2
IN THE FALL OF 2017, a friend and I visited an exhibit at the Renwick Gallery in Washington, DC, that seemed to capture a certain popular cultural fascination with forensic science.1 A few blocks from the White House, the Renwick doesn’t attract the same crowds as the museums that line the National Mall—the Air and Space Museum, the National Gallery, or the recently opened National Museum of African American History and Culture, for example. The day was cold and rainy, the kind of weather that persuades people to stay put, and so I was surprised to find the first room of the exhibit packed with visitors. Peering over one another’s shoulders and angling for better views of the individual displays, many using small flashlights to aid in their sleuthing, they were enthralled. They were investigating. They were playing forensic detective for the day.
The exhibit, Murder Is Her Hobby: Frances Glessner Lee and The Nutshell Studies of Unexplained Death, paid homage to Lee, known as the “mother of forensic science” and the country’s first female police captain, by showcasing one of her most important contributions to the field: eighteen dollhouse-sized dioramas depicting various crime scenes that Lee used while serving as an instructor at the Department of Legal Medicine at Harvard University to train budding forensic investigators.2 In the “kitchen,” the figure of an apron-clad woman lies prostrate on the floor near the oven, a tiny Bundt cake resting on its open door. An old woman hangs from the rafters in the “attic,” with papers strewn about and an upturned cane back chair at her feet. In the “three-room dwelling,” a man lies face-down on a bedroom floor, his pajamas stained with blood, while his wife lies in the adjacent miniature bed, her head resting on a blood-soaked pillow.
Moving from one tiny “crime scene” to the next in the darkened exhibit hall, visitors speculated aloud—sometimes to themselves and with a striking level of assurance—about what clues they detected and what crimes of passion or negligence or avarice they thus deduced. Rather than focusing on the intricate handiwork, indeed artistry, of Lee’s tableaus or the paradoxical rendering of such gruesome details through the classic medium of children’s make-believe worlds, most people seemed to revel in the chance to discern the truth and solve the crimes. If only for a couple of hours on a rainy Sunday afternoon, they delved into the world of forensic science, albeit a fabricated and miniaturized one.
Writing about forensic fiction, particularly novels like those of Kathy Reichs, on which the television series Bones is based, anthropologist Zoë Crossland notes how “forensic work is commonly explained as a form of intellectual puzzle.”3 Assembling the facts and interpreting the signs, including those gleaned from the corpse itself, allow the scientist as puzzle solver to piece together the story, whether of violent death, guilt or innocence, or the identity of a nameless victim.
In some important ways, what the US military attempts through its MIA accounting resembles this crime-scene investigation puzzle-solving endeavor. They too seek to tell a story through the assemblage of forensic facts. They too bring to bear an array of investigatory tools to solve the puzzle. But unlike the protagonists of fictional TV series or forensic novels or even Frances Glessner Lee’s trainees, the forensic scientists tasked with MIA accounting deal with vastly different sets of scales—of time, space, and, most significantly, numbers. Their scenes of investigation and recovery are sites of long-past wars. Their examinations focus less on the how or the when than on the where and, most pressingly, the who.
Then there are the numbers: some eighty-two thousand unaccounted-for US service members from the major conflicts of the past century (World War II, the Korean War, the Cold War, and the Vietnam War), at least half of which are cases requiring detailed archival investigation and scientific analysis—tens of thousands of “nutshell studies,” enough to fill the Renwick Gallery exhibit halls at least two thousand times over, each dependent on multiple lines of evidence to discern the individual identity of a set or fragment of remains.
Finally, there is the simple fact that the forensic work of recovering and identifying remains is not just about investigation or resolution. It is about consecration. In restoring individual identity to a set of remains of an unaccounted-for service member, the process recognizes an individual life sacrificed to the nation. It is an act of setting apart the sacred.
I hadn’t fully understood the scale or weight of that endeavor until I spent time in the forensic laboratory on Joint Base Pearl Harbor-Hickam. It was there that I learned of the scientific, archival, and operational facets of the accounting mission, both in the past and the present—how it started, how it changed over time, and what obstacles it continued to face. And it was there, through observing the fine grain of the forensic work of MIA accounting, that I came to appreciate how the efforts themselves to recover and identify remains form part of the rituals of remembrance and mourning attached to this country’s missing war dead.
From Mortuary to Laboratory
The laboratory where I conducted my research in 2011 and 2012 and visited again in 2014 was not the same facility that currently serves as the hub for the forensic scientific work of the Defense POW / MIA Accounting Agency (DPAA).4 That lab is now housed in a brand-new three-story building on the other side of the Joint Base Pearl Harbor-Hickam. It’s an impressive structure, but in some ways its newness obscures the history behind the mission—namely its evolution from an endeavor defined primarily by mortuary practice to one dominated by forensic science. The evolution has had its growing pains, some of them with lasting effect.
It took time for me to learn this history, and in many ways, I first perceived it through the physical spaces of the laboratory, the Joint POW / MIA Accounting Command, the joint base, and the location itself—on the island of Oahu. The impression that history—and geography—matter came across right away through the setting of the two bases, Pearl Harbor Naval Station and Hickam Air Force Base, which were combined into the joint base in 2010. Driving through the art deco main gates of the Hickam entrance and along the winding roads fringed by palm trees and lush vegetation, past the roundabouts with fighter planes from various conflicts, past the buildings that still bear the scars of the December 7, 1941, aerial attack by the Imperial Japanese aircraft, I felt at times as if I were wending my way through a movie set. In fact, the two historic bases have lent their scenes to several Hollywood productions over the years—to films such as Aloha and Battleship and episodes of Hawaii Five-O (though, interestingly, not the 2001 blockbuster Pearl Harbor). Against that backdrop, the old lab was located in an unassuming, cream-colored, single-story building, its entrance announced by, unsurprisingly, the black-and-white POW / MIA flag. In the years since the 1976 transfer of the US Army’s mortuary facility for the conflict in Southeast Asia from its temporary location in Thailand to Hawaii, a handful of brown trailers were gradually added as the original facility expanded from an initial staff of twenty-five to over four hundred personnel.5
The expansion reflected in part the geographic concentration of the previous century’s absent war dead: of the more than eighty-two thousand unaccounted-for service members, 75 percent of the losses are located in the Asia-Pacific region.6 Thus, Oahu, Hawaii, served as a convenient midpoint for remains retrieved from Southeast Asia, the South Pacific, and the Korean Peninsula. Furthermore, buried on the island in the National Memorial Cemetery of the Pacific, known as the Punchbowl, were also hundreds of Korean War unknowns and the unnamed dead from the Pearl Harbor bombings, whose remains would become critical to the expanded mandate and a congressional push for increased annual identifications beginning in 2010.7
Before setting foot in the CIL, I got a taste for the current organizational logic of the MIA accounting mission’s scientific branch: standard operating procedure, or SOP. An SOP is a lab’s playbook and regulations—it lays out the specific protocols for every element of the scientific work. For as much as they seem fixed in form, SOPs index changes—modifications, refinements, additions—that are integrated as knowledge evolves and practice improves. It’s a “living document,” one of the lab managers explained. “It constantly grows.” To my uninitiated eye, the lab’s SOPs were awash in detail. A few weeks before I flew out to Hawaii, a package arrived in the mail, a hard copy of certain sections of the various SOPs, which I was to review and be tested on before I could observe the laboratory. In addition to providing a DNA sample and my fingerprints as part of my induction into the world of the US military’s forensic science, I also had to demonstrate my mastery of the fundamental rules for safeguarding evidence, accessing examination space, and keeping information secure.
