The Anasazi "Disappearance": A Migration, Not a Vanishing.

What actually happened, in a paragraph.

Between approximately 1130 and 1300 CE, the Ancestral Puebloan populations of the Four Corners region of the American Southwest — the central Mesa Verde region of southwestern Colorado, the Chaco Canyon and surrounding region of northwestern New Mexico, the Kayenta region of northeastern Arizona, and the southeastern Utah canyons — underwent a sustained reorganization that culminated in the effectively complete depopulation of the region by approximately 1300 CE. The reorganization had multiple stages: an initial decline of the Chaco Canyon ceremonial-political center (c. 1130–1180), a subsequent population concentration into the Mesa Verde and Kayenta core areas accompanied by the construction of the famous cliff-dwelling settlements (c. 1180–1280), and a final rapid abandonment of those settlements over a period of approximately two decades (c. 1280–1300). The depopulation coincided with the "Great Drought" of 1276–1299 CE, a multi-decadal precipitation deficit documented with annual precision by the Douglass tree-ring chronology and subsequent dendroclimatological work. Evidence for internal conflict during the late period of occupation is substantial: defensive features added to many sites, the construction of high-perched cliff dwellings difficult to approach by hostile parties, and direct evidence of violence at several sites including the Castle Rock and Sand Canyon massacres analyzed by Kuckelman and colleagues. The Cowboy Wash site in Mancos Canyon, Colorado, produced direct osteological evidence of cannibalism in the late twelfth century (Christy Turner and successors), which has been variously interpreted as evidence of severe stress, of intergroup violence, or of ritual practice. The populations that left the Four Corners did not vanish: they migrated south and southeast and were absorbed into or contributed substantially to the Pueblo populations of the Rio Grande valley in New Mexico and the Hopi Mesas in northeastern Arizona, where their direct descendants live today. Hopi, Zuni, and Rio Grande Pueblo oral traditions independently describe this migration. The case is "open" in this archive in the sense that the framing as a mysterious disappearance — while historiographically inaccurate — remains the form in which most popular readers first encounter the topic, and the file's purpose is to walk that framing toward the considerably better-supported migration-with-stress account.

The documented record.

Terminology: from "Anasazi" to "Ancestral Puebloan"

The term "Anasazi" was introduced into American archaeological literature in the 1930s, particularly through the work of Alfred Vincent Kidder and his students, drawing on a Diné (Navajo) word recorded in the field. Verified The Diné word anaasázi has been variously translated as "ancient ones," "ancient strangers," "ancient enemies," or "ancestors of enemies" depending on context. The contemporary Pueblo descendants of the populations the term refers to have for decades asked that it not be used, both because of the etymological connotations and because it labels their ancestors with an outsider's term. From the 1990s onward, the U.S. National Park Service, most academic publications, and most descendant communities have adopted "Ancestral Puebloan" as the preferred term [1].

This file uses "Ancestral Puebloan" preferentially; the title and case-meta block retain "Anasazi" because that is the form in which the question is most-often searched and which most popular readers will recognize. The persistent use of "Anasazi disappearance" in popular sources is itself part of what this file's framing addresses.

The Pueblo III period and the late occupation

The Ancestral Puebloan occupation of the Four Corners region is conventionally divided by the Pecos Classification (Kidder 1927, with subsequent refinements) into periods: Basketmaker II–III (c. 50 BCE–750 CE); Pueblo I (c. 750–900); Pueblo II (c. 900–1150); and Pueblo III (c. 1150–1300). The Pueblo III period is the one ending with the abandonment. Verified [2]

The Pueblo II period in the Four Corners is dominated by the rise of Chaco Canyon as a regional ceremonial-political center, with the construction of the great houses (Pueblo Bonito, Chetro Ketl, others) and the radial road system extending up to 250 km from the central canyon. Chaco's apparent peak is conventionally placed at approximately 1050–1130 CE. By about 1130–1180 the Chaco system was substantially diminished; the population had not left the broader region but had reorganized around new local centers.

The cliff dwellings and the defensive shift

From approximately 1190 CE onward, populations across the Four Corners moved from open-mesa-top and valley-bottom settlements to defensive locations: cliff alcoves accessible only by climbing routes, tower-house complexes with restricted sight-lines, and consolidated multi-family settlements with limited access points. Verified [3]

The most famous cliff dwellings — Cliff Palace and Spruce Tree House at Mesa Verde, Keet Seel and Betatakin at Kayenta, the Long House at Bandelier, and others — were constructed during this period and were occupied for only one or two human generations before final abandonment. The pattern is consistent across the region: a clear shift to defensive architecture in the late twelfth century, occupation through the thirteenth century, and abandonment by approximately 1300.

