WE ARE BEGINNING TO DETECT A SHORTFALL IN ... OUR DATA BASE.
- Cable sent by the U.N. Special Representative in the Balkans, Yasushi Akashi, on 13 July 1995, two days into the gendercidal massacres at Srebrenica in Bosnia.(2)
From the opening hours of the 1999 war in Kosovo, an overriding tactic was evident in Serb military strategy: the gender-selective detention and mass killing of ethnic-Albanian men, especially those of "battle age." Although the Milosevic regime's genocidal assault on Kosovar society swept up all other sectors of the population, killing many and expelling hundreds of thousands to neighbouring countries, the most systematic and severe atrocities and abuses were inflicted disproportionately or overwhelmingly upon non-combatant men. The Organization for Security and Cooperation in Europe (OSCE) was emphatic in its recent report, the most comprehensive available:
Young men were the group that was by far the most targeted in the conflict in Kosovo ... Clearly, there were many young men involved in the UCK [Kosovo Liberation Army] ... but every young Kosovo Albanian man was suspected of being a terrorist. If apprehended by Serbian forces -- VJ [Yugoslav army], police or paramilitary -- the young men were at risk, more than any other group of Kosovo society, of grave human rights violations. Many were executed on the spot, on occasion after horrendous torture. Sometimes they would be arrested and taken to prisons or other detention centres, where, as described afterwards by men released from such detention, they would be tortured and ill-treated, while others would simply not be seen again. Others were taken for use as human shields or as forced labour. Many young men "disappeared" following abduction.(3)
Similarly, although much remains unclear at the time of writing, a large body of refugee testimony and growing forensic evidence from East Timor suggests a systematic targeting of younger males for dismemberment by machete, mass execution, and torture to death.(4) A Timorese woman in a camp for refugees/abductees said she was told by a militia man: "You may have got your country but it will be a land full of widows."(5) As in Kosovo, fragmentary but cumulatively powerful reports have spoken of younger men being pulled from refugee groups and executed or "disappeared," and of a massive underrepresentation of men in refugee communities, this time in the militia-run camps of West Timor.(6)
The present article seeks to place such acts of "gendercide" in comparative and global-historical perspective. It argues that gendercide -- inclusively defined as gender-selective mass killing -- is a frequent and often defining feature of human conflict, and perhaps of human social organization, extending  back to antiquity. I contend as well that gendercide is a regular, even ubiquitous feature of contemporary politico-military conflicts worldwide.
A theme of the article is that gendercide, at least when it targets males, has attracted virtually no attention at the level of scholarship or public policy. As such, it can be classed as one of the great "taboo" subjects of the contemporary age. I hold, nonetheless, that an inclusive understanding of gendercide carries powerful implications for the emerging field of comparative genocide studies. For present (not necessarily all) purposes, "gender" can be defined primarily, if not exclusively, in terms of biological sex.
The term "gendercide" was first coined by Mary Anne Warren in her thought-provoking 1985 book, Gendercide: The Implications of Sex Selection. Warren drew
an analogy between the concept of genocide and what I call gendercide. The Oxford American Dictionary defines genocide as "the deliberate extermination of a race of people." By analogy, gendercide would be the deliberate extermination of persons of a particular sex (or gender). Other terms, such as "gynocide" and "femicide," have been used to refer to the wrongful killing of girls and women. But "gendercide" is a sex-neutral term, in that the victims may be either male or female. There is a need for such a sex-neutral term, since sexually discriminatory killing is just as wrong when the victims happen to be male. The term also calls attention to the fact that gender roles have often had lethal consequences, and that these are in important respects analogous to the lethal consequences of racial, religious, and class prejudice.(7)
Warren gives the analyst a great deal to work with here. There is the gender-inclusive framing of "sexually discriminatory killing"; the normative injunction that all such killing is equally "wrong"; and the sophisticated linking of the gender variable to "racial, religious, and class prejudice." But although she pledges to explore the issue in detail in her chapter on "Gendercidal Precedents," the promise of the theoretical framing is never fulfilled -- or rather, it is only partially fulfilled. The chapter concerns itself exclusively with female-selective killing: female infanticide, the witch-hunts in Europe, suttee or widow-burning in India, female genital mutilation, "the denial of reproductive freedom" (to women), and "misogynist ideologies." Much of the remainder of her book is devoted to the subject of the sex-selection of children as a form of gendercide against women. Whatever the merits of extending the framework this far (or to the genital mutilation of women or men), gendercide, for all practical purposes, is limited in Warren's analysis to "anti-female gendercide."(8)
In fact, however, non-combatant men have been and continue to be the most frequent targets of mass killing and genocidal slaughter, as well as a host of lesser atrocities and abuses. The mass killing of males, particularly of "battle-age" men,(9) has roots deep in the history of conflict between human communities. "From antiquity, [patriarchy] has had an inherent problem with men not covered  by the bonds of kinship or culture and has traditionally sought to marginalise them through diverse means," writes the Jamaican scholar Errol Miller in a little-known and stimulating book. "The practices of killing all male captives, of castrating the men whose lives have been spared, and of offering men less opportunities for manumission from slavery, all show that men's domination of men outside the bonds of kinship and community has been more severe and brutal than men's domination of women within or outside the kin or ethnic group."(10)