Though an obvious artifact of any scientific laboratory, the SOPs of the Central Identification Laboratory had their own telling history. Like the Joint POW / MIA Accounting Command’s (JPAC) evolving physical space, they mapped the transition from mortuary practice to a more forensic, scientifically driven mission. Created in the mid-2000s, they also flagged an initiative by the lab to maximize transparency and ensure the scientific integrity of the identifications it produced.8 The subtext here was that more than just keeping apace with standardized practice in the field, the SOPs sought to prevent problems of the past—past mistakes, flawed approaches, and strained relations between civilian scientists and military personnel that had characterized the early years of the Vietnam War–era MIA accounting. Underwriting these aims of transparency and integrity was the less tangible but equally vital issue of trust. People both inside and outside the laboratory needed to have faith in the facts being produced; they needed to see the basic processes of recovery and identification as trustworthy.9
Early into my research at the lab, I learned that there was a time when the forensic work of MIA accounting had been publicly called into question. The details of the episode were recounted to me by five leading figures in forensic anthropology and odontology (dentistry), who as academics and / or practitioners had served as external consultants, that is, peer reviewers for the CIL. Two of them had been working with the lab since the 1980s; three had provided the external reviews for the identification of 1st Lt. Michael J. Blassie, the former Unknown Soldier of the Vietnam War. My first month of research at JPAC coincided with their annual visit to the laboratory, and I was able to meet with them as a group. As we sat in the lab’s library, surrounded by hundreds of identification packets (closed cases), their plastic-comb-bound volumes lining the shelves, the overwhelming majority of which these men had personally peer reviewed, they chronicled the changes they had witnessed over the past two decades. They started with what brought the lab to their attention and to that of their colleagues in the wider academic and professional community: mistakes. At that time, the mid-1980s, the scientific director—in fact, the lone “scientist” at the lab—was Tadao Furue, who held a bachelor’s degree in science from the University of Tokyo but no advanced degree in a relevant field of forensic science.10 He had, nevertheless, worked for the US Army in Japan from 1951 to 1977 processing Korean War remains (those unidentified during the policy of concurrent return) and later remains from the war in Southeast Asia. In 1977, he joined the laboratory in Hawaii.11 To be fair, there was little appetite within the Department of Defense in those early years to throw resources at developing forensic expertise, let alone hiring personnel with proper scientific qualifications. Again, it was still a mission largely governed by military mortuary practice—recovering remains and trying to make identifications based on archival, circumstantial, dental, and limited forensic anthropological evidence.
By the early 1980s, rumors of unsupportable identifications had begun to surface—around the same time that the remains of the Vietnam War Unknown Soldier were selected for interment at the Tomb of the Unknowns at Arlington National Cemetery—culminating in a lawsuit filed by Anne Hart, wife of an MIA. Hart rejected the US military’s identification of her husband’s remains and demanded that they be examined by an external scientist.12 Michael Charney, who was the director of the Center for Human Identification at Colorado State University, undertook the review of the remains, a mere seven fragments, none more than six inches in length, and denounced the military’s identification as “incompetence of the worst sort.” He explained that “the fragments were so minute, there was no way they could be identified as Lt. Col. Hart. The things Furue claimed to detect from the bones—age, sex, race—were just not possible.”13
By that point, the external consultants explained to me, Congress was paying attention. In December 1985, three independent forensic scientists, leading figures in their respective fields—William Maples, Ellis Kerley, and Lowell Levine, two anthropologists and one odontologist—were sent to evaluate the lab. (Levine, the odontologist, was one of the five consultants filling me in on the lab’s early history.) Tasked with “conduct[ing] an in-depth review and analysis of identification procedures and associated documentation used in the [lab],” the independent reviewers found a pattern of errors with data collection and analysis.14 They reported their findings to Congress and made twenty-three recommendations, including that “positive identifications” be made by direct fingerprints or X-ray comparisons; “unidentified commingled, fragmented remains should be presented as such and recommended for mass burial”; a “nationally or internationally known anthropologist” become lab director; the Armed Services Graves Registration Office’s review board include two forensic anthropologists, two odontologists, a lawyer, and forensic pathologists; and the lab’s scientists be encouraged and financially supported to attend scientific meetings on the mainland.15 The recommendations underscored the need to address the scientific integrity of the MIA accounting process by ensuring good science carried out by qualified scientists—from the top down, personnel not only trained in the relevant fields but knowledgeable about evolving practice. In short, they needed to reestablish trust in the government’s capacity to care for its fallen service members. The US Army accepted all twenty-three recommendations, and the evaluation proved a watershed moment. By 1987, Ellis Kerley—one of the anthropologists—became the lab’s scientific director, whose “standing in the field helped the lab gain needed credibility in the scientific community,” members of which he brought out to Hawaii as visiting scientists and external consultants.16
A Civilian-Military Gap
Kerley’s directorship also heralded an important shift in how the scientific component of the MIA accounting mission was conceived and structured. Gradually, the CIL became more of a civilian rather than military-led operation. As Lowell Levine, a veteran himself, explained in less diplomatic language, “You had to divorce the science from the tree suits, the mean green machine [the army],” to ensure the integrity of the forensic efforts—that is, CILHI needed to demonstrate that the science wasn’t beholden to extrinsic interests. Though change was slow and resources lacking (despite the 1985 recommendations, the lab still had only one computer in 1990, available to staff for thirty-minute slots), the consultants noticed a subtle transformation in the language of the reports—from the formerly definitive statements that nevertheless lacked supporting evidence to more cautious scientific analysis that acknowledged the limits of the available data.
