Sebastian Bondzio, Lukas Hennies, Christoph Rass
On Wednesday, March 23rd, 2022, the New York Times published an animated map showing Russian attacks on civil targets in Ukraine. Up to this date, the invading Russian forces had hit and destroyed about 1.500 “civilian buildings, structures, and vehicles”. More than 953 civilians had been killed, with an unknown number of unreported civil casualties to add. A month into Russia’s invasion of Ukraine, estimates of the number of Russian soldiers who died in the war ranged between 10.000 and 15.000, the number of dead among the Ukrainian soldiers defending their country remains unknown. At the end of March 2022, 10 million people had been displaced by the Russian attack on Ukraine and weeks of brutal and criminal warfare. About three million Ukrainians have fled across an international border and thus became refugees. Displacement, destruction, injury, trauma, and death are the dimensions of suffering in yet another war that produces the loss of human life – killing and dying – as its primary and irreversible outcome.
The impact of the production of death in war on societies in history is a long-standing field of research at NGHM@UOS. Our projects have looked at how the Nazi Regime built its war machine and produced a community, the “Volksgemeinschaft”, ready to wage a war that turned into a war of annihilation, the Holocaust, and genocides, while the “Third Reich” was relying on a murderous war economy driven by millions of forced- and slave laborers. We study forced displacement and the struggle of survivors of Nazi persecution – the “displaced persons” of the early post-war years are a striking example of the efforts and resources needed, over decades, to cope with caring for the survivors of mass displacement by war and genocide.
We work with our partners in Belarus and Ukraine to detect, document, and preserve sites of mass killing where thousands were murdered during the Holocaust and the German war of annihilation. Our projects on 2nd World War “battlefields” and German concentration camps contribute to a critical culture of memory as a cornerstone of public history and help uncover the hidden traces of the war and Nazi terror underground. In working with DPAA, as part of our research on the Huertgenwald battlefield, we assist in searching for still missing U.S. Servicemen whom the Wehrmacht killed in the bitter fighting there in 1944/45. Out of our joint project with the Esterwegen camp system memorial (Gedenkstätte Esterwegen) grew a search for yet unknown graves of Soviet POWs in the area.
Our research on how mounting military casualties impacted civil society (“Heimatfront”) in Germany during the First World War expanded our interest in the history of military casualties from the killing grounds of modern warfare to the social contexts where the loss of life was experienced emotionally: the homes, families, and everyday lives from where men were drafted into armies – or volunteered to serve –, left to fight and kill – and die – in a war. The homes to which later not a traumatized veteran returned, but a letter was delivered announcing the death of a loved one.
We began to wonder how the death of hundreds of thousands of soldiers, mostly men, at “the front” was mirrored by the dispersion of casualties across their neighborhoods, hometowns, countries. Where did the soldiers from a specific town or city die? In what patterns was their death experienced at home?
What were the temporal, spatial, and social dynamics of the impact of war at home? Where did the “soldiers” actually come from, who would miss and mourn them, and – on a macro level – how are military casualties dispersed across society? Would the patterns of how military casualties hit the “homefront” help to explain how societies managed to keep waging war? Would we learn something about how families, collectives, societies coped with the fear of losing a son, father, brother, husband and receiving the dreaded notice? How did societies deal with the contesting needs evoked by the presence of different and often extreme emotions? What cultural changes did the losses and emotions lead to, and how did emerging war cultures help societies at home to remain able to wage war? Would the production of death and its impact and the coping mechanisms we would find help us understand how war-making was sustained and when societies reached a breaking point and decided to end the killing?
Far from being able to provide convincing answers (please find some in our more extensive publications), this short essay draws together some of our explorations in historical big data on military casualties we have used to model the production of death of military personnel in war throughout the 20th century.
Data on men killed as soldiers during modern wars is abundant. Armies and bureaucracies have left giant tomes of files created to manage millions of men in mass armies while keeping track of enlistments, the number and structure of casualties, to determine how much “human material” stood ready to be sent into battle, how many died or got wounded and how many replacements would have to be mobilized to keep up the fight. Creating honor lists and administering benefits to veterans and their families required more bookkeeping on the loss of human lives when a war ends.