Gerda Lerner writes of the Middle East in classical times that "There is overwhelming historical evidence for the preponderance of the practice of killing or mutilating male prisoners and for the large-scale enslavement and rape of female prisoners." Citing Lerner, Barbara Ehrenreich speculates that in the gendercide of males may lie, by a curious twist, the origins of misogyny: "In the situation Lerner describes, where enemy males were killed and enemy females enslaved, the only surviving adult representatives of the defeated enemy would of course be female, and the psychological equation would have been established, over time, between femaleness and the enemy 'Other.'"(11)
Such gender-selective strategies were by no means fully consistent or universal, even to the limited extent that historians have been able to test the classical accounts. Leo Kuper agrees that "it was common enough practice to destroy besieged cities and to slaughter their inhabitants, or their male defenders while taking the women and children into slavery." But he implicitly contrasts events such as the classical "destruction of Troy and its defenders, and the carrying off into slavery of the women (as described in the legendary accounts and the Greek tragedies which have come down to us)," with the genus of "root and branch extermination, expressed in the slaughter of men, women and children."(12)
It is nonetheless remarkable how regularly one comes across references, in the literature on modern mass killing, to staggering demographic disproportions of adult males versus adult females -- that is, a wildly skewed underrepresentation of adult men. One of the best-documented cases, though its strict designation as a "gendercide" is disputable, is the Stalinist purges of the 1930s and 1940s. "As early as August 1937, they were shooting seventy men a day," wrote Robert Conquest in his study of The Great Terror.(13) In a fascinating addendum, Conquest used 1959 census figures to argue that the Soviet population "was some 20 million lower than Western observers had expected after making allowance for war losses." "But the main point," he noted, "arises from a consideration of the figures for males and females in the different age groups." He then unveiled a striking table indicating that whereas age cohorts up to 25-29 displayed the usual 51-to-49 percent split of women to men, from 30-34 the gap widened to 55 to 45 percent. Thereafter, the disparity became massive, reflecting the generations of males caught up in the purges and the Great Patriotic War. From 35-39, women outnumbered men by 61 to 39 percent;  from 40-54, the figure was 62 to 38 percent; in the 55-59 age group, 67 to 33 percent; from 60-69, 65 to 35 percent; and 70 or older, 68 to 32 percent. Conquest summarized the findings as follows:
Many women died as a result of the war and the purges. But in both cases the great bulk of the victims was certainly male. From neither cause should there be much distinction in the figures for the sexes for the under-30 age groups in 1959. Nor is there. For the 30-34 block the[re] ... is a comparatively small difference, presumably indicating the losses of the young Army men in their late teens during the war. In the 35-39 group, which could have been expected to take the major war losses, we find figures of 391 to 609 women. One would have thought that these men, in their early twenties in the war, would have had the highest losses. But the proportion then gets worse still, and for the 40-44, 45-49 and 50-54 [cohorts] remains a set 384 to 616. Even more striking, the worst proportion of all comes for the 55-59 age group (334 to 666: in fact in this group alone there are almost exactly twice as many women as men). The figures for the 60-69 group (349 to 691) and for the 70 and over group (319 to 681) are also much worse than the soldiers' groups. Now all authorities agree that the Purge struck in the main at people "between thirty and fifty-five"; "generally, arrested people are all thirty or over. That's the dangerous age: you can remember things." There were few young or old, most of them being "in the prime of life." Add twenty years for the 1959 position.
Precise deductions are not possible. Older men died as soldiers in the war. But on the other hand, the mass dispatch to labour camps of prisoners of war returned from Nazi hands in 1945 must have led to an extra, and non-military, death rate among the younger males. So must the guerrilla fighting in the Baltic States and the Western Ukraine, which lasted for years after the war; and so must the deportations from the Caucasus and the general renewal of Purge activities in the post-war period. But in any case, the general effect of the figures is clear enough. The wastage of millions of males in the older age groups is too great to be masked, whatever saving assumptions we may make. We here have, frozen into the census figures, a striking indication of the magnitude of the losses inflicted in the Purge
-- and, we might add, a striking indication of perhaps the worst gender-specific slaughter in human history. But Stalin has not lacked for competitors among the architects of twentieth-century genocide and mass killing:
No territory-wide census was taken in the Congo until long after the rubber terror [of 1890-1910] was over. But Daniel Vangroenweghe, a Belgian anthropologist who worked in a former rubber area in the 1970s, found persuasive demographic evidence that large numbers of men had been worked to death as rubber slaves or killed in punitive raids -- and he discovered the evidence in the [Belgian] regime's own statistics. No other explanation accounts for the curious pattern that threads through the village-by-village headcounts taken in the colony long before the first territorial census. These local headcounts consistently show far more women than men. At Inongo in 1907, for example, there were 309 children, 402 adult women, but only 275 adult men. ... At nearby Iboko in 1908 there were 322 children, 543 adult women, but only 262 adult men. Statistics from numerous other villages show the same pattern. Sifting such figures today is like sifting the ruins of an Auschwitz crematorium. They do not tell you precise death tolls, but they reek of mass murder.(14)
I was ... told that in Cerewek, Gabus, and Sulur [Indonesia, after the 1965-66 genocide] 70 percent of the population are widows. Some people even said that in Banjardowo it was very hard to find a single adult male. Where could they have gone to?(15) 
