Chimú Empire

The empire's wealth came from water engineering and from the sea. Its rulers commanded canals that moved Andean meltwater across valley floors and between valleys, and its friezes are crowded with fish, pelicans and nets because the Humboldt Current fed the city as surely as the river did. Its smiths worked silver in quantities no other Andean culture matched, its weavers made textiles that survive as some of the finest pre-Columbian work anywhere, and its potters deliberately turned their backs on the imagery of their Moche predecessors.

Chimor was conquered by the Inca between about 1465 and 1475, and that conquest is the reason we can say anything at all about its kings. The Inca kept the dynasty as clients, took the reigning ruler to Cuzco, and moved thousands of Chimú craftsmen and farmers to other parts of their empire. Sixty years later the Spaniards arrived, founded Trujillo five kilometres from Chan Chan, and began digging up the dead.

The Chimú left no writing of their own. Everything in this entry comes from archaeology and from Spanish chronicles written a century after the events they describe, and the two often disagree. That makes Chimor a good test case for how pre-Columbian history is actually made: not from documents, but from adobe walls, canal gradients, isotope ratios in children's teeth, and notarial records of treasure-hunting contracts.

At a glance

Where did the Chimú come from?

The north coast had been urban for more than a thousand years before Chan Chan. The Moche had built pyramids and canal systems in the same valleys from about AD 200, and their huacas — Huaca del Sol and Huaca de la Luna — stand less than ten kilometres from the Chimú capital. After the Moche order broke down in the eighth and ninth centuries, and after the highland Wari state's influence withdrew, the coast fragmented into valley-sized polities. One of them, in the Lambayeque region to the north, produced the extraordinarily wealthy Sicán or Lambayeque culture, with its capital at Batán Grande and its furnaces at Cerro Huaringa.

Chimor grew out of that fragmentation. Chan Chan was founded around AD 900–1000 in a part of the Moche valley that had not been intensively occupied, and it grew with unusual speed and deliberation. The usual date in the literature is "about AD 1000", and the map's own series begins at AD 911 — a defensible early bound rather than a founding date. Whenever the first adobes were laid, the city was new: there was no earlier centre to remodel, no conquered population to accommodate, and no inherited sacred architecture to work around.

The origin story the Chimú told about themselves is preserved in Spanish transcriptions. A lord called Taycanamo arrived from the sea on a raft of balsa logs, sent — he said — by a great lord across the water to govern the land, and established himself in the lower valley. His son Guacricaur and his grandson Ñançenpinco extended control over the whole valley. The story is almost certainly a political genealogy rather than a record of migration. Its close parallel with the Lambayeque foundation myth of Naymlap, who also arrived by balsa raft with an entourage and founded a dynasty, suggests that the Chimú adopted and adapted a northern myth of legitimate seaborne kingship as they absorbed Sicán territory. Bioarchaeology gives no sign of a population replacement on the coast; the Chimú were the same people as their predecessors, organised differently.

What changed was ideology. Where Moche art is full of sacrificed prisoners and warrior lords, Chimú art turns to the sea, to textile patterns, and to the person of the ruler. Archaeologists working at Chan Chan describe this as a conscious break: a new dynasty inventing a visual language that owed nothing to Moche iconography and everything to the claim that the king was a different kind of being. The Chimú myth of origins made that explicit, with the first nobles emerging from a gold egg, noblewomen from a silver egg, and everybody else from copper.

Why build a capital in the desert?

Chan Chan sits in one of the driest places on earth, close enough to the ocean to hear it and far enough from the river to need engineering. The valley receives less than three millimetres of rain a year. Everything the city ate, drank and built with depended on water brought down from the Andes in the Moche and Chicama rivers, and on water found underground.

The Chimú answer was a hydraulic system of remarkable scale. Canals ran from the rivers along the valley margins, irrigating the flats; the Moche valley network reached its greatest extent around AD 1000 and was later cut back by 30–40 per cent. Within the city, water came from walk-in wells and from sunken gardens called huachaques, dug down to the water table; Kent Day observed that some of these gardens probably began as borrow pits, quarries for the adobe that built the city, and were then planted when the digging reached moisture. That is a nicely unsentimental piece of urban ecology: the capital's gardens were the holes left by its walls.