Tree-ring chronology and the Great Drought

The American Southwest is the region in which dendrochronology was developed as a precise dating tool, through the work of Andrew Ellicott Douglass at the University of Arizona from approximately 1904 onward. Verified Douglass and his successors constructed a continuous tree-ring chronology for southwestern wood that now extends back more than 2,500 years, with annual precision and the ability to identify individual years of low or high precipitation [4].

The "Great Drought" of 1276–1299 CE is the most-cited single climatic event in this record. Verified Twenty-three of the twenty-four years in that period show ring-width values below the long-term mean across multiple chronologies in the Mesa Verde and Kayenta regions; eleven of those years are well below the mean. The drought was not the longest dry period in the broader Holocene record, but it was the most severe sustained dry period during the relevant centuries of Ancestral Puebloan occupation.

The Wallace and Bowannie multi-decadal precipitation reconstruction (2007) and the subsequent work of Salzer and others have extended the resolution of the record. The Great Drought is now seen against a background of substantial multi-decadal climate variability in the eleventh through thirteenth centuries: the so-called Medieval Climate Anomaly produced both wet decades (which had supported population expansion in the early Pueblo III period) and dry decades (which produced the cumulative stress that the 1276–1299 drought finalized) [5].

Defensive features and direct evidence of violence

The shift to defensive architecture in the late twelfth and thirteenth centuries is not in itself evidence of actual violence; defensive architecture can also reflect perceived threat. Verified However, several Late Pueblo III sites in the central Mesa Verde region have produced direct osteological evidence of violent death events:

• Castle Rock Pueblo (southwestern Colorado), excavated by the Crow Canyon Archaeological Center, produced evidence of a final-occupation massacre. Skeletal remains showing perimortem trauma consistent with violent death were recovered in non-burial contexts (i.e., the bodies were not buried in the normal way), with burned structures, abandoned ground stone, and other indicators of catastrophic site termination. Dating places the event at approximately 1280 CE (Kuckelman et al. 2002, 2010) [6].
• Sand Canyon Pueblo, similarly excavated by Crow Canyon, also produced late-occupation evidence of violent death and abandonment under apparent attack, dating to roughly the same period [6].

The evidence at Castle Rock and Sand Canyon is taken by most current Southwestern archaeologists to indicate that intergroup violence was a real component of the final abandonment phase, alongside the climatic stress. Whether the violence was internal to the Ancestral Puebloan communities or involved external groups (Numic-speaking populations moving into the region from the Great Basin, for example) is not fully resolved.

The Cowboy Wash cannibalism site

The Cowboy Wash site (5MT10010) in Mancos Canyon, southwestern Colorado, was excavated in 1994 in advance of a Soil Conservation Service project. Verified The excavation produced disarticulated human remains in pueblo room fills, with cut marks, percussion fractures, and patterns of bone treatment consistent with butchering for consumption. Christy Turner and colleagues, who had argued for the existence of multiple cannibalism episodes in the Anasazi record across earlier work, identified Cowboy Wash as a particularly clear case. The dating places the event at approximately 1150 CE, in the early phase of the Pueblo III period and the early phase of the broader regional reorganization [7].

The Cowboy Wash evidence has been reinforced by the 2000 Marlar et al. publication in Nature, which reported the recovery of human myoglobin in a coprolite (preserved fecal material) from the site — direct biochemical evidence that the cannibalistic processing was followed by consumption [8]. The Marlar paper is one of the strongest single pieces of evidence for cannibalism anywhere in the global archaeological record.

The interpretation of Cowboy Wash is contested in the descendant community. The Hopi Tribe and other Pueblo descendant communities have objected to the framing of their ancestors as cannibals; some critics (notably Kantner 1999, in a careful review of the evidence) accept the osteological evidence but emphasize that one site does not constitute a regional pattern. The mainstream archaeological position is that the Cowboy Wash evidence is real and is one of probably several discrete instances of extreme-stress cannibalism during the period of regional stress, but that it is not characteristic of routine Ancestral Puebloan practice [9].

The migration and the descendants

The populations that left the Four Corners did not vanish. Verified They migrated south and southeast and were absorbed into the existing Pueblo communities of the Rio Grande valley in New Mexico (the modern Tewa, Tiwa, Towa, and Keresan-speaking pueblos) and the Hopi villages of northeastern Arizona, and contributed to the formation of the Zuni community of west-central New Mexico [10][11].