All through the liberation war [of 1971], able-bodied young men [in East Pakistan/Bangladesh] were suspected of being actual or potential freedom fighters. Thousands were arrested, tortured, and killed. Eventually cities and towns became bereft of young males who either took refuge in India or joined the liberation war.(16)
The major long-term demographic result [of Pol Pot's 1975-79 genocide] is the preponderance of women in modern Cambodia. Women, including large numbers of widows, make up 60 to 80 percent of the adult population in various parts of the country, as well as among Cambodians abroad.(17)
Rwanda has become a country of women. It is currently estimated that 70 percent of the population is female and that 50 percent of all households are headed by women.(18)
That the gender-selective mass killing and "disappearance" of males, especially "battle-age" males, remains a pervasive feature of contemporary conflict is not open to dispute. Indeed, its frequency across cultures and conflict types marks it as a possibly definitional element of contemporary warfare, state terrorism, mob violence, and paramilitary brigandage:
Kosovo, 1999. "Shortly before dawn on April 27, according to locals, a large contingent of Yugoslav army troops garrisoned in Junik started moving eastward through the valley, dragging men from their houses and pushing them into trucks. 'Go to Albania!' they screamed at the women before driving on to the next town with their prisoners. By the time they got to Meja they had collected as many as 300 men. The regular army took up positions around the town while the militia and paramilitaries went through the houses grabbing the last few villagers and shoving them out into the road. The men were surrounded by fields most of them had worked in their whole lives, and they could look up and see mountains they'd admired since they were children. Around noon the first group was led to the compost heap, gunned down, and burned under piles of cornhusks. A few minutes later a group of about 70 were forced to lie down in three neat rows and were machine-gunned in the back. The rest -- about 35 men -- were taken to a farmhouse along the Gjakove road, pushed into one of the rooms, and then shot through the windows at point-blank range. The militiamen who did this then stepped inside, finished them off with shots to the head, and burned the house down. They walked away singing."(19)
Jammu and Kashmir, 1999: "Since 1990, some 700 to 800 people have 'disappeared' after being arrested by police or armed paramilitary forces ... The victims have included men of all ages, including juveniles and the very old, and all professions, including businessmen, lawyers, labourers and many teachers. Many of them appear to be ordinary citizens picked up at random, without any connection to the armed struggle ... Their relatives still live in unbearable uncertainty about the fate of their loved ones."(20)
Colombia, 1998: "Rightist militiamen using chainsaws on some of their victims killed 11 peasants and kidnapped 13 others, accusing them of collaborating with leftist guerrillas ... Police said the chainsaws were used to torture and behead several of the victims. Others were shot to death. Ten of the victims were men, police reported. The slain woman, a minor, was killed by militiamen seeking her husband, who was not at home."(21)
Rwanda, 1997. "I've been a member of the RPF [Tutsi-dominated Rwandan Patriotic Front] since April 1991. I know a lot about the massacres committed by the Rwandan Patriotic Army [in the course of invading Rwanda and ending the genocide]. The RPF  wanted to establish its supremacy, and to do so they had to eliminate any potential rival. In many cases the Army came for men, ages 18 to 55, and took them away by night, never to be seen again. Their families search for them in vain, in the prisons of Rwanda, but they all died at the hands of the Rwandan Patriotic Army. The general pattern was to select youth and men who were still active, as well as leaders, teachers, farming instructors -- anyone who played a role, any community leader was particularly singled out."(22)
Bosnia-Herzegovina, 1992. "They were shelling our village [while] I was in a shelter. Some men got away. Those who were in their homes were beaten, tortured and killed by the Cetniks [Serbs]. ... We came out of the shelter. They were looking for men. They got them all together. We saw them beating the men. We heard the sounds of the shooting. One man survived the executions. They killed his brother and father. Afterwards the women buried the men."(23)
Sri Lanka, 1991. "Soldiers have searched for rebel suspects in some of the 30 refugee camps that now ring Trincomalee. A leader of one camp, near the village of Nilaveli north of Trincomalee, said that 84 men were detained in a series of Army sweeps last year. One refugee who returned said he was locked in a prison that held about 350 men. 'They tied my hands behind my back and kept a blindfold around my eyes. I hardly ate for two weeks,' he said. 'I was finally released with 15 [men]. We don't know what happened to the others.' The camp leader, who requested anonymity, said only 45 out of the 84 detainees are accounted for. The others 'disappeared or have been killed. We don't know.'"(24)
Peru, 1990. "The Peruvian Army occasionally reacts to ambushes and attacks by invading a community and killing dozens of young and old males, sometimes in full view of relatives."(25)
Delhi, October-November 1984 [following the assassination of Prime Minister Indira Gandhi by her Sikh bodyguards]: "The nature of the attacks confirm that there was a deliberate plan to kill as many Sikh men as possible, hence nothing was left to chance. That also explains why in almost all cases, after hitting or stabbing, the victims were doused with kerosene or petrol and burnt, so as to leave no possibility of their surviving. Between October 31 and November 4, more than 2,500 men were murdered in different parts of Delhi, according to several careful unofficial estimates. There have been very few cases of women being killed except when they got trapped in houses which were set on fire. Almost all of the women interviewed described how men and young boys were special targets. They were dragged out of the houses, attacked with stones and rods, and set on fire ... When women tried to protect the men of their families, they were given a few blows and forcibly separated from the men. Even when they clung to the men, trying to save them, they were hardly ever attacked the way men were. I have not yet heard of a case of a woman being assaulted and then burnt to death by the mob."(26)
