New Evidence Reveals Violent Final Days at Arizona’s Montezuma Castle

It’s one of Arizona’s most famous landmarks: a pair of 900-year-old limestone cliff dwellings whose sudden abandonment centuries ago has proven to be one of the Southwest’s most enduring mysteries.

New evidence suggests that the site — now part of Montezuma Castle National Monument — was not simply evacuated by its inhabitants, as archaeologists have believed for more than 80 years.

Instead, recent research shows that its final days were likely fraught with violent conflict and death — an account corroborated by Native American oral histories of the site’s collapse some 600 years ago.

“It changed the conventional thinking [about the site],” said Matt Guebard, archaeologist with the U.S. National Park Service, about his research into the cliff dwellings’ fate.

“The conventional thinking left a lot of room for speculation about when and why people might have abandoned the area.

“The new study doesn’t explain why violence occurred at the site, but it does provide a mechanism for the abandonment.”

Guebard’s research is part of a renewed effort to better understand how the two cliff dwellings — located in central Arizona’s Verde Valley — were built, used, and eventually deserted by the culture known as the Southern Sinagua, a group that later gave rise to many modern Hopi clans.

montezuma-castle arizona
The Bearstrap Clain of the Hopi trace their ancestry to the architects of Montezuma Castle, who built the cliff dwellings around the year 1100 and abandoned it in the late 1300s. (Photo by Lucas Hoedl, courtesy of Matthew Guebard. May not be used without permission.)

(See a striking discovery from Sinagua history: “Thousands of Ancient Petroglyphs, ‘Dramatic’ Solar Calendar Reported in N. Arizona“)

“Aside from a small research project conducted in the 1980s, very little archaeological work had been done since the 1930s,” Guebard said.

When the site was originally excavated some 80 years ago, archaeologists found abundant evidence of a massive fire in both structures, from charred wall plaster to burned roof beams.

One set of buildings, erected inside the limestone rockshelter high above the canyon floor, became known as Montezuma Castle. The other, constructed on the ground against the canyon’s north wall, was dubbed Castle A.

Their charred remains led many 20th century archaeologists to conclude that the buildings burned, for reasons unknown, some time after they had been abandonment.

But later, the thinking changed, and some observers came to believe that the sites were intentionally burned by the Sinagua before they left, as part of a so-called decommissioning ritual.

In order to bring a fresh perspective to the question of why these dwellings were deserted, Guebard and his colleagues revisited the notes of the original excavations.

They also studied artifacts and features from these sites, including ceramics, roof beams, and other architectural elements.

Plus, Guebard solicited input from several tribal groups whose ancestors lived in the Verde Valley.

“Generally speaking, we spend a lot of time and effort working closely with culturally associated Native American tribal governments,” Guebard said.

“They review many of our projects and provide important feedback that helps to develop project objectives.”

Using these various lines of evidence, an entirely new picture of the cliff dwellings’ final days started to come into focus, Guebard said.

“As I worked through field notes and photographs [from the 1930s excavations], it became increasingly clear to me that the fire at Castle A was probably the result of a violent event,” he said.

For example, he said, new dates from charred wall plaster inside the structure revealed that the building burned between the years 1375 and 1395.

Meanwhile, an analysis of the pottery found inside the dwelling showed that they were made in styles — like Jeddito Yellow and Jeddito White — that coincide with that same period. This suggested that Castle A was inhabited right up until the time it burned.

Perhaps more persuasively, the remains of four people had been excavated from Castle A in the 1930s. Initially, archaeologists thought that the dead had been buried there long before the fire struck.

But a closer examination of previous research done on those remains revealed that the dead had sustained trauma before their deaths, as evidenced by cut marks on their bones, burn marks, and fractures in three of the four skulls.

“We learned that the interior portion of each fracture displayed evidence of singeing on live bone,” Guebard said.

“So, the sequence of events seems to be blunt trauma followed closely by burning.

“It is also interesting to note that all of the remains with reported evidence of trauma and burning were found in a single grave.”

(Learn about new insights into violence in the pre-contact West: “Mass Grave Found in California Reveals Prehistoric Violence Against ‘Outsiders’“)

All told, the archaeological evidence seemed to indicate that the site was burned suddenly, and without warning, as part of a violent attack, Guebard said.

And, he added, this was corroborated by accounts given by several members of Native American clans, whose oral histories describe an attack on Castle A by rival bands that sought to drive the Sinagua out.

“Years before starting the project, I had heard from tribal members that there were stories about violence at the site,” Guebard said.

“So as I started to realize that the archaeological evidence was supporting violence at Castle A, I felt it would be helpful to get their unique perspectives on the project and on my interpretation of the data.”

