Chaco regional system

At a glance

Where is Chaco, and why there?

Chaco Canyon is a 32 km gash in the sandstone of the San Juan Basin, 500 to 1,000 metres wide, cut up to 180 metres into Upper Cretaceous rock, running south-east to north-west. It is not symmetrical. The bedrock dips gently north-east, so the south side stands higher, debris fans spill in from the south, and the vertical cliffs line the north wall. That asymmetry is why every great house in the canyon sits on the north side facing south: the builders were using the cliff as a windbreak and the open southern exposure as a solar collector. Pueblo Bonito's four-storey rear wall stands about ten metres from the cliff, and its lower storeys face the sun.

The climate is marginal for farming and was marginal then. The canyon floor lies at about 1,980 metres (6,200 ft). Annual precipitation averages roughly 220 mm (8.5 in), but the variance is brutal: between 1950 and 1980 the yearly total ran from 85 mm to 350 mm, and the median is lower than the mean. The growing season averages 120 frost-free days, give or take 30. Rain arrives mostly as violent summer thunderstorms that sheet off the rock and vanish; the Chaco Wash is intermittent, and the canyon floor has repeatedly cut and filled with arroyo channels over its history. There is no permanent river, only a few seeps and springs in the box canyons.

The one abundant resource was building stone. The cliff-forming unit is the Cliff House Sandstone, laid down about 85 to 75 million years ago, and its lowest member splits into flat, bedded slabs — easy to quarry, easy to lay, and easy to erode, which is why the walls that survive look as though they have melted. What the canyon lacked was timber. And timber is what the great houses needed most.

What is a great house?

A great house is not a big pueblo. The park service's definition is precise: these buildings were planned from the start rather than added to room by room; room size tripled or quadrupled; ceilings doubled in height; second, third and fourth storeys were built; specialised kivas appeared — great, elevated, tower and tri-walled; and the walls were built with a thick rubble core faced with finely coursed sandstone veneer. The result is architecture that looks like masonry furniture, with courses so regular that early visitors assumed it was Mexican work.

Pueblo Bonito is the type specimen: D-shaped, its south wall nearly 175 metres long, its north arc rising four storeys. The room count depends entirely on what you count. The Chaco Research Archive gives "over 350 ground-floor rooms, 32 kivas and three great kivas"; the park's brochure says more than 600 rooms and 40 kivas; other papers give 650, and some 800. The difference is upper storeys — rooms that existed, collapsed, and were reconstructed into the totals — plus kivas counted by roofed chambers or by floor features. Write the range: roughly 650 rooms, with estimates up to 800, and say what is being counted.

Chetro Ketl is second: about 400 rooms, the largest surface area of any Chacoan building at more than 2.3 hectares, with two great kivas in its plaza. Pueblo del Arroyo, the fourth largest, is the only tri-walled structure in the canyon and is oriented east rather than south. Kin Bineola is E-shaped rather than D-shaped, three storeys at the rear, with a great kiva about 15 metres across. Una Vida and Hungo Pavi are largely unexcavated. Casa Rinconada is a single isolated great kiva, 19.2 metres across the interior, with 34 niches around the wall and a 12-metre underground passage entering from the north — so that a person could appear in the kiva floor without being seen.

The masonry itself is a lesson in labour. The walls are two skins of shaped stone around a rubble fill, with mud mortar, and the stone was quarried from the cliffs within the canyon. The beams came from elsewhere: roughly 200,000 timbers in the park's count, more than 240,000 in the newer tree-ring study. Most were ponderosa pine or spruce/fir, and tree-ring sourcing has traced about 70 per cent of the sampled beams to the Chuska and Zuni mountains, more than 75 km away. Before AD 1020 nearly all came from the Zunis; by 1060 the Chuskas had taken over. The park notes that the ten largest ruins alone required roughly 45,000 high-elevation trees. No ancient society in North America moved more wood further by muscle alone.

When was Chaco built?

For a place this famous, the chronology is surprisingly soft at both ends. The conventional start is AD 850, but that is a phase boundary in Lekson's synthesis, not a foundation date: Early Bonito 850–1040, Classic Bonito 1040–1110, Late Bonito 1090–1140. Construction is documented in the mid- to late ninth century at Pueblo Bonito, Una Vida and Peñasco Blanco; Chetro Ketl's main build, on its own tree-ring dating, is 1030–1090. The "850" is best read as "by at least the middle AD 800s", which is how the park phrases it.

