Multiple Lines of Evidence: Searching for the Sand Creek Massacre Site

Detail from The Sand Creek Massacre, elk hide painting by Eagle Robe, Eugene J. Ridgely Sr., Northern Arapaho Tribe.

In 1864, the U. S. Army carried out a surprise attack on a village of about 600 Cheyenne and Arapaho Indians camped on the boundary of their reservation along the Big Sandy Creek in present day southeastern Colorado. The unprovoked attack on a non-combatant winter encampment took the lives of about 160 Cheyenne and Arapaho tribal members, two-thirds of whom were women, children, and others elderly or infirmed. Today, Sand Creek is among our nation’s most profound historic places. The events that took place here remain powerful and, to some, sacred symbols of sacrifice and struggle.

To preserve the memory and sacred space of this tragic event, the Sand Creek Massacre National Historic Site was established in 2007, the culmination of a decade-long process. U.S. Senator Ben Nighthorse Campbell initiated efforts in the late 1990s to protect the site as a national park unit. Before the national historic site could be established, however, a locational study was needed. The project, congressionally mandated by the Sand Creek Massacre National Historic Site Study Act of 1998, was a multi-disciplinary effort designed to identify the location of the attack. Because of the emotion associated with the site, there was potential for conflict in interpretation of the events of the massacre. Instead, in an unusual spirit of acknowledgement, the study accommodated multiple perspectives on the location of the massacre.

Tensions between Indian Tribes and European Americans in Colorado

In the mid-nineteenth century, European American settlers pushed through the Great Plains and into the Colorado Territory. As Colorado Territory grew, so to did the tension and fear between settlers and Indian tribes. The settlers’ desire for land and prosperity conflicted with the needs of Indian nations who lived, hunted, and traveled across the same areas. In 1861, a reservation was established that was insufficient in size and was but a fraction of what had been promised ten years earlier to the nomadic Cheyenne and Arapaho.

The seeds of conflict in 1864 began with the murder of Cheyenne Chief Lean Bear. Negotiations between leaders – governors, agents and officers on one-side, chiefs and headmen on the other, failed. Assurances of amnesty for “friendly” tribesmen were tempered by a proclamation to “kill and destroy” hostile Indians. The burning of several Cheyenne camps by Colorado cavalrymen near the South Platte, and claims that troops had initiated conflict near a place called Fremont’s Orchard foreshadowed war. A September sojourn by Chiefs to meet with Colorado Territory Governor John Evans and U.S. Army Colonel John Chivington accomplished little.

The Sand Creek Massacre

On November 24, 1864, a force of some 600 troops under the command of Colonel John Chivington, rode from Camp Fillmore towards Fort Lyon, some 120 miles down the Arkansas River. At Lyon, the expedition was joined by an additional battalion. An all-night ride on November 28 brought the troops to the edge of the Indian reservation. Here, at dawn, the troops got their first sight of the Arapaho and Cheyenne camps of Black Kettle, White Antelope, Left Hand, and other leaders. Within moments, a barrage of arms fire and howitzer bombardment was leveled against the tipis. Some people used driftwood and underbrush as defense, while others crawled into pits along the banks of Sand Creek. Big Head, Howling Wolf, Spotted Horse, and other small groups of warriors fought the troops in desperate struggles near the creek. Troops pursued Cheyenne and Arapaho people for several miles up Sand Creek. By afternoon, about 160 members of the tribes lay dead, the majority women and children. Before departing the soldiers ransacked and burned the village.

Dozens of wounded troopers and their dead comrades were taken to Fort Lyon for care and burial. Here, protestations against the attack were lodged by some officers and civilians – sparking bitter indignation that would continue for years. Following two congressional and one military investigation in 1865, the attack was quickly condemned as a massacre. John Chivington the U.S. Army Colonel who led the attack and John Evans, the territorial governor of Colorado who condoned it were found responsible. Evans was asked to resign, but soon won the territory’s senate seat. Chivington voluntarily resigned his commission and entered the freighting business. Not one officer, soldier, or government representative was ever indicted, tried, or sentenced for actions taken or associated with the massacre. “Since the day it happened, the Sand Creek Massacre has maintained its station as one of the most emotionally charged and controversial events in American history, a seemingly senseless frontier tragedy reflective of its time and place”(Greene and Scott 2004:4).

Relocating the massacre site

Despite sustained controversy for nearly a century and half, within a few decades of the massacre the specific site became obscured and, finally, unrecognizable. By the time U.S. Senator Ben Nighthorse Campbell initiated efforts to protect the site as a national park unit, the actual location was unknown. Congress required the NPS to collaborate with the Cheyenne and Arapaho Tribes and the State of Colorado to positively identify its location. In 1998 the NPS assembled a multi-disciplinary team to begin the site location process (see NPS 2000 and Greene and Scott 2004 for a complete discussion). The project included archeology; archival research; aerial photography analysis; geomorphology; oral histories among land owners, local collectors, and tribal descendents; and tribal traditional methods to locate the massacre site.