A member of the Bearstrap Clan of the Hopi, for example, which traces its history back to the construction of the cliff dwellings, recounted tales of Sinagua villagers fleeing into Montezuma Castle and pulling up the ladders, isolating themselves inside while invaders set it on fire.

Meanwhile, modern members of the Dil zhe’e, or Tonto Apache, gave accounts of Apachean ancestors striking up an allegiance with the neighboring Yavapai, who together evicted the Sinagua from the cliff dwellings by “burning them out.”

“What was interesting is how close the oral histories match up with the archaeological evidence,” Guebard observed, “but also how each story provided a unique and different perspective of the event represented by the archaeological evidence.

montezuma-castle-arizona interior
The interior of Montezuma Castle still shows signs of burning, including on the wall plaster and on the primary and secondary wooden beams, to the right of the scale card. (Photo by Matthew Guebard. May not be used without permission.)

“This is surprising and doesn’t happen very often.”

Together, the archaeological evidence and oral histories paint a picture of diverse cultures in the Verde Valley in the 14th century — a scenario with significant implications for our understanding of what life in pre-contact Arizona was really like, Guebard noted.

For one thing, many anthropologists believe that the ancestors of the modern Apache and Yavapai didn’t migrate to the American Southwest until the 16th century.

In recent years, new evidence has begun to mount that suggests Apachean ancestors did indeed arrive as early as the 1300s, and Guebard noted that his research may add to it.

“I hope that the paper elicits some discussion about the possibility that ancestral Apache and Yavapai people were living in Central Arizona much earlier than archaeologists previously thought,” he said.

(See some of this recent evidence: “Long-Hidden Sites Discovered in the Southwest May Change Views of Ancient Migrations“)

Moreover, he added, his research into the fate of Montezuma Castle, Castle A, and their inhabitants shed new light on the value of oral histories in clarifying a historical record that’s often dismissed as being impenetrably “prehistoric.”

“Many archaeologists have been skeptical about the use of oral histories, and rightfully so,” Guebard said.

“It is often difficult for non-tribal researchers to incorporate information that may seem mythical or anecdotal in nature.

“In this case, the oral histories and the archaeological data fit together really well. So I think this study shows that oral histories can supplement information in the archaeological record.

“That doesn’t mean that every oral tradition will be useful for a researcher, but it does suggest that the collection of oral histories should become a standardized part of every archaeological investigation.

“In my experience, tribal representatives have insight that can’t be acquired through archaeological data collection alone.”

The effort to understand what happened to one of central Arizona’s most important landmarks is by no means over, Guebard added.

If violence did precede the evacuation of the Montezuma Castle cliff dwellings, more work still needs to be done in order to understand what happened next.

“One of the conclusions of the study is that violence caused the abandonment of the Montezuma Castle village, not necessarily the Verde Valley as a whole,” Guebard said.

“Future archaeologists working at other sites in the Verde Valley will need to contribute to that discussion.”

Guebard published his findings in the journal Kiva. Guebard, M. (2016). During the Migration Time: Oral History, Violence, and Identity in the Prehistoric Verde Valley KIVA, 82 (3), 259-277 DOI: 10.1080/00231940.2016.1208632

Leave a Reply to michael shea Cancel Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.


  1. Myra Larsen

    This very interesting article states there was violence and a fire. Four bodies show physical trauma and they were buried. Who had time to bury them in the midst of the attack? Later? Why? Did those who escaped into Montezuma Castle return to do that? Why would they only bury four bodies? Was any pottery found buried with them?

    What about those who were not buried and were just left where they fell? They are not mentioned in the article.

    It seems every answered question leads to more unanswered questions.


  2. michael shea

    burning was used as paint . the pictures of the inside, on the right side is a severed head of a animal (enemy?) in the corner facing left.

  3. McG

    Here are answers to some of Ms. Larsen’s questions using information already published for this project in a few journal articles.

    There were a total of 28 burials found at the site, but only four had evidence of violent trauma. These four remains were found intermixed within a single burial shaft. Because very little information was collected during excavation of the burial shaft (conducted in 1933-34), it is difficult to speculate about why or when the burials actually occurred. Similarly, pottery was found within the burial shaft, but it was not possible to determine if it was directly associated with these burials. The assumption made, however, is that human remains with evidence of trauma were buried shortly after the fire/attack.

    Archaeologists also discovered a skeleton beneath collapsed and burned roof material at Castle A. This suggests that at least one person was trapped within the dwelling during the fire.

  4. Rena Pikyavit

    Oral history has always been used in my culture…why wouldn’t be useful in understanding or explaining what happened at these sites….sometimes it’s frustrating for me because some of the archeologist forget we are still here…