But getting good scientists, especially well-trained forensic anthropologists, not only to move to Hawaii, so distant from the mainland, but also to work for the military, wasn’t an easy task. In the wake of the Vietnam War, anthropologists, including forensic anthropologists, were skeptical of the US government, and especially the military. As Tom Holland explained, “the relationship between universities and the military was so tainted by the Vietnam War. Coming out of WWII is the exact opposite. The scientists were seen as the saviors. The scientists were the ones that gave us the atom bomb that ended the war. But that was a whole different world.”17 Four years after assuming the position of the lab’s scientific director, Ellis Kerley, then president of the American Academy of Forensic Sciences, was fired by Lieutenant Colonel Johnie Webb, deputy to the commander at CILHI. Kerley’s successor, Kimberly Schneider, lasted even less time. Holland recalled arriving at the island and his post as one of three newly hired forensic anthropologists only to find its management in turmoil. “About a month after I was here, she [Schneider] had a blow up with the Command concerning what she felt were intrusions into scientific integrity brought about by the Joint Task Force-Full Accounting, JTF-FA. At that point, JTF-FA was run by a two-star general by the name of Thomas Needham, whose nickname was Nuke ‘em Needham. He was a real piece of work and didn’t have much regard for scientists. So I think Kim probably had a very legitimate concern.”18 Holland took over the permanent position of scientific director in 1994. Despite Kerley’s efforts to reform the laboratory, the military emphasis on expediency still extended to field operations. Hired the same year as Holland, anthropologist Bob Mann spent much of his first few years deployed to Southeast Asia, where, under General Needham, the military’s “gauge of success was the number of investigations that were done, the number of sites that were excavated and closed. Emphasis on closed.” The civilian anthropologists were only advisors on excavation sites, and the attitude was “get in there and get it done.”19
To protect the integrity of both the archaeological work of recovery and anthropological work of identification, Holland sought an external guarantor for quality assurance and transparency. He achieved this by getting the lab accredited. Accreditation is used to verify a laboratory’s quality management system and assure that it can perform certain test methods properly.20 “Given what they do, with life and liberty at stake,” he explained, “forensic laboratories probably more than any other laboratory need maximum transparency.” In December 2000, the lab’s management began their campaign to make the CIL the first forensic anthropology laboratory to be accredited by the American Society of Crime Laboratory Directors / Laboratory Accreditation Board (ASCLD / LAB). They petitioned for the accreditation not on the basis of human identification, but rather on trace evidence.21 Holland used the analogy of a car crash to convince the ASCLD / LAB reviewers of the parallel. Reassembling and analyzing the shards of a broken headlight from a car crash was like reassembling and analyzing a human skeleton that had been recovered at an air crash site, he argued. It worked. In 2003, they secured the accreditation, and in 2008, CIL succeeded in becoming the second federal laboratory to pass the ASCLD / LAB-International program (based on the International Organization for Standardization’s general requirements for the competence of testing and calibration laboratories, or ISO 17025).
To a layperson, the lab’s accreditation might seem like a lot of extra bureaucracy—from quality assurance to peer reviews and SOP revisions. But it did critical work. The achievement raised the lab’s profile both nationally and internationally—CIL, its facilities and scientists, soon led rather than trailed the field in postmortem identification, especially in the context of mass fatalities—and, equally significant, the move effectively placed a buffer between the scientific wing and the military operations within the Command (Levine’s “mean green machine”) and the larger accounting mission. In that sense, the forensic efforts of MIA identification had a new master to serve—namely, the standardized protocols demanded by an external scientific accreditation agency, not military leaders. In sum, accreditation raised the bar: it required the consistent execution of exacting science, an element integral to the ethos of exceptional care.
Although accreditation increased transparency and ensured scientific integrity, at times it compounded tensions between civilian scientific and military expertise within the Command.22 Lab staff often squared off with military leadership over questions of authority—who knew better, whose purview it was, who was the expert in the room. Part of the problem was that résumés didn’t translate. The military personnel in the Command had their achievements and experience fixed to their chests: bars, pins, and medals. Intelligible to one another, such records of service made sense in what one of the lab’s archaeologists described as the “last true rank society,” where individuals are ranked—that is, have more social status or prestige—according to their genealogical distance from the chief. But the scientists at the lab could only gesture to degrees. “Scientific smarts don’t count,” remarked one anthropologist. “The military folks can’t evaluate how well you know your science.” Whether a scientist had earned her PhD at a nationally top-ranked forensic anthropology program or was considered a leading figure in the field didn’t register because—like the bars and pins—it only fit within its own value regime, recognized by other anthropologists and archaeologists. Further compounding that disconnect was the fact that the military staff typically rotated every two to three years through the official process known as permanent change of station. Such constant hitting of the reset button undermined respect for corporate knowledge, including the specialized knowledge of scientists—anthropologists and archaeologists—deployed on recovery missions.
While there were scientists at JPAC who had previously served in the military or came from military families, they were still seen as personnel from the civilian-dominated laboratory. Archaeologist Bill Belcher’s father served his entire career in the US Army. He fought at the tail end of World War II, throughout the Korean War, and completed four year-long tours in the Vietnam War. “I grew up with the sensation of knowing—and my mom would talk about—[how] he might not come back. He might get killed.” He remembered “her fear of a car coming into the driveway and a specific kind of car—a government car—that was black, a black sedan, because she said that she knew at that time, her husband, my father, would be either dead or missing in action.” He brought that sensibility and respect for the military to the mission. But sometimes it didn’t feel reciprocated. “It boils down to they want us to respect them and their position and experience, which I think is great. And I think we should. But they don’t want to do it back. They don’t want to respect our education and our experiences outside of here.… It takes a lot of work to get a PhD.”23
Echoing this impression, but from the other side, a former JPAC military staff member explained what he saw among his colleagues, many of whom had deployed to Iraq or Afghanistan (himself included):
They get that mentality of working up to being a first class or a first sergeant or whatever. They don’t want to take orders from civilians because, I think, you know, coming into the military out of high school, you get all this praise for being in the military or working hard.… I remember from basic training that civilians—it was drilled into our heads—that civilians, I don’t want to say, [were] a piece of shit, but it was basically that they didn’t do anything when they got out of high school. They’re just at home worthless doing nothing. You’re going to go back and they’re still going to be doing the same thing but look what you’ve done. You’ve sacrificed yourself, your family, and time.24
For many members of the military at the Command, a JPAC deployment was a way-stop along a career dedicated to either training for war or going to war, and while they wholeheartedly supported the goals of MIA accounting, working alongside civilians who did not fall within their same chain of command could be challenging, especially during recovery missions. No matter how exacting, no SOP could manage expectations about clashing authority when the metrics of expertise were so divergent.
