forest and cut by few permanent rivers. Rain is not scarce in aggregate — Tikal receives about 1,945 mm a year — but it falls almost entirely between June and December, and the four or five months that follow are dry enough to kill a maize crop. The bedrock is porous karst, so rainwater vanishes underground instead of collecting in streams. In the central Petén there is almost no surface water at all in the dry season except in bajos, the low, seasonal swamps that hold mud in April and grass in September, and in the reservoirs that people built.

Settlement followed water and good soil rather than rivers. The first villages appear by about 1000 BC, growing maize, beans and squash in a system that would remain the basis of the Maya diet for three thousand years. By the Middle Preclassic, around 600 BC, larger centres with plastered platforms and carved monuments had appeared at Nakbe and El Mirador in the far north of the Petén. El Mirador was enormous for its date: its La Danta complex rises some 70 metres, built largely before the Christian era. It was abandoned, along with its neighbours, by about AD 150, in a Preclassic collapse that has never been fully explained but which looks uncomfortably like a rehearsal for the better-known one.

What followed was not a rebuilding of the same thing. From about AD 250 a denser network of smaller, more numerous kingdoms appears, each with its own dynasty and emblem glyph, spread across the Petén, the Belize river valley and the Usumacinta. The reasons are partly ecological — the collapse of the Mirador basin's overstretched polity freed land and routes — and partly political: the model of kingship that emerged in the Preclassic proved portable. A lord who could command labour, store water, and marry into the right family could found a dynasty in a place with no particular natural advantages, and dozens did.

Cities here were not compact in the Mediterranean sense. A Maya capital was a core of temples, palaces and ballcourts surrounded by sprawling residential settlement, much of it barely visible under forest, with house plots, kitchen gardens and orchards between the platforms. The political order was correspondingly decentralised in space even when it was autocratic in principle: a king ruled a landscape of farms and hamlets through subordinates, and the capital was less a walled town than a ceremonial and administrative focus for a populous countryside.

What did the Classic kings inherit from the Preclassic?

They inherited the ideology of divine kingship, and almost nothing else intact. The Preclassic had already produced monumental masks, ballcourts, stucco façades, and rulers depicted with the trappings of gods; the idea that a king was the pivot between human society and cosmic order was old by AD 250. What the Classic added was a machinery of legitimation that could be written down and repeated: the Long Count date, the emblem glyph, the royal name compounded with k'uhul, and the stela — a slab of limestone carved with a portrait of the king and a text that fixed him in time.

The Early Classic, roughly AD 250 to 600, is the period when these conventions hardened. Tikal's dynasty traced itself back to a founder remembered as Yax Ehb' Xook, placed by later scribes around AD 90, and its Stela 29, carved in AD 292, is among the earliest Long Count monuments in the lowlands. Copán, far to the south-east in Honduras, was refounded in AD 426 by a warrior named K'uk' Mo' Ajaw — Yax K'uk' Mo', "First Quetzal Macaw" — who arrived, according to the city's own later texts, with backing from Tikal. That single event shows how the system worked: dynasties were made and remade by external patrons, and a kingdom's founding story was also a statement about which great power it belonged to.

The Early Classic was also a period of contact with the greatest city in Mesoamerica. Teotihuacan, in the Basin of Mexico, had no rival in size or monumentality, and its influence appears at Maya sites in architecture, pottery, obsidian and iconography. How much of that influence was conquest, how much trade, and how much the fashion of a cosmopolitan elite is one of the liveliest arguments in Maya studies. What is not in doubt is that in the fourth century a group connected to Teotihuacan intervened directly in the politics of the Petén, and that the intervention reset the balance of power for two hundred years.

How did divine kingship actually work?

At the top sat the k'uhul ajaw, the "holy lord". Ajaw alone means "lord" and could be held by many nobles; k'uhul made the title sacred and marked a paramount ruler. Above even that, some kings claimed the title kalomte, whose meaning is uncertain but which seems to have denoted an overlord of overlords. Below the king were sajalob, subordinate lords who governed second- and third-tier sites, raised troops, took captives and were depicted on the monuments of their masters. The title sajal has been translated as "feared one", and the inscriptions usually attach it to war.