The study began with research in archival sources for information about the location and extent of the massacre site. An initial review of published literature and extensive archival research by historians and other knowledgeable people included military reports, diaries, eyewitness accounts, historical maps, aerial photos, and land records. Research concentrated on evidence related to the Big Sandy Creek and its tributaries, troop travel, trails, camps, and other evidence of the location of the attack.

Concurrent with the archival research, geomorphological analyses of changes in landform, analysis of aerial photographs, oral history documentation, and tribal traditional methods assessed all available evidence of geographic references to the extent and boundaries of the massacre. Evidence of specific locations within the massacre site boundaries also was sought. As results from these lines of inquiry were assembled into a scenario of the extent and boundaries of the massacre site and the locations of individual events within it, the team continuously evaluated the findings. At the completion of the archival research, the team suggested a scenario that the massacre site was five-and-a-half mile long by one-and-a-half mile wide. Within it, they postulated the locations of the Cheyenne/Arapaho encampments, sand pits which figured in the survival of some during the massacre, route of approach and subsequent positioning of military units, tribal horse herds, routes along which survivors fled, military bivouacs, spring used for drinking water, and other features. These proposed locations gave focus to the primary efforts of the archeology and geophysical remote-sensing crews.

Based on the preliminary site location scenario, the archeological investigations concentrated on the use of metal detectors as a systematic inventory tool. The inventory consisted of three phases – survey, recovery and recording.

... artifact finds were located and marked by either metal detecting the ground surface or by visual inventory. The survey team lined up and walked designated transects…until an area was completed. The recovery crew followed and carefully uncovered subsurface finds, leaving them in place. The recording team then plotted individual artifact locations using a total station transit and electronic field book to collect location information on each object or feature found, assigned field-specimen numbers, and collected the specimens. (Greene and Scott 2004:72)

Soon after beginning remote-sensing inventory activities, the team began to find significant 1860s-era artifact concentrations. The artifact locations, as related to the locations predicted from examination of the historical documentation, however, was not readily apparent. The archeological findings required the working hypotheses about event locations to be questioned and refined.

Based on questions posed by the new evidence, the archeologists turned their efforts to locating the village site. Shortly thereafter, the most compelling, publicly visible, and controversial archeological finding came as nearly 400 1860s-era artifacts, consisting of both military and domestic items, were uncovered. The team interpreted these artifacts as marking the location of the Cheyenne and Arapaho village. The artifact clusters were, however, approximately one-quarter of a mile downstream from the proposed location based on the historical record. With the identification of the village site in place, the locations of other elements of the attack and its aftermath could be extrapolated with some confidence. The archeological evidence confirmed, with relatively high confidence, the locations of individual features within the massacre site that were predicted by the archival sources.

Archeological fieldwork, May 1999. The archeological crew, including members of the Cheyenne and Arapaho Tribes, uncovered artifacts found through metal detection.

Photo by Arnie Thallheimer.

Multiple lines of evidence for determining site location

The site location project involved collaboration among the NPS, the Northern and Southern Cheyenne and Arapaho tribes, the Colorado Historical Society, landowners, and volunteers. While multiple stakeholders participated in every facet of the planning, field work, analysis, and reporting, their participation did not preclude diverse and divergent perspectives on the research methods, the veracity of various lines of evidence, and the conclusions generated. Public interest remained intense throughout the project, which maintained a widespread fascination in the process and reopened to public discourse the subject of the Sand Creek Massacre.

While all parties did agree on the exterior boundaries of the entire massacre site, determinations of the location of the village site – and by extension, the routes of military approach and the sand pits – remained at the core of individual stakeholders’ differing interpretations. To the NPS, archeological confirmation of historical and archival evidence was the cornerstone of the determination of the massacre site. For many tribal participants, however, archeological evidence was but one factor to consider among a suite of other methods and lines of evidence, including oral traditions and spiritual ways of knowing about locations of certain events. While not all representatives of the four tribes came to the same conclusion about the location of the village site, many concluded that it was about three-quarters of a mile downstream from the location determined by the NPS.

One fundamental basis for difference between the NPS conclusion about the village site location and that of some tribal representatives was the credibility each group gave to particular historical maps. For decades scholars of the Sand Creek Massacre site have utilized a series of maps created between 1904 and 1915 by George Bent, a half-Cheyenne survivor of the attack. Two of the maps show how groups of lodges and other features were positioned within a great bend in Big Sandy Creek. The Bent maps were considered by NPS team to be generally correct, but because they were not to scale, certain features could not be located fully within the large bend in the creek as indicated.

