A Misplaced Massacre: Struggling over the Memory of Sand Creek

by Ari Kelman

  Some of the descendants disagreed. From the first, several of the Cheyenne tribal representatives resented Limerick’s presence at the meeting. They wondered why a newcomer had been introduced into the memorialization process at the eleventh hour. Limerick, after all, had not earned their trust by working with them on commemorating the massacre. Quite the contrary, years earlier she had inadvertently run afoul of David Halaas, the historian who remained a close friend and advisor to the Cheyenne descendants. Upon hearing word in 2000 of the discovery of Silas Soule’s letters, which turned up in a Denver-area attic on the eve of Senator Campbell’s Capitol Hill hearings to determine whether the Sand Creek memorial should be established as a national historic site, Limerick initially had expressed ambivalence about the documents’ contents. Her muted reaction in the Denver papers had angered Halaas. Even after Limerick had clarified that a reporter had mischaracterized her position, and that she actually had not yet read the correspondence, some bad feelings lingered through the years. Beyond professional friction, though, the Cheyenne descendants were not overawed by Limerick’s scholarly credentials. Instead, they viewed her MacArthur “genius grant” and her international reputation as potential challenges to their cultural authority. As Otto Braided Hair explained, “We’ve got a good handle on our own history. We don’t need an outsider to tell us about Sand Creek.”12

  Looking back on that meeting, Alexa Roberts acknowledged: “We [the NPS personnel] made some mistakes.” The NPS had “scheduled that consultation for immediately after the burial of human remains, so everyone’s emotions were running really hot.” Not to mention that “Patty Limerick and Tom Thomas were new to the project and were relatively unknown quantities to the descendants.” Roberts had not yet fully explained to Limerick and Thomas “the relationship between the process of putting together the general management plan, including how we would interpret the massacre, and everything that had already happened with the memorialization of Sand Creek.” Roberts later recalled, “We eventually wanted to sit down and figure out with a group of people, including the tribal representatives and some independent scholars [in this case, historians unaffiliated with the NPS], to identify a strategy that would allow us to choose some documentary sources over other documentary sources.” She remained certain at the time that the tribal oral histories would occupy a central place in the site’s interpretation. The issue, then, was that hard experience had taught Roberts, “The public would want to know how we chose our sources.” She added, “The meeting with Patty Limerick was intended to lay the groundwork for dealing with that issue in the future. Unfortunately, we dropped the ball by not preparing the descendants for what we were trying to do.”13

  From Roberts’s perspective, adding new elements into the mix had destabilized the memorialization process’s volatile chemistry. As in the past, tension emerged from a struggle to control the production and distribution of history and memory. The tribal representatives had fought for years for some say in how the NPS would depict the violence, and they were unwilling to let an interloper become an arbiter of ongoing disputes. They rejected having Limerick, in Roberts’s words, “choose between conflicting documentary sources and assess the validity of one perspective versus another perspective.” Given the long struggles over incommensurable Sand Creek narratives, Otto Braided Hair believed that his people should select from among competing accounts. As he said, the descendants understood their past. They understood, too, that whoever crafted the massacre story related at the Sand Creek site would determine who would be cast as an insider or an outsider, a perpetrator or a victim. For a century and a half after Sand Creek, Arapahos and Cheyennes typically were written out of American history, more often forgotten than remembered. The Sand Creek memorial would provide them with a platform from which they could tell their stories at a national historic site—stories of tragedy and betrayal, of loss and heartbreak, but also of survival and persistence.14