The archaeological evidence: distinct ceramic, architectural, and ritual traditions from the Mesa Verde and Kayenta regions appear in the Rio Grande and Hopi areas during the late thirteenth and early fourteenth centuries, in patterns consistent with the arrival of large groups of immigrants who joined existing communities. Population estimates of the Rio Grande pueblos increase substantially in this period in ways the local populations could not have produced through natural increase alone.

The oral evidence: Hopi clan migration traditions describe the arrival of clans from the north and the west; the traditions are specific about routes, intermediate stopping points, and the order of clan arrival at the Hopi Mesas. Zuni oral traditions similarly describe ancestral arrivals from the Mesa Verde region. The Rio Grande pueblos have parallel traditions [10].

The framing of the Ancestral Puebloan "disappearance" as a mystery is the framing of nineteenth-century Anglo-American observers who arrived at the empty cliff dwellings and either did not consult or did not credit the descendants who were living a few hundred miles south. Twentieth- and twenty-first-century archaeology has reincorporated the descendant testimony into the standard account, and the result is the relatively secure migration-and-absorption picture now taught.

The interpretive positions.

The "climate-driven migration" mainstream consensus

The current mainstream Southwestern archaeological consensus treats the Four Corners abandonment as a multi-causal event in which sustained climatic stress (culminating in the 1276–1299 drought, against a background of multi-decadal precipitation variability), accompanied by intergroup violence at the local scale, drove the systematic migration of populations to better-watered regions to the south and southeast. The receiving Rio Grande pueblos, the Hopi villages, and (slightly later) the Zuni communities absorbed the migrants. The framing of "disappearance" is a misreading by Anglo-American observers unfamiliar with the descendant communities. Verified as the consensus position.

The "internal conflict" emphasis (Turner, LeBlanc)

A subset of scholars has emphasized the role of internal violence and warfare, with Steven LeBlanc's Prehistoric Warfare in the American Southwest (1999) the most-cited single statement. The position: that warfare and the threat of warfare drove the defensive-architecture shift, that the climate stress would have been survivable without the social breakdown that produced raiding and counter-raiding, and that the final abandonment was a response to violence as much as to drought [12]. Claimed; partially supported by the Castle Rock and Sand Canyon evidence.

The "ritual reorganization" hypothesis (Lekson, Kohler)

Stephen Lekson, Tim Kohler, and others have argued that the abandonment is best understood as a deliberate religious-political reorganization in which the Chaco system — whose collapse preceded the final abandonment by more than a century — was followed by a sequence of attempts to re-found Chaco-like ritual centers (at Aztec Ruins, at Mesa Verde, and eventually at the Rio Grande), with the final migration representing the culmination of this reorganization rather than a refugee flight. The drought and the violence are real, in this framing, but the migration is read as a planned response to a longer-running social-religious cycle [13]. Claimed; consistent with the Hopi and Pueblo oral traditions of structured clan migration.

The "alien" / "lost civilization" fringe

A persistent fringe claim, advanced in various ancient-astronaut and lost-civilization popular sources, treats the Ancestral Puebloans as either a non-indigenous "lost" people or as a people whose disappearance involved supernatural or extraterrestrial intervention. Claimed

Why this fails: the descendant populations exist; the genetic continuity has been demonstrated by aDNA studies; the oral traditions track to specific archaeological sites; the technology of the Ancestral Puebloans is fully continuous with the broader pre-Columbian Southwest. The framing as a "lost" people is a Western misreading produced by the linguistic and cultural distance between Anglo-American observers and the Pueblo descendants. Disputed by the framing's proponents; the strong consensus is that the migration-and-absorption account is fully sufficient.

The remaining open questions.

The specific migration routes and timings

The general direction of the migration (south and southeast) is well-established. The specific routes individual groups took, the timing of their arrivals at receiving communities, and the social negotiations through which they were absorbed, are partially recoverable from ceramic and architectural traces but are not fully known. Continued archaeological work in the intermediate zones — the Acoma region, the Zuni-Cibola district, the upper Rio Grande tributaries — is gradually filling these in.

Whether external populations contributed to the late-period violence

The hypothesis that Numic-speaking populations (ancestral Ute, Paiute, Shoshone) moving south and east from the Great Basin contributed to the violent pressure on the late Ancestral Puebloan communities is plausible — the Numic linguistic expansion is dated to approximately the right period — but is not securely demonstrated by direct archaeological evidence. Whether the violence at Castle Rock and Sand Canyon is internal Pueblo-on-Pueblo violence or involves external attackers cannot be determined from the osteological evidence alone.

