The military’s purposes in counting the dead had also influenced the reliability of military records. Casualty lists were not compiled because of concern about accounting for the individual lives lost, as the absence of any formal procedure for notifying kin made apparent. Counting the dead had been largely an issue of assessing military resources, of seeing who was left alive to fight. A commander needed to know his military strength. Union general George McClellan had been famously obsessed with both his own numbers and those of his enemy, consistently overestimating the number of Confederates arrayed against him by two-or threefold and essentially incapacitating himself through this statistical fixation. For William Tecumseh Sherman, a man of action and decision rather than crippling reflection, numbers became a language in which to express and assess battle’s challenges and achievements. In his postwar memoirs he accompanied his description of each engagement with a summary of losses presented in the form that the War Department had prescribed for casualty reports. After his discussion of the Battle of Atlanta, for example, Sherman filled several pages with detailed numbers of losses and concluded, “I have no doubt that the Southern officers flattered themselves that they had killed and crippled of us two and even six to one…but they were simply mistaken, and I herewith submit official tabular statements made up from the archives of the War Department, in proof thereof.” For Sherman, as literary scholar James Dawes has observed, counting represented “the epistemology of war.” He could best understand the war and explain his military virtuosity by translating his experience into numbers of dead.5
But if a general needed to know his own strength, so too he hoped to conceal it from the enemy, and these tactical misrepresentations could distort the permanent historical record. In May 1863 General Robert E. Lee had issued a general order criticizing prevailing custom in reporting casualties for encouraging “our enemies, by giving false impressions as to the extent of our losses.” Inflated estimates and a tendency to report minor wounds as casualties, he believed, had resulted from commanders’ pride in losses as “an indication of the service performed or perils encountered.” After Gettysburg Lee himself pursued quite a different strategy but one equally inimical to the accurate reporting of losses: he seems to have quite systematically and intentionally undercounted his casualties in order to conceal the battle’s devastating impact on his army.6
After the war, as the immediacy of death receded, the pride in sacrifice that Lee had identified as a source of dangerously inflated numbers grew even stronger. “Claims to gallant conduct,” William Fox complained, “are very apt to be based upon the size of the casualty list.” Regiments competed for having sustained the greatest losses and thus, by implication, for having exhibited the greatest valor. Fox found that “there have been too many careless, extravagant statements made regarding losses in action. Officers have claimed losses for their regiments, which are sadly at variance with the records which they certified as correct at the close of the war.” In this postwar battle for glory, deaths became a measure not of defeat but of victory.7
The effort to compile definitive death statistics became a preoccupation in the years after Appomattox. But as with the reinterment movement, Yankees and Confederates possessed very different resources to direct to the task. Northerners would employ the expanding bureaucracy of the triumphant nation-state not just to rebury the dead but to count them. The census of interments requested by the quartermaster general at the end of the war made its contribution, ultimately resulting, as we have seen, in the reburial movement and in the twenty-seven installments of the Roll of Honor, which in their enumerations of graves provided one approximation of the totals of Union dead. Military officials had also ordered that before disbanding each Union regiment submit a “muster-out” roll including the name and fate—wounded, killed, died of disease, deserted, captured, discharged—of every man who had served at any point during the conflict. The War Department supplied large sheets, one yard square, printed with appropriate headings to be completed in multiple copies. William Fox largely relied on these documents for his compilation of casualties, and he believed they showed “clearly and accurately the mortuary losses of the regiments to which they pertain.”8
Union officer–turned–writer John W. De Forest suggests reason for some skepticism. In his popular 1867 novel Miss Ravenel’s Conversion from Secession to Loyalty, he provided a vivid portrait of the challenges one officer faced in completing the muster rolls at the end of the war. Dulled by “clouds of fever and morphine” and confronted by “a mass of company records,” Captain Edward Colbourne nevertheless struggles to do his assigned duty within the three days allotted before the troops disband. He is, he observes, the only man in his unit who has been present since its origin and thus is the only one with the requisite memory. At the end of a long night of labor, he submits the completed document to others to copy, faints, and is confined to bed for forty-eight hours. One cannot but wonder if William Fox ever read De Forest’s novel or recognized that his own data rested on such contingencies of memory and circumstance.9
Between 1865 and 1870 the War Department acknowledged the deficiencies in its records, issuing reports that presented three different—and ever-increasing—numbers of Union losses. In 1866 the Final Report of the Provost Marshal General to the Secretary of War counted 279,689 dead, but in early 1869 the adjutant general revised that number to 294,416 and then a year later, in response to an inquiry from the surgeon general, reported a total of 303,504. In 1885 Joseph Kirkley, who held the newly established post of statistician of the War Department, offered a further revision, reporting 359,528 Union deaths. A small subsequent adjustment, deriving from new information about deaths in Confederate prisons, added 694 to this sum, yielding what has come to be the most widely accepted count of 360,222.10
These constant revisions resulted in large part from information gradually brought forward by individuals seeking back pay of deceased kin or applying for federal pension and survivors’ benefits, which were established in 1862 and expanded steadily through the rest of the century. The creation of this extensive pension system for Union veterans made systematic and accurate data about military service necessary. The array of muster rolls, strength reports, hospital records, and casualty lists kept during the war did not create a coherent personnel record for any individual soldier and thus left no easily accessible file to support a pension claim. To rectify this situation, the federal government worked to create from the mass of wartime documentation a set of records that would detail the experiences of individual men. These came eventually to be known as the Compiled Military Service Records, and after 1903 they included Confederate as well as Union soldiers. Ultimately nearly thirty million northern and more than six million southern entries—each documenting the appearance of a name on a muster roll, a hospital census, a casualty list, or other official form—were inscribed on index cards and sorted into individual soldiers’ files. The scale of the effort required a small army of clerks, and the literal weight of this history inflicted its own postwar casualties. In 1893 the overcrowding of workers and documents in offices in the ill-fated Ford’s Theatre—the site of Lincoln’s assassination twenty-eight years earlier—caused two floors to collapse and kill twenty-two employees.11
But both public and private efforts to account for the dead preceded and paralleled those specifically related to pension claims. Almost every state in the North and many in the South had endeavored to produce counts and rosters well before expanding pension provisions required substantial federal involvement. Even during the conflict, individual states had authorized “Rolls of Honor” and other lists of names of those who had served and died. Many of these efforts foundered amid wartime exigencies, but in the years after Appomattox nearly every northern state renewed its effort to produce a roster.