  After protesting Limerick’s involvement, the descendants began questioning exactly how the GMP would be crafted. Did Senator Campbell’s Sand Creek legislation not make clear the tribal representatives’ prerogatives in the process? Had they not secured for themselves, through a series of negotiations during the site search, a seat at the planning table, including the right to shape how the massacre would be interpreted for the visiting public? The sense of distrust in the conference room ran so deep that Norma Gorneau, one of the Northern Cheyenne Sand Creek delegates and a longtime friend of Alexa Roberts, wondered aloud if the federal government had betrayed the Cheyennes, if the NPS would now insist on recapitulating the fight over nomenclature. Would the bloodshed, Gorneau asked, be presented as a battle rather than a massacre? At its end, the meeting produced more questions than answers. The NPS personnel asked themselves how they could interpret such a politically fraught episode as Sand Creek and if they had irreparably damaged their relationship with the descendants. Some of the descendants, in turn, asked themselves if the NPS had been acting in bad faith all along and if they had mistakenly placed a site sacred to their people in the hands of bureaucrats committed to “protecting their jobs instead of the Cheyennes’ and Arapahos’ heritage.”15

  Following the consultation at the Cow Palace, NPS personnel, including Alexa Roberts and Tom Thomas, decided to table questions of interpretation temporarily as they continued working on the GMP. Roberts remembered, “After that meeting we reluctantly put interpretation on the back burner.” The NPS representatives believed at the time that they did not have any option; they had to rebuild trust with the descendants, again proving their good intentions before broaching discussions of how to present the massacre’s history to the public visiting the memorial. The problem, though, was that the NPS relied on what Steve Brady had years earlier dubbed its “sacred playbook,” a series of guidelines for systematizing agency tasks, including the planning of new units of the National Park System. Roberts admitted, “[This] was not how we did things in the NPS. And so we found ourselves having to explain, internally, why we didn’t have long-range interpretative planning as part of the GMP process. It was sort of uncomfortable in that moment but absolutely necessary over the long term, Tom [Thomas] and I believed.” Casting about for subjects that they hoped would not generate more ill will, Roberts and Thomas made what seemed, in light of the history of the site search, a surprising choice. “We turned to mapping,” Roberts recalled, “which, ironically, brought us back to questions of interpretation.”16

  The descendants remained wary of the NPS and stepped up their efforts to shape how the massacre would be represented at the site, and how, by extension, it might be remembered collectively by visitors to the memorial. On December 7, 2010, Otto Braided Hair, Joe Big Medicine, and David Halaas arrived at the NPS’s Denver offices, where they met with Alexa Roberts and Tom Thomas, as well as other NPS staffers, including representatives from the organization’s Geographic Information Systems division. The group planned to talk about an upcoming cartography meeting. Roberts remembered, “It was just supposed to be a technical, nuts-and-bolts discussion.” Trying to avoid a repeat of the debacle following the burial ceremony, she hoped to “lay some groundwork for a future mapping workshop we had in mind.” Around the noon hour, Roberts introduced Jeff Campbell, a volunteer at the Sand Creek site. Campbell, Roberts explained, had done some cartographic research on his own time, and had generated a hypothesis that might allow competing constituencies to compromise, perhaps finally putting to rest the village controversy. Aware that all eyes in the room were on him, Campbell “worried that [he] was going to be on the stand, facing a hostile cross-examination.” He recalled, “But that wasn’t what happened.” Instead, as he spun out his theory, rooted in an array of primary sources, the response was positive.17

  Jeff Campbell looked a bit like the Marlboro Man—not the Marlboro Man of the advertisements that last ran in the United States in the 1980s and 1990s, but how that model might have appeared in 2010 (assuming that lung cancer had not
long since caught up with him). Campbell often wore a white cowboy hat, a Western shirt tucked into Wrangler jeans, work boots that had seen their share of dusty miles, and a belt featuring a buckle whose size typically fell, on a scale that only ranch hands and Brooklyn hipsters seemed to understand, somewhere between dressy and ironic. He had a full head of graying hair and a bushy mustache. Raised in Phoenix, Arizona, and El Paso, Texas, in a home where history was central—“I was just nursed on the stuff. Growing up, most of our family vacations, usually camping trips, were to historic sites.”—Campbell went on to major in the subject, with an emphasis on the American West, at the University of Texas at El Paso. After graduating, he moved to Albuquerque, New Mexico, where he worked for decades as a criminal investigator, eventually serving as a special agent in the state attorney general’s office. He focused on cases that had gone cold and grew expert at pulling together intelligence sources, which, he noted, meant, “working the archives.” Campbell brought his research skills with him after he retired from law enforcement, moved to Eads, and turned his attention to Sand Creek.18