Iraqi Kurdistan, 1983. "In August 1983, Iraqi security troops rounded up the men of the Barzani tribe from four resettlement camps near Arbil. These people were not engaged in any antigovernment activities. ... Two of Barzani's sons at that time led the Kurdistan Democratic Party and were engaged in guerrilla activities against the Baghdad government, but only a part of the tribe was with them. The entire area of Barzan had, along with many other parts of Kurdistan, been evacuated by the government, and the Barzanis who had opted no longer to oppose the government had been moved to resettlement camps. All eight thousand men of this group, then, were taken from their families and transported to  southern Iraq. Thereafter they disappeared. All efforts to find out what happened to them or where they had gone, including diplomatic inquiries by several European countries, failed. It is feared that they are dead. The KDP [Kurdish Democratic Party] has received consistent reports from sources within the military that at least part of this group has been used as guinea pigs to test the effects of various chemical agents."(27)
What explanations can be advanced for this apparent predominance of males among the victims of genocide and mass killing? Most of the variables are, in fact, fairly intuitive. First, there is a military logic to the destruction of the "battle-age" portion of a targeted community, whether as a sufficient measure in itself or as a prelude to "root-and-branch" extermination of the community. Honig and Both's examination of the gendercidal massacres at the Bosnian city of Srebrenica in July 1995 conveys this military logic well:
... If it is not as simple an answer as "revenge," and if the suggestion of [the victims being] "combat casualties" is patently false, why did the Bosnian Serbs kill the men of Srebrenica in cold blood? Much of it seemed to have to do with the character of the war in Bosnia. Muslim men posed a pointed threat to the Serbs, who were educated in the traditions of people's war and who aimed to create an ethnically pure Serb territory. And, as the police chief of Serb-controlled Banja Luka commented in 1992, "In ethnic warfare the enemy doesn't wear a uniform or carry a gun. Everyone is the enemy." Civilians were considered indistinguishable from soldiers. The long-term success of ethnic cleansing depended on killing off the Muslim men, without whom the population's women and children would have no means of returning to their birthplaces. And since Srebrenica had been one of the main refuge places for Muslims from eastern Bosnia, executing the men would give the Serbs a more secure hold over the region. Both the requirements of people's war and the dictates of ethnic purity demanded that potential resistance, now and in the future, be minimised. As many men as possible had to be killed.(28)
As well, many genocides contain strong overtones of "elitocide." Societal elites, like "battle-age" males, may be targeted in isolation or as part of a phased assault on an entire people. And males overwhelmingly constitute the public face, at the very least, of those elite sectors. (The Burundi genocide of 1972, for instance, targeted mainly Hutus who were senior students, prominent church workers, and soldiers. These are all institutions in which the analyst would expect to find a strong male predominance.) Moreover, since most elites are to most appearances mostly male, it is not a great leap to the proposition that male equals elite -- just as men's "potential" as combatants may be enough to secure them death in a typical counter-insurgency sweep.
Lastly, there are the cultural codes and humanitarian biases that have been pervasive throughout cultures. Even many highly warlike societies have chosen to limit the scale of their physical destruction of an opposed population through enslavement, concubinage, or outright freeing of women (and children). Certain ingrained norms have sometimes obtained in dealing with the elderly and infirm, though there is also ample evidence of specific savageries directed against this group. Regardless, and crucially, the most vulnerable and consistently targeted population group, through time and around the world today, is non-combatant men of a"battle age," roughly 15 to 55 years old. They are nearly universally  perceived as the group posing the greatest danger to the conquering force, and are the group most likely to have the repressive apparatus of the state directed against them. The "non-combatant" distinction is also vital. Unlike their armed brethren, these men have no means of defending themselves, and can be detained and exterminated by the thousands or millions. The gender of mass killing, moreover, likely extends beyond the age range specified. Elderly males are probably more prone than elderly women to be caught up in the "malestrom" of war; and modern warfare, with its relentless press-ganging and criminality, extends ever further down the age ladder in the hunt for child soldiers and street thugs -- overwhelmingly boys.
In all of these actions there are institutional, material, political, and cultural interests and variables underlying the systematic targeting of males. We are not, in other words, talking about an abstract "hatred of men" as lying at the root of these genocides and genocidal massacres, in the way that Nazi mass murder was clearly founded on an ideological hatred of Jews (and others). But the frequent and often massive correlation between male victimization and the most annihilatory genocidal excesses may merit a fundamental rethinking of the prevailing "gendered" framing of many of these issues.
If gendercide and mass killings of males is to some degree definitional of modern conflict, we may also be able to isolate an essential if not universal ritual of gendercide against men. It is the physical act of separating men from women as a prelude to consigning men to death. The ritual is enacted with great frequency the world over, although it is not always explicit in the above examples. Nonetheless, as Hochschild likened the evidence of gendercide in the grotesquely-misnamed Congo Free State to the "ruins of an Auschwitz crematorium" (see the excerpt above), we should see in our minds the camp commander and his henchmen on the platform, systematically and dispassionately "culling" part of a group (here, the male part) and consigning those selected to rapid extermination.