The Archaeology and Anthropology of MIA Accounting
When Lance Corporal Merlin Raye Allen returned to his hometown of Bayfield, Wisconsin, few if any of those who lined the streets to welcome him home might have imagined what little of his physical body remained—a single tooth. Nor could many have guessed the painstaking efforts to locate the proper site to excavate—where to dig for his remains—and once recovered, how to identify them as definitively belonging to him. Yet those underlying, undetected aspects were among the most important insights I gained from my time at the lab: that the processes of recovery and identification are complex and each case is unique, with its own set of complications and contingencies, and that definitive, scientifically sound identifications depend on the convergence of multiple lines of evidence, which may take years to compile. In short, there’s a lot more to the forensic science of accounting than meets the eye.
I would also add that the remains of war dead from long-past conflicts aren’t what you might picture. In 2011 and 2012, the bones laid out on the examination tables at CIL were rarely complete—a far cry from the articulated skeleton that hangs in an anatomy classroom or as a prop on the set of a forensic detective show. Much more often fragmented and partial, they defied assumptions about what lies within the flag-draped coffins photographed coming home to Dover Air Force Base or interred at Arlington National Cemetery. Instead, these heterogeneous collections of skeletal remains were all that was left, or all that could be recovered, of young lives cut short, of once vital bodies, of human beings sent off to fight for their country. In looking out at those tables, you see war’s destruction and a limited, if rigorous, attempt to answer its violence.
The science of identifying human remains starts with accession, the process by which remains are formally received at the laboratory and entered into its system of analysis. But to be accessioned, remains have to come from some place. They must first be recovered, which means their original loss must first be researched. In most instances, years of archival analysis, intelligence gathering, and field investigations drive recovery missions. “Historical investigation,” DPAA historian Michael Dolski explains, “is more of an art than a science,” and yet “historical analysis provides the starting point for the scientific staff.”25 Coupled with “human intelligence”—for example, statements by local villagers concerning a loss incident or the burial of remains, collected by DOD analysts working in Southeast Asia—historical investigations help determine where to seek remains (which site to excavate) or assist in evaluating the context that has resulted in remains arriving to the laboratory.
The cases laid out on the tables in CIL’s inner sanctum of the forensic anthropology examination room come from a finite number of sources. In the past, the majority of the remains accessioned to the lab resulted from a recovery effort, for example an excavation of an air crash site or a burial site in Southeast Asia, such as that which unearthed Merl Allen’s tooth. They may also come from a unilateral turnover, where a third party has handed over remains to the US military. While unilateral turnovers have been the source for multiple Vietnam War identifications, the most significant example to date relates to the Korean War, when in the early 1990s, the North Korean government handed over 208 boxes of “individual” remains. The 208 cases in fact turned out to represent approximately six hundred individuals whose bones were commingled and which the lab’s scientists are still working to segregate and identify.26 Finally, since 2015, the DOD has aggressively pursued a program to disinter unknowns—service members buried as unknowns in national military cemeteries, principally in Hawaii at the National Memorial Cemetery of the Pacific—the Punchbowl—from the Korean War and World War II.27 In the case of unilateral turnovers and disinterred unknowns, provenience may be uncertain—that is, the precise location of where the remains were originally recovered might be unknown, or at least less certain. Just as an art historian might seek to authenticate or verify the origin of a painting, so too the analysts at the lab try to ascertain the original associated loss site through archival and historical research. Even when it is known, such as a crash site in Vietnam or Laos or Cambodia, the link between the remains and the site is still best understood as putative until all evidence is aggregated and analyzed. In that instance, the archaeological investigation of the location, its physical features and its material remnants, provides the first depths to plumb for evidence of identity.
The field recoveries that the US military undertakes as a part of the MIA accounting mission aren’t the fine brush-and-trowel digs of classical archaeological sites. Because of limited time and resources—in Laos, for example, US recovery teams were for years allowed in country for a maximum of only thirty days—they operate at a breakneck speed. That said, they still adhere to the fundamental tenets of archaeological theory, including that an excavation is an inherently destructive process; the site is destroyed in pursuit of remains and any correlating material evidence—artifacts such as dog tags or blood chits, personal effects, aircraft data plates, military-issue equipment, and so on. All the more reason to document it carefully, to draft “soil profiles” tracing out the color, consistency, and compactness of the earth to define the boundaries of an impact crater or a burial site. As Greg Fox, a veteran archaeologist and former manager at the lab explained to me, archaeological evidence often reveals distinctions between war and peace, between expediency and ritual; wartime burials are usually “not actual interments but rather field sanitation,” where graves are hastily dug and bodies haphazardly placed.28 He likened recovery and identification efforts to processing metal. The excavations, he explained, were never guaranteed success. The recovery team mined veins to yield raw material. But it was the laboratory, in its process of “smelting the ore,” that was expected to produce the “zero-defect” product within a system of inevitable defects, no matter how slight.