  Starting in 2002, Campbell began studying the massacre, poring first over the NPS’s Site Location Study. The hardened cop in him worried that “there was too much certainty” in the document. “As a criminal investigator” he was “always suspicious of that kind of thing, when an event is so terribly complicated and yet the people recounting it are so completely confident about their stories.” He “decided to launch [his] own investigation” into the violence. Campbell learned quickly from the secondary literature that “none of the major authors agreed.” Then, like so many others before him “hooked” by the mysteries of Sand Creek, he dove into the primary sources. As he got deeper and deeper into the archives, he eventually produced two book-length document readers of his own, one devoted to the history of the 1st Colorado Regiment, the other to the 3rd Colorado Volunteers. At the time, the NPS wanted to acquire Bill Dawson’s property, and the village controversy burned at its brightest. Campbell, consequently, “realized that there was this whole other issue: the site.” As he remembered it, “from where I sat, the site was a crime scene and had to be investigated like that.”19

  Using methods that he had mastered while still working at the New Mexico attorney general’s office, Campbell tacked a large sheet of butcher’s paper to the wall of his home in Eads and began sketching out the massacre, creating what amounted to a chronological map, starting with the moment Chivington’s men departed from Fort Lyon, through the slaughter itself, and concluding with the Colorado volunteers bivouacked near the killing field. Campbell planned to “diagram every prominent feature of Sand Creek” and then, as he would have done with a cold case, “line up that drawing with witness statements”—in this instance the primary accounts, drawn from the federal inquiries, and also the various tribal oral histories collected through the years—of the bloodshed, “to determine if the witnesses could have seen what they said they had seen.” As Campbell continued sifting through testimony, “sorting the wheat from the chaff,” he realized that several eyewitnesses suggested that at the time of the attack, “the stream channel wasn’t running at a 90-degree angle. It was more like a 120-degree angle. And then, George Bent’s letters also said that the bend was about 120 degrees, rather than 90 degrees, which, of course, is why Laird Cometsevah thought that.” Campbell finally suspected that the so-called Big Bend of Sand Creek, known during the site search as the Dawson South Bend or the traditional site, had been shaped differently at the time of the massacre.20

  In 2008, Campbell completed most of his research and arrived at a working theory. Early in the twentieth century, he learned, the Chivington Canal Company had constructed a large irrigation ditch running between Sand Creek and the nearby town of Brandon, Colorado. When the creek ran high during the wetter winter and spring months, impoundment reservoirs would gather its floodwaters, where they would wait to be used by local ranchers during the dry summer and fall seasons. That canal, Campbell believed, had eventually “shifted the stream’s course,” moving its bed approximately a mile southward over time, toward the base of what would become the monument overlook, and causing the Dawson South Bend to assume a 90-degree angle.21

  As Jeff Campbell related his hypothesis at the consultation meeting on December 7, 2010, the NPS personnel and the descendants understood that if fieldwork confirmed Campbell’s suppositions, many of the contentious issues plaguing the memorialization process would disappear. Doug Scott and Jerry Greene, who had argued for years that the concentration of artifacts unearthed in 1999 marked the location of Black Kettle’s village, could be right. So, too, could Laird Cometsevah, who had insisted that George Bent had accurately mapped the massacre, that the Big Bend had embraced Black Kettle’s people’s lodges. If the stream had moved since the era of the bloodshed, there might be no discrepancy between those intractable positions or others: between Samuel Bonsall’s depiction of the massacre site and Bent’s, between firsthand accounts offered by perpetrators and oral histories collected from survivors, between social-scientific methods that had yielded a bounty of relics during the archeological component of the search and traditional Native ways of understanding the past. The next day, the descendants traveled to the memorial site with Campbell, Alexa Roberts, Tom Thomas, and several other NPS officials. They did not have time to survey the area under suspicion, but Otto Braided Hair and Joe Big Medicine remarked, taking Campbell’s ideas into account, “everything made so much more sense.” The group agreed to return to the site as soon as possible.22

  Jeff C. Campbell’s theory of the projected course of Sand Creek in 1864. (Adapted from a map provided courtesy of Jeff C. Campbell.)