The Pennsylvania legislature, for example, had authorized the creation of a comprehensive roster of soldiers in 1864, but the project was not launched until 1866. Samuel Bates,
Pennsylvania state historian, found his assignment no easy task, at first locating only a partial file of muster rolls in the state adjutant general’s office as a basis for his work. His “only recourse,” he recognized, was to contact individual officers and interrogate them about the history of their units.12
Massachusetts made two separate efforts to assemble a complete list of soldiers’ service and their fates. Rosters kept by the state adjutant general’s office were printed in 1868 and 1869, but twenty years later the legislature created the post of State Military and Naval Historian and directed its first incumbent, Thomas Wentworth Higginson, to compile an index list of Massachusetts soldiers and sailors that would incorporate the more accurate information that had been collected by the federal pension office. Together with lists of men, Higginson published statistical summaries of casualties by unit and engagement. Massachusetts, he concluded, had sent 113,835 men to the war and had lost a total of 13,498. Higginson’s skepticism about history’s precision seems to have been well founded: a 1997 compilation of Massachusetts soldiers based on records in the National Archives counts 146,738.13
Confederates, who after 1865 had no nation-state, no government bureaucracy, and no expectation of federal pensions, turned state and private resources to a similar effort to document and honor soldiers’ lives and deaths. But the incompleteness of Confederate records posed special challenges. The figure of 258,000 Confederate military deaths commonly cited by historians today can at best be regarded as an educated guess. The disintegration of the Confederate army made the collection of comprehensive data at war’s end impossible, and the movement of the Confederate Archives between the evacuation of Richmond, their capture in Charlotte, North Carolina, and their eventual acquisition by the U.S. War Department meant that a significant number of regimental casualty lists and other official records are missing. There are, for example, almost no muster rolls at all of Alabama troops, and all records after the end of 1864 are very fragmentary. Nevertheless, in the South as well as the North, most states tried to compile and publish rosters of those who had served and those who had died, and these volumes continued to appear into the second decade of the twentieth century.14
In 1862 the South Carolina legislature had passed a measure calling for a comprehensive Record Book “as a token of respect” to Carolina’s dead. The report that resulted was riddled with errors. In 1864 Professor William Rives of South Carolina College was appointed to undertake a second effort, and he struggled with military devastation, interruptions of mail service, and inadequate financial support. By advertising in newspapers for information, scanning obituaries, interviewing veterans, enlisting the help of tax collectors, and filling notebooks of “coarse brown paper” with data, he had by 1870 collected the names of twelve thousand South Carolina soldiers who had died in Confederate service. But, he stated, “I could not complete the work to my satisfaction.” In 1912 the Historical Commission of South Carolina took up the task once again, and A. S. Salley, commission secretary, published three volumes covering five infantry regiments in 1913.15
In North Carolina, John W. Moore overcame obstacles presented by the incomplete and erroneous data available within the state by turning to Confederate records that were in the hands of the U.S. War Department. But he found these official reports inadequate as well. “Scarcely one had full account of the casualties,” he wrote. “Unlettered orderly sergeants” produced “spelling that was really wonderful” although likely, he feared, to astonish those whose names were so creatively rendered. Nevertheless, Moore was confident that the four volumes he published in 1882 represented the most accurate presentation of North Carolina statistics possible.16
Other southern initiatives extended beyond individual states and addressed explicitly sectionalist purposes. The newly formed Southern Historical Society, established in 1869 and committed to “vindicate the truth” of Confederate history, sought to provide an accurate count of southern losses. In 1869 its secretary, the distinguished physician Joseph Jones, shared his estimate of casualties with the former Confederate adjutant general Samuel Cooper. Jones believed that one-third of all men actively engaged on the southern side had died in the conflict. Cooper affirmed that these numbers were “nearly…correct” but believed that a fuller search of Confederate records now in the hands of the federal government would provide greater detail and accuracy. For both Cooper and Jones, establishing the totals of troops North and South and documenting the extensive Confederate losses promised to provide an explanation—and justification—for the defeat as well as irrefutable evidence of the Confederate soldier’s “resolution, unsurpassed bravery and skill.”17