Kingship was performed rather than administered. A ruler acceded by seating himself on a jaguar throne and receiving the crown and the sceptre; the date was recorded to the day, and the legitimacy of everything he did afterwards depended on it. He bled himself, drawing blood from tongue or penis and burning it, and the resulting smoke was said to open a passage to the ancestral world — the vision serpent on Yaxchilán's lintels, with the king's wife kneeling and pulling a thorn-studded rope through her tongue, is the most vivid surviving image of the rite. He played ball, danced in the costume of the maize god, took captives, and presided over the period-ending ceremonies that tied his reign to the turning of time.

This was not a priesthood that happened to rule. The king was the axis between the community and the cosmos, and his body was the instrument: his blood fed the gods, his war fed the ancestors, his accession restarted time. The great funerary temples — Pakal's Temple of the Inscriptions at Palenque, with its sarcophagus lid carved as a map of the underworld, or Tikal's Temple I, raised over the tomb of Jasaw Chan K'awiil — were machines for turning a dead ruler into an eternal presence in the city's centre.

Yet divine kingship was not unlimited, and it weakened over time. By the Late Classic the title k'uhul ajaw was being claimed by rulers of quite modest towns, while great kings increasingly ruled through sajalob and marriage alliances rather than direct command. Palaces grew into sprawling compounds of courtyards and storerooms, the households of queens, princes and officials; the epigraphy records counsellors and regents, and at least one woman, the Lady of Tikal, seems to have ruled in her own right in the sixth century. The system was never a simple despotism, and in its last century the aristocracy around the throne appears to have grown faster than the throne's ability to feed it.

Why did Tikal and Calakmul become enemies?

Geography set the board. Tikal sat on a low ridge between two bajos in the central Petén, with good soil, dependable rainfall and control of routes running north to the Yucatán and south to the Motagua. Calakmul lay some hundred kilometres to the north-west, in the Campeche forest, beside a huge seasonal swamp that could be engineered into a water supply. Both were large; Calakmul has 6,750 recorded structures and a pyramid over 45 metres high. Neither could destroy the other's landscape, and between them lay a belt of smaller kingdoms whose loyalty was the real prize.

The decisive event came from outside. On 16 January AD 378 — the date varies by two days depending on which correlation constant is used — a man named Siyaj K'ak', "Fire is Born", arrived at Tikal. He had passed through El Perú, west of Tikal, eight days earlier. On the day of his arrival the reigning king, Chak Tok Ich'aak, died. Within a year a new ruler, Yax Nuun Ahiin I, was installed; his monuments show him in Teotihuacan war gear, with the goggle-eyed storm god on his shield, and Tikal's architecture thereafter includes the talud-tablero profile characteristic of the Mexican city. Tikal's Stela 31, carved a generation later, presents the new dynasty as the legitimate heir of both Tikal and the west, with Siyaj K'ak' named as "Lord of the West".

Whether this was a Teotihuacan conquest, a coup by a Tikal faction with foreign backing, or the arrival of a mercenary captain who made himself indispensable, historians do not agree. The older picture of a Teotihuacan empire reaching into the jungle has been qualified: there is little evidence of sustained administrative control from central Mexico, and Teotihuacan itself was in trouble by the sixth century. What is clear is that the entrada of AD 378 installed a dynasty with foreign prestige, that Tikal became the dominant power in the central lowlands for the next century and a half, and that its rulers used the connection to Teotihuacan the way later European dynasties used Rome — as a source of legitimacy that cost nothing to invoke.

Calakmul's answer was patience and marriage. The kingdom of the Snake, Kaan, appears in the inscriptions with an emblem glyph of a snake's head; its dynasty may have begun at Dzibanche, further east, before moving to Calakmul in the seventh century. From there it built a network of clients that encircled Tikal: Naranjo, Dos Pilas, El Perú, Cancuén, and a string of lesser places whose rulers took Calakmul's princesses and Calakmul's orders. The two great cities never fought a war of annexation. They fought over the allegiance of the same small kingdoms, year after year, and the smaller kingdoms changed sides whenever it paid.

What did Caracol do to Tikal in 562?

Caracol lies in the Belize river valley, and it is the best demonstration that size in the Maya world was not a matter of stone alone. The site covers about 177 square kilometres, with settlement so dense that its excavated house counts imply a population above 100,000; its central palace, Caana, rises 43 metres, still the tallest building in Belize. Its early history is obscure, but by the mid-sixth century it was a client of Tikal: its king Yajaw Te' K'inich II, "Lord Water", acceded in AD 553 under Tikal's auspices.