While I learned the principles of field archaeology in the abstract during that first period of research at the lab, I came to appreciate the work of forensic anthropology through up-close observation and occasional hands-on experience. When the opportunity arose, I joined forensic anthropologists at the CIL as they analyzed and reviewed assigned cases, either producing their own report or peer reviewing that of another scientist. In shadowing them, I began to recognize interconnections across the scientific component of the mission, including how knowledge gained from one conflict’s unaccounted for advanced efforts to identify the missing from others. Observing casework also helped underscore differences in conditions of recovered remains among those conflicts, especially between those excavated from crash sites versus the disinterred unknowns from the national military cemeteries. For example, one of the few times I encountered a complete (or nearly complete) skeleton was when I joined two interns at the lab in assisting an anthropologist as he prepared the remains of a Korean War unknown for analysis. The case was one of two disinterments on the same day. Unearthed from the nearby Punchbowl, the coffins were hammered open and their contents revealed and recorded. They were then transferred to the lab’s autopsy room, where, led by the forensic anthropologist, the interns and I set to work “cleaning” the remains. We gently removed the cotton batting and cloth still wrapped around or stuck to the skeleton. We felt around the wool blanket that had ensconced the remains, patting and smoothing down its folds in search of smaller bones or teeth that might be hiding in the clumps of a chalky substance, a formaldehyde-based embalming powder. With the remains freed from the fabric, we began to wash them. Slowly, bone by bone, dipping tooth brushes and occasionally hands into the flowing water, we gently bathed and coaxed the powder off the skeleton. Its varied colors emerged, the effect of long-term contact with different elements, such as metal or soil: a darkened femur, coal black on one side, shades of brown on the other; rust-colored tones, light and dark, for most of the other bones. After several minutes of careful labor, the layers of powdery dust that had protected this individual’s remains for the last sixty years dissipated. Now clean, the bones were arranged in anatomical order on an examination table, ready for the next step in unearthing the unknown service member’s identity.
In examining remains, forensic anthropologists sort and classify; they measure, assess, and interpret the physical evidence of a skeletal element, or set of elements—that is, a bone, a bone fragment, or a collection of bones / fragments. They read the skeletal elements for clues to identity. This phase happens in the “blind”; the anthropologist is not provided with any information about the case or who it is suspected of being. The veiling is done to ensure that the analyst is not subconsciously biased to draw a conclusion or to see a feature that may not exist. To begin with, they must determine how many individuals are represented by the remains arrayed on the table—a calculus of the principle of “minimum number of individuals,” the basic idea of which is to “avoid counting the same individual twice.”29 Once remains are segregated into discrete assemblages or a single set of remains (and human remains at that), the forensic anthropologists seek to build a “biological profile” based on four features: age, sex, stature, and ancestry.30 Presented with a fully articulated, complete set of remains (again, think of the anatomy classroom skeleton), an anthropologist can come up with an almost definitive assessment of sex; a very close calculation of age and height; and a relatively strong estimation of ancestry (for example, whether the individual’s ancestral background is European, Asian, or African).31 Certain bones are instrumental in indicating particular aspects of this profile. Unsurprisingly, the pelvic bones help differentiate sex, and the surface of one area, the pubic symphysis, provides an indication of age, as do signs of growth with long bones, such as the femur and humerus, as well as clavicles and teeth. Long bones are also used to calculate height, and metric and morphological analysis of the skull, including the teeth, are typically the basis for estimations of ancestry.
I once asked one of the anthropologists if after so many years of generating biological profiles, he started to look at people, living people, differently, imagining their skeletal structure, like Ismail Kadare’s protagonist in The General of the Dead Army: “ ‘What a damned business this is we’ve got on our hands,’ the general said. ‘I can’t even pass anyone on the street or see anyone in a café now without automatically checking to see what type his skull is.’ ”32 He assured me he didn’t. Prior to joining the CIL, he had gained extensive experience studying and identifying human remains, including remains that had to be macerated (the flesh removed) in order to analyze the skeleton. If he didn’t assess the skulls of strangers passing him on the street, did he do the reverse? Did he imagine features traced out from the skeleton before him? No. “I take skin off, but I don’t put skin on.” His task was analyzing bones, not imagining flesh.
In addition to the four features of a biological profile, the anthropologist looks for traces of events and conditions during life, what Laqueur calls “the marks of the life that once clothed them”: an injury such as a healed-over broken bone; illness, diet, and nutrition; repetitive physical activity; and, key for the CIL investigation, trauma, both peri- and postmortem (at the time of and after death).33 Taphonomy, the study of the various factors affecting the preservation, condition, and recovery of skeletal remains, also provides telltale signs of the posthumous life of the remains—were they exposed to the elements, stored above ground, buried in acidic soil?
But what biological profile can emerge from a handful of bone fragments? Recall Michael Charney’s scathing critique of the US military’s identification of Anne Hart’s husband: “There was no way they could be identified as Lt. Col. Hart. The things Furue claimed to detect from the bones—age, sex, race—were just not possible.” Thus, only when the available evidence supports it, forensic anthropologists at the lab present their estimations of age, sex, stature, and ancestry, acknowledging in each instance where that assessment falls along a continuum of probability. Indeed, the notion of probability—that is, probable versus absolute terms—is critical to how remains are analyzed in forensic anthropology. As one anthropologist explained, though the “moment of death freezes the skeleton in a particular instant of development,” there is nevertheless a certain “slushiness,” a gray area between the definitive and probable, that the scientists invariably encounter. Their work therefore employs calculable error rates that can be translated into a probability of certitude. Those conclusions undergo scrutiny by peer reviewers and case managers before they are accepted as fact. “There is no room for error,” stressed another anthropologist.
Old Versus New Proof: Dental Versus DNA
Like archaeological findings and material artifacts, forensic anthropology constitutes only one line of evidence. There are other vital means of determining individual identity. In the field of forensic science applied in incidents of mass fatality—whether missing persons, victims of state-sponsored violence such as in Argentina or Bosnia and Herzegovina, the September 11, 2001, World Trade Center attacks, or victims of natural disasters—much has been made of the probative strength of DNA analysis.34 Forensic genetics does indeed present a very powerful tool. But in the case of the US military’s MIA accounting efforts, dental analysis often proves equally effective. This has to do with recordkeeping. The military recognized at the start of the 1900s that dental records offered an expedient way to identify its service members. Military dental records were kept with an expectation that they could be used for that purpose—not unlike how DNA samples are now collected from all service members upon entering the military.35 Even when X-rays were not taken, there were graphic “odontograms” made with great care and detail.36 Thus, if teeth are among the recovered remains, forensic odontologists review the case and compare the X-rays of the recovered teeth against dental records. This dental comparison typically operates on the principle of concurrence—that is, on evidence that concurs, as opposed to definitively identifies or excludes. Take, for example, a case in which three teeth have been recovered, two of which are unrestored and one restored. The two unrestored teeth could merely concur with dental records (as the examining dentist would not have marked anything on the chart for teeth that were normal); the restored tooth, on the other hand, could be used as evidence for identification or exclusion, because the dentist would have noted the specific type of restoration and material used.
