Your goal: Master what we actually know about Caral and the Caral-Supe (Norte Chico) civilization — the 5,000-year-old cities of coastal Peru — and learn, from zero, the archaeology needed to weigh the evidence: how sites are dated, how diets are read from bones, and how rival theories about the birth of civilization are tested against dirt.

Drive about three hours north of Lima, turn inland up a green ribbon of valley squeezed between bone-dry hills, and you arrive at a cluster of stone platform mounds the size of city blocks. Nothing about the place looks tropical or jungle-clad — this is one of the driest deserts on Earth. The mounds are pyramids, there are six of them arranged around open plazas, and when the site was inhabited there was not a single pot, metal tool, or written word anywhere in the Americas.

Caral sits 182 km north of Lima and 23 km in from the coast, on an alluvial desert terrace at the start of the Supe Valley's central section, 350 m above sea level.

Its age is staggering: while the pyramid of Saqqara was being built in Egypt, and the later ones were going up at Giza, and the Sumerian cities of Mesopotamia were being occupied — between 3000 and 2500 B.C. — Caral's pyramidal buildings were already being constructed and remodelled.

"Civilization" sounds like a compliment, but in archaeology it is a technical checklist. Haas and Creamer, two of the leading (and, as you will see, combative) researchers of this region, give the standard version: a civilization is one of those few exceptional cultures that develops formal institutions of government, urban centers, organized religion and art, monumental construction, marked social stratification — meaning built-in ranks of rich and poor — and a highly productive agricultural economy. Keep that checklist; the whole course is an argument about which boxes Caral ticks.

By wide agreement the Andes are one of the world's six major independent civilizations — Mesopotamia, Egypt, India, China, Mesoamerica, and the Andes.

Independent is the key word. Rome learned from Greece, which learned from Egypt and the Near East. But in six places — and only six — people invented cities, government, and monumental religion from scratch, with no model to copy. Archaeologists call such a case pristine, and it is the closest thing social science has to a natural experiment: six independent runs of the same experiment, "how does complex society begin?"

The Caral-Supe society developed in complete isolation, without contact with the civilizations of the Old World or the New — and it flourished at least 1,800 years before the Olmec society, long considered the mother civilization of the Americas.

Archaeologists slice prehistory by the technology lying in the ground, and pottery is their favorite clock: styles change fast, shards are indestructible, and nearly every ancient dump is full of them. Which is why a great site without pottery is so disorienting — there is nothing to read the date from.

Peruvian chronology calls Caral's era the Late Archaic — the Cotton Preceramic — running from about 5000 to 3750 years before the present (“B.P.”, radiocarbon's zero point; Module 3 explains that clock).

The same era is commonly called Preceramic VI, or the Late Preceramic, by North American archaeologists — 'preceramic' meaning simply: before pottery existed here at all.

The trap claimed its first victim early: the archaeologist Paul Kosok, one of the first to report monumental settlements in the inner Supe Valley, visited the Chupacigarro complex at the end of the 1940s and concluded that the absence of ceramics made it virtually impossible to calculate its relative chronological position.

Sit with the irony. Kosok was staring at the answer: the site had no pottery because it is older than pottery in Peru. The very clue that should have screamed impossibly ancient instead filed Caral under "undatable," and the first city of the Americas went back to sleep for another half century.

The site had even been recorded under another name — Chupacigarro Grande — until Ruth Shady's project renamed it Caral, from local toponymy, to end the confusion.

The real significance of Supe's early society was only demonstrated when Shady began excavations at Caral in 1996.

What her team and rivals established by the 2001 survey: Caral was one of 18 large preceramic sites then known in the small Supe Valley alone.

Hold that number loosely. As survey continued, later reports raised the Supe tally to about 20 settlements (the figure you meet in Module 2) — the count has crept up with each season of fieldwork, a first hint that this landscape is still being mapped.

The city itself covers 66 hectares, organized into a nuclear zone of 32 public structures plus residential complexes, and a peripheral zone bordering the valley.

Radiocarbon dates place the occupation of Caral in a roughly 600-year span between 2627 and 2020 calibrated years B.C. (how those dates are measured — and what “calibrated” hedges — is Module 3's full story).

Here is the puzzle this module solves. Civilizations are supposed to be born in generous river basins — the Nile, the Euphrates, the Indus. Caral sits in a desert so dry that in many years it effectively does not rain at all. Yet this coast fed cities. To see how, you need the region’s three-part anatomy: a desert shore, a wall of mountains, and a jungle behind it — packed astonishingly close together.

Civilization was achieved in north-central Peru during the 3rd millennium B.C., between the Santa and Chillón valleys and extending inland to the Marañón and Huallaga rivers — a region that consists of coast, highlands and Andean forest.

Those two rivers are the frame, not the focus: the Santa marks the northern edge of this stretch of coast and the Chillón (just north of modern Lima) the southern edge — so the valleys this course lives in, Supe and Huaura, sit comfortably inside the span, not at its rim.

The area hangs together physically: its nine western valleys are connected through the high-Andean plateau where their rivers rise, which links them to the Callejón de Huaylas and, to the east, to the Marañón and Huallaga basins — gateways to the Amazon.

Hold that image: a ladder of small green valleys crossing a desert, each rung climbing from the Pacific to the high plateau and dropping into the jungle beyond. People, goods, and ideas could move across the ladder — sea to sierra to selva — and Shady will argue that this transversal traffic is exactly what a young civilization runs on.

Within this region, the star is an almost comically modest valley. Supe is not a great river basin; it is a thread of green a few kilometers wide.

The Supe River runs almost at the level of its floodplain, so the fields beside it can be irrigated easily, just by digging simple channels — no sophisticated technology required.

And yet, despite topographic conditions offering little tillable land and a river dry for most of the year, a minimum of 18 urban centers were erected within 45 km of the coast.

Around 5000–4000 years before the present the Supe society held a privileged, high-prestige position in the north-central area, expressed in the great extent and complexity of its many settlements.

How many settlements exactly? Excavation has so far confirmed on the order of 18 large Late Archaic sites in the Supe Valley — the figure from the 2001 survey; by 2006 the registered total had edged up to about 20. The count keeps drifting upward because fieldwork keeps finding more.

Whatever the exact tally, read the order of magnitude: roughly a score of monumental settlements packed into one small valley.

Notice what a strange signature that is: maximum monuments, minimum farmland. Whatever powered the Supe Valley, it was not agricultural abundance alone. That tension — too many cities for such a small valley — is a clue you should carry through the whole course.

The other half of the answer is offshore: Supe's population had easy access to the resources of one of the world's most productive oceans, plus lomas (fog-fed vegetation), the valley's riverine plants and animals, and lands easily irrigated by river and spring water.

The same current that feeds this coast also, every few years, betrays it. A warm-water reversal called El Niño caps the upwelling: the anchovy shoals crash and the fishery starves, while the normally rainless desert is lashed with rains and floods that tear at fields and canals — the same violent flows a Peruvian calls a huayco. Both halves of the economy, sea and land, can fail at once and without warning. A society settling here had to be built around that boom-and-bust rhythm — a point that returns when we ask what religion was for and what finally ended the Supe world (Modules 10 and 14).

So redraw your mental map of "cradle of civilization." Not one broad river plain, but a compressed stack of environments: the richest sea on the planet, a thin irrigable valley, fog meadows, high plateau, and jungle — all within a few days’ walk. Poverty of rain, wealth of everything else. The next question is how we know people organized all this 5,000 years ago — which means learning to read the radiocarbon clock.

Module 1 left a mystery: with no pottery to read, how did anyone ever pin a date on Caral? The answer is the tool that revolutionized archaeology in the twentieth century — and it was built for exactly this situation, because it reads dates not from artifact styles but from anything that once lived.

Now watch the tool used well. When Shady teamed up with Jonathan Haas and Winifred Creamer to date Caral for the journal Science, the first problem was choosing what to date. In a desert, wood can survive for centuries before being built into a wall — date an old beam and you date the tree, not the building.

The Caral team sidestepped that trap: most samples dated were fibers from plants with a life-span of about one year, eliminating the problems associated with “old” wood in a desert environment.

The samples went to two independent laboratories — six to Beta Analytic in Miami and twelve to the Illinois State Geological Survey.

The samples were also spread deliberately: they came from a range of different spots and contexts across the site — stratified trash, floors, wall and platform fill — not one lucky deposit. Keep that in mind; it is what will make the result so hard to wave away.

The verdict, published in Science in 2001: monumental corporate architecture, urban settlement, and irrigation agriculture began in the Americas by 4090 years B.P. — 2627 calibrated years B.C.

Read that date against Module 1’s timeline: 2627 cal B.C. is squarely in the age of Egypt’s pyramid-builders and the Sumerian cities. The uncalibrated range at Caral ran from 4090 down to 3640 B.P. — in calendar terms, the city’s roughly 600-year occupation between 2627 and 2020 cal B.C. that you met in Module 1. One paper, eighteen dates, and the Americas were suddenly 1,500 years older than the textbooks said.

A quick word on the round numbers you will hear, since they seem to clash. Popular accounts call Caral “5,000 years old” or date it to “3000 B.C.,” yet the measured occupation of Caral-the-city is the tighter 2627–2020 cal B.C. There is no contradiction: the bigger, rounder figure belongs to the whole civilization, not this one site. As the next section shows, the Norte Chico region emerges around 3000 B.C.; “five thousand years ago” is that regional headline, while 2627 B.C. is this particular city’s measured birthday.

A single dated city could still be a freak. The next step — and the beginning of a rivalry that would turn famously bitter (we pick it up in a moment) — was Haas and Creamer striking out from Caral to date the neighboring valleys of the Norte Chico, publishing in Nature in 2004.

A geography check first, because a new name is about to appear. Everything so far — Caral, Áspero, the whole score of sites — lies in the Supe Valley. But Supe is just one of four adjacent valleys that make up the Norte Chico: from north to south, Fortaleza, Pativilca, Supe, and Huaura — all four nested inside the broader Santa–Chillón north-central area you met in Module 2, of which the Norte Chico is the tight coastal core (the name itself was Haas and Creamer’s coinage, part of what nettled Shady). That is the point of the move. To test whether Caral’s precocity was a fluke of one lucky valley, Haas and Creamer deliberately went next door — dating sites in Pativilca and Fortaleza, not Supe. Finding the same early monumental pattern there is exactly what turns "an astonishing site in Supe" into "a civilization across a region" (the four valleys get their full treatment in Module 12).

Their team presented 95 new radiocarbon dates from 13 of the more than 20 large early sites of the Fortaleza and Pativilca valleys — sites sharing monumental architecture, extensive residential architecture, and a lack of ceramics.

The 95 dates confirmed the emergence and development of a major cultural complex in the region during the Late Archaic, between 3000 and 1800 calibrated years B.C.

A few dates ran deeper still — ten samples fell before 3500 cal B.C., and one outlier of 9210 cal B.C. hints at an Early Archaic human presence long before the cities.

Put together, the dates establish that the people of this region were a significant force on the Andean cultural landscape for more than 1,200 years.

So by 2004 the picture was no longer "one precocious site" but a crowded landscape: dozens of monumental centers across four valleys, rising around 3000 B.C. and running twelve centuries.

Keep that human story in view as Part III lays out their scientific disagreement: the question of who built the first Andean state is argued, in part, by people who spent two decades also arguing over who deserves credit for finding it. With the dates settled and the feud noted, we stop measuring the city and finally walk into it.

Enough measuring from the outside. Walk in. Caral spreads across a desert terrace, its buildings arranged not at random but to a plan — and reading that plan is how archaeologists infer a society they can never interview.

The arrangement of Caral's structures implies a spatial ordering that preceded construction — a planned design recognizing social criteria like hierarchical strata and symbolic divisions.

Be careful with that inference, because it is easy to overstate. Planning by itself shows coordination — a single authority able to impose one design on the whole site, rather than a village accreting house by house. It does not, on its own, prove rulers: a community could in principle plan cooperatively. The step to rulers comes because the plan is not neutral — as Shady reads it, the layout deliberately encodes hierarchy, ranked social strata and the upper/lower division below — and that built-in hierarchy is then confirmed in the dirt by the sharply unequal houses and refuse you will meet later in this module. Planning alone buys you a directing authority; planned hierarchy plus unequal homes buys you a ranked society with an elite.

The nuclear zone splits into two great halves: an upper half with the most impressive pyramids and a lower half of smaller structures — a layout that likely expresses the traditional Andean dual division into hanan and hurin (upper and lower).

The central zone holds six large platform mounds, numerous smaller ones, two sunken circular plazas, and extensive residential architecture — over 65 hectares in all.

The dominant structure is the Great Pyramid (Sector E), the largest complex in the city, measuring 170.8 m east-to-west by 149.7 m, its mound rising to nearly 30 m — fronted by a sunken circular court 35.5 m across.

Stand on its summit and, in Shady’s description, the whole city and valley open below you while only the sky and the framing hills close the horizon — a stage built for whoever was allowed to stand on top. Around it cluster the Central, Quarry and Lesser pyramids to the west, and the pyramids of the Gallery and the Huanca to the east.

Between the two eastern pyramids stood a huanca — a tall upright stone, 2.15 m high — a monolith set as the axis linking them.

To fix the scale in mind, here is the whole cast of principal structures with their measured dimensions (from Shady’s excavation report). Notice the sheer range — from a mound the size of a stadium down to houses smaller than a modern living room — and that the elite residence has roughly four times the floor of a commoner’s.

The lower half has no great pyramids but one unique structure: the Temple of the Amphitheater, named for the largest circular sunken court in the city, ringed by tiered seating like a theater.

It is worth pausing on what these sunken circular plazas are: round courts dug below ground level, entered by stairways, where people gathered facing inward and down toward a central space. They are stages for collective ritual — and, as you’ll see in Module 11, one of them gave up the oldest musical instruments in the Americas.

The clinching evidence for rank is domestic. Excavators compared where people lived, how their homes were built, how big they were, and what they threw away — and the gap is stark on every axis.

Elite residences sat right beside the pyramids, solidly built of cut stone — coarse-grained diorite — with clay-plastered walls painted red, white, and yellow.

One such elite house by the Quarry Pyramid was spacious, measuring 16.0 x 12.9 m and holding nine rooms with platforms, benches, and hearths.

Its refuse marked elite consumption: on top of the everyday fare lay prized sea lion and deer bones.

What was that "everyday fare" the whole city shared? Overwhelmingly the sea, plus the valley’s crops.

The staple diet, common to rich and poor alike, was marine: anchovies, sardines, clams, and mussels, eaten with valley-grown plants like squash, beans, and sweet potato.

The lower half's Small Residential Group, by contrast, were cramped cane-and-reed houses on leveled terraces — some scarcely 49 sq m, roughly a quarter the floor of that elite house — and their occupants lived on the common marine-and-plant fare without the sea lion and deer.

So the hierarchy is written in floor plans, building stone, and garbage alike: large diorite-walled houses beside the temples, their middens carrying prized sea lion and deer on top of the common fare, versus small cane houses on the margins whose people ate that common fare alone. Same staple base, unequal extras — that contrast is the raw material for the fierce debate, in Part III, over whether Caral was a true state.

The people who built Caral had no metal tools, no draft animals, no wheel, and no writing. Everything you saw in Module 4 was raised by human muscle and basketry. How they did it turns out to be the same question as how we date it — through a humble woven bag.

Mound fill was carried in shicra bags — open-mesh bags made of reeds — which were filled with stones and then placed, bag and all, inside the retaining walls.

Picture a work gang: cut-stone retaining walls define a mound, and the hollow is packed with thousands of stone-filled reed bags, like sandbags in a flood wall. It is a technique perfectly suited to a society rich in reeds and manpower but poor in metal and machines — and it left archaeologists two gifts.

The first gift is dating. A shicra bag is woven from plants that lived about a year (recall Module 3’s logic), and it can only have been made to be buried in the mound — so a radiocarbon date on a bag dates the construction itself, tightly. Bags are now a preferred sample for dating the monuments.

At Caral the mounds grew by burial: Late-period rooms were raised by burying earlier ones under voluminous rock fill in shicras, and three main periods of change — Early, Middle, and Late — can be distinguished in the Great Pyramid.

That is why the pyramids are layered like onions. Rather than demolish an old temple, they entombed it in stone-filled bags and built the next one on its roof. Each burial seals a moment in time — floors, offerings, even wall paint whose colors changed from one period to the next — for an excavator to peel back centuries later.

How many building phases a mound took is not a dry detail — it measures how much labor a ruler could command at once, which is a measure of political power. And here Caral’s excavators and their Norte Chico rivals disagree.

Shady argued the main platform mound at Caral — roughly 150,000 cubic meters — was built in only two major construction phases, implying a leadership able to mobilize enormous labor in a single push.

Test excavations by Haas and Creamer in the Pativilca and Fortaleza valleys revealed two different patterns of mound construction — some raised in one event, others accreted in stages — complicating any single story of how power worked.

Keep this alive: "few big phases" suggests concentrated, commanding power; "many small phases" suggests power that could only nibble labor a little at a time. The same mounds, read through stratigraphy, are the battlefield for the question of how strong Caral’s rulers really were — the theme of Part III.

Caral sits 23 km inland. But the Supe story has a twin at the shoreline, and the relationship between them frames one of the oldest debates in Andean archaeology: did civilization here rise from the sea or from the soil?

Áspero is a coastal site at the mouth of the Supe Valley, just 500 meters from the Pacific, set among sandy hills, wetlands, and beach.

Its people lived by the sea — extracting fish such as anchovy and sardines and shellfish — and its Late Archaic monumental component covers about 18.8 hectares with twenty-two buildings reflecting social differentiation.

Áspero was the flagship site for a bold theory. Most civilizations are said to rest on farming; Michael Moseley proposed that this one did not.

Moseley's Maritime Foundations of Andean Civilization hypothesis holds that organizing the procurement and distribution of marine resources — not agriculture — was central to the first complex societies here.

This was heresy in the best sense. Every other pristine civilization — Mesopotamia, Egypt, China — was built on cereal grains. Moseley argued the Andes were unique: the Humboldt fishery was so productive that fish, not farms, could underwrite the surplus and organization a civilization needs. The anchovy shoals from Module 2 were, on this view, the engine.

Here is the elegant hinge on which the whole regional economy turns — and it is not fish or corn but cotton.

Effective exploitation of marine resources is inextricably related to the production of cotton at the inland sites: you cannot fish the shoals at scale without nets and lines, and those were made of cotton grown in the valley.

So the coast and the valley locked into exchange: Áspero traded fish and shellfish to Caral, and in return obtained agricultural products, especially cotton fibre, needed for making fishing nets and lines.

Feel how neat that is. The fishers cannot fish without the farmers’ cotton; the farmers want the fishers’ protein. Neither the pure "sea" story nor the pure "soil" story survives on its own — coast and valley are one interdependent system, stitched together by a plant you cannot eat. The next two modules ask what that system actually put in people’s stomachs.

Module 6 left farming and fishing locked together. Now we dig into the farming half — and discover it is far older, and far stranger, than "they grew food to eat." The first surprise: irrigation in Peru predates the cities by two thousand years.

In the Zaña Valley of northern Peru, Tom Dillehay's team found buried gravity canals directly dated to at least 5,400 radiocarbon years ago — and possibly as old as 6,700 years — the earliest irrigation known in the Andes.

These were small-scale canals: Canal 1 ran about 4 km from its intake, feeding farm plots on low terraces at the foot of alluvial fans — modest engineering, but real irrigation.

This matters for the whole argument. If Andean people were digging canals and tending fields two millennia before Caral, then agriculture was not a late add-on to a fishing society — it was a deep, developing skill that the Norte Chico cities inherited and scaled up. The Supe farmers were heirs to a long tradition.

Now the strangeness. The single most important crop in this economy was not food at all.

The effective exploitation of the sea was inextricably tied to inland cotton production — because nets, lines, and bags were made of it — making an inedible fibre the keystone of the regional economy.

Hold Module 6’s exchange in mind and the logic completes: cotton is what the fishers must buy and the farmers can uniquely supply. A civilization whose flagship crop is worn and knotted rather than eaten is a strong hint that its economy ran on production and exchange, not just filling bellies — a point the state debate in Part III leans on hard.

Caral's plant remains include squash, beans, lucuma, guava, pacay, camote (sweet potato), and cotton — but corn, Zea mays, is absent.

That absence is a bombshell. Every schoolchild "knows" ancient American civilization means maize — the Maya, the Aztec, endless fields of corn. Caral had essentially none. Its farmers grew fruits, legumes, tubers, and industrial cotton, not a cereal-grain staple. So this civilization was not built on the grain that built Mesoamerica, which sharpens the mystery of what, exactly, fed and financed it — the question stable isotopes will finally answer.

The debate between sea and soil could go in circles forever on indirect clues — nets here, canals there. What finally cut through was a method that reads diet directly from the bodies of the people themselves. It rests on a simple slogan: you are what you eat, chemically.

Recent research — From here on, this gold badge marks findings that come from the newest wave of studies (2021–2026), which post-date most of what this course draws on and are still actively revising the picture. Everything without the badge rests on the foundational 2001–2008 scholarship. Watching which is which is part of thinking like a specialist: the framework is older, the instruments testing it are new.

In 2021 a team including Shady put 52 individuals from Caral and Áspero through exactly this analysis, using Bayesian models to turn the isotope numbers into diet proportions. The results were pointed.

The reconstruction found high C3-carbohydrate consumption at both sites — 55–68% of total calories at Áspero and over 70% at Caral — meaning plants, probably tubers, formed the dietary foundation.

Marine protein mattered more at coastal Áspero than at inland Caral, and at Caral marine consumption actually decreased over time — the opposite of a purely sea-based trajectory.

Read that against Moseley. Even at the fishing city, and even more at the valley capital, calories came mainly from plants, not fish. Fish supplied important protein — especially on the coast — but the caloric engine was C3 crops. The sea fed the people; the land powered them.

A larger study reanalyzed stable isotopes from 572 individuals across 39 Central Andean sites, spanning roughly 7000 B.C. to A.D. 200, to reconstruct diet region-wide in probabilistic terms.

Its conclusion: in coastal and middle-valley settlements it was plant cultivation, not fishing, that fueled the development of the earliest complex societies from about 3000 B.C.

So the pendulum that Moseley pushed toward the sea has swung back toward the soil — though, as Module 6 insisted, the honest verdict is a coupled system: plants for calories, fish for protein, cotton binding the two. The isotopes don’t crown a single winner; they show a balanced economy, with the agricultural side carrying more of the caloric load than the maritime theory supposed. With the economy mapped, Part III turns to the hardest question: who was in charge?

Everything so far — planned city, ranked houses, monuments, surplus economy — circles one loaded word: state. Was Caral a genuine state, the oldest in the Americas? This is the most consequential claim in the whole field, and it is fiercely contested. To weigh it you first need to know what the word means to an archaeologist.

Shady defines a state as a society that produces an economic surplus, whose members are organized in ranked social strata, and which is directed by permanent authorities with coercive power to uphold their decisions.

Notice the third clause does the real work. Surplus and ranking you have already seen in the dirt. But permanent authorities with coercive power — rulers who can compel obedience — is a much harder thing to prove from ruins. That is where the argument lives.

On this definition Shady calls Caral the capital of the first pristine state in the Andes — a state that arose with no older state to copy.

Haas and Creamer — who, despite their coming disagreements with Shady, also reject the modest "chiefdom" label — make a striking quantitative case that the Norte Chico was no ordinary tribal landscape.

The Supe, Pativilca, and Fortaleza valleys cover only about 1,800 km2, yet hold more than 30 large sites, all with monumental architecture and all occupied in the Late Archaic.

The sites sit 1 to 10 km apart with no smaller settlements between them — a packing unlike any known chiefdom society.

Their communal mounds exceed 50,000 cubic meters in volume, dwarfing the monuments of ethnographically known chiefdoms in Polynesia, Africa, and the Caribbean.

The logic: chiefdoms simply do not produce this — dozens of monumental centers crammed into a small area, each raising mounds bigger than any chiefdom on record. Something more organized was at work. Whether you call that "state" or reach for a new term, the modest labels clearly fail.

One more twist complicates the picture: residential architecture at Caral makes up less than 5% of the site area — these were overwhelmingly ceremonial and public places, not dense dormitory cities.

That last figure is a slow-burning fuse. If the great centers were mostly temples and plazas with few permanent residents, then who built them, and where did the people live? Hold the question — it detonates in the next module, where Haas and Creamer explain how power could work without a resident population and, remarkably, without war.

Module 9 left a puzzle: monumental centers with few permanent residents, so who built them and how were they made to? The textbook answer for early states is blunt — conquest. Haas and Creamer’s Norte Chico tells a different story, and its most startling feature is what is missing.

Across the Norte Chico there is an absence of warfare: no defensive walls, no weapons, no burned settlements, and sites placed in undefended locations.

If not force, then what? Haas and Creamer locate power in the economy you learned in Part II — specifically in a chokehold on two indispensable goods.

Power rested on controlling the production of the two things everyone needed: cotton, without which the coast could not make nets, and domesticated plant foods, without which no one had a balanced diet.

Recall the coupled economy: fishers cannot fish without cotton nets; everyone needs the valley’s crops for calories. Whoever controlled the inland fields controlled the one input the whole system could not do without — a far subtler lever than an army.

Inland power-holders then drew in coastal labor through positive and negative sanctions — providing or withholding those vital resources.

The builders themselves likely came seasonally: with small resident populations inland, coastal fishing communities came to the Norte Chico during the July–August lull to trade, worship, and work.

And they were fed for it — feasting was built into construction, with fire-altered cooking rock scattered through the mounds at about one rock per cubic meter of fill.

Picture the whole machine: no soldiers, but a calendar of gatherings where coastal families come inland in the dry season, feast, worship at the pyramids, and add a season’s labor to the mounds — then go home with cotton and crops. Power here looks less like a fist and more like a host who controls the pantry and throws the only feast in town.

The final binding force was ideology: religion was the tool the governing class used to strengthen social cohesion and, at the same time, to exercise control and justify its own privileges.

So the Norte Chico answer to "why do we obey?" is not "or we die" but a braid of economic dependence, seasonal feasting, and shared religion — power that people walked toward rather than fled. Whether that braid amounts to a "state" is the unsettled question; that it worked, for over a thousand years and without war, is the remarkable finding. Next we look at the religion itself.

Religion held the Norte Chico together (Module 10). But "religion" in the dirt is not a belief you can excavate — it is a set of objects and spaces made for no practical purpose. Caral left a remarkable trove of them, and the star find rewrites how early the Andes were keeping records.

From the Temple of the Amphitheater came a set of 32 flutes, carved from condor and pelican bone.

Nearby lay 38 cornets, made from camelid and deer bone.

Pause on the method. A single flute could be a toy; thirty-two flutes and thirty-eight cornets cached together in a ceremonial court is a different kind of fact — it implies organized, collective performance, planned and equipped. Quantity and context turn objects into evidence for institutions.

Caral’s builders repeatedly buried valuable things on purpose — a behavior that only makes sense as ritual, since the objects were put permanently beyond use.

Recall from Module 5 that temples were entombed and rebuilt in phases; those burials often included deliberate offerings — bundles of willow, burnt textiles, figurines, food — sealed into the fill. Later Andean peoples would do the same for millennia. The very act of building was an act of worship.

The single most consequential ritual find is a humble-looking tangle of knotted cords.

In the Gallery Pyramid, a multiple offering on a pallet included an ancient quipu — implying that a coded system for recording information was already in use at Caral.

Sit with that span. The knot-record technology that would one day administer the largest empire in the pre-Columbian Americas appears already, in embryo, in a Caral offering. It is the strongest single thread (literally) tying this first civilization to everything that follows — the theme of Module 16. First, though, two modules that widen the lens from Caral to its whole world.

We have stood inside Caral long enough to mistake it for the whole story. It is not. Caral is one node in a crowded regional network — and how you read that network decides one of the field’s sharpest disputes: was Caral a capital that ruled, or simply the most famous member of a league of peers?

The Norte Chico proper is four adjacent river valleys north of Lima — Huaura, Supe, Pativilca, and Fortaleza.

Across these valleys are more than 20 large early sites sharing the same signature — monumental architecture, extensive residences, and no ceramics.

This is the regional fact that upgraded "oldest city" to "oldest civilization" back in Module 3: not one precocious site but a whole synchronized landscape of them, rising together around 3000 B.C. Civilization is a regional phenomenon, and here the region is dense with it.

Caral was not even the largest: Era de Pando covers about 79.74 hectares to Caral’s 66.

Here the two camps split — and notice it is a disagreement about politics, read from the same stones.

Shady argues Caral was the capital — the seat of government of a Supe Valley state whose political hegemony and prestige radiated across the region, marked by the shared pyramid-and-sunken-plaza template.

On her reading, the repeated architectural signature across the valleys is the footprint of Caral’s influence — a center projecting a template outward, the way a capital does.

Haas and Creamer read the same density differently: a regional system of many interacting centers, integrated economically and ritually, without needing one site to rule the rest.

Don’t rush to a winner. The honest position is that the region was unmistakably integrated — shared architecture, shared economy, shared ritual — while the political question of central rule versus confederation remains open. What is certain is that no single valley invented this alone; it was a regional achievement, which is why the next module turns to how that whole society was organized and planned.

A civilization is not just monuments and economy; it is a way of organizing people. With no texts, how do we reconstruct the social order of a society dead for 4,000 years? Partly from the layout and objects — and partly by reading them through the well-documented Andes of later times.

Beyond farmers and fishers, Caral supported specialists: curacas (chiefs) who managed settlements, astronomers who set the ritual calendar, quipu record-keepers, canal managers, physicians, and craftsmen.

A roster like that is itself an argument for complexity: only a society producing reliable surplus can free people from food-getting to specialize in astronomy, medicine, or record-keeping. Specialization and surplus are two sides of one coin — and both are checklist items from Module 1.

Recall the hanan/hurin split of the city from Module 4. That upper/lower division was not just architecture; it appears to organize the settlements themselves.

Across the Supe Valley the settlements were distributed by sections, with a distinct capital zone and a marked duality — paired sites on both banks of the river, mirroring the upper/lower halves within Caral.

So the same dual template scales up and down: within a city, and across the valley. This nested dualism is one of the most Andean things about Caral, and a strong thread to the societies that follow.

Finally, Caral was no isolated valley. Its offerings and refuse contain materials that could only have come from far away — proof of exchange networks reaching across ecological zones and hundreds of kilometers.

Long-distance exchange brought Spondylus shell to Caral from the warm waters off Ecuador, prized for predicting climate change and for ritual.

Other imports reveal the reach: from the Andes came condors and the lloque plant; from the Amazon came the giant snail, annatto, and the feathers of macaws and howler-monkey fur.

Trace those goods on a map and Caral sits at the center of a web pulling in the sea, the high Andes, and the Amazon — exactly the coast–highland–jungle connectivity of Module 2, now realized as actual objects in actual offerings. The transverse management of the territory was not a theory; it was Spondylus from Ecuador and monkey fur from the jungle, buried in a desert temple. With the society mapped, Part IV asks how it ended — and what of it never did.

Civilizations are supposed to fall dramatically — sacked, buried, forgotten. Caral’s ending is quieter and more interesting: not a catastrophe but a slow loss of centrality, as the spotlight of Andean history moved elsewhere. Understanding why is the last piece of the puzzle.

By 1800 cal B.C. the Norte Chico was losing its status as a focal point on the Andean landscape, after more than 1,200 years of preeminence.

It was eclipsed as larger polities rose to the north and south — new centers of development in the Casma and Rímac basins would take its place.

This is decline as relegation, not annihilation. The cities were not destroyed in war (there wasn’t any, Module 10); rather, after a millennium as the most prestigious society in the region, the Norte Chico was overtaken by bigger, later centers. The torch passed; the runner didn’t die.

If not war, what drove the decline? The honest answer is a mix, but the environment — the same double-edged coast from Module 2 — is a leading suspect, and it is exactly here that the newest research has most to add.

One physical clue is written in the landscape: coastal dunes migrated inland, a signature of El Niño-driven climatic shifts that may have contributed to the dispersal of the Caral-Supe population.

A 2022 study turned that general suspicion into a concrete physical story. Instead of just asserting "climate change," it reconstructed, step by step, how the coast itself was reshaped.

The engine is ENSO — the El Niño Southern Oscillation — whose swings between drought, flood, and wind-blown sand repeatedly battered the valley's farmland and fisheries.

The study even modelled the floods with Computational Fluid Dynamics, the same physics used to design aircraft, to reconstruct how slurry-laden El Niño runoff rebuilt the coastline.

The turning point was a giant El Niño that raised the Medio Mundo beach ridge around 1600 BC — a wall of flood-borne sand that sealed off the Supe bay.

Follow the consequences, because this is the whole collapse in one chain. With the bay sealed, wind-blown sand buried the lower valley’s best farmland and the once-teeming bays silted up; farming retreated to thin terraces near Caral, food production shrank, and the coast–valley exchange network from Part II came apart.

As those food and marine resources dwindled, the study concludes, the conditions made life untenable past about 1600 BC.

For a visceral sense of the force: the study notes that in the 1989 El Niño, a single Supe farmer watched 50 hectares of his bottomland wash away in one event.

Multiply that across a valley and repeat it every few years, and you can see how a civilization built on a narrow strip of irrigable land could be worn out — not conquered, but silted, flooded, and buried. That is the recent picture: the same El Niño rhythm that Module 2 said the Supe economy was always exposed to eventually, over centuries, closed the valley down.

Occupation did not simply stop: at Vichama, a fishing settlement in the Huaura Valley that had been part of the Caral-Supe network, monuments from the Late Archaic were reused and built upon in later periods.

And the successor cultures carried the tradition forward — the societies that replaced Caral continued its patterns, a line that runs on to Chavín and beyond.

So the ending reframes the whole story. Caral did not vanish; it was succeeded. Its people dispersed and regrouped, its cities were overtaken, but its ways of building, worshipping, and organizing were inherited by everything that came after. We turn next to that inheritance made concrete — first to Peñico, the successor city that rose in the Supe Valley at the very moment Caral fell, and then to the deeper legacy that runs all the way to the Inca.

Module 14 ended on an abstraction: Caral did not die, it was succeeded. Now meet the successor — in the same valley, and excavated so recently that its book appeared in 2025. If Caral is the birth of Andean civilization, Peñico is the first chapter of what came next.

Recent research — This whole module rests on the newest source in the course: the 2025 monograph Peñico: City of Social Integration in the Supe Valley during the Early Formative, by Ruth Shady and her colleagues — our clearest window onto the Supe Valley after Caral.

Peñico was founded in 1800 B.C., right after the loss of prestige of the first great Supe cities — the Sacred City of Caral among them.

It belongs to the early Formativo Temprano (Early Formative), flourishing between 1800 and 1500 B.C. — beginning exactly where Caral’s Late Archaic left off.

Sit with that overlap. In Module 14 the Norte Chico lost its centre by about 1800 B.C.; here, at 1800 B.C., a new city rises in the same valley. The spotlight did not go dark — it moved a few kilometres and changed hands.

Crucially, Peñico’s builders worked on the base of the traditions of their antepasados caralinos — their Caral ancestors — while adding architectural, mural, and sculptural art of their own.

The book’s title calls Peñico a city of social integration, and that is a claim about geography and economy, not a compliment.

Peñico stood at a strategic node of communication between the Supe and Huaura valleys, linking the peoples of the coast, the highlands, and the eastern (Amazon-facing) slopes.

Its urban core covered 16.6 hectares, set at 600 m above sea level on a terrace that made its buildings loom over the valley while lifting them above the floods.

Recall Module 2’s ladder of nine valleys joining coast, sierra, and jungle. Peñico is that ladder made into a town: positioned precisely where two valleys meet, it made its living as a hinge in the exchange web that Caral had built — inheriting the network even as it inherited the culture.

Peñico is a study in continuity-with-variation. The Caral template is unmistakable — monumental public buildings, a sunken circular plaza, ranked architecture implying social strata — but the details are new.

Music carried straight over from Caral: among the finds is a conch-shell ocarina, a wind instrument fashioned from a sea snail.

Set that beside Module 11’s 32 bone flutes at Caral: a thousand years on, in a new city, ceremony still meant organized music — the same cultural DNA, a different instrument.

They decorated the inner walls of the Salón Ceremonial Principal (Main Ceremonial Hall) with reliefs of parallel horizontal lines — the mural art the introduction flagged as Peñico’s own.

And they modelled figures in unfired clay, the most charming a small sculpture of a baby howler monkey — an unmistakably Amazonian animal.

That little monkey is quietly profound. Recall Module 13, where howler-monkey fur travelled from the Amazon to Caral through the exchange network. Here the jungle turns up again, now sculpted into Peñico’s art — proof that this integrating city still reached across the continent’s ecological zones, just as its ancestors had.

Peñico’s most distinctive economic signature is something Caral’s story only hinted at: a trade in colour.

A likely engine of Peñico’s prestige was hematite — iron oxide, the mineral ground to make red pigment — which the city extracted and circulated through the exchange network.

Red was no ordinary colour. Recall the temple walls of Caral, plastered and painted red (Modules 4–5); red carried deep symbolic weight in Andean cosmology. A city that controlled the source of that sacred pigment held something both economically tradeable and ritually charged — a neat emblem of how Peñico fused exchange and belief, the two threads this course has followed since Module 1.

We end where the significance actually lies. Caral matters less as a ruin than as a source — the point at which a whole family of Andean patterns first appears. This module makes that legacy concrete and then shows why the excavation, quite literally, has barely begun.

Basic architectural and organizational patterns first appearing in the Norte Chico in the third millennium B.C. provided a foundation for the similarities that define the Andean region across the succeeding 4,000 years.

The legacy is not vague "influence" but a set of concrete, recognizable features you have already met in this course:

• Dual organization — the hanan/hurin moiety split (Modules 4 and 13) structures Andean society right up to Inca Cusco.
• The quipu — the knotted-cord record (Module 11) becomes the administrative backbone of the Inca empire.
• Monumental forms — platform mounds and sunken circular plazas (Module 4) recur across later Andean ceremonial centers.
• The pilgrimage-and-feasting economy and reciprocal labor (Module 10) prefigure the way later Andean states mobilized work.

It would be a mistake to think Caral is a solved, dug-out site. The opposite is true — and new tools are making that vivid.

Modern surveys apply super-resolution radar (SAR) and elevation data processed with deep learning to hunt for buried structures across the Supe Valley and coast.

Their conclusion is humbling: many artificial structures in Caral-Supe remain undiscovered.

Even Caral’s size is an estimate — perhaps 3,000 residents in the city and up to 20,000 across the region — a reminder of how much basic demography is still being worked out.

Tie that back to Module 9’s finding that under 5% of Caral was residential: if the great centers were mostly ceremonial and much of the region is still unmapped, then our picture of where people actually lived — and how many — is genuinely provisional. The first city of the Americas is still being found. That is the honest note to end the evidence on, before we synthesize the whole course in the capstone.

You began this course able to say almost nothing about Caral. You can now weigh the evidence like a specialist. This capstone doesn’t add facts — it assembles the ones you have into the handful of big arguments that organize the whole field, and marks honestly where the arguments are settled and where they aren’t.

A large amount, hard-won from dirt and physics. Around 3000–1800 B.C., in four desert valleys north of Lima, a society built dozens of monumental centers — platform mounds and sunken plazas — with no pottery, no metal, no wheel, and no writing. It was dated by radiocarbon on short-lived fibers and shicra bags (Modules 3, 5), and it arose independently, one of the world’s six pristine civilizations and the oldest in the Americas by well over a millennium (Module 1). Its economy coupled valley farming (including the inedible keystone, cotton) with one of the planet’s richest fisheries (Modules 6–8); its society was ranked, planned, specialized, and dual-organized (Modules 4, 13); and it ran, remarkably, without warfare (Module 10).

1. Sea or soil? Moseley’s Maritime Foundations said the fishery built civilization here; the canals, the cotton chokehold, and above all the isotope diets (plants supplying most calories) have swung the verdict to a coupled economy — fish for protein, plants for energy, cotton binding them (Modules 6–8).

2. City or ceremonial center? Caral is monumental but under 5% residential (Module 9), so it sits between a true dwelling-city and a pilgrimage temple-complex — and remote sensing says much is still unmapped (Module 16).

3. State or network? Shady reads a pristine state with Caral as capital; Haas and Creamer read an integrated but decentralized regional system — both from the same 20-plus sites (Modules 9, 12). Integration is certain; centralized rule is not.

Caral is not just old. Because it is pristine — civilization invented from scratch — it is one of only six natural experiments the human species has run on the question that opened Module 1’s neighbor disciplines: why do we get government, hierarchy, and cities at all? Caral’s particular answer — surplus and exchange and religion, without conquest — is a data point the other five cradles cannot supply, and it is still the youngest-discovered and least-exhausted of them. That is why a dry valley in Peru rewrote the textbooks, and why the work is nowhere near done.

You now hold the framework and its open edges together. That is what it means to have mastered what we know — and what we don’t yet know — about Caral.