The role of soil degradation

Some studies (Force et al., Van West) have proposed that long-term agricultural soil degradation in the Four Corners contributed to declining yields independent of climate, making the system more vulnerable when the Great Drought arrived. The specific magnitude of this contribution relative to climate is partially but not fully constrained.

The Chaco collapse and its delayed effects

The Chaco system collapsed approximately a century before the final Four Corners abandonment. The exact relation between the Chaco collapse and the subsequent stresses — whether the loss of the Chaco regional integration left the local communities more vulnerable, whether refugees from the Chaco collapse contributed to the population pressure that produced the conflict-and-stress pattern, or whether the two events are largely independent — continues to be worked out.

The persistent "disappearance" framing

The continued use of "Anasazi disappearance" in popular sources, school curricula, and tourist materials at the major National Park sites is itself a research subject — specifically, a question about how Indigenous descendant testimony has been integrated (or not) into the public narrative about Indigenous archaeology. The Hopi Tribe, the Pueblo of Acoma, the Pueblo of Zuni, and other descendant nations have engaged in extensive public outreach to correct the framing; the framing nonetheless persists in many sources, including in the title of this file.

Primary material.

• The Ancestral Puebloan archaeological sites themselves: Mesa Verde National Park (Colorado), Hovenweep National Monument (Colorado/Utah), Canyon de Chelly National Monument and Navajo National Monument (Arizona), Chaco Culture National Historical Park and Aztec Ruins National Monument (New Mexico), Bandelier National Monument (New Mexico), and many smaller sites under state and tribal jurisdiction.
• The dendrochronological record, maintained by the Laboratory of Tree-Ring Research, University of Arizona.
• Excavation reports of the Crow Canyon Archaeological Center, particularly the Castle Rock Pueblo and Sand Canyon Pueblo projects.
• The Cowboy Wash excavation reports (Soil Conservation Service contract reports, 1994–1996; subsequent academic publications).
• The Marlar et al. Nature paper on the Cowboy Wash coprolite myoglobin analysis.
• Hopi clan migration traditions, recorded with tribal consultation; partial documentation in the Bureau of American Ethnology reports (early twentieth century) and in continuing tribal cultural-resource publications.
• Zuni and Rio Grande Pueblo oral traditions, documented with similar tribal consultation processes.

The sequence.

1. c. 750–900 CE (Pueblo I) Initial development of large above-ground habitation structures in the Four Corners.
2. c. 900–1050 Early Pueblo II. Chaco Canyon develops as a regional ceremonial-political center.
3. c. 1050–1130 Chaco peak. The Great Houses (Pueblo Bonito, Chetro Ketl, Pueblo del Arroyo, others) reach their maximum extent. The Chaco road system links outlying "outlier" communities up to 250 km away.
4. c. 1130–1180 Chaco system declines. Regional ceremonial center function shifts to Aztec Ruins, north of the original Chaco core.
5. c. 1150 Cowboy Wash event. Direct evidence of cannibalism in Mancos Canyon (Turner et al.; Marlar et al.).
6. c. 1180–1250 Population concentrates into defensive locations. Construction of the major cliff dwellings at Mesa Verde (Cliff Palace, Spruce Tree House, others) and Kayenta (Keet Seel, Betatakin) begins.
7. c. 1250–1276 Peak occupation of the cliff dwellings. Substantial population in the Mesa Verde and Kayenta core regions.
8. 1276–1299 The Great Drought, documented by tree-ring chronology.
9. c. 1280 Castle Rock and Sand Canyon massacre events (Kuckelman et al.). Late-occupation violent abandonments.
10. c. 1280–1300 Rapid final abandonment of the Four Corners. Migration south and southeast to the Rio Grande pueblos and the Hopi Mesas. Substantial population increases in the receiving communities consistent with absorption of migrants.
11. c. 1300–1500 The Hopi, Zuni, and Rio Grande Pueblo communities consolidate into their late-prehistoric and protohistoric forms.
12. 1540 Coronado expedition encounters the Pueblo communities (the Spanish arrivals do not encounter "disappeared" populations; they encounter the descendant pueblos, mostly in their current locations).
13. 1880s–1900s Anglo-American discovery and publicization of the Mesa Verde cliff dwellings (the Wetherill family; Gustaf Nordenskiöld's 1893 publication). The framing as a "vanished" or "mysterious" people develops in this period.
14. 1906 Antiquities Act and the establishment of Mesa Verde National Park.
15. 1929 A.E. Douglass establishes precise tree-ring dating for the Southwest.
16. 1994–1996 Cowboy Wash excavation.
17. 1999 Christy Turner et al., Man Corn, published. Steven LeBlanc's Prehistoric Warfare in the American Southwest published.
18. 2000 Marlar et al. Nature paper on the Cowboy Wash coprolite.
19. 2000s Crow Canyon Archaeological Center reports on Castle Rock and Sand Canyon (Kuckelman et al.).
20. 2002–present Adoption of "Ancestral Puebloan" by the National Park Service and most academic publications. Continued tribal-archaeological collaboration on the framing.