In Vietnam War cases, teeth can present a twenty-first-century dilemma of scientific promise. Encased in enamel, the crown is the most durable element of the human skeleton—the remains that best defy the decaying properties of the acidic soil found in Southeast Asia—and they endure in ways that bones don’t. Given the military’s archive of radiographic and dental chart records, dental analysis is therefore not only a comparatively inexpensive tool, but also a highly effective one.37 In the popular imagination, however, dental analysis has taken a back seat to DNA testing. MIA families often want the “certitude” of genetic evidence, even when the odontological analysis, coupled with the archaeological record and material evidence, provides sufficient proof for a definitive identification. I saw this dynamic firsthand at a regional family update, when a relative insisted on speaking with the scientific director, Tom Holland, because she wanted an explanation as to why the lab hadn’t ordered DNA testing on the recovered remains. Holland was summoned, and over the next several minutes, they reviewed the case, step by step, from the excavation to the recovered remains. Among the tiny fragments recovered was a single tooth that had been conclusively shown to match the missing man’s dental radiographs. It was the only element large enough to undergo DNA testing, but the sampling and extraction process would take additional time and would use up some portion of the sample (i.e., the dentin, the boney tissue that makes up the bulk of the tooth).38 Doing so to obtain unnecessary DNA confirmation of a conclusive identification, Holland gently explained, could mean that the family would have less to bury.
With epithets like “gold standard,” “truth machine,” or “God’s signature”—not to mention the touchstone event of 1st Lt. Michael J. Blassie’s identification—it is understandable how DNA testing has raised MIA families’ expectations regarding what evidence should constitute the principal proof of their missing relative’s identity.39 Indeed, DNA in some instances is the only tool that can segregate commingled bones or identify highly fragmented remains, and the technology has improved significantly from the days of 1st Lt. Blassie’s identification. For example, whereas in the past, forensic geneticists at the Armed Forces DNA Identification Laboratory (AFDIL) needed a minimum of 2.5 grams of bone to produce a DNA profile, since 2006, they are able to do so with .25 gram.40 Still, much depends on the conditions of the remains: “The majority of remains submitted to AFDIL for testing have been subjected to harsh environmental elements for at least 30 years. Some individuals were involved in aircraft incidents, which contributed to the highly fragmented nature of the remains. Some remains have been subjected to years of burial in highly acidic or basic soils or trapped in pockets of jet fuel or saltwater.”41 Either cut from bones or as fragments themselves, bone samples are sent from the CIL in Hawaii to AFDIL, located on Dover Air Force Base in Delaware.42 There, they undergo mitochondrial DNA (mtDNA, transmitted along the maternal line) testing at first, with additional nuclear DNA—either Y-chromosome (Y-STR, from the paternal line) or autosomal (auSTR, both lines)—as necessary and if the relevant references exist. Nuclear DNA is found in the nucleus of a cell, while mtDNA comes from the mitochondria, the “power generator” organelles found in the cytoplasm of most cells.
For DNA testing to succeed in identifying a missing person, there needs to be something to compare it with—that is, the genetic profile from a bone sample must be compared against DNA sequences collected from surviving relatives or, if possible, from “self-references” such as old envelopes (e.g., where the flap or stamp was licked), baby teeth, a hair brush, or some object that might yield a viable DNA sample of the missing person.43 Since the early 1990s, AFDIL has sought “family reference samples” from relevant donors among MIA families.44 Historically, they have emphasized donors from the maternal line, given that mtDNA is more prolific (most human cells have only a single copy of nuclear DNA but an average of two hundred copies of mtDNA per cell) and more durable, meaning less susceptible to degradation from the environment and over time. But the preference for mtDNA also had to do with another set of related factors: first, the limitations of extraction and testing methodologies; and second, the demographics of the missing.45 To begin with, the overwhelming majority of the missing war dead are young men who did not have children. To use nuclear autosomal DNA, family reference samples are required from one or both parents, siblings, or children of the missing service member.46 Given the amount of time passed since death (several decades), finding the required nuclear DNA donors can be difficult. But with mtDNA or Y-chromosomal DNA, AFDIL needs only one viable sample from the maternal or paternal line.
AFDIL’s early emphasis on mtDNA also reflected the times. Short tandem repeat (STR) technology, which analyzes repeating sequences of DNA that are highly variable from individual to individual, was in its infancy in the 1990s, and due to the limitations in the extraction process, which failed to digest all bone material, not enough nuclear DNA was recovered to allow STR systems to work. It wasn’t until 2003 that Y-STR kits were available. In 2006, AFDIL devised a method to overcome the problem of chemically treated bones—called the demineralization protocol—which allowed for all bone material to be digested (the same method that reduced the required sample size from 2.5 grams to .25 gram). The advance freed up all nuclear DNA and mtDNA for purification, and thus, with additional modifications, enabled AFDIL to test for nuclear DNA.47
As the genetic technology improved, so too did AFDIL’s outreach program. Its Family Reference Collection Form maps the vast potential network of donors, expanding outward from the “missing individual” to relatives as distant as great grandnieces and fourth cousins. The numbers of references collected reflect family mobilization and awareness across the various conflicts. As of 2018, an mtDNA, autosomal, or Y-chromosome sample, provided by a relevant donor, was on file for 85 percent of the missing personnel from the Vietnam War, 92 percent from the Korean War, 85 percent from the Cold War, and 6 percent from those missing from World War II.48 Notably, despite their role in driving the US military’s MIA accounting mission and their decades-long engagement, the Vietnam War families in fact have a comparatively low participation rate, at least relative to the historically less organized and less politically active Korean War families.
Family Reference Sample chart.
Two Cases Compared
While I also saw cases from World War II and the Korean War, it was the examples from Southeast Asia, in particular excavations of air crash sites in Vietnam and Laos, that drove home the complex, contingent course of investigation, recovery, and forensic analysis that leads to an identification. In those cases, recovery missions typically yield highly fragmented and partial remains, bits of bone and teeth that may barely fill your palm. Their examples also made clear how identifications necessarily derive from converging lines of evidence—the various forms of proof, from archaeological and material evidence to anthropological, dental, and DNA analyses, that come together to point definitively to an individual service member. There is no single evidentiary silver bullet, though DNA and dental analysis often provide the data that tip the scales of probability and exclusion.
Seen side by side, two cases I reviewed in my first year at the lab—REFNO (reference number) 1895 and REFNO 0738—stand out as particularly good illustrations of these points. Their specific histories map the arc of accounting over decades and across investigations that unfolded in fits and starts, shutting down and circling back. In them we can see the ethos of exceptional care gradually taking shape and coming to fruition as it insisted finally on pinpointing sites of loss whose surfaces then had to be scoured to retrieve the scant traces of the almost entirely disappeared. In them we also glimpse the labor required to “smelt the ore” of recovery to produce an identification and return remains to surviving kin.