The resulting lists serve administrative or commemorative purposes. They also allow researchers to build detailed models of the one thing war delivers: death. Far from neglecting the price always paid by civilian populations in military conflicts, this essay will focus on several examples where military records were used to look at military casualties from a different perspective: How did military operations translate to societies at home through the experience of losing someone who died as a soldier?
The examples presented here cover the period from the First World War to the Vietnam War and represent a different angle: we look at mobilization, casualties from a single city, the death toll of a major battle, the dead and wounded from individual military units and the losses sufferd by an entire society in two wars.
Missouri goes to war – 1917.
The first example provides an idea of how modern societies mobilize for war, based on a dataset containing information on about 150.000 U.S. Army Veterans from Missouri. The dataset begins in the late 1870s showing only very few men enlisting. This only changes in the 1910s and gained more and more momentum after the war began in Europe during the summer of 1914. Mobilization reached its climax 1917/18 reflecting a gigantic buildup of the American armed forces. About 4.7 million men and women would eventually serve during the First World war and the Armed forces of the United States would lose 54.402 of its personnel in combat and an additional 63.114 non-battle casualties.
Our model thus demonstrates the immense dynamics of mobilization as the USA entered the war in 1917. Geocoded by addresses and the date of enlistment, the data re-draws the map of cities like St. Louis and Kansas City, both ranked among the ten largest cities in the USA and mirrors the increasing speed and dimensions of enlistment across the state. Only a fraction of the men who went into the army from Missouri were deployed overseas. The National Archives list a total of 2.485 soldiers from the state who lost their lives in the “Great War”, representing roughly 1.6 percent of the enlistees in the entire dataset. This fascinating dataset could, of course, be explored in greater depth to look at the social profile of the soldiers, their pathway into the army and war, or the injuries they suffered.
Experiencing loss: a German town 1914-1918
The number of casualties recorded for the state of Missouri roughly matches the number of men from the German town of Osnabrück killed between 1914 and 1918 as soldiers of the Prussian Army. Two million men from the German Empire died as soldiers during the “Great War”. About fifteen percent of families in a town like Osnabrück had to confront losing one or more relatives in a city which sent an estimated 15.000 men went to war, at least one from each extended family.
While dying and killing in war are often discussed with reference to the killing fields at the front (“Massensterben”), we decided to look at how the casualties were dispersed when projected back onto the soldier’s homes.
When doing so, it is essential to know that armies count their soldiers by units and, throughout the 20th century, military units in most European armies were recruited from specific, smaller, or larger geographical regions into specific units attached to those areas. Military units thus often, at least at the beginning of a war, had a distinct regional character. However, men from a specific recruitment area or district were put into a relatively large number of different units assigned to that particular area, and eventually, regional cohesion would become harder and harder to maintain when losses were mounting and finding replacements became more difficult and chaotic.
Such mechanisms of mobilization and personnel management in modern military organizations allow us to take three different perspectives on the data:
(a) In battle, casualties are often listed by the military units deployed – in fact, most historical documents on casualties are produced following a military logic as part of tracking losses, determining strength, and the need for replacements, on the one hand, keeping track of “fallen soldiers” to organize care for their remaining families on the other; (b) casualties from individual units can often not just be related to specific battles but also be projected back onto their recruitment bases. Such data will then represent a fraction of the losses suffered by the people living there since it still mirrors the unit-centered military logic. (c) This changes when we look at all military casualties from a specific area, town or village – those usually come from a multitude of different units -, or an entire country to understand how the death of military personnel impacts a war-waging society on all levels.
Thus, depending on the mode of recruitment, the way military units are organized, and the casualties they suffer are logged, the production of death looks very different when framed by a “battle”, by the losses of an individual military unit or the losses suffered in a specific region at the “homefront”. Most historical documents reproduce the military’s point of view. But does that protocol of death represent the way that such losses impact a war-waging society?
A case study on three regiments of the Prussian Army fighting in the First World War provides a visual example of how casualties of individual units linked back to their recruitment base. Infantry Regiment (IR) 78, Reserve Infantry Regiment (RIR) 92, and RIR 229 drafted most of their soldiers from Northern Germany from three overlapping districts moving from West to East. At the same time, while most of the men came from these districts, the birthplaces of the soldiers were dispersed all over the German Empire and even beyond its borders. For example, RIR 229 was first mustered in Breslau at the end of 1914 and, in 1915, became reinforced by a replacement battalion of the IR 78. This merger drastically changed the regional composition of the regiment and threw men from very different regional and cultural backgrounds together in one unit.