Private citizens in the North also set to counting wartime casualties and deaths. Frederick Phisterer, a German immigrant who had received the Medal of Honor for heroism at Stones River in 1862, published a Statistical Record of the Armies of the United States in 1883 as a supplement to Scribner’s popular series of thirteen volumes entitled Campaigns of the Civil War. Phisterer’s work included chapters on “Losses” and “Officers Deceased While in Service.” William Fox claimed that his monumental Regimental Losses in the American Civil War, published in 1889, offered “full and exhaustive” numbers from both Union and Confederate units. Thomas Livermore, who had served as a major in the New Hampshire volunteers, attempted to amplify and correct Fox’s conclusions in Numbers and Losses in the Civil War in America, which grew from an essay read before the Military History Society of Massachusetts in 1897 into a book that appeared in 1900. Frederick Dyer undertook an even more comprehensive effort in what became his 1908 Compendium of the War of the Rebellion in 1,796 pages, based, he assured his readers, on “authentic information from all reliable and available sources.” The government’s Official Records of the War of the Rebellion, begun in 1874 and ultimately published in 128 volumes, was, Dyer proclaimed, “woefully deficient,” thus rendering his work imperative. Dyer’s first volume opens with a summary of Union enlistments and losses that lists the number of those killed in action and those dying of wounds, disease, suicide, and even sunstroke. But Dyer perpetuated errors in government data that Fox had corrected nearly two decades earlier.18
Americans North and South, in official capacities and as private citizens, proliferated enumerations of the war dead but remained far from establishing a definitive count. The specificity, rather than the accuracy, of these totals attracted Americans seeking consolation in the comprehensive and comprehensible character of numbers. A figure might begin to grasp the entirety of so many dead and communicate the enormity of war’s toll.
Yet even as they counted, Americans speculated about what the numbers they so eagerly amassed actually meant. Joseph Jones counted soldiers and their deaths both to demonstrate southern valor and to explain the defeat of the hopelessly outnumbered Confederacy. Regimental commanders counted to tell the story of “how well [their unit had] stood” and to be remembered among those whose losses, and thus whose courage, was greatest. States in both North and South enumerated the dead to honor the slain. A name upon a list was like a name upon a grave, a repository of memory, a gesture of immortality for those who had made the supreme sacrifice. And the hundreds of thousands of Civil War dead who remained unnamed could at least be counted. Names might remain unknown, but numbers need not be.19
Americans counted in order to define the emerging notion of the Civil War Dead as a describable and shared national loss that transcended individual bereavements. They counted to establish the dimensions of the war’s sacrifice and the price of freedom and national unity. They counted because numbers offered an illusion of certitude and control in the aftermath of a conflict that had transformed the apparent limits of human brutality. They counted, too, because there were just so many bodies to count. Numbers seemed the only way to capture what was most dramatically new about this war: the very size of the cataclysm and its human cost.
But as numbers solved some problems of understanding, so they presented others. William
Fox worried that the sheer magnitude of the war’s death toll rendered it incomprehensible. “As the numbers become great,” he wrote, “they convey no different idea, whether they be doubled or trebled.” His proffered solution was to reduce the numbers to what he regarded as a more human scale, by considering casualties on the level of the regiment. “It has a well known limit of size, and its losses are intelligible.” Fox urged his readers not to “grow impatient at these statistics.” The numbers, he assured them, were not “like ordinary figures” but instead were “statistics every unit of which stands for the pale, upturned face of a dead soldier.” These were not cold abstractions but numbers that literally, he argued, possessed a human face.20
The muster roll that served as the source for his statistical analysis was an aggregation, but its lists, Fox found, offered far more than just numbers. Its brief entries invoked “sad pictures” of individual deaths and lives. A world lay behind every name. “There are no war stories that can equal the story of the muster-out roll,” he insisted:
“Killed, May 3, 1863 at Marye’s Heights;” and the compiler lays down his pencil to dream again of that fierce charge which swept upward over the sloping fields of Fredericksburg.
“Wounded and missing, May 6, 1864, at the Wilderness,” suggests a nameless grave marked, if at all, by a Government head-stone bearing the short, sad epitaph, “Unknown.”
“Killed at Malvern Hill, July 1, 1862;” and there rises a picture of an artilleryman lying dead at the wheels of his gun…
“Died of fever at Young’s Point, Miss.,” reminds one of the campaigns in the bayous and poisonous swamps, with the men falling in scores before a foe more deadly and remorseless than the bullet.21
This Republic of Suffering Page 27