Cases on this archive that connect.

Göbekli Tepe (File) — a different case in which a once-impressive site was abandoned (or in Göbekli Tepe's case, deliberately buried) and the descendant populations went elsewhere. Both cases involve careful distinctions between "the site was abandoned" and "the people disappeared."

Atlantis (File 057) — the contrasting case in which a "vanished civilization" exists only in textual tradition, where the Ancestral Puebloan abandonment is a real, documented, well-understood reorganization that has only been called "vanishing" because the framing was set by observers unfamiliar with the descendants.

The Library of Alexandria (File 058) — another case in which the framing as a single dramatic loss conceals a longer history of gradual reorganization and partial continuity.

Planned: the Maya Terminal Classic collapse (File 061, the next file in this Ancient series, on a structurally similar Mesoamerican case); the Chaco regional system; the Bears Ears co-management arrangements with the Pueblo and Diné tribes; the Mississippian Cahokia abandonment.

Full bibliography.

1. Cordell, Linda S., and Maxine E. McBrinn. Archaeology of the Southwest. Third edition. Left Coast Press, 2012. The standard textbook treatment, including the terminology discussion.
2. Kidder, Alfred Vincent. "Southwestern archaeological conference." Science 66 (1927): 489–491. (Original Pecos Classification.)
3. Lipe, William D., Mark D. Varien, and Richard H. Wilshusen, editors. Colorado Prehistory: A Context for the Southern Colorado River Basin. Colorado Council of Professional Archaeologists, 1999.
4. Douglass, Andrew Ellicott. "The secret of the Southwest solved by talkative tree-rings." National Geographic 56 (1929): 736–770. The original announcement of southwestern tree-ring dating.
5. Salzer, Matthew W., and Kurt F. Kipfmueller. "Reconstructed temperature and precipitation on a millennial timescale from tree-rings in the southern Colorado Plateau, USA." Climatic Change 70, no. 3 (2005): 465–487. The Wallace and Bowannie (2007) reconstruction is in the same dendroclimatological series.
6. Kuckelman, Kristin A., Ricky R. Lightfoot, and Debra L. Martin. "The bioarchaeology and taphonomy of violence at Castle Rock and Sand Canyon Pueblos, southwestern Colorado." American Antiquity 67, no. 3 (2002): 486–513. With Kuckelman's subsequent 2010 paper, "The depopulation of Sand Canyon Pueblo, a large Ancestral Pueblo village in southwestern Colorado." American Antiquity 75 (2010): 497–525.
7. Turner, Christy G. II, and Jacqueline A. Turner. Man Corn: Cannibalism and Violence in the Prehistoric American Southwest. University of Utah Press, 1999.
8. Marlar, Richard A., Banks L. Leonard, Brian R. Billman, Patricia M. Lambert, and Jennifer E. Marlar. "Biochemical evidence of cannibalism at a prehistoric Puebloan site in southwestern Colorado." Nature 407 (September 2000): 74–78.
9. Kantner, John. "Anasazi mutilation and cannibalism in the American Southwest." In The Anthropology of Cannibalism, ed. Laurence Goldman, Bergin & Garvey, 1999.
10. Ferguson, T.J., and E. Richard Hart. A Zuni Atlas. University of Oklahoma Press, 1985. (For the Zuni oral-tradition migration record.)
11. Bernardini, Wesley. Hopi Oral Tradition and the Archaeology of Identity. University of Arizona Press, 2005.
12. LeBlanc, Steven A. Prehistoric Warfare in the American Southwest. University of Utah Press, 1999.
13. Lekson, Stephen H. The Chaco Meridian: Centers of Political Power in the Ancient Southwest. AltaMira Press, 1999; revised second edition Rowman & Littlefield, 2015.
14. Kohler, Timothy A., Mark D. Varien, et al. Emergence and Collapse of Early Villages: Models of Central Mesa Verde Archaeology. University of California Press, 2012.
15. Nordenskiöld, Gustaf. The Cliff Dwellers of the Mesa Verde, Southwestern Colorado: Their Pottery and Implements. Stockholm, 1893. (The principal early publication that established the Anglo-American framing of the cliff dwellings.)