Then the relationship turned. In AD 556 Tikal attacked Caracol in what the glyphs call a ch'ak war, an "axe" war, and defeated it. Six years later, on 29 April AD 562, Caracol struck back in an event recorded as a star war. Tikal's king, Wak Chan K'awiil, was captured; the Caracol text implies he was sacrificed. Tikal did not build another dated monument for over a century. Its public sculpture was deliberately mutilated, its palaces burned, and its population fell. The archaeologists who found the smashed monuments in the 1950s had no idea what had happened; it took the decipherment of Caracol's Altar 21, found in 1986, to supply the missing century.

The star war glyph shows a star — probably Venus — pouring liquid onto the earth, and it seems to mark a particular kind of conflict with cosmic sanction, one that aimed at the enemy's dynasty rather than his fields. Not every Maya war was a star war; most were raids for captives, tribute and prestige. But the star war of AD 562 was decisive in the way that mattered: it did not annex Caracol's rival, it broke the capacity of Tikal's royal line to rule for two generations. Caracol itself did not become a hegemon. It had removed a patron and gained a rival, and within seventy years Calakmul was managing the region instead.

There is a caution in the source. Altar 21 was carved in AD 633, seventy-one years after the event it describes, under a king, K'an II, who had his own reasons for presenting his dynasty as the humbler of Tikal. The archaeology — the hiatus, the mutilated stelae, the burnt palace — corroborates the defeat. The details of how it happened come from a monument commissioned three generations later, and should be read as a political memory as much as a record.

How did Calakmul rule an empire without occupying one?

Calakmul's system is the closest thing the Classic Maya had to an empire, and it worked through people rather than garrisons. A king who submitted became a vassal: he kept his throne, his local gods and most of his revenue, and in return attended his overlord's accession, sent troops, supplied labour, and married where he was told. His sons might be educated at Calakmul; his daughters might marry into the Snake dynasty; his heirs might be the sons of a Calakmul princess. The great city's power is visible in the distribution of its emblem glyph across the lowlands, stamped on the monuments of client kings who wanted it known whose friend they were.

The system's architect was Yuknoom Ch'een II, "Yuknoom the Great", who reigned from AD 636 to 686 and whose name appears in the texts of half a dozen subordinate kingdoms. His most spectacular instrument was a woman. In 682 Calakmul sent Wak Chanil Ajaw, known as Lady Six Sky, to Naranjo, a kingdom whose dynasty had been broken by Caracol; she arrived as a princess of Dos Pilas and the daughter of its king, re-established the dynasty, and ruled for two decades as its effective sovereign, commissioning monuments that show her in the regalia of a king. Naranjo became a loyal client, and Calakmul's western flank was secured without a battle.

The counterstroke came from Tikal. Jasaw Chan K'awiil I, who took the throne in AD 682, rebuilt his city's fortunes and in AD 695 defeated Calakmul in a battle that the Tikal inscriptions present as a reversal of 562. Calakmul's king, Yuknoom Yich'aak K'ak', was killed or captured. The great city never recovered its hegemony, though it survived and continued to build. Tikal's victory was celebrated in Temple I, where the king was buried thirty-nine years later, and in a series of monuments that recast the previous century as an interruption rather than a defeat.

Even at its height, Calakmul's reach was not territorial. It could not tax a province or appoint a governor, and its clients rebelled when it weakened. That is why the Maya political system looks, from a distance, like the later Aztec Triple Alliance: a tribute-and-prestige network around one or two great cities, with dozens of small kingdoms calculating their advantage. The difference is scale and writing. The Aztecs left accounts written by their subjects and their enemies; the Maya left accounts written by their kings, in a script designed to make the king the centre of the story.

Why did the western kingdoms fight so relentlessly?

West of the Petén, along the Usumacinta and its tributaries, the political geometry was different. There was no single dominant power and no room for one: the river was the highway, the valley sides were the farmland, and the city-states sat on the terraces above the water in a line that made them natural rivals. Palenque, in the hills of Chiapas, Yaxchilán and Piedras Negras on opposite banks of the Usumacinta, Bonampak in the forest behind Yaxchilán, and Toniná in its own high valley to the west, fought for centuries over crossings, tribute and captives.

Palenque, whose people called their kingdom B'aakal, was sacked twice by Calakmul's allies, in AD 599 and again in 611, and the second defeat seems to have broken the dynasty for a generation. Its recovery came under K'inich Janaab' Pakal, who acceded in 615 at the age of twelve and ruled for sixty-eight years. Pakal built the Temple of the Inscriptions, the most elaborate funerary monument in the Maya world, and his son K'inich Kan Bahlam II completed the group and the texts that made Palenque's dynasty look older and more sacred than it was. Their programme was not only architecture: it was a rewriting of history, with the kingdom's origins pushed back into myth and its gods arranged in a triad that mirrored the order of the cosmos.