REFNO 1895:49 On July 10, 1972, Commander Frank C. Green Jr., pilot of an A-4F Skyhawk, failed to recover from an attack dive and crashed into a high karst mountaintop in North Vietnam.50 Because of the crash site’s location in hostile territory, friendly forces were unable to return, and Cdr. Green was designated killed in action / body not recovered. After initial confusion regarding the exact location of the crash, several investigations and field surveys were conducted at the presumed site, with United States and Vietnamese (Socialist Republic of Vietnam or SRV) joint teams traveling there four times between 1994 and 1997 to interview potential witnesses, survey the site, and assess conditions for recovery. In 1997, the site was deemed too hazardous for a recovery mission, and for good reason: the potential recovery locations entailed an area of rock fall at the base of a karst outcropping and an area near its summit. In 2005, however, a US / SRV joint team returned to resurvey the location, and in 2007, they returned to resurvey again. By 2008, excavations began. The first mission (October–November 2008) tackled the base of the karst outcropping; the second covered the much more dangerous terrain of the summit itself. The archaeological report spells out the challenging physical conditions: “The primary excavation area during JFA 10-1VM, Section II, is located near the summit of the karst outcropping that is approximately 133 [meters above mean sea level]. Section II is accessible via an approximate 100 m near-vertical to vertical ascent from Section I requiring the use of fixed ropes and ladders. The recovery scene is located on a 30° to 45° north-facing slope with numerous small to large limestone boulders.”51 The report includes photographs of team members clipped into rappelling harnesses and attempting to move large boulders.
Despite the formidable terrain, the two missions were successful. In addition to locating correlating wreckage and pilot-related equipment, both teams recovered human remains: in aggregate, there were two tooth fragments and “numerous fragmentary elements in fair to poor condition,” including portions of the skull and fragments of long bones, vertebrae, and other elements, as well as “numerous non-diagnostic bone fragments” and “a small quantity of minute bone particles mixed with sediment.”52 Although it may sound like a lot of remains, in fact, as the identification memorandum explains, their size and condition “preclude[d] the determination of any individualizing biological characteristics.” In other words, nothing definitive could be gleaned from an anthropological examination of the recovered bits of bone that could associate them with Cdr. Green. The same was also true of the two tooth fragments; not even the tooth type of either could be determined. DNA, on the other hand, proved the key line of evidence. Though only two samples cut from the skeletal remains yielded mtDNA profiles, they matched family reference samples provided by two maternal-line relatives (cousins) of Cdr. Green. Thus, in the end, the archaeological and material evidence of the crash site, coupled with the DNA analysis, “allow[ed] for the remains to be attributed to CDR Frank C. GREEN, Jr., to the exclusion of all other reasonable possibilities.”
REFNO 0738:53 On June 21, 1967, Captain Darrell J. Spinler, pilot of an A-1E Skyraider, crashed along the banks of the Xe Kong River in Xekong Province, Laos. Though a second pilot remained in the area for two hours searching for any signs that Capt. Spinler had survived, he was eventually given the status of killed in action (body not recovered). In October 1993, a joint US / Lao People’s Democratic Republic (LPDR) team investigated the purported crash site and interviewed two witnesses who claimed to have seen the pilot’s body on the nearby river bank; they “speculated that the remains likely had been washed away during the ensuing rainy season(s).” During that same investigation, the team surveyed the site and located wreckage consistent with an A-1E Skyraider. A year and a half later, the US government evaluated the case for its recovery potential, and the case was officially designated “no further pursuit,” effectively suspending future investigations or site surveys. But as happens periodically with Vietnam War–era cases, eventually the government’s decision was revisited, and another team was sent to investigate the site in 1999. Four years later, the no further pursuit designation was reaffirmed. For six more years, the case lay dormant, until a joint US / LPDR team working in the same province returned to the site once more and located additional A-1E Skyraider wreckage. After sixteen years of tacking between pursuit / no pursuit, these latest findings pushed the site up the priority list for excavations. From October 31, 2010, to November 11, 2010—just under two short weeks—another joint US / LPDR team excavated the site and recovered “human remains, personal effects, and pilot-related equipment from amid the sparse wreckage field.”
Remains in this case consisted of small fragments of three “disarticulated” teeth (meaning disarticulated from the mandible, i.e., the jaw bone): #10, #18, and #28 in the Universal Numbering System used in dentistry. Their condition precluded the use of DNA testing, “given the current state of that technology,” the identification memorandum explained. But dental comparison, the historical mainstay of the US military’s forensic accounting, carried the day. One of the three teeth (#18) had a restoration—a “mesial-occlusal amalgam filling.” Though Capt. Spinler’s dental records had no radiographs, they did include an odontogram that documented that same type of restoration on tooth #18. Notably, with this particular case, the lab also had to exclude the possibility that the dental remains might belong to another A-1E pilot—that is, that the site might be correlated with another Skyraider loss in the area. In fact, there were two other such pilots unaccounted for within fifty kilometers, but both of those sites had been located, and the recovered tooth fragment—that of tooth #18—did not match the dental records for either of the other two pilots. Here, the aggregated evidence, as partial as it was, had to exclude other potential unaccounted-for service members in order to attest definitely to Capt. Spinler’s identity. The material evidence recovered at the site played an equally weighty role—specifically, the recovered wreckage including blue canopy glass that correlated exclusively to an A-1E Skyraider. Thus, the dental analysis, coupled with the associated “non-biological material” recovered at the site, allowed “the remains to be attributed to Capt Darrell J. SPINLER to the exclusion of all other reasonable possibilities.” Identified on January 5, 2011, Captain Spinler would return to his family on Father’s Day weekend of that same year.
Encounters and Expectations
There are families who, for a variety of reasons, have not followed the MIA accounting issue, the science, or even their individual relative’s case over the intervening years. For them, news of identification—be it driven by DNA, forensic anthropology, odontology, material artifacts, or some combination of those lines of evidence—may come seemingly out of the blue. In 2011 and again in 2012, I saw this reaction repeatedly during family visits to the lab, when relatives of a recently identified service member traveled from the mainland to Hawaii in order to take custody of the remains and escort them home.54 World War II, Korean War, and Vietnam War families arrived at the lab with different levels of knowledge about their individual cases—about the recovery efforts and the identification itself. Some had obviously pored over the identification packet, the bound collection of documents that include the official memorandum of identification and supporting scientific and historical reports, and knew its details intimately. When asked whether he had read through his father’s packet, the son of a Vietnam MIA nodded his head earnestly: “Absolutely every word.”
Met at the front desk by a photographer, families were first shown to a wooden plaque bearing the names of each individual identified since 1976 and given the opportunity to have their picture taken beside their relative’s newly inscribed brass plate. Then they headed down the hall to a conference room to meet with members of the External Relations Office within the Command. With boxes of tissue and coffee service in fine china at the ready, they viewed the JPAC informational video, which, set to sweeping orchestral music, extolled the personnel, technology, and lengths taken to recover and identify the remains of the unaccounted for. They had a chance to ask questions, often of Johnie Webb, who long since had left the position of deputy commander of CILHI to head JPAC’s External Relations Office and had perfected the art of offering soothing, if somewhat stock, condolences. “You take your loved one home and give him the proper burial that he as a hero deserves,” he would say, and as his face lit up, “This is the moment we live for.”
From there, visiting family members proceeded to the lab, where they were given a brief informational tour by one of the forensic anthropologists—JPAC science on display. It began just outside the forensic anthropology examination room, where glass walls offered a clear view onto the work of analyzing remains. The anthropologist gave an overview of the forensic science of recovery and identification, referring periodically to various props laid out on a nearby table—plastic bones and model skulls, laminated photographs of excavation sites, dental X-rays, even a diagram of a cell with the mitochondria, cytoplasm, and nucleus marked. Most visitors readily admitted they knew little about the science, and their questions often related to information they had come across in their identification packet.
The welcome photograph, the video, the tour, all set the scene for the main event, when families were invited to step into the family visitation room to see the remains of their missing—now recovered and identified—service member. The first time I accompanied a family into the room, I was reminded of Bosnia and that singular moment when the coffins of the identified victims of the Srebrenica genocide are unloaded one after the other and placed in symmetrical rows stretching the length of a stark gray warehouse, ready for burial the next day. Women aren’t allowed to assist with carrying the coffins, and so instead, they look on, some trying to spot their loved one’s name on the identification placard at the foot of the slender pine box draped in green cloth. Some walk along the rows stooped over, checking each coffin until they find their son, husband, father, or brother. I heard a mother once remark as she watched two men swiftly, without strain, hoist a coffin above their heads, “they are so light.” There was so little left of the person, just a handful of bones, that there was barely any weight to shoulder. These memories came back when I watched the niece of a Korean War MIA encounter her uncle in the quiet space of the family visitation room. She was three years old when he left for the war; her siblings never knew him. There on the table was a clear plastic bag—an evidence bag—laid neatly atop a green wool army blanket. Its sight brought tears to her eyes; the rest of us—her husband, JPAC staff, the military escort, and I—stepped back into the shadows of the dimly lit room.
She reached out to touch the bag, her fingers tracing the bones within. After a moment, she wiped away tears and remarked wistfully at how it was just “one person in such a small bag.” She asked about teeth, as she knew her uncle had had special bridge work done when he was seventeen years old, but the teeth were tucked beneath the rest of the skeletal elements in a separate bag and out of sight. Did she want to touch the remains, to have them brought out of the bag? She shook her head. Did she want time alone with the remains? Again, she declined. But something had caught her eye instead: a handwritten number on the white slip of paper attached at the bag’s seal was unclear—was it 2007 or 2001 (the year of accession)? The evidence coordinator left the room to prepare a more carefully printed version. When he returned, he gingerly sliced open one end of the bag and replaced the slip, extracting a smaller plastic pouch from inside, the one containing the teeth. The niece held the tiny packet in her hands for a moment.
Finally, she was ready to take her leave. The evidence coordinator began the final step of the lab’s care for the remains, one of the most symbolically striking acts of the entire recovery and identification process. Sometimes assisted by families or a military escort, sometimes on his own, the evidence coordinator, a young man named Ben, would fold the blanket around the remains, already secured in the evidence bag, and then fasten the edges of the thick wool fabric together with six large golden safety pins. This was no slapdash effort. There is a strict protocol to the position of each pin: all facing the same direction, four along the long edge and one on each end. Ben learned firsthand that there was no room for error. He had once allowed someone else to attach the pins but hadn’t double-checked the work. The person had made a mistake, and not long afterward, the Army Mortuary Affairs Office on island got a call from Arlington National Cemetery, reprimanding them for the slip.
The ritual swaddling of the remains signaled the end of the scientific inquiry into individual identity and the beginning of the military rites of homecoming. It also enfolded the remains into the longer history of the United States military’s care for war dead. The same woolen blankets have been used to wrap the dead since World War I. In a laboratory where remains from all three major conflicts—World War II, the Korean War, and the Vietnam War—might be laid out on the tables in the forensic anthropology examination room at any given moment, the blankets and their uniform ritual leveled—if only temporarily—distinctions in conflict, branch of service, rank, and martial valor. Named and known once again, all were dressed in the plain cloth. All were headed home, some to the astonishment of surviving kin.
“You know, we didn’t expect any of this.” Having watched her uncle’s remains fastened atop a funeral home stretcher, and wheeled outside the building to the waiting hearse, the Korean War MIA niece turned to the anthropologist who had given the tour to say her thanks. It was a response I would witness repeatedly during such family visits—expressions of gratitude for events wholly unexpected. “We didn’t expect any of this,” echoed the daughter of an identified Vietnam War pilot a week later. She had come with her husband to take custody of her father’s remains and bring them home. “We got the call out of the blue, the day after Presidents’ Day.” Her husband had been the first to hear the voicemail message from the service casualty officer and told her to listen to it.55 “I thought at first it was a hoax. And then I said, ‘I think this is about my dad.’ ” “We never expected this day to come,” yet another daughter of a Vietnam War MIA explained. She was six years old when her father was shot down, and her brother just four. “We had resigned ourselves to the fact that it would be left to our imagination.”
Encountering the physical remains of their missing relatives after such prolonged absence, families thus grappled with this curious amalgam of memory and imagination, often expressed in terms of what they did and did not expect. As we will see in the following chapter, it is such expectations, with all their emotional and political force, that both drive and frustrate the US military’s scientific processes of “fullest possible accounting.”