  Five months later, on May 17, 2011, nearly two decades after fortune hunters had touched off the modern memorialization effort by informing David Halaas that the Sand Creek site had gone missing, and almost a century and a half after the massacre itself, Northern Arapaho descendants and the NPS personnel charged with producing the GMP met in the historic site’s conference center, the renovated shell of what had been Bill Dawson’s ranch maintenance facility. Alexa Roberts remembered that after Gail Ridgely offered a prayer, “everyone gave opening statements about what had happened to that point in our efforts, focusing on the discrepancies between Bent’s maps and the artifact find, and how we couldn’t do anything about interpretation because of those differences.” Jeff Campbell next recounted his theory, explaining the potential impact of the Chivington Canal. The group then walked the site, talking about how “human activities had modified the landscape over time and conceivably had resulted in a change in the course of the stream channel through the years.” The day, Roberts noted, “turned into a fabulous concurrence of different ideas. It turned out that everyone, people who had been at odds for years or even longer than that, had all been right in some important ways.”23

  Over the following two days, the Cheyenne descendants arrived at the site. On May 19, Jeff Campbell stood atop a ridge, upstream from the monument overlook, and explained his research to a rapt audience. David Halaas was delighted: “If the Big Bend of Sand Creek was crescent shaped back then, not the L shape of the Dawson South Bend today, that makes George Bent’s map right! There’s no more basis for dismissing it!” Gary Roberts then tied up another loose thread. Roberts, who had studied the massacre for four decades, referred to the rite performed in 1978 by the Sacred Arrow Keeper and Laird Cometsevah: “If Jeff [Campbell] is right, they made Cheyenne earth in the correct place, because the spot where the monument is now is the spot where the troops began their attack. The people in the village would have first seen the troops there.” In other words, “the making of Cheyenne earth at that spot,” which, along with the Bent map, had always informed Cometsevah’s understanding of the massacre, “made perfect sense.” Finally, late in the afternoon, a hydrologist and geomorphologist allowed that Campbell’s theories did not necessarily contradict the surveys completed during the site search. Earlier geomorpho
logical research had never concluded that the stream had been immobile through the years, only that it had not moved outside the confines of its expansive flood plain. Campbell’s theories fit with these assertions and could, the two scientists noted, be tested. But confirmation would have to wait for another day.24

  In the meantime, the NPS personnel celebrated, albeit, because of past history, tentatively. Alexa Roberts said of Jeff Campbell’s methods and background, “This was the biggest mass murder he ever worked.” For his part, Campbell seemed surprised by the reception that his theories had received. “All I’ve done for ten years is put stickers on maps,” he said, “letting the evidence guide me. I never thought my research would bring everyone together onto the same page.” Still, he counseled caution: “I’d never say I’m 100 percent sure about this. More than 140 years have passed since the massacre, remember.”25

  Campbell’s careful perspective was born of a career spent studying violent encounters, an appreciation for the uncertain nature of memory, and an understanding of the massacre’s history and the modern effort to commemorate Sand Creek. Years of working with incompatible testimony, recounting crime scenes awash in blood, had taught him to proceed deliberately before reaching conclusions. As for the massacre, Campbell suggested, based on the archival sources and tribal oral histories, that it might be possible to know what had happened on the banks of Sand Creek. But even then, he admitted to lingering doubts: “I think I know what I know. But what I know is still pretty limited.” What Campbell meant is that he had only tried to determine some relatively basic information about the massacre, including the site where it took place. Put another way, he was reasonably confident about the question of where, but deeper analytical queries—How could such a thing have happened? Who should be held accountable? What were the massacre’s lasting implications?—were still, he acknowledged, “a matter of interpretation.” And of course there was the rub.26