Women and Gendercide
The manner in which women are targeted in genocidal slaughters may also amount to gendercide. There is no doubt that the term should be applied to all cases of mass rape followed by murder. In certain historical circumstances -- including relatively recent ones (Bangladesh, Nanking, Berlin) -- women have been targeted en masse for combined rape and killing, or raping to death. This must surely rank among the most excruciating deaths known to humankind, and much the same "culling" process may be evident as in the case of gender-selective mass killings of men. In the contemporary era, a further deadly element has been added to the mix: AIDS. Most of the current cases of largescale rape in conflicts are in sub-Saharan Africa (Sierra Leone, the Democratic Republic of Congo, Angola), in areas that also lie at the heart of the "AIDS belt."(29) For women in these conflicts, rape thus carries with it the realistic threat, not only of death on the spot, but of a lingering and agonizing wasting by disease. 
Moreover, the century's best-studied genocides -- the holocausts against the Armenians and the Jews -- also featured acts of gendercide against women that (like those against males) were analytically significant, if subsidiary. These ranged from individual actions to fully-fledged gender-selective policies and institutions. Two examples of the latter in the case of the Jewish holocaust are the female "work" camps, and the genocidal death marches from their gates at the end of the war. (In both cases, the agents of the gendercide were also predominantly female.)(30) However, there seem to be very few such cases of women being separated from men and marked off for execution while men are preserved, even temporarily. Women (and children, and the elderly) tend to be targeted as part of "root-and-branch" exterminations that target all members of the community.
We should by no means limit our framing to the traditional politico-military one. In first deploying the term "gendercide," Mary Anne Warren examined infanticide in history, making plain how pervasive and exterminatory towards girl children in particular this policy has been over the ages. "There are very few cultures in which male infants are more apt to be killed than females," Warren writes. She finds evidence across civilizations, from Arab societies where "the birth of a daughter was regarded ... as a humiliating calamity -- and often still is"; to northern Indian tribes "that "killed virtually all female infants at birth," to nineteenth and early twentieth-century western Europe, where the murder of girl children (and sometimes boys) "was publicly condemned but practiced covertly, in ways that made it appear accidental or inadvertant [sic]."(31) R. J. Rummel may be not far off in his suspicion that "the death toll from infanticide must exceed that from mass sacrifice and perhaps even outright mass murder":
In many cultures, government permitted, if not encouraged, the killing of handicapped or female infants or otherwise unwanted children. In the Greece of 200 B.C., for example, the murder of female infants was so common that among 6,000 families living in Delphi no more than 1 percent had two daughters. Among 79 families, nearly as many had one child as two. Among all there were only 28 daughters to 118 sons. In India ... young girls were murdered as a matter of course. When demographic statistics were first collected in the nineteenth century, it was discovered that in "some villages, no girl babies were found at all; in a total of thirty others, there were 343 boys to 54 girls. ... In Bombay, the number of girls alive in 1834 was 603."
Rummel adds: "Instances of infanticide ... are usually singular events; they do not happen en masse. But the accumulation of such officially sanctioned or demanded murders comprises, in effect, serial massacre."(32) And though the infanticide phenomenon (like the ritual sacrificing of children or adults) is very far from an exclusively female phenomenon, specifically female infanticide seems closely enough identified with the subordination of women and most things "feminine" in history to constitute a gendercide against women. Fortunately, it is a phenomenon that may today have something of the status of slavery: largely eradicated in its classic form, though with lingering traces (China) and more muted offshoots still apparent (such as sex-selective abortions for upper-class families in India and elsewhere).
Feminist scholars have frequently cited the trial, condemnation, and slaughter of tens of thousands of women for "witchcraft" in early modern Europe as an example of the gender-selective killing of women. Christina Larner's equation of "witch-hunts" with "woman-hunts" seems well-grounded, given that the gender disproportion was of the order of four to one against women, at least in the Scottish data she cites.(33) This "identification of the relationship of witch-hunting to woman-hunting" is necessary, Larner writes,
to concentrate attention on such questions as why women were criminalized on a large scale for the first time in this period, and whether there is any significance in the simultaneous rise of prosecutions for witchcraft (old women) and infanticide (young women); whether there was any change in the socio-economic position of women in this period; why [the idea of] a female secret society should seem particularly threatening at this juncture, and to what extend the popularization of Christianity, a patriarchal form of religion, was a factor.
"This does not mean that simple overt sex war is treated as a satisfactory explanation for witch-hunting, or that the 20 per cent or so of men who were accused are not to be taken into account," Larner stresses. "It means that the fact that the accused were overwhelmingly female should form a major part of any analysis."(34)
Men: Gendercidal Institutions
But the opposite of Larner's formulation is also true. In fact, I can think of no better defence of a focus on the mass killing of males than a simple paraphrase of her argument: "The fact that those targeted were overwhelmingly male should form a major part of any analysis." Are there areas outside the extermination of non-combatant men in conflict situations that might also be amenable to a "gendercide" analysis?