Infantry Regiment 78 of the Prussian Army had its battalions garrisoned in different towns across Northwestern Germany, with one of those garrisons being Osnabrück. Many soldiers from the city served in this regiment during the First World War. In “their” regiment, men from Osnabrück represented a larger group but by no means a majority. An animated map of the casualties of the entire regiment over time and space shows the vast geographic and social space the regiment’s soldiers were hailing from. Besides the spatial dimension, casualties also had a specific temporal pattern. Whenever the regiment went into battle and losses occurred (red dots indicate places where (multiple) deaths occurred), these were transmitted home (blue dots) and occurred there in distinct spatial and temporal clusters.
In-between battles, when the regiment was moved between the Western and the Eastern front, different sections of the front lines or went into recreational spaces for refitting, almost no casualties were reported back home for days, weeks, or even occasionally months at a time. On a regimental level, deaths thus occurred in high-intensity clusters determined by events at the front line with calm periods in-between dictated by the rhythms of modern trench warfare. The losses drew distinct spatial and temporal patterns on the battlefields but also in its recruitment area.
Concerning the home front, however, this is only half the picture. Shifting the focus to individual cities or towns, like Osnabrück, a regional economic and administrational center and thus also a typical Prussian garrison with several large barracks billeting units from several regiments, puts the production of death into another perspective.
The towns register of war dead from World War One, preserved in the State Archives of Lower Saxony, provides detailed information on more than 2.200 men who served in the Prussian Army (and the Imperial Navy) and died as soldiers during the war or in hospitals in the immediate post-war period. They were drafted into dozens of military units, which sometimes went into action simultaneously, but they were on separate trajectories for most of the war. Units engaged in the same battle could also suffer distinctly different numbers of casualties depending on the specific sector and phase in which they saw action. Projecting all the men from Osnabrück onto their home addresses in town listed in their recruitment papers thus shows how these different circumstances across dozens of military units resulted in the pattern of losses experienced by families at Osnabrück.
Between 1914 and 1918, on average, every 16 hours, a soldier died, and a death notice had to be sent to one family’s home in Osnabrück. There were peaks during the spring and summer months when significant battles were fought, and lesser numbers during the autumns and winters. In Osnabrück, the casualties did not only arrive in a specific temporal pattern, they were also spatially distributed across the complex social topography of the city – with different patterns in the upscale quarters of the bourgeois in the West and the densely settled working-class neighborhoods in the southern and eastern parts of town.
Interestingly, extreme heights in the production of death as documented by the aggregate casualty statistics of the Prussian Army were often cut off locally by the mix of different units recruiting men from the town and their specific patterns of high and low-intensity casualties that depended on the when and where of their combat missions. When IR 78 suffered high losses, only a fraction of those losses would actually link back to Osnabrück. While soldiers from one unit died together in their sector of deployment, the losses were dispersed across the unit’s entire recruiting area and thus never became visible in one single location or space in their entirety. By the same token, local casualties could exceed the overall trend in military losses when, by chance, multiple units related to an area suffered medium or high losses at the same time.
Soldiers originating from one town or place died in many different units in many different places and only represented a fraction of the casualties suffered in the field by their particular military unit. This could cause a certain cushioning effect which made the losses even of massive death-producing events like the fighting at Verdun in 1916 – where hundreds of thousands of deaths occurred in a distinct area within only a few months of fighting – being felt “at home” in a different way.
Of course, there were places in the German Reich that suffered high casualties from the Verdun, but for a lot of other places – like Osnabrück – the dispersal of the dead across thousands of different places made the impact of this particular event less hard. The focus on losses in a specific battle thus misses the complexities of how losses at the frontline translated to home – as the experience of loss and grief -, while the focus on individual experiences and mourning misses the macro-effect of war casualties hitting home. We propose to apply such reflected approaches to avoid the reproduction the military logic of data collection and knowledge production when talking about the impact of war through casualties at home.