Palenque's fall came from a neighbour it had underestimated. In AD 711 Toniná, a hilltop city with a pyramid 74 metres high, defeated Palenque and took its king, K'inich K'an Joy Chitam II, captive. Toniná's Monument 122 shows him bound; Palenque's own inscriptions record a ten-year gap in the dynasty; the defeat itself is not commemorated there. The city survived, built, and declined slowly through the eighth century, with no new elite construction after about 800.

Yaxchilán's great age began with Itzamnaaj Bahlam III, "Shield Jaguar", who took the throne in AD 681 and ruled for sixty-one years, dying in 742 in his nineties. His monuments line the terraces above the river: lintels showing him with captives, his wife Lady Xoc in the bloodletting rite, and a hieroglyphic stairway that was rebuilt by his son Bird Jaguar IV in 752 to make his own accession look inevitable. Bird Jaguar's reign is one of the best documented in Maya history and one of the most obviously staged. In 808, another Yaxchilán king captured the last ruler of Piedras Negras, an event that effectively ended that city as a political force.

Copán, in the hills of western Honduras, controlled the jade and obsidian routes from the Motagua valley and built one of the most sculpturally inventive cities in the Maya world. Its greatest king, Uaxaclajuun Ub'aah K'awiil — "18 Rabbit" — reigned from AD 695 and was captured and beheaded in AD 738 by K'ak' Tiliw Chan Yopaat of Quiriguá, a small site downriver that had been Copán's vassal. The reversal probably had Calakmul's encouragement behind it; it certainly had the effect of removing a Tikal ally. Copán recovered, but never again dominated its region. Its last recorded king, Ukit Took', was enthroned in 822 and left an altar unfinished, two of its four faces blank.

Piedras Negras, on the Guatemalan bank of the Usumacinta, was a smaller but unusually well recorded kingdom, with a dynasty traceable from AD 297 and a run of kings whose monuments record victories over Palenque in 628 and a long rivalry with Yaxchilán. Its sculptures are among the finest of the Late Classic, and its collapse is among the best dated: the last king was taken in 808, and the site was abandoned within a generation. The pattern across the west is consistent — a century of intensifying war, then a rapid unravelling in the 800s, with the smaller kingdoms going first.

What did writing and the calendar do for the state?

Maya writing is a logosyllabic script: some signs stand for words, others for syllables, and most inscriptions mix the two, with a single word often written several ways. There are roughly 800 signs in the known repertoire, of which perhaps 200 are syllabic. The earliest certain examples come from the Preclassic — the painted texts of San Bartolo, in the Petén, are around 2,300 years old — and the script was fully formed by the time the Classic monuments were being carved. It was not the only script in Mesoamerica — the Zapotec of Monte Albán had been writing for centuries — but it is the only American script that recorded a complete spoken language, and it was used for dynastic history, ritual, astronomy, tribute lists, and the captions on painted pottery that record the names of owners and the gods they served.

The script survived the conquest just long enough to be destroyed. In 1566 the Franciscan bishop Diego de Landa, conducting a campaign against Maya religion, burned a library of codices at Maní; only four Maya books survive, the Dresden, Madrid, Paris and Grolier codices, all late and all astronomical or divinatory. Landa also recorded what he took to be a Maya alphabet, which was in fact a list of signs with their sounds and names, and for four centuries it was dismissed as a misunderstanding. In 1952 the Soviet linguist Yuri Knorozov argued that the glyphs were phonetic, and in 1959–60 Tatiana Proskouriakoff showed that the dates on Piedras Negras's monuments clustered around the lifespans of individual rulers rather than the movements of the heavens. Between them they turned the inscriptions back into history.

The Long Count was the instrument that made that history datable. It counts days from a fixed era date — 11 August 3114 BC in the proleptic Gregorian calendar, on the standard correlation — in units of 20 and 18: k'in (day), winal (20 days), tun (360 days), k'atun (7,200 days) and b'ak'tun (144,000 days). A date such as 9.6.8.4.2 fixes an event to a single day, and can be converted to the Gregorian calendar with an error of at most a few days. The system was not a pure base-20 count, and the names b'ak'tun and k'atun are modern coinages for units the Maya themselves labelled differently, but the arithmetic is exact and the inscriptions use it obsessively.