We must acknowledge first that the line between combatants and non-combatants is often blurred. Consider the fate of press-ganged Ethiopian conscripts flung into battle against Eritrea in early 1999:
Nearly two months [30 May 1999] after the battle of Tsorona, the bloodiest yet of this desert "border war," Ethiopian soldiers still lie unburied on the baking plain, just metres from Eritrean trenches. ... The Ethiopian commanders' strategy was simple. Deploying tens of thousands of barely trained recruits along a 5km front, they drove them forward, wave upon wave, with the aim of blowing them up on minefields until they had cleared a path to the Eritrean front line for better trained infantry, mechanised forces and armour. ... It didn't work: the soldiers hardly raised their weapons, but linked hands in communal solace in the face of certain death from mines, the trenches, perfectly aimed artillery and their own officers, who shot them if they turned and ran. This was the horror of which Ms. Haile [an Eritrean woman soldier] and her companions spoke, of mowing down the horde till their Kalashnikovs were too hot to hold. 
The result in the Ethiopian/Eritrean conflict is a slaughter that, if not gendercide strictly viewed, must still be seen as a profoundly gendered atrocity -- one of the worst of the post-World War II era, in fact.(35) The "press-gang" trend could obviously be traced back through history. Rummel, for example, gives an estimate of three million Chinese men who died out of some 14 million conscripted by Chiang Kai-shek's Nationalist forces during the Second World War -- "close to the total military battle dead for Germany in all of World War II, and twice the number for Japan."(36)
The mass murder of prisoners-of-war might also be considered an act of gendercide against non-combatant males. This group has received "particularly lethal treatment ... throughout history," according to Rummel. "If their lives were spared they were often sent to work as slaves in mines, on galley ships, in swamps, or at other labor that killed them off rapidly. The Mongols used their prisoners in the front ranks when attacking fortified cities and towns, and forced them to fill in moats or prepare catapults close to the dangerous walls. If not turned into slaves, prisoners of war were often simply killed, captured garrisons massacred."(37)
The picture of modern times is distinguished only by the greater scale of the killing. Indeed, the worst gender-selective slaughter in human history -- perhaps the most concentrated mass killing of any kind -- targeted male prisoners-of-war barely half a century ago. I refer to the mass liquidation by Nazi forces of Soviet POWs captured or rounded up behind the lines in the first year or so of Operation Barbarossa. Daniel Goldhagen gives a figure of "2.8 million young, healthy Soviet POWs" killed by the Germans, "mainly by starvation ... in less than eight months" of 1941-42, before "the decimation of Soviet POWs ... was stopped" and the Germans "began to use them as laborers."(38) Nor should it be assumed that the POW designation referred to men who had, in fact, seen military service. Nazi policy in the rear areas was very explicit: "Men between the ages of 15 and 65 were to be treated as POWs ... [and] taken to POW camps." The 18th Panzer Division studied by Omer Bartov had "orders to arrest all men of military age and [']send them to the rear[']," which rapidly became a euphemism for mass murder by execution, starvation, and exposure. In his detailed study of German occupation policies in the USSR, Alexander Dallin devoted a chapter to the fate of these Soviet prisoners of "war," whatever their particular route to detention and death:
Testimony is eloquent and prolific on the abandonment of entire divisions under the open sky. Epidemics and epidemic diseases decimated the camps. Beatings and abuse by the guards were commonplace. Millions spent weeks without food or shelter. Carloads of prisoners were dead when they arrived at their destination. Casualty figures varied considerably but almost nowhere amounted to less than 30 percent in the winter of 1941-2, and sometimes went as high as 95 per cent.(39)
Bartov states the consensus view that "some 5,700,000 Russian soldiers fell into German hands [during the entire war], of whom about 3,300,000 died in captivity." This outright genocidal and gendercidal assault was a direct result of  "the ideological concepts of the Nazi regime, which strove physically to eliminate the 'Bolshevik Untermenschen' [sub-human]."(40) The fact that the sub-humans specifically targeted for death among the population of the occupied zone were almost exclusively male has tended to pass -- as it were -- unmenschened.
What of institutions other than the military that target men for genocidal atrocity? Corvée labour is an obvious one, and one of the most time-honoured of state-building and empire-building strategies. Rummel estimates that in colonial times, "at a rock-bottom minimum, 10 million ... forced laborers must have died thusly," and "the true toll may have been several times this number."(41) The estimate is indeed a conservative one: Adam Hochschild claims the death-toll in the Congo "rubber terror" alone numbered many millions, though male forced labourers were far from the only Congolese who suffered and died.(42) According to Jessica Carew Kraft, "the Congo terror ceased only after the population declined so dramatically that forced labor became unprofitable";(43) we have already seen demographic evidence for the gender disproportion of that decline. Jonassohn and Björnson place French rule in "their" part of the Congo around this time in the same category of cruelty. The French Governor-General Antonetti, planning construction of a railway to the coast, was "frank about the human cost" of the project. "Either accept the sacrifice of six to eight thousand men, or renounce the railways," he declared, and later: "I need 10,000 dead [men] for my railways."(44) In general, if such corvée groups across cultures and throughout history were less than 90 or 95 percent male, one would be surprised. The primacy of the gender variable in the equation -- the overwhelming tendency of despots to select males for forced labour leading rapidly to extinction -- might reasonably warrant the designation of certain corvée institutions as gendercidal.