The different unit’s soldiers from Osnabrück fought in, and these units’ different patterns of battle deployments made us aware of the complexity of the production of death during the First World War. The dynamic map showing the geocoded places of death for soldiers from Osnabrück draws a picture of the front lines, East and West, between the Entente and the Middle Powers. At the other end, a map of the city of Osnabrück gets sprinkled with a steady stream of death notices arriving at the homes of men KIA (“killed in action”).
A single battle: Gelibolu 1915/16
Analyzing the impact of one single large-scale battle of the First World War in detail yields further insights. Turning to the meso-level, the Canakkale/Gelibolu/Gallipoli campaign provides a gruesome example.
To establish a maritime support route for the struggling Russian Empire, in 1915, Allied forces tried to deal the Ottoman Empire a lethal blow by invading the Gelibolu peninsula at the entrance to the Dardanelles, attempting a push towards Constantinople. The ensuing battle raged from February 1915 to January 1916. According to published research, the British, French, Australian, and New Zealand units deployed in the campaign suffered a total of 56.707 fatal casualties (the all over death toll including missing and non-recovered soldiers, men who died of wounds sustained at Gellibolu elsewhere, or died in POW-camps will probably be higher than this figure). The Ottoman Army suffered 56.643 battle casualties, with a very large number of other deaths attributed to missing soldiers, men who died of diseases or their injuries but not instantly on the battlefield probably not being included in that number.
We have geocoded a dataset with information on the Ottoman casualties produced by the Army of the Republic of Turkey to show the place of birth for each casualty. The dataset lists 50.688 soldiers of the Ottoman Army killed in the Gelibolu campaign. While the campaign is usually dated from 2/1915 to 1/1916, the dataset spans from April 15th, 1915, the beginning of the Allied landing and ground assault, to December 12th, 1917, which probably reflects the death of soldiers from wounds suffered in the battle at later times (dates were converted from the Maliye Calendar to the Gregorian Calendar).
A geocoded analysis leads to a map showing how one battle fought over roughly one year during the First World War produced a pattern of loss and mourning across the territory of modern Turkey and adjacent regions then still part of the crumbling Ottoman Empire. A similar analysis would probably be possible based on the casualty data collected in New Zealand or Australia, for instance.
Animated to show how the casualties occurred over time, losses move across Turkey in waves corresponding with the more intense phases of the battle – the initial invasion in April 1915 or the Allied offensive in August – while pauses indicate the emerging trench warfare and variations in the magnitude of fighting and killing.
When color-coded by the different Army Corps deployed in the battle, the spatial analysis reveals geographical recruitment patterns in the Ottoman Army. Each army corps had a distinct core zone of recruiting and took in men from provinces outside these zones, albeit in lesser numbers. As the fighting at Gelibolu stretched over months after month and Ottoman units suffered high losses, the dispersion of the casualties across time indicates that various Army Corps were present in different phases of the battle in an overlapping sequence. This, in turn, means that in terms of news about deaths at Gelibolu reaching the soldier’s families at home, the battle also impacted different regions of Turkey with varying intensity at other times.
And again, while at Gelibolu today, monuments and graveyards, like on so many historical conflict landscapes, essentialize dying in war on a battlefield, projecting the killed soldiers onto the places of their civilian lives helps to turn the perspective from the production of heroes to the loss of loved ones at home. This dimension of the war experience is mirrored by the way the main cemetery of the Ottoman Army on the Gelibolu Peninsular is laid out: The graves are organized by major cities and provinces. At the same time, it is difficult to trace individual families, as the Ottoman Empire did not know “family names” which were only introduced in the early years of the Turkish Republic.
In the path of death: Wehrmacht units 1939-1945
To study the impact of losses in a German infantry unit during the Second World War, we went back to the micro-level of one single company. We chose the 1st Company of Infantry Regiment 464, which was part of the 253rd Infantry Division of the German Wehrmacht. A rather average unit whose men waged Hitler’s criminal war in all its dark shades from 1939 to 1945.
This war of annihilation, as we have recently shown in a study on about 9.103 “burned villages” that the Germans left in ruins in Belarus between 1941 and 1944 with countless civilian victims, made unspeakable violence against the civilian population an essential part of its everyday strategies.