The completion of thirteen b'ak'tuns on 21 December 2012 became, in the early 2000s, a global industry of doom. It rests on a misunderstanding of how the calendar works. The Long Count is cyclical: 13.0.0.0.0 is followed by 13.0.0.0.1, and the world does not end, any more than it ended at the previous completion. Only two inscriptions in the entire corpus mention the current b'ak'tun's end: Tortuguero Monument 6, a fragmentary text about the descent of a deity, and a block from La Corona. Neither predicts catastrophe. What they show is a Maya interest in deep time — the same interest that produced distance dates running millions of years into the past and future, and tables predicting eclipses and Venus's appearances centuries ahead.

Writing and calendar together were instruments of government. A stela did not merely record a king; it placed him at a precise point in a cosmic sequence, on a day whose significance his scribes had calculated, in front of an audience that could not read the text but could see the image and hear the date proclaimed. The monopoly on this knowledge was never total — scribes were nobles, and literacy spread among the elite — but it was concentrated enough that when the courts collapsed, the script went with them.

How many people lived in the lowlands, and how were they fed?

Until recently the answer depended on how much of the forest you could see through. The 2016 PACUNAM LiDAR survey, published in 2018, mapped 2,144 square kilometres of the Maya Biosphere Reserve in northern Guatemala and identified 61,480 ancient structures — a density of 29 per square kilometre, including causeways, terraces, canals and fortifications that no one had known existed. From that the authors estimated 80 to 120 people per square kilometre at the height of the Late Classic, which implies a population in the millions rather than the hundreds of thousands that older textbooks assumed.

Site-level estimates vary widely and are worth treating with caution. Tikal has been estimated at anywhere between 10,000 and 90,000 inhabitants; Caracol's excavated structure counts imply something over 100,000 across its 177 square kilometres; Calakmul is usually put at around 50,000. The figures are not commensurable, because different scholars count different things — the palace core, the sustaining area, the whole polity — and because the boundary between town and farmland is precisely what Maya settlement dissolves.

The economy that supported those numbers was intensive and varied. Maize was the staple, grown in milpa fields that were cleared, planted and fallowed, but also on terraces cut into hillsides and in raised fields beside wetlands, with canals for irrigation and fish. Beans, squash, chilli, avocado, papaya and cacao filled out the diet; cacao was a luxury and a currency, and its cultivation shaped politics in the wet lowlands where it grew. Dogs and turkeys were kept; deer, peccary, tapir and turtles were hunted; the ramón tree has been proposed as a staple starch, though the argument is contested.

Water was the binding constraint. In the central Petén, cities depended on reservoirs lined with clay and fed by plastered catchment areas, with silt traps and filtration systems that required constant maintenance. Tikal's reservoirs could hold months of supply for tens of thousands of people; Caracol's system of aguadas and check dams did the same. In the north, where the water table is higher, the Yucatán's cities relied on cenotes, natural sinkholes that gave access to the aquifer. This difference in hydrology would matter enormously when the rains failed.

Long-distance trade moved the goods that the lowlands could not produce. Obsidian came from the volcanic highlands at El Chayal and Ixtepeque, jade from the Motagua valley, salt from the north coast, marine shell and quetzal feathers from the south, cotton and cacao from the piedmont. Much of it travelled by canoe along the coast and up the Usumacinta, and the control of portages and river crossings explains a great deal of Maya warfare. The same routes carried ideas: Teotihuacan's war gear, the talud-tablero façade, and the ballgame cult of the Classic Veracruz coast all arrived along them.

Why did the southern cities empty in the ninth century?

The sequence is unusually well dated, and it does not look like a single event. Dos Pilas, a predator state that had fed on its neighbours, was abandoned in AD 761 after its subjects rebelled. Piedras Negras lost its king to Yaxchilán in 808. Copán's last king was enthroned in 822 and left his altar unfinished. Tikal's last dated monument belongs to the 860s. Toniná, in the Chiapas highlands, carved the last known Long Count date in AD 909. By 950 the central Petén was largely empty of the courts, craftsmen and monument-makers who had defined the Classic order, though farmers remained in some places for generations.

The climate record is the strongest single line of evidence. A titanium record from the Cariaco Basin off Venezuela, published in 2003, shows a century-scale drying trend in the Terminal Classic, punctuated by severe multiyear droughts centred on approximately AD 810, 860 and 910. A stalagmite record from Yok Balum cave in Belize, published in 2012, gives a year-by-year rainfall history for the last two millennia and shows unusually wet conditions between 440 and 660 — the period of fastest population growth and political expansion — followed by a drying trend from 660 to 1000, and a final extreme drought between 1020 and 1100. The wet phase matches the Classic florescence; the dry phase matches its fragmentation.

But drought alone does not explain the pattern. The northern Yucatán suffered the same rainfall trends and did not collapse; Chichén Itzá and Uxmal flourished in the ninth and tenth centuries. The southern cities depended on reservoirs that could be managed through a normal dry season but not through a run of failures, and their populations were large enough that a shortfall meant famine rather than hunger. What the climate did was convert a political crisis into a demographic one: once kings could no longer guarantee water and maize, the ideological basis of their rule — that the cosmos answered to them — collapsed with the harvest.

War made it worse. The eighth century saw an intensification of conflict across the lowlands, with fortified sites, burned palaces and captured kings. Some of this was the ordinary competition of the system; some was the specific consequence of Calakmul's defeat in AD 695, which left its clients without a patron and their neighbours without a check. The collapse of trade networks followed the collapse of the courts that maintained them, and the loss of scribal and craft specialists meant that the knowledge needed to run a reservoir or read a calendar died with the institutions that employed it.

The debate now is less about whether drought mattered than about how to weight it. Some archaeologists argue for a cascade in which climate stress triggered elite competition, which triggered war, which triggered the withdrawal of peasant support from a kingship that had failed to deliver. Others emphasise internal political contradictions — over-mighty nobles, over-built palaces, an ideology that could not survive a bad century. A third group questions the word "collapse" altogether, pointing out that Maya society continued in the north and the highlands, that the population of the Yucatán grew, and that what ended was a particular political order, not a people. All three positions can be reconciled, and the honest answer is that the ninth century was a convergence: a political system that had become fragile, a climate that turned against it, and a war that finished it.

What survived, and who are the Maya now?

The north did not collapse. Chichén Itzá rose in the ninth and tenth centuries to dominate the Yucatán, with an architecture and iconography that drew on central Mexican models — warrior columns, chacmool figures, a feathered serpent — and a political order that seems to have been less personal than the Classic divine kingship, perhaps governed by a council of lineages. Its interpretation is contested: the old story of a Toltec conquest from Tula has been challenged by scholars who see shared iconography and trade rather than invasion, and the "Toltec" label may describe a style and a legend rather than an army. After Chichén Itzá's decline, Mayapán held the north until the fifteenth century, and the highland kingdoms of the K'iche' and Kaqchikel built their own states in Guatemala.

The Spanish arrived in the 1520s, and the last independent Maya kingdom, the Itza state centred on Nojpetén in the Petén lakes, fell only in 1697. The colonial period destroyed the codices, the priesthood and the courts, but not the languages or the communities. The decipherment of the script in the twentieth century, and the archaeology of the last fifty years, have made the Classic Maya the best-known ancient American civilisation, and they have also made them modern: the 2012 phenomenon, the tourist economy of Chichén Itzá and Tikal, and the political organising of Maya communities in Guatemala and Chiapas all draw on the same history.

For a long time the Maya were presented as a peaceful people of astronomer-priests, a useful counterweight to the Aztecs in the imagination of the nineteenth century. That picture was demolished by the same decipherment that revealed the wars, the captures, the sacrifices and the dynastic murders. What has replaced it is more interesting: a civilisation of competitive kingdoms, intensely literate, obsessively concerned with time, capable of engineering on a landscape scale, and unable to stop the competition that eventually consumed it. The Maya did not vanish, and they did not predict the end of the world; they built one of the most sophisticated political systems in the ancient Americas and then, over two centuries, took it apart.

Common myths about the Maya city-states

Timeline of the Maya city-states

What the map shows next

At AD 800 on the map the constellation is at its densest and most fragile: Tikal and Calakmul are still standing, Palenque and Copán are in their last building phases, and the small kingdoms of the Usumacinta are already in crisis. Pull the slider back to AD 300 on the map and the map shows the Early Classic world that the entrada of AD 378 would overturn, with Tikal's dynasty newly installed and Teotihuacan's shadow falling across the Petén.

At AD 849 on the map the southern lowlands are still drawn, but the last dated monument at Toniná is sixty years away and the political order that produced it has already failed. The next entry, later Maya city-states, follows the survivors north into the Yucatán; the Itza hold out in the Petén until 1697.