The most famous piece of Chimú engineering is also the most misunderstood. The La Cumbre, or Chicama–Moche intervalley canal, was designed to carry water south from the Chicama river, across the desert divide, to the piedmont north of Chan Chan — a channel measured at more than 54 kilometres by the archaeologists who traced it, and described as some 80 kilometres long in the site's World Heritage documentation. It was surveyed and built to gradients that later engineers have found impressive, with a cross-section calculated to keep sediment moving. It was also never finished. Charles Ortloff, Michael Moseley and Robert Feldman concluded from their hydraulic modelling that construction was abandoned before completion, at the point where tectonic movement had altered the ground slope, and that it therefore never supplied the Moche valley with Chicama water at all.

That conclusion has been contested for forty years. James Kus argued in 1984 that tectonic uplift was probably not the main cause of abandonment, that the canal did function in a limited way, and that it should be read as a political response to drought stress in the Moche valley — a grand public work undertaken when the water ran short and dropped when the drought broke. Others have blamed straightforward engineering error. The dispute matters because it changes what kind of state Chimor was: a far-sighted hydraulic bureaucracy, or a court that spent enormous labour on a scheme that never worked.

The sea, at least, never failed. Huanchaco, the Chimú port, is still a fishing village, and the reed boats its fishermen use — the curving craft usually called caballitos de totora — are the same design as the reed vessels shown in Chimú art. Fish, shellfish, seabirds and seaweed supplied protein to a population that had little room for pasture, and Spondylus shell, fetched from Ecuadorian waters, supplied the prestige item that Chimú lords scattered as powder in front of them when they walked.

How did Chimor become an empire?

Expansion came in two bursts, according to the chronicle-derived chronology that Gordon McEwan and others have summarised. The first is associated with Ñançenpinco, in the middle of the fourteenth century: having secured the Moche valley, he pushed north to the Jequetepeque and south to the Santa. The second belongs to the last reign, that of Minchancaman, and by about 1450 had carried Chimor to its greatest extent, from Tumbes or the Lambayeque valley in the north to the Chillón or Pativilca in the south. The precise frontiers are disputed — the older literature, including the 1985 ICOMOS evaluation of Chan Chan, extends the empire to the Gulf of Guayaquil, which almost certainly overstates it — but the shape is not: a long, thin state that controlled river mouths and valley floors and left the highlands alone.

Conquest was only one instrument. Excavation at Collambay, a village in the chaupiyunga zone on the eastern edge of Chimor where coca can be grown, found a ceramic assemblage mixing Chimú, highland and local styles, and graves that followed highland practice with no Chimú-style burials at all. The reading is that the Chimú did not garrison or colonise such places; they negotiated, and took coca, feathers and ore in exchange for access and protection. Coca was not a luxury. It was needed for ritual, for diplomacy and for labour, and control of the chaupiyunga was worth more than another stretch of empty desert.

Where the Chimú did rule directly, they built administrative centres: Farfán in the Jequetepeque valley, Manchan in the Casma. These were not copies of Chan Chan but planned compounds with storerooms, courts and workshops, staffed by officials who collected and redistributed produce. The state ran on labour rather than money. Households paid their tax in work — on canals, roads, temples, the army — and were fed and provisioned while they did it, which is why the archaeology of the Chimú countryside looks so thin: the population was drawn into the capital and into state projects rather than scattered across farmsteads.

The economics of all this are visible in the storerooms. The ciudadelas at Chan Chan are full of long, narrow, ventilated chambers, and the provincial centres replicate them. What the Chimú state did, above all, was store: maize, dried fish, cotton, cloth, beans, chicha. Control of storage is control of distribution, and distribution is what a ruler does for his clients.

Who ruled Chimor, and by what right?

Chimú kings were divine, or near enough that the distinction did not matter politically. They were held to have a separate origin from commoners; their persons were approached through intermediaries; their images and emblems carried the crescent of Si, the moon deity, the emblem that appears on most Chimú jewellery. They did not inherit their predecessor's wealth.