In the First World War, fighting – at least on the western front – was centered around massive trench systems that emerged and became fortified after the short initial phase of mobile warfare had failed to bring military success. Based on the experience of seemingly endless trench warfare, German military strategy in the Second World War was deliberately designed to avoid another stalemate. The dominant mode of operation was thus movement, in advance or retreat, broken by shorter or longer periods of static fighting or occupation.
This is mirrored in the movements of Wehrmacht infantry divisions. The 253rd Infantry Division, a unit of the 4th (1939) wave of mobilization – and thus an average Division – did not take part in the assault on Poland but was used to attack in the West in May 1940. It quickly went through Belgium and France, served briefly on occupation duties there before it was moved into occupied Poland to prepare for the invasion of the Soviet Union.
In June 1941, it crossed the line of demarcation between the parts of Poland occupied by Germany and Russia to become part of the German war of annihilation against the Soviet Union, advanced with the Army Group Center to the city of Reshew, and, after the winter of 1940/41, would spend the next four years in a series of defensive battles, retreats and holding operations in various sectors of the eastern front – while building a brutal regime of occupation and exploitation in its rear areas whenever possible. In March of 1944, for example, the infantry division was – like all divisions and army corps of the Wehrmacht’s 9th Army – part of the massive deportation of civilians to the Ozarichi camp system in Belarus, one of the most significant and most brutal war crimes committed by the Wehrmacht.
Following one infantry company from the 253rd Infantry Divisions 464rd Infantry Regiment mirrors this pathway of death and destruction.
We gathered data from the “Erkennungsmarkenverzeichnis” (a rough equivalent of the “morning report” produced by units of the U.S. Army) of 1./464 company which is available from 1939 to early 1945, providing data on every soldier transferred in or out of that unit. To sustain an average strength of 100 to 150 men throughout most of the war, more than 1.500 individual soldiers went in and out of 1./464 between late 1939 and early 1945.
The casualty reports of 1./464 list a total of 936 men who died or were transferred out of the unit wounded or sick, 233 soldiers were listed as “killed in action”. Mapping the data on wounded and killed soldiers shows the sequence of advances and retreats and fighting with a sharp difference in intensity between the invasion of France and the attack on the Soviet Union and a gruesome crescendo as the war turned against Nazi Germany. Also, the map shows the ratio of wounded and killed men as a rough share of 25 percent of the casualties being fatal or permanent (killed, missing, prisoner of war) and the rest being wounded or sick (who would eventually return to duty).
This diagram shows each incidence horizontally when a soldier was transferred in (blue) and out (red) of 1./464 and time on the vertical axis. We thus see the dynamics of how many soldiers were needed to sustain a single company of combat infantry. An animated map shows the hometowns of soldiers drafted into 1./464 over time. This animation provides some interesting findings: In the early stages of the war, most men came from the recruitment base of the 253rd Infantry Division in the Rhineland.
As the war progressed and losses mounted, the regional composition of the unit began to lose spatial cohesion as soldiers from all over the “Third Reich” were shoved into the ranks to maintain the unit’s operability. Furthermore, as this map shows all the units soldiers and not only those who died in the war, it points out that only a small fraction of the soldiers in a particular unit were killed. Most of them survived the war.
Beyond the World Wars: US casualties from the wars in Korea and Vietnam
For our last two examples, we move to the macro-level of entire wars and use records on military personnel killed in action for two major wars in the United States’ history of the 20th century kept by the National Archives and Records Administration of the USA in Alexandria, VA (NARA).
While available American data on the First and the Second World War does not include the hometowns or -counties of individual soldiers, data for the war in Korea and the war in Vietnam list where the men came from. For the Korean War, lists show the soldier’s home counties, for the Vietnam War, actual places of birth or residence were recorded. This allowed for visualizing the dynamics of the spatial impact of both wars on American society.
US casualties from the Korean War
The NARA-Data on the Korean War lists 36.573 cases dating from 1950 to 1953 (one case from 1954). The dispersion of the war dead across the USA shows how the densely populated Eastern and Midwest Areas were hit hard in 1950 and 1951, as were the larger cities on the West Coast. Later in the war, the density of losses diminishes, and only towards the end do casualties mount again.
We thus can observe the different phases of the war. Furthermore, it becomes clear how costly the initial phase of U.S. troop deployment war was and how stalemates tended to stretch the duration of the conflict with relatively low casualties. For future research, it is relevant to understand the relation between the actual losses and how American society talked about this production of death. Including public discourses in such analysis leads to a mixed-method approach that allows for a deeper understanding of how particular societies perceive and cope with death in war and how much of it they were willing to bear to reach a political objective. This way, historians might better understand the relevant factors that enabled a society to sustain prolonged engagement in a conflict and which coping mechanisms allowed them to bear the mounting loss of lives.
US casualties from the Vietnam War
The 2nd Vietnam War is usually dated from 1954/55 to 1975. In contrast to the war in Korea, where the highest losses were experienced during the first year, this war produced a small number of US-American casualties during its first decade. The casualties were scattered in a loose spatial pattern across the USA. Beginning with an intensified U.S. engagement, especially after the Gulf of Tonkin Resolution of August 1964, the casualties suffered by U.S armed forces increased sharply. The NARA dataset on personnel from the U.S. armed forces who were killed in action or died of other conflict-related reasons in this war lists 58.219 cases, dating from 1955 to 1975 (with a handful of deaths occurring in 1987-2006).
A buildup to higher casualty figures across the country occurred in the mid-1960s, when the U.S. officially declared war on North Vietnam, reaching its climax towards the end of the decade while about half a million U.S. soldiers were present in Vietnam. The fast and numerous turnover of personnel nescenssary to maintain this troop strengh due to the “tour of duty”-system, made the Vietnam war a “veteran-producing” conflict. The return of large numbers of traumatized men and women from the conflict over a prolonged period while the conflict was draging on, and the increasingly critical media coverage of the violence and American casualties abroad played a significant role in breaking the will to continue this war. This points to the importance of how war victims and casualties are translated into public, media, political and also historical discourse, how death in war is framed, explained, made visible or invisible, is subject to free debate of supressed.
As U.S. troops began to disengage, withdraw, and “Vietnamize” the conflict in the early 1970s, the number/density of the losses declined until the war officially ended in 1975.
Looking at the nexus between military casualties, the experience of loss and grief at home, discourses on military engagement and its price, and the perception of outcomes as “victory” or “defeat” helps to better understand why wars are ended or carried on. A comparison between the Korean War and the Vietnam War from that perspective could also contribute to deconstructing the emergence of different narratives about wars.
Without going deeper into the analysis of the social profile of the military casualties based on the information available in the documents used in this essay to create simple digital models of how the production of death in war impacted societies from a chronological and spatial perspective, and without a broader discussion of the temporal and geographic patterns the data showed, and also without zooming into the micro-level of individual counties, towns, villages, streets or families, we aimed at demonstrating how the droves of (historical big) data collected on dead soldiers can be used to turn our perspective on wars: Instead of focussing on reconstructing the killing and dying on battlefields, it is worthwhile to model the impact of death as the product of war onto war-making societies as well.
The fact that we can quickly gather mass data on “fallen soldiers” does however put up another question: It is to be asked whose death is deemed relevant and gets recorded. While most armies keep track of their casualties and many societies use such data, amongst other things, to prepare “lists of honor” and to put names of killed soldiers on monuments, it is a lot harder to find similar data on the loss of civilian lives beyond rough statistics.
Historical sources make dead soldiers very visible – most of the time, we are able to find out more documents about the dead than about the survivors – while other groups, civilians as victims of war and of war crimes, remain rather invisible as individuals. The efforts to collect individual data on the victims of the Shoah and other Nazi crimes amply demonstrate the difficulties to reconstruct biographies of civilian victims of genocide, war, and mass violence.
In many cases, there are much fewer documents with systematic collections of names and circumstances listing civil victims of war and violence on an individual case level for entire societies or conflicts, which points to questions of how the production of knowledge by states creates visibilities and invisibility that historians have to reflect when doing data-driven research.
The conclusion, however, stays the same: the main product of war is death, and it is time to stop viewing soldiers – who wage violence in war and suffer it – as heroes or victims on battlefields alone. They were and are people with homes, lives, families who are put in uniform and sent to kill and die. Their death makes a difference beyond the mere functioning or failure of a military apparatus. Societies forced to defend themselves, like Ukraine today, endure great sacrifice. Countries bringing themselves to attack and invade others have to ask themselves, today and as they look back at their history, if the production of death they deliver and suffer is ever worth their ill causes.