Predictably, the historical record shows a considerable overlap between corvée and associated institutions of imprisonment and incarceration -- again, a victimization experience that is and always has been a male near-monopoly the world over. (In Canada, the federal prison population is 99 percent male.) Integrating this phenomenon properly into a discussion of state-directed mass killing throws surprising light on the "accepted" gendering of historical events. Taking an example already cited, it allows us to set the witch-hunts of early modern Europe against the backdrop of "the harshest period of capital punishment in European history" -- knowing what we know of the gendering of capital punishment almost everywhere throughout history. And what happens when we gender the "witch-hunts" and mass incarcerations of the modern era -- e.g., Stalin's purges, or state crackdown on "Reds," "terrorists," and "agitators"? These historical and sociological questions have barely been hinted at, let alone coherently framed, in any mainstream discourse; and there is no space or specialist's competence to explore them here. My ambition, as throughout, is simply to "engender" the debate.(45) With this objective in mind, I turn now to consider the implications of the gendercide framework for the field of comparative genocide studies. 
Gendercide and Genocide
In the light of this record, one is prompted to wonder as to why the historical recurrence of a social phenomenon failed to register as a critical social problem, particularly among social thinkers and social scientists.
- Vahakn Dadrian(46)
One of the most promising developments in the academy's engagement with issues of conflict and human rights is the recent emergence of a school of "genocide studies." This dates from Leo Kuper's short, seminal book Genocide: Its Political Use in the Twentieth Century in 1981, though antecedents can be traced back through Vahakn Dadrian and Hannah Arendt to Raphael Lemkin, who coined the term "genocide" in 1944.(47) Currently, the school -- which includes such scholars as Kuper, Frank Chalk, Kurt Jonassohn, Yves Ternon, and R.J. Rummel -- has sought to develop a comparative history, sociology, and typology of genocide, as well as attempting to isolate "early warning" signs and propose ameliorative strategies. Though still in its early stages, the work, now buttressed by institutions like the Montreal Institute for Genocide and Human Rights Studies, has produced a flourishing case-study literature that has added inestimably to our understanding of mass killing throughout history.
It is remarkable, though, how invisible or barely-visible the gender variable remains in this emerging literature, at least as far as non-combatant males are concerned. One is put in the position of piecing together disconnected fragments. Kuper, in his field-defining work, mentions in passing that "While unarmed men seem fair game, the killing of women and children arouses general revulsion"; but it is a specific reference to the Northern Ireland conflict, which has not (in Kuper's view and others') reached a genocidal scale.(48) The obfuscation of the variable in the wider literature may reflect the fact that it is non-combatant males who tend overwhelmingly to be the victims of gender-selective mass killing, and this remains a powerful taboo in the feminist-dominated discussion of gender.
Some sense of the theoretical confusion that can result is suggested in Jonassohn and Björnson's analysis of genocides in antiquity. They write: "The Old Testament contains a number of cases that today would be considered as genocides -- not because of the casualties of warfare, but because of the killing of noncombatant women and children."(49) Later, they state of the destruction of Sybaris in southern Italy that "Since the killing was not limited to members of the army, we can call this a genocidal massacre."(50) What is seeping in here, albeit subtly, is a cultural bias that defines an entire population group by the activities of some of its members -- a phenomenon that ought to be of particular concern to scholars of genocide. Jonassohn and Björnson seem to be suggesting, first, that the death of non-combatant women and children is definitional to a genocide; and in the second place that the mass killing of civilians constitutes a genocide. The strategy, as is so often the case, makes "womenandchildren" coequal with the civilian population, while expunging non-combatant males  from the framing. Thus, a group that constitutes a prime target of genocidal assault through history, continuing to the present day, drops out of the analysis.
It could be argued that the line between militarized and non-militarized males is difficult to draw in many situations of social strife and military conflict. But in many of the cases Jonassohn discusses (such as besieged cities in classical and medieval times), such distinctions would also have been debatable in the case of women, the elderly, and even children. Such total wars in microcosm tend to enlist all members of the "civilian" population in military or quasi-military activities -- which may explain the determination of many conquerors to engage in "root-and-branch" extermination of the entire city or community. There seems little reason, prima facie, to consign males to the "military" category in this manner -- in effect, marking them off as expendable and non-definitional to genocidal killing -- while granting blanket exemption to "womenandchildren," the destruction of whom, it is implied, defines really egregious (i.e., genocidal) acts of mass murder.
I have come across just one mainstream definition of "genocide" that accords gender first-among-equals status as a primary category of victims. It is simultaneously so broad and so confining a framework, however, that it cannot be adopted without substantial modification. The definition was proposed by Steven Katz, building on the crucial ingredient of the perpetrator's perception and intention stressed by predecessors like Chalk, Jonassohn, and Dadrian.(51) The term "genocide," Katz argues, can be applied only to "the actualization of the intent, however successfully carried out, to murder in its totality any national, ethnic, racial, religious, political, social, gender or economic group, as these groups are defined by the perpetrator, by whatever means."