That last point is the core of the most influential model in Chimú studies. In 1981 Geoffrey Conrad argued that Andean kingship on the north coast operated by "split inheritance": the successor took the office and its political powers, while the dead ruler's estate — palaces, lands, herds, treasure, retainers — passed to a corporation of other heirs whose job was to maintain the dead king's cult for ever. The consequence is structural. A new king begins his reign with authority but no revenue, and the only way to fund his own court, his own tomb and his own posthumous cult is to tax the provinces harder or conquer new ones. Conrad argued that this legal principle, not ecological necessity, drove both Chimú and Inca expansion, and that the administrative stresses it created eventually helped break both empires.

The model has been attacked hard. Allison Paulsen's 1981 reply in the same journal was titled "The Archaeology of the Absurd" and argued that Conrad presented "evidence that is not evidence in support of a weak and untestable hypothesis", by an author who had ignored the climatic and materialist literature. Conrad answered his critics in the same issue, and the argument has never been settled. The striking thing is how well the buildings fit the model: ten compounds, ten kings, each with its own storerooms, its own kitchens, its own well and its own burial platform, each sealed and maintained after the ruler's death.

Below the king sat a hereditary nobility with its own smaller compounds, then officials, then craft specialists with privileges — the right to wear ear spools, the right to live near the nobility — and then the mass of farmers, fisherfolk and labourers. The law was severe: theft was treated as an offence against the gods as well as against persons. The state's language was Quingnam, a coastal tongue of the Yunga family that is known today from a handful of words, place names and colonial comments, and was already fading when the Spaniards arrived. Chimor's records were not written. Administration ran on counting, storing and moving things, and on the memory of officials; no archive survives because none was made in a form that could.

What were the ciudadelas for?

The monumental core of Chan Chan covers about six square kilometres and consists of large rectangular enclosures inside high adobe walls — nine according to UNESCO, ten according to Kent Day and most archaeological accounts, eleven in some counts. Each enclosure has a single entrance, opening into a corridor system designed to be traversed in one way only. Inside are courts, U-shaped structures that Day called audiencias and interpreted as control points for administrators, storerooms, kitchens, walk-in wells, open plazas for feasting and, in seven of the compounds, a burial platform.

The architecture is a machine for regulating access. Walls up to nine metres high, one door, corridors that turn and narrow: a visitor to the king passed through a sequence that told him, physically, how unimportant he was. The friezes on the walls — fish, pelicans, nets, waves, human figures in procession — carried the same message in a different register. Day read the compounds as the administrative headquarters of the empire, with the audiencias as the desks where tribute was booked in.

Geoffrey Conrad read them as tombs with palaces attached. His study of the burial platforms — thoroughly looted, but recoverable in plan and sequence — found forecourts, cells, and evidence of repeated ritual, human sacrifice and the maintenance of a cult of the dead king. On his account the compound was built as a palace, converted at the ruler's death into a mausoleum, and thereafter served by a staff whose job was to keep feeding and clothing a corpse.

Recent excavation has pushed the argument further. Henry Gayoso Rullier and Nadia Gamarra Carranza, working on Gran Chimú, the largest of the compounds, showed that the niches holding wooden statues at the entrances were late additions, and argued that the statues represent figures in funerary ritual and the ancestor cult centred on the king's mummified body. Their conclusion is that the compounds were palaces that became royal mausoleums — which is not quite either earlier position, but closer to Conrad's.

Around the palaces lived everybody else. The old model divided the city into royal compounds, elite compounds and four neighbourhoods of wattle-and-daub rooms where artisans supposedly worked full time for the crown. Gabriel Prieto's excavations from 2022 in the southwest neighbourhood have upset that. The houses he opened contained macaw feathers, Spondylus shell, fine textiles and metalwork, but no sustained evidence of workshops. His reading is that these were the residences of an intermediate elite, not of a proletariat of craftsmen — and that the ceramics they used came from Casma, Jequetepeque and Lambayeque, which suggests that people from conquered provinces were living in the capital.

The dead tell a harder story. Mass sacrifices of children are known from several Chimú sites, including Pampa la Cruz and El Pollo near Huanchaco. Oxygen isotope analysis of thirty of those children found that 63 per cent were not born in the Moche valley. They had been brought in, from other valleys and possibly from further inland, to be killed. Whatever else the Chimú state did, it could command the bodies of children from communities it ruled.

What did the empire live on?

The subsistence base was irrigated agriculture plus the sea. The fields grew maize, cotton, squash, lima beans, chilli peppers, gourds, sweet potatoes, peanuts and avocados, and a fruit the excavators identified as guanábana or soursop, apparently a favourite. Llamas and guinea pigs supplied meat; the ocean supplied everything else. Cotton mattered twice over, as food crop and as the raw material for the textiles that were the state's principal currency of prestige.

The textile tradition is one of the great achievements of Andean craft. Chimú weavers worked cotton and camelid wool into tapestry of extraordinary fineness, and recent excavation at Chan Chan has recovered painted cotton garments and camelid-fibre tapestry from ordinary elite houses, not only from tombs. Featherwork was equally important: macaws were bred in the city, and macaw and tanager feathers were imported from the Amazon slope to make crowns and ornaments. A sacrificed child at Pampa la Cruz was buried wearing one.

Metal was the other glory. The Chimú worked more metal than any other Peruvian culture, and they preferred silver to gold — a taste that says something about the value system, since silver is harder to obtain, needing to be mined from veins and smelted rather than found as nugget and flake. Chimú smiths worked in repoussé, hammered, incised, cut, made filigree and cast. Their output was vessels, jewellery, ceremonial weapons, shield covers, figurines inlaid with mother-of-pearl and turquoise. The crescent-moon figure of Si appears on most of the jewellery, which is a reminder that the best metalwork was not decoration but theology.

The scale of production was industrial. John Topic estimated that perhaps 12,000 of Chan Chan's 40,000 inhabitants were engaged in craft production of some kind. If that figure is roughly right, Chimor's capital was less a city of farmers than a city of workshops supplied by farmers — a court economy in which the ruler's demand for cloth, silver and feathers organised the labour of tens of thousands of people across a thousand kilometres of coast.

Feasting bound it together. The huge silver beakers in museum collections — capacious, plain, made in sets — were drinking vessels for chicha, and their numbers imply large, inclusive gatherings rather than small elite rituals. Joanne Pillsbury's reading of this material is that drinking with, and to, dead kings was a central ritual of Chimú rulership: the vessels served the living and the dead in the same act, and the ancestor cult was a banquet.

What did the Chimú believe?

The moon was the emblem that mattered. Si, the moon deity, gives her crescent to most of the surviving Chimú jewellery, and the friezes of Chan Chan are a marine bestiary — fish, crustaceans, seabirds, fishing nets, waves. The two currents that meet off Huanchaco, the cold Humboldt and the warm El Niño, are sometimes read as the subject of the walls of the Tschudi compound, which would make it a rare case of Andean art depicting an oceanographic fact.

Chimú religion was an ancestor religion with a strong emphasis on the corpse. The king's mummy was kept, dressed and fed; his compound became his shrine; the rituals around him were repeated, not concluded. Human sacrifice accompanied royal funerals and, in the case of the child sacrifices at Pampa la Cruz and El Pollo, seems to have been mass and repeated — a state ritual, organised at scale, drawing victims from across the empire.

There are limits to what can be said. No Chimú liturgical text exists. The sixteenth-century chroniclers recorded fragments of oral tradition about kings and dynasties, but almost nothing about ritual; the archaeologists have the paraphernalia and the bodies and not the words. Statements about Chimú belief are inferences from objects, and should be read as such.

How did the Inca take Chimor?

The conquest came late in the Inca expansion and it was not quick. Tupac Yupanqui, son and heir of Pachacuti, campaigned north along the coast in the 1460s and 1470s; the conventional date for the fall of Chimor is about 1470, though the sources allow a window from roughly 1465 to 1475, and some accounts put the final submission later. The Chimú fought, lost a series of engagements in the northern valleys, and then lost their king: Minchancaman was taken to Cuzco and kept there as a hostage-guest.

The arrangements afterwards are unusually well documented for a pre-literate polity, because the Spaniards inherited them. Chan Chan was sacked, but it was not destroyed. The Inca ruled the north coast through the surviving dynasty: Minchancaman's son Chumun Caur governed the former kingdom as the Inca's client. The dynasty kept its rank; the Inca took what they wanted, which was above all the metalworkers.

The resettlement was systematic. Chimú silversmiths were moved to Cuzco; others were sent to the Lake Titicaca region. Farmers and whole communities were moved as mitima, colonies transplanted to break up the political geography of the conquered north and to supply labour elsewhere. The strategic logic was the standard Inca one: a conquered people dispersed cannot revolt, and a conquered craft tradition concentrated at court enriches the conqueror. Much of the gold the Spaniards later took from the Inca had been made by Chimú hands.

The economic damage was severe in a different way. The Inca state reorganised the north coast's irrigation and tribute systems to its own specifications, and the Moche valley's canal network had already contracted long before the Inca arrived. Whether the Inca deliberately cut the water off, as some popular accounts have it, is not established; what is clear is that the late fifteenth century saw a shift of wealth and population away from Chan Chan towards Inca administrative centres.

What did the Spaniards destroy?

Francisco Pizarro's expedition reached the north coast in 1532, and the Chimú welcomed the Spaniards as the lesser of two masters. The ICOMOS evaluation of Chan Chan puts it bluntly: the conquistadores were welcomed by the Chimú "out of hate of the Incas". In 1535 Pizarro founded Trujillo, named for his birthplace in Extremadura, five kilometres from Chan Chan, and the old capital's fate was settled — not by a battle but by the founding of a rival town with a Spanish legal charter, a plaza, a cabildo and a supply of Indian labour.

Chan Chan was not abandoned overnight. It was quarried. The huacas were treated as banks, and the old city's walls as a supply of ready-made mud. Castilian law required anyone finding Indian tombs or shrines to register them and pay the crown's share; from 1537 the paperwork exists, and by the 1550s the business had become a corporate one. Rocío Delibes Mateos has reconstructed the "compañías de huaca" from the notarial registers of Trujillo: groups of vecinos who pooled capital, labour and licences to dig. In 1559 a company was formed to dig the Huaca Grande del río — the Huaca del Sol — by Alonso Riero, holder of a viceregal licence, and doña Ana Pizarro, widow of the conquistador Diego de Mora; she contributed Indians with tools, he his personal supervision. Another company worked the Huaca Pachacamac with 22 shares at 600 pesos each. Francisco de Toledo's ordinance of 27 August 1571 tried to regulate the traffic by requiring licences from the viceroy, which is a fair indication of how much of it there was.

The result at Huaca del Sol is measurable. C. Mansfield Hastings and Michael Moseley, studying the adobes in the 1970s, recorded that colonial looters had used the Moche river itself to wash the mound away, and that at least two-thirds of it was gone; the surviving cut face, 28 metres high, exposed eight construction stages. Huaca del Sol was a Moche building, not a Chimú one, but it stood in the Chimú heartland and it was mined as part of the same estate.

Then came the resettlement. From 1569 Francisco de Toledo's general inspection of the viceroyalty produced the reducciones: the forced concentration of scattered Andean communities into grid-planned Spanish-style towns, on the argument that dispersed settlement made Christian instruction and tribute collection impossible. Jeremy Mumford's study of the programme describes a resettlement that moved more than a million people. For the Moche valley the effect was to drain the remaining population out of the old settlement pattern and into new towns under Spanish eyes. Chan Chan's last inhabitants left, and the site became a quarry, a landmark and a rumour of treasure.

Popular accounts often compress all this into a single catastrophe — a great flood, or the Inca sack, or the Spanish conquest. The evidence points to a sequence: conquest in 1470, a client dynasty, a Spanish town founded alongside in 1535, licensed looting from the 1530s to the 1570s, forced resettlement in the 1570s, and then three centuries of El Niño rains, wind and weeds working on unfired mud. Adobe is a material that fails slowly and completely. The city did not fall; it dissolved.

How do we know any of this?

Almost nothing about Chimor comes from Chimú documents, because there were none in a form we can read. The written evidence is Spanish and late: Miguel Cabello de Balboa's Miscelánea Antártica, compiled around 1586, which preserves the dynastic list beginning with Taycanamo; the anonymous History of Trujillo of 1604, which records oral testimony about the kings; and the reports the clergy compiled from their parishes. These are genealogies collected a century after the Inca conquest, from informants whose grandparents had lived under the empire, and they disagree with each other — Minchancaman appears as the seventh, eighth or eleventh king depending on which account you follow. They are evidence for what sixteenth-century coastal people believed about their past, which is not the same thing as a chronicle.

Everything else is material. The first systematic work was antiquarian: the ruined city was mapped and drawn in the nineteenth century by travellers including Rivero, Tschudi, Squier, Middendorf and Bandelier, whose plans and elevations are still useful because they record walls that have since collapsed. Scientific archaeology arrived with the Chan Chan–Moche Valley Project, a Harvard-led mapping and excavation programme directed by Michael Moseley and Kent Day between 1969 and 1974, which produced the first real city plan, the first ceramic sequence and the first modern analyses of the ciudadelas, the irrigation system and the burial platforms. Moseley's own synthesis argued for a landscape shaped by tectonic uplift, river incision and episodic El Niño flooding, an argument that has been refined and disputed ever since.

The current phase belongs to Peruvian institutions. The Ministerio de Cultura's Proyecto Especial Complejo Arqueológico Chan Chan has excavated continuously since 2006, mostly in the compounds and their burial platforms, recovering tapestries, ceramics and burials from Huaca Toledo. University-led projects have reopened the neighbourhoods, and archaeological science has changed the questions: oxygen and strontium isotopes on the sacrificed children, residue analysis on vessels, radiocarbon dating of canal sediments, drone survey of the canal systems. After five centuries of looting, the recent excavation of burial platforms that still held tapestries, ceramics and human remains has been a reminder of how much of the site is still untouched and how little of it has been dug.

The historiography has a political edge that is worth naming. The Chan Chan–Moche Valley Project was foreign-funded and foreign-directed; for decades afterwards, as Peruvian archaeologists have pointed out, the Moche and the Lambayeque attracted most of the research money and the Chimú were comparatively neglected. Prieto's phrase for it is that the Chimú "haven't been invited" to the party of Peruvian prehistory. That is changing, but it explains why the standard English-language account of Chimor still rests on work done between 1969 and 1982.

Why does Chan Chan still matter?

Chan Chan was inscribed on the UNESCO World Heritage List in 1986, under criteria (i) and (iii), as the largest earthen-architecture city in pre-Columbian America. It was placed on the List of World Heritage in Danger in the same year, and it has stayed there. The reason is the material: unbaked mud, in a desert that is now wetter than it was, with a water table that rises during El Niño years and salt that climbs the walls. A strong El Niño sends water running down adobe faces and dissolves the mud from the base up; one night of rain can cost a frieze that survived four centuries.

The conservation problem is genuinely hard. Adobe that has been buried for centuries is stable; adobe exposed to air, sun and occasional rain is not. Every excavation creates an obligation to build a roof, and roofs change the thing being conserved. The site's managers have to choose, continually, between leaving structures buried and invisible and exposing them to the public and the weather. Add to that the ordinary pressures of a growing city on the site's edges, and the persistent trade in looted ceramics and textiles.

What survives is not a ruin in the romantic sense but a plan. The ciudadelas give the political structure of a non-literate empire in built form — one compound per reign, one entrance per compound, one platform per tomb. The canals give its economy. The workshops and storerooms give its administration. Very few ancient states anywhere leave their constitution lying around in mud.

Common myths about the Chimú Empire

Timeline of the Chimú Empire

What the map shows next

At 1200 on the map the coast is a patchwork: Chan Chan is a young city consolidating the Moche valley while the Sicán lords of Batán Grande still hold the north. By 1375 on the map the picture has changed — Lambayeque has fallen to Chimor, provincial centres at Farfán and Manchan administer the conquered valleys, and the empire runs from the Lambayeque valley to the Santa.

At 1470 on the map the Chimú territory is being absorbed into the Inca empire, and by 1481 on the map the coast is governed from Cuzco through a client dynasty at Chan Chan. The next chapter belongs to the Inca, and after them to the Spanish empire, whose town of Trujillo would finish what the conquest began. For the later country that inherited this coast, see Peru.