During the research for the site location study, a previously unknown military map, created by Lt. Samuel Bonsall in 1868, four years after the massacre, was discovered in the National Archives. On the basis of this map, the team’s historians projected the location of the village site downstream of the location according to the Bent map. The Bonsall map was considered highly reliable not only because it was recorded only four years after the event, but because of the standardized “strip map and journal” methodology used at the time to record “every point of practical importance” (Greene and Scott 2004:101; National Park Service 2000:43).

To many Cheyennes, however, lending more weight to the military map than the Bent map implied a disregard of Cheyenne ethics or “tribal law” regarding accuracy in memory and oral history, regardless of the number of years or generations across which information in transmitted. Therefore, many tribal representatives and a number of non-Indian scholars hold that the Bent map is more credible than the Bonsall map. Differing perspectives on the credibility of each map lead to different conclusions about the location of the village site and, by implication, routes of the military approach and position, as well as the survival pits.

A second major factor contributing to conclusions about the location of the village is a 1978 visit to the massacre site by the Southern Cheyenne Keeper of the Sacred Arrows, the tribe’s highest spiritual authority. Based on his ceremonial knowledge, the Keeper consecrated the ground within the bend in the creek as Cheyenne Earth, thereby indisputably identifying the spot for tribal participants that many descendants have always known to be the massacre site.

Sand Creek Massacre Survivors, painting based on a story of a woman and her granddaughter who survived the massacre and were guided by a wolf from their hiding place to the safety of a Cheyenne camp.

By Donald Hollowbreast.

Additional confirmation of the location of the massacre site, as depicted in the Bent map, for many tribal members comes from a spiritual presence, often experienced as the voices, footsteps, or transient embodiment of women and children. Sometimes the spiritual presence or proof of the efficacy of prayers is evidenced by birds, eagles, badgers, and other animals.

Ceremonial methods of identifying places where ancestors were killed, as well as praying for them and blessing the site, include Pipe Ceremonies, Cloth Ceremonies, food offerings and others. Prayer, above all, is the overriding determinant, not only for understanding the site in the past, but in caring for it and the spirits that reside there in the present and future. The use of archeology, in the context of spiritual ways of knowing, is secondary to traditional knowledge. As explained by Southern Cheyenne Chief Laird Cometsevah:

Cheyennes have a different view than other Indians or white men. White men call it a sixth sense, maybe the Cheyennes have an extra sense where they can feel or see spirits or areas where spirits are present. Sometimes they see their ancestors, in daylight or night; they have this extra gift that was given to the by the Almighty. Before going up to this area Joe Big Medicine, Luke Brady, and I had a spiritual feeling. So we put on our red paint and when we put it on that’s when they found the area with the 75 musket balls…they found them when we put on our paint. Also, east at the Kern’s property we did the same thing and that’s why they found some more. This is the Indian way of doing things, this is the traditional method. It still works and it still has its powers (National Park Service 2000:281).

This is not to say that archeological evidence is regarded as insignificant in a tribal perspective that lends more weight to ceremonial and oral historical evidence. To the contrary, the actual items carried by the people at Sand Creek and the bullets that ended their lives are regarded nearly as sacred. The archeological record is extremely important – important enough to be prayed for – but what it represents may have many alternative interpretations.

These interpretations of the events of the Sand Creek Massacre based on the combined historical, archeological, geophysical, and other lines of evidence were included in the study. Based on their conclusions, the team submitted to Congress its determinations regarding the location and extent of the Sand Creek Massacre site as required by the 1998 Site Location Study Act, leading to its establishment as a national historic site.

Ultimately, what mattered most in the locational study was consensus about the boundaries of the massacre site in its entirety among all the partners. What will matter most for the future is the site’s stewardship – both physical and spiritual – and its ability to educate, to provide solace and, perhaps, even to heal. Multiple interpretations of the archeological record enrich our understanding of the massacre site, rather than detract from it. In the end, the Site Location Project team acknowledged and valued the multiple perspectives on the Sand Creek landscape (National Park Service 2000:287-290), which will form the basis for interpretive programs that help to educate a global public about an American tragedy.

References

Greene, Jerome A. and Douglas D. Scott
2004 Finding Sand Creek: History Archeology, and the 1864 Massacre Site. University of Oklahoma Press, Norman, Oklahoma.

National Park Service
2000 Sand Creek Massacre Project. Volume One: Site Location Study. Intermountain Region, Denver, Colorado.

By Alexa Roberts, Ph.D.