In one of his Talmudic footnotes, Katz pursues the point further, critiquing the work of Henry Huttenbach for discounting the possibility of sexual -- "what I prefer to call gender" -- groups being targeted for extermination. Huttenbach contended that this "is an error because empirically neither homosexuals nor women under Hitler were the targets of genocide. I agree with the facts -- that is, I concede that neither homosexuals nor women (qua women) were the victims of genocide in World War II -- but this empirical argument does not count against the logical argument that homosexuals or women could, under other circumstances, be the targets of genocide. Both of these possibilities are logically conceivable."(52)
From an inclusive gender perspective, Katz's definition can be seen as seriously deficient in its inability to comprehend that a group beyond women and homosexuals might be attacked on gender grounds (that is, "qua men"). An overriding problem, though -- as Jonassohn and Björnson point out -- is that Katz's definition of genocide is so restrictive that not even his main subject, the Jewish holocaust, would qualify: "It is quite clear that Hitler did not intend to murder the Jews in their totality. There were many exceptions to this intent."(53) This brings up the critical debate over necessary scale that preoccupies analysts of mass killing. Can the term "genocide" (and thus "gendercide") legitimately be  applied to acts of mass murder that are not total in their effect, or even in their intention?(54)
"Genocide," in common usage, does tend to carry totalising implications. But it is far from the case that this framing reflects a consensus among scholars, let alone international legal theorists and policymakers.(55) Citing the United Nations definition of genocide (1948),(56) John B. Allcock carries the argument to the opposite extreme: "It is often assumed that in order to qualify as genocide, killing must take place on a very large scale, with perhaps thousands if not millions of victims. It is important to note, however, that within the terms of the UN Convention, no account is taken of the number of victims. The execution of a handful of villagers for reasons of national, ethnical, racial, or religious identity might be legitimately regarded as an act of genocide."(57) Mary Anne Warren also puts the argument succinctly, in a passage from Gendercide:
The concept of genocide, as it is commonly understood, does not apply only to those actions which result in the complete extermination of a race of people ... Sometimes it is appropriate to speak of certain actions as genocidal atrocities, even though many members of the victimized race or culture survive. ... Furthermore, not all instances of genocide involve direct or deliberate killing. Deaths or cultural disintegration deliberately or negligently brought about through starvation, disease or neglect may also be genocidal. Indeed, some acts of genocide do not involve any deaths at all, but rather consist in the wrongful denial of the right to reproduce.
Accordingly, Warren "suggest[s] that an action, law or policy should be regarded as genocidal if (1) it results in an absolute or relative reduction in the number of persons of a particular racial or cultural group; and (2) the means whereby this result is brought about are morally objectionable for independent reasons -- e.g., because they violate certain individuals' right to life, liberty, or security against wrongful assault."(58) All these formulations are useful in constructing a gendered analysis of genocide, even if Warren's exploration is exclusively women-focused, and even if she applies her framework to phenomena (such as female genital mutilation) that detract from the force of her argument.
I join with proponents of the trend in arguing that the partial destruction of a wider group (ethnic, religious, etc.) is sufficient to warrant the designation "genocide," though I share Kuper's view that "the charge of genocide would not be preferred unless there were a 'substantial' or an 'appreciable' number of victims."(59) If this argument is accepted, we can eliminate Katz's requirement of "murder in its totality" from our definition of genocide. This is far from a cosmetic alteration, since it undermines the Holocaust exceptionalism that lies at the heart of Katz's thesis. But I nonetheless choose to rework his definition as follows: "the actualization of the intent, however successfully carried out, to murder in whole or in substantial part any national, ethnic, racial, religious, political, social, gender or economic group, as these groups are defined by the perpetrator, by whatever means."
Gendercides against men and women -- but particularly men -- may be seen in this light as one of the more common forms of genocide. Can we go a step further and hypothesize what types of genocides or genocidal massacres are most likely to exhibit a gender-specific and/or gender-selective dimension? Female infanticide, the rape-killings of women through history, and mass murders for witchcraft (of which the European case stands alone in history) should probably be so characterized. On the other hand, mass purges and "politocides" such as Stalin's massacres and the Cultural Revolution in China, could be expected to be weighted disproportionately or overwhelmingly against males. The related phenomenon of "elitocide" could be similarly classed. Finally, the most militarized genocides -- those carried out against a backdrop of partisan or rebel activity, or heavily masculinized dissidence -- seem to exhibit the most pronounced gendering against male victims. A correlation is often evident with "patriarchal" culture, as this might be manifested in patterns of community organization and family roles. In his richly insightful study of "blood-feud" institutions in Montenegro, Christopher Boehm gives a vivid picture of the almost surreal lengths to which the gendering of such conflicts can be carried:
In the old days, women were free to come and go as they chose under feuding conditions, since taking their blood did nothing to help the blood score and also counted as a dishonor, morally speaking. Thus, their normal daily activities could continue. But men were sorely pressed when it came to doing any work other than herding, which allowed them to stay under cover with a rifle ready at all times. In 1965 [at the time of field research] it was for this reason that women still did so much of the heavier work in the fields, so I was told by the slightly apologetic Montenegrin "male chauvinists," who viewed this as a once-necessary custom formed in an earlier era. During the traditional period it also made sense, from the standpoint of economics, for Montenegrin men to go raiding a great deal. By engaging in this activity they were contributing to the household economy, and in a raiding party, whatever its risks, they at least were safe from attacks motivated by blood revenge. But whatever might happen to the men during a feud, the women were always free to keep the household economy going because the rules of feuding were taken so seriously by the opposing party.
With respect to the sanctity of women, it was even possible for them to enter directly into combat during the first stage of a feud, when the killer's clan shut itself in and the victim's clan attacked the fortified stone farmhouse, which had loopholes [for firing rifles] everywhere. With no fear of being harmed, women could carry straw and firebrands up to the house to try to burn it. Also, women of a besieged house could go outside at night carrying torches, to light up the enemy so that their own men could shoot at them. This exemplifies the strength of these particular rