Norte Chico (Caral-Supe)

← Encyclopedia

On a dry terrace above the green floor of the Supe valley, some twenty-three kilometers in from the Pacific and a little under two hundred north of where Lima now stands, six earthen mounds rise out of the desert. The largest of them, the Pirámide Mayor, lifts about eighteen meters above its base and spreads across an area near the size of four football fields. There is no mortared stone facing in the Egyptian manner, no fired brick, no inscription. The fill that gives the platforms their bulk is held in woven reed bags — net sacks of plaited fiber packed with quarried stones and stacked course on course inside retaining walls. The people who carried those bags up the slope had no pottery. They had no metal weapons, no city wall, no carved scene of a captive or a battle. They raised the oldest monumental architecture in the Americas, and they raised it in a thousand-year-old silence about war.

This is Norte Chico — the “Little North,” the swath of coastal river valleys north of Lima where the earliest complex society of the New World took shape — known equally by the name of its greatest center and its home valley, the Caral-Supe civilization. It flourished in the Late Archaic, the preceramic horizon, from roughly 3500 BCE through its waning around 1800 BCE. For the length of that span it was contemporary with the first cities of Sumer and the broader world of Mesopotamia, with the pyramid-building Old Kingdom of ancient Egypt, and with the planned brick cities of the Indus valley. It belongs to that small company of places where social complexity arose on its own ground, owing nothing to any older civilization — a pristine cradle, and the only one in the western hemisphere.

The river valleys and the type-site

Norte Chico is not a single city but a region densely settled with them. The complex occupies four adjacent river valleys that descend from the western Andes to the Pacific within a short stretch of the north-central coast — the Huaura, the Supe, the Pativilca, and the Fortaleza. Across them lie more than twenty large early sites; popular counts run to about thirty. Each is built to a recognizable pattern: one or more raised platform mounds, sunken circular plazas, residential terraces, and upright standing stones. The shared architectural grammar across so many neighboring settlements is what marks Norte Chico as a true cultural complex rather than an isolated experiment.

The type-site and the grandest of them is the city in the Supe valley whose name the civilization carries. Its central zone covers some sixty-six hectares of public and residential building. Six large platform mounds frame a great open space; an upper sector adds further pyramids and a pair of sunken circular courts. The plan is deliberate — the monuments are not scattered but composed, oriented, and proportioned, with later study finding solstitial, lunar, and stellar alignments built into the principal structures. Down at the mouth of the river, where the Supe meets the sea, sat a coastal companion settlement, a fishing town that supplied the inland center; smaller monumental sites such as Bandurria and Vichama lie along the same coast. The whole forms a system rather than a capital — a landscape organized for labor and assembly.

Building without pottery, without war

Two absences define the civilization as sharply as its monuments do, and both have forced a rethinking of what is required to build at this scale.

The first is the absence of ceramics. Norte Chico is aceramic — preceramic, in the language of Andean archaeology — and this is not a gap in the record but a fact about the culture. These builders had no fired clay vessels at all. For most of the twentieth century the equation ran the other way: pottery was taken as a marker of settled complexity, so a site without it was assumed to be early, small, and simple. The mounds of the Supe valley were noticed and then set aside on exactly that reasoning. That they were instead enormous, planned, and old overturned the assumption. Storage, cooking, fermentation, and transport were managed with gourds, woven baskets, stone bowls, and the net bags that also built the pyramids. A society can be monumental, coordinated, and durable while remaining, in the technical sense, “primitive.”

The second absence is more arresting. Across the whole of Norte Chico there are no fortifications, no caches of weapons, no depictions of combat, no mutilated bodies, no burned-and-sacked destruction layers — none of the iconography of dominance through violence that runs through so much early state art elsewhere. The platforms were not citadels. The plazas were not parade grounds. Whatever held this society together and moved its labor up the mound, it does not appear to have been the threat of the spear or the celebration of the conqueror. The contrast is pointed when set beside the militarized self-image of later Old World powers, and it makes Norte Chico a standing case study in how authority can be founded on something other than force — on ceremony, on the calendar, on the management of water and exchange, on the binding power of shared construction.

The maritime-agricultural economy

The engine of all this was a partnership between the sea and the irrigated valley floor, and the precise balance of that partnership is the civilization’s central scholarly debate.

The coastal towns drew on one of the richest fisheries on earth, the cold upwelling current off the Peruvian shore, dense with anchovy and sardine. The inland centers commanded the river and the cultivable land along it, where the people grew not a cereal staple but an unusual mix: squash, beans, guava, achira, sweet potato — and, above all, cotton. Cotton was the keystone crop, and it was an industrial one rather than a food. Spun and netted, it became the fishing nets that harvested the sea; with gourds for floats, it equipped the coastal economy. So the two halves locked together. The fishers sent dried fish up the valley; the farmers sent cotton down to be made into the nets that caught more fish. Each side fed the other, and the surplus from the whole arrangement is what freed the hands that built the mounds.

How to weight the two halves was framed by the archaeologist Michael Moseley, whose 1975 thesis on the maritime foundations of Andean civilization argued that the marine harvest — extraordinarily abundant, available without farming — supplied the caloric base for sedentism, population growth, and monumental building before intensive irrigation agriculture took hold. It was a provocative claim, because it located the origin of New World complexity in the sea rather than in the field, breaking the usual rule that great architecture waits on the grain surplus. Later work refined and complicated it. The inland agricultural centers, on this reading, were not satellites of the coast but the seat of economic and political authority, with control of cotton — the indispensable input to net-fishing — as the lever. The two positions are not as far apart as they first seem: cotton is the hinge, an agricultural product whose whole purpose was to make the maritime harvest possible, and the establishment of inland centers may itself have been an effort to grow more of it. Whether one calls the result a state, a set of allied chiefdoms, or a “civilization” without a single capital remains contested, but the integration of coast and valley into one working order is not.

Ritual, sound, and the recording of number

The mounds were not tombs and not granaries. They were stages for assembly and rite. The sunken circular plazas, set below the surrounding ground so that participants descended into them, organized gatherings around a focal point; the largest such court ran some forty meters across, the amphitheater plaza large enough to hold hundreds. At the heart of the ceremonial complex sat an altar built for fire, set apart within its own enclosing wall and fed by ducted draft — a hearth for offerings that consumed what was given to it.

Sound belonged to these spaces. Buried within the sunken plaza of the amphitheater were thirty-two flutes carved from the wing bones of condor and pelican, and a set of cornets fashioned from the bones of deer and llama — an instrumentarium that places music, and presumably procession and chant, at the center of the rite. The recurrence of standing stones, the huancas, at sites across the region adds a vertical, marker-like element to the same ceremonial vocabulary, plausibly tied to the astronomical sightlines built into the mounds.

One find reaches forward across four thousand years of Andean history. On the steps of one of the pyramids the excavators recovered a quipu — an arrangement of knotted cords, the device by which the later Andean world, down to the empire that the Spanish met, kept its accounts and records by knot and color and position. The Caral cord is among the oldest known, and it implies that the management of this economy — the count of tribute, of labor, of stored goods — already ran through a knotted-string notation millennia before any of the cultures that made the technique famous. It is record-keeping without writing, number held in the hand.

Discovery and the dating that moved a millennium

The mounds had been seen before they were understood. They were noted by mid-twentieth-century surveyors and dismissed, again, because the surface yielded no pottery. The reappraisal came when the Peruvian archaeologist Ruth Shady began sustained excavation in the Supe valley in the 1990s and grasped that monumentality without ceramics meant the site was not merely early but extraordinarily so.

The proof was radiocarbon. Because the platform fill was built from organic reed bags, the construction itself could be dated directly. Samples of those reeds, analyzed with the North American archaeologists Jonathan Haas and Winifred Creamer, returned dates in the late third millennium BCE; the results, published in Science in 2001 under the title Dating Caral, a Preceramic Site in the Supe Valley on the Central Coast of Peru, pushed the threshold of urban monumental society in the Americas back by more than a thousand years. A second campaign across the wider region, the Proyecto Arqueológico Norte Chico, gathered ninety-five new dates from a sample of more than a dozen sites; reported by Haas, Creamer, and Alvaro Ruiz in Nature in 2004 as Dating the Late Archaic occupation of the Norte Chico region in Peru, it placed the florescence of the whole complex — stepped pyramids, sunken circular plazas, and standing monoliths across the four valleys — firmly between 3000 and 1800 BCE. The civilization was inscribed as a UNESCO World Heritage Site, the Sacred City of Caral-Supe, in 2009.

Scholarship and sources

The modern understanding of Norte Chico rests on a compact and recent literature, since the field is barely a generation old as a recognized subject.

• The foundational dating paper is Ruth Shady Solís, Jonathan Haas, and Winifred Creamer, “Dating Caral, a Preceramic Site in the Supe Valley on the Central Coast of Peru,” Science 292, no. 5517 (2001): 723-726 — doi:10.1126/science.1059519 — the article that established the late-third-millennium age of the type-site.
• The regional synthesis is Jonathan Haas, Winifred Creamer, and Alvaro Ruiz, “Dating the Late Archaic occupation of the Norte Chico region in Peru,” Nature 432 (2004): 1020-1023 — doi:10.1038/nature03146 — which extended the chronology across the four valleys with ninety-five radiocarbon determinations and described the shared architectural complex.
• The theoretical frame for the economy descends from Michael E. Moseley, The Maritime Foundations of Andean Civilization (Cummings, 1975), whose argument that the sea, not the field, underwrote early Andean complexity remains the axis of debate; later refinements emphasizing cotton and plant-fiber technology continue the conversation, among them the analysis collected at the Refining the Maritime Foundations study.
• The institutional record of the regional survey is held by the Field Museum’s Proyecto Arqueológico Norte Chico, and the conservation and management framing by UNESCO’s Sacred City of Caral-Supe listing.

Read against the older cradles, Norte Chico has the look of a controlled experiment. Where the script-bearing societies of the Near East left their accounts in clay, here the count was tied into cord; where the great river kingdoms recorded their kings in stone — the realm of the Pyramid Texts and the gods of the Mesopotamian pantheon — here the mounds carry no royal name at all. It is the same family of achievement reached by a different route, and it stands a full millennium before the Olmec heads of the Gulf coast or the palatial walls of Mycenaean Greece gathered their own worlds into shape.

The long drought

The waning came near 1800 BCE, and it appears to have come not by conquest but by the climate that the whole economy depended on. The same band of centuries that strained the Old World — the dry, unstable horizon known as the 4.2-kiloyear event, the interval in which the irrigation kingdoms of the Near East and the Nile passed through crisis — reached the Peruvian coast as a prolonged drought layered over the swings of El Niño and La Niña. The fisheries shifted offshore and out of reach; wind-driven sand crept inland and choked the canals; the cotton and the cultivated plants failed with the water. The settlements were not stormed and burned. They were left.

The leaving carried the order outward rather than ending it. People moved up the valleys toward the high ground and the glacial runoff, and along the coast to fresh sites, taking the mound-and-plaza plan with them — the sunken circular court, the platform raised by hand, the assembly without a wall. A society that had spent a thousand years coordinating thousands of pairs of hands around fire, water, cotton, and the sea met the failure of all four the way it had met everything else: not by taking up arms, but by picking up the carrying-baskets and building somewhere the river still ran.

→ Related: Olmec · Mycenaean Greece · Sumer · Mesopotamia · Mesopotamian Religion Sumerian Akkadian · Ancient Egypt · Indus Valley Civilization · Pyramid Texts · Neolithic