What changed the subject was dendrochronology. A. E. Douglass applied tree-ring dating to Pueblo Bonito in 1928–29, and the results were a shock: the great houses dated not to some late prehistoric climax but to the Pueblo II period — the same era as the small ordinary houses in the same canyon. The great houses are not the endpoint of a long development. They are contemporary with the villages around them, which means their size is a social fact, not an evolutionary one. That single result created Chaco archaeology as a problem.

The boom came in the eleventh century. Most accounts put the canyon's peak between 1080 and 1100, with "seven new great houses in the Chaco Core" added as the timber supply shifted from the Zunis to the Chuskas after 1060. The regional system — roads, outliers, shared architecture — is later than the first great houses: the park dates Chaco's rise to regional centre to about 1050, and the classic integration to 1040–1140.

The end is better dated than the beginning. Great-house construction stops in the 1130s; the canyon was almost completely depopulated by about 1150; the last ceremonial use runs to roughly that date. UNESCO's "850 to 1250" is a century too late for the canyon proper — it covers the wider World Heritage property, which includes Aztec Ruins.

How many people lived at Chaco?

This is a genuine, unresolved argument, and the honest answer is a range with named camps. The most widely cited figure for the canyon at its height is 2,100 to 2,700 people, from Lekson and from Plog, Heitman and Watson, as summarised by Lentz and colleagues. The High Altitude Observatory gives 2,000 to 5,000. Archaeology Southwest has argued for "hundreds, not thousands", since many great-house rooms were not lived in. Kohler, Brumbaugh and Bocinsky argued in 2025 that even the highest estimates for the population served by the system are "somewhat too low".

The spread has four causes, and the article should name them: whether roofed rooms count as dwellings; whether the canyon's population was permanent or episodic, swollen by pilgrims and labour gangs; whether you count the canyon or the whole basin; and whether you assume local maize fed everyone or food came in from outside. The park's own summary is the safest sentence available: "Even simple questions like resident population size have not yet been answered."

For the regional population, the current range is enormous: 10,000 to 200,000 people in the entire Chaco World at its peak, with most estimates between 20,000 and 50,000. That spread is the story. Anyone who gives you a single number is choosing a model, not reporting a fact.

How did they farm in a canyon with no river?

Badly, or cleverly, depending on which school you follow — and this is one of the sharpest live arguments in Southwestern archaeology. The case for local farming rests on real engineering: multiple canal systems, mostly on the north side of the canyon, with masonry diversion gates, gridded fields and check dams. Gordon Vivian recorded a gridded field system near Chetro Ketl in the 1960s. In 2018 a team using optically stimulated luminescence dating and lidar traced Bonito-phase canals at the far west end of the canyon, near the confluence of the Chaco and Escavada washes, and strontium isotopes confirmed that water came from more than one source. Some geologists think the water table then lay only about a metre below the surface, which would have made hand-dug wells possible.

The case against is equally concrete. The park service itself says that "although complex water catchment and control structures were installed throughout Chaco Canyon, there is limited evidence for large-scale agricultural efforts", and adds that some researchers think the water-control features were for making mud mortar and plaster rather than for farming. Experimental plots planted in 1977 at Pueblo Alto, the Chetro Ketl fields and elsewhere produced poor results; the conclusion was that anyone growing crops here "clearly had to pay extremely close attention to their crops and were experts". Isotope studies have also challenged the old idea that maize was shipped into the canyon, while supporting the hunting of large animals in the uplands.

The defensible position is that Chaco's farmers were skilled, water control was real, and the canyon's capacity was still far below its monumental ambitions. That gap is the point: a place that could not feed itself was importing timber, turquoise and labour on an industrial scale.

Why did Chaco run on turquoise?

Because turquoise was the one thing everyone wanted and almost nobody had. It does not occur in Chaco Canyon. The nearest major source is the Cerrillos Hills near modern Santa Fe, more than 200 km away, including Mount Chalchihuitl, the vast prehistoric pit recorded by William Blake in 1858. Isotopic work on museum collections has since shown that Chaco's turquoise came from several source regions, with different procurement patterns at different great houses — so Cerrillos is the nearest and most-cited source, not the only one.

The volumes are staggering. Pueblo Bonito's early excavations recovered more than 100,000 finished ornaments of turquoise, marine shell, jet and shale — all materials combined, not turquoise alone. Room 33, the burial crypt, held as much worked and unworked turquoise as any site in the Southwest. Burial 14 in that room wore 11,200 turquoise beads and 3,300 shell beads; the burial beneath it had 5,800 turquoise beads. One compilation puts the canyon's total at more than 200,000 pieces.

The manufacturing evidence is the most interesting recent development. The old model had production in small houses and consumption in the great houses — a "corporate" economy in which elites coordinated rather than commanded. But a 2025 reanalysis by Hannah Mattson counted 758 lapidary tools at Pueblo Bonito, almost eight times the number at the small-house site 29SJ629 and more than three times the combined assemblages from all Chaco Project sites. Her conclusion is that significant jewellery-making happened inside Pueblo Bonito, partly within elite households. That breaks the premise of the corporate model and is the strongest new challenge to the old consensus.

The workshop debris is everywhere, which is the detail that makes the economy vivid. One floor at 29SJ629 held more than 3,000 pieces of turquoise, with microdrills of silicified wood and abraders, dated AD 925–1050. A site north of Fajada Butte produced unfinished ornaments and what looks like a jeweller's toolkit. Survey work found turquoise debris on anthills at 94 per cent of Early Bonito-phase sites in the canyon, and beads broken during drilling at more than 40 per cent. You can still find the swarf on the ground.

Did Mesoamerica reach Chaco?

Yes, but not in the way the old popular accounts implied. There is no evidence of Mexican colonists, garrisons or missionaries. What there is, is objects and knowledge.

The strongest single case is cacao. In 2009 Crown and Hurst ran HPLC-MS on absorbed residues from five sherds from Pueblo Bonito's trash mounds and found theobromine — the cacao biomarker — in three of them, all probable cylinder jars, dated AD 1000–1125. It was the first demonstrated use of cacao north of the Mexican border. The vessel evidence matches: fewer than 200 cylinder jars are known from the whole Southwest, and 166 of them come from Pueblo Bonito, with 111 in a single cache in one room. The nearest cacao cultivation at Spanish contact was in central Mexico, including parts of northern Veracruz and Colima.

Scarlet macaws tell a similar story with feathers attached. More than 400 macaw skeletons have been recovered from the Southwest, 35 of them at Chaco, with direct dates showing birds acquired persistently between AD 900 and 1150 — and the earliest group at AD 885–990, fifty to a hundred and fifty years earlier than scholars had assumed. Room 38 at Pueblo Bonito held twelve macaws in a 25 cm layer of guano. The birds' historic range reached Tamaulipas, near Tampico, about 1,800 km away. Whether they were bred locally is disputed: Watson and colleagues found no nestlings under four months, no eggshell and no pens, and concluded they were not bred in the Southwest; a 2018 ancient-DNA study found remarkably low genetic diversity consistent with breeding a small founder population somewhere in the region, at a site nobody has found.

Copper bells complete the set: 672 of them from at least 113 Southwestern sites between AD 900 and 1450, with no archaeological evidence of local metallurgy, which points to West Mexico. What the exotics prove is not colonisation but procurement, and the transfer of esoteric knowledge. As Crown and Hurst put it, the growth of Pueblo Bonito "was coupled to the ability of some individuals to acquire Mesoamerican objects, cacao, and macaws, and perhaps their esoteric knowledge".

What were the roads for?

Chaco's roads are not paved highways and they are not visible from the ground. They were made by clearing vegetation and topsoil to create depressed, straight pathways 9 to 10 metres wide, sometimes lined with berms or low walls. Most have been found from the air, and lidar is still adding segments: one study found 3.2 km of road that no one had detected on foot.

The numbers are peculiar. The Great North Road runs about 50 km north from Pueblo Alto and stops abruptly at the edge of Kutz Canyon, going nowhere. The South Road runs about 55 km towards the Red Mesa Valley. Some regional roads reach 57 km. But most are short: in Roney's synthesis, 92 per cent are less than 5 km long and 57 per cent are less than 1 km. They connect great houses to shrines, kivas and landscape features, not to each other in a continuous network. The structures along them are mostly herradura shrines, not houses.

There are two readings, and the park service sets both out. The practical one: experiments show that walking a surviving road segment uses considerably less energy than walking the same distance on unmodified ground, and with 200,000 beams to move, a prepared surface would have been useful. The ritual one: "in a society that had neither pack animals nor wheeled vehicles it is unclear what, if any, practical need could have required roads thirty feet wide". Zuni and Acoma sacred trails still use surviving prehistoric road segments in places, which is about as direct a link between ancient roads and living practice as archaeology can offer.

The middle ground is probably right: roads made movement easier and made a statement at the same time. The Great North Road, running fifty kilometres to a canyon edge and stopping, is the single fact that carries the whole argument.

Who ruled Chaco?

Five answers are on the table, and the honest article names them. The ritual-centre model, associated with Renfrew, Van Dyke and Kantner, sees great houses as places of "high devotional expression", funded by offerings from pilgrims in a sacred economy. The administrative-palace model, argued by Lekson and Earle, sees elite residences, staple finance and attached specialists. The corporate-chiefdom model, from Blanton, Feinman and Hagstrum, sees production dispersed in small houses and consumption centralised at great houses, with elites coordinating rather than commanding. The feasting model, from Crown and Mills, sees great houses as stages for hospitality and drinking that created status. And Lekson's maximalist version sees Chaco as "the one state-like polity the U.S. Southwest produced", a derivative of Mesoamerican city-states.

The hard evidence has moved against the simplest versions of all of them. In 2017 Kennett and colleagues showed that nine individuals in Pueblo Bonito's Room 33 crypt share identical mitochondrial genomes, with mother–daughter and grandmother–grandson relationships, forming an elite matriline that lasted about 330 years, from AD 800 to 1130. That is a hereditary dynasty, which the corporate model struggles to explain. Meanwhile Mattson's lapidary-tool study puts workshop-scale production inside the great house, which the "production in small houses" model struggles to explain. And Mills has said plainly that Chaco's degree of hierarchy "was different from later Pueblos", and that this is one of the most important changes in Chaco scholarship in recent years, "including my own".

The violence is real but thin, and mostly not at Chaco. The Middle period, AD 900–1150, was by Southwestern standards unusually peaceful: few homes burned, violent trauma rare, settlements in non-defensive locations. Lekson calls it "Pax Chaco". The exceptions are spectacular. Turner reported a charnel pit at a ruin he called Small House in Chaco Canyon, dated about AD 900, containing the disarticulated and deliberately damaged remains of eight people, with cut marks, burning and missing vertebrae — his evidence for cannibalism. Turner and Turner's wider survey found damaged remains at more than 40 sites, with 32 assemblages meeting their criteria and a minimum of 295 individuals. But the cannibalism claim is genuinely contested on method: the same bone signature can be produced by secondary burial and deliberate defleshing, and the distinction requires cut-mark micromorphology that has only recently been developed. Lekson, who is not squeamish, wrote that "claims of cannibalism remain controversial".

The most famous "burned great house" is a trap. Room 28 at Pueblo Bonito was burned, but the fire was set under a shelf, the doors were sealed, some with vessels, and 102 cylinder jars were destroyed around AD 1100. That is deliberate ritual termination, not an enemy attack. Chaco's violence is not a massacre; it is a handful of damaged bodies, a dynastic crypt, and a room full of jars that someone burned on purpose.

Did Chaco watch the sky?

Some of it, demonstrably; more of it, probably; and a famous part of it, no longer visible.

The demonstrated part is geometry. Pueblo Bonito's main dividing wall runs within 15 minutes of arc of true north–south, and the southern wall west of it is close to east–west. Casa Rinconada's symmetry axis, defined by its two T-shaped doors, is aligned north–south within 20 minutes of arc; its 34 niches are spaced so that lines between opposing pairs pass within 10 cm of the centre; and the sockets for its four roof posts form a square centred within 10 cm, oriented to within 30 minutes of the cardinal directions. That is not accidental. It is also not, by itself, astronomy.

The famous case is Fajada Butte, where three sandstone slabs, 2 to 3 metres high and weighing about two tonnes each, stand a few centimetres apart in front of two pecked spirals. Anna Sofaer and colleagues reported in 1979 that at summer solstice a narrow vertical shaft of light moves down through the centre of the larger spiral, with corresponding marks at equinox and winter solstice, and concluded that placing the slabs in that position "involved deliberate human intervention". The site was recognised in 1977. In 1982 Newman, Mark and Frehlich argued that the slabs were a natural rockfall, and that the petroglyphs were made afterwards by people who had observed the light — a critique that concedes the observation while denying the construction. Then, in 1989, two of the slabs shifted, and the light no longer falls as it did. The park service attributes the shift partly to erosion accelerated by human traffic, and access to the butte has been restricted since 1982.

Casa Rinconada has a similar story. It is often described as a solstice observatory, with a beam of light entering a lone window and illuminating a niche on the west wall. The High Altitude Observatory's own account lists three reasons to doubt it: the upper wall was reconstructed after excavation, so the window's original position and shape are unknown; the north-west roof post would have blocked the beam; and a small room was later built against the outer wall where the window now is. "All this makes it rather unlikely that the now popular solstitial 'light show' was purposefully built into Casa Rinconada."

The lunar claim is the largest and the weakest. Sofaer and colleagues have argued that twelve major Chacoan buildings are oriented to the extremes and midpoints of the moon's 18.6-year cycle, forming a regional pattern across about 5,000 km², and that the Great North Road commemorates the relationship between the canyon and celestial north. The 1979 paper itself only says that the Fajada site "may have been used" for lunar observation, and its lunar reconstruction is explicitly speculative. Treat the standstill alignments as Sofaer's argument, not as settled science — while noting that the solar and cardinal alignments are measured facts, whatever they meant.

Why did Chaco end?

Not for one reason, and probably not quickly. The archaeological end-date is clear: construction stops in the 1130s, the canyon empties by about 1150. The explanations come in layers.

Climate is the sharpest. A reconstruction of Colorado River flow covering AD 762–2005 shows its most extreme low-frequency feature in the mid-1100s: a hydrological drought with a decrease of more than 15 per cent in mean annual flow averaged over 25 years, and no high flows for about six decades. Tree-ring reconstructions put warm drought windows at AD 990–1060, AD 1135–1170 and AD 1276–1297. Benson and colleagues argue that maize depended on winter precipitation for germination and monsoon rain for growth, and that both failed; they describe two droughts, the first emptying most great houses of the central San Juan Basin, the second clearing the Four Corners. Note the dating: the drought is 1135–1170, not "the 1130s–1150s" that older accounts give.

Timber and landscape pressure are real, though the classic chain is shakier than it sounds. The canyon could not supply its own roof beams: 240,000 timbers, 70 per cent from more than 75 km away. That implies sustained pressure on the wider landscape. But the familiar story — deforestation, erosion, a falling water table, collapse — is a long-standing hypothesis associated with Vivian and Dean rather than a demonstrated sequence, and at least one team (Wills and colleagues) argues there is "no substantive evidence for deforestation at Chaco". Keep the timber hauling, which is documented, and hedge the erosion story, which is not.

Social reorganisation matters as much as climate. The end was not a simple evacuation but a shift: Chaco's influence continued at Aztec, Mesa Verde and the Chuskas, and the networks contracted before the population moved. Warfare is the weakest strand and the most sensationalised. The "Chaco death squads" thesis — an elite using public violence and cannibalism as political control — was proposed by Turner and Turner and criticised on chronology and method; it is a hypothesis in the collapse debate, not a finding.

The most interesting and most disputed idea is Lekson's Chaco Meridian: that Chaco, Aztec and Paquimé were deliberately placed on the same line of longitude, and that a small elite moved the capital north to Aztec around 1125 and then south to Paquimé about 150 years later. The supporting facts are real — all three were the grandest centres of their eras and sit within a few hundred metres of a true north–south axis partly traced by the Great North and South roads. The rebuttals are also real. David Phillips of the University of New Mexico called the argument "novel, provocative, and wrong", noting that the statistical case depends on narrowing the search cone from a half-circle to about twenty degrees, turning a one-in-a-hundred-and-eighty coincidence into a one-in-twenty one. And Lipe and Varien point out that Aztec looks partly abandoned exactly when the elite is supposed to be running an empire from it.

Where did the Chacoans go?

They went in all four directions, and they are still there. The park service states that archaeologists and tribal representatives agree the Chacoans left towards the Zuni–Acoma area to the south, the Mesa Verde region to the north, the Hopi mesas to the west, and the Rio Grande valley to the east, joining communities that already existed.

The first move was north, to the Totah — the confluence of the San Juan and Animas rivers — where the Aztec regional system replicated Chacoan great houses, planned layouts and kivas. Aztec West was built in the decade 1110–1120, with about 400 masonry rooms, 12 kivas and a great kiva, and is the classic Chacoan D-shape. Salmon Ruin was founded earlier, at the end of the drought of AD 990–1060. The conventional story runs Aztec from 1110 to 1275, but that is really two occupations: a Chacoan foundation, a decline or hiatus from the 1130s or 1140s, and a Mesa Verdean reoccupation in the 1200s that ended around 1275.

Then came the wider collapse. The Great Drought of the 1270s and 1280s — 1276–1297 in the tree-ring window — and the social stress around it emptied the northern San Juan. The best evidence is the disappearance of dated beams: beam-cutting fell from a high level in the 1260s and 1270s to almost none in the 1280s, and the region was largely abandoned by 1300 in favour of the Zuni–Acoma area, the Hopi mesas and the Rio Grande.

Twenty-six tribes are traditionally associated with Chaco today: 21 Pueblo nations plus the Hopi, and five non-Pueblo nations — Navajo, Jicarilla Apache, Mescalero Apache, Southern Ute and Ute Mountain Ute. The park's own statement is the one to quote: "Although not all clans or kin groups in modern-day Pueblos migrated from Chaco Canyon, there are groups in almost all Pueblos that did... Thus, there is no one-to-one direct affiliation of a single present-day tribe with the inhabitants of Chaco Canyon." Pueblo descendants describe Chaco as a gathering place where peoples and clans converged; the Hopi name for the canyon is Yupkoyvi, "the place beyond the horizon"; the Acoma name means "place of greasewood". At least ten modern Navajo clans are, by oral history, descended from Chacoans.

The reburial dispute is part of this story, not a footnote. In May 1999 the park published a notice of inventory completion covering 265 Native American remains and 743 funerary objects, affiliated with twenty tribes. The Hopi objected that consultation had not been tribe-by-tribe and that affiliation had been decided for the park as a whole rather than object by object. In February 2000 the federal Native American Graves Protection and Repatriation Review Committee found that the Hopi's complaints "have merit" and recommended that the park withdraw its notice and reassess. The park reburied the remains and objects in its possession in 2005.

What is left, and who owns it?

Chaco Canyon National Monument was proclaimed by Theodore Roosevelt on 11 March 1907, under the Antiquities Act, because the ruins were "of extraordinary interest because of their number and their great size". In December 1980 it became Chaco Culture National Historical Park, with 13,000 acres added and a system of 39 Chaco Protection Sites on surrounding federal, state and Navajo land. On 8 December 1987 it was inscribed as a UNESCO World Heritage Site, jointly with Aztec Ruins National Monument and smaller outlying sites, under criterion (iii) alone: the sites "graphically illustrate the architectural and engineering achievements of the Chacoan people, who overcame the harshness of the environment of the southwestern United States to found a culture that dominated the area for more than four centuries". The listing has no buffer zone, and UNESCO's own statement concedes that road traces and outlier communities extend well beyond the property.

The living argument is about what surrounds the park. In June 2023 the Interior Department withdrew about 336,404 acres of public land within a ten-mile radius of Chaco from new mining and oil and gas leasing for twenty years — Public Land Order 7923 — on the ground that the buffer protects more than 4,700 known archaeological sites outside the park. The withdrawal applies only to federal land and minerals, not to private, state or tribal minerals, and Navajo allottees can still lease.

That protection is now under review. In November 2025 the Bureau of Land Management began consultation to revoke the order entirely, following an executive order on energy. In July 2026 it published a draft environmental assessment with three options: keep the ten-mile buffer, revoke it in full, or revoke everything beyond five miles. The comment period closed on 29 July 2026, after objections that only 14 days had been allowed. The Hopi Tribe formally opposed revocation and asked for a full environmental impact statement. The National Trust for Historic Preservation put the Greater Chaco Cultural Landscape on its 2026 list of America's 11 Most Endangered Historic Places, for the second time. As of 19 October 2026 no final decision had been published: the order stands, and the fight is live.

The political tension inside that fight is worth stating plainly, because it is often flattened. The Pueblo nations and the Hopi want the ten-mile buffer. The Navajo Nation Council has opposed it, asking for five miles, because Navajo allottees inside the larger buffer depend on lease income. Chaco is not a single tribal cause; it is a place where different Indigenous nations disagree about what protection means.

Common myths about Chaco

Timeline of the Chaco regional system

What the map shows next

At 1228 on the map the centre of gravity has moved north: the Aztec regional system holds the Totah, while Norse Greenland is at its height far to the north-east and the Chacoan world has dissolved into its successor communities.

At 1616 on the map the Pueblo world is under European pressure and the Haudenosaunee confederacy dominates the north-east. The canyon itself has been empty of farmers for four and a half centuries, and the people who built it are the pueblos of the Rio Grande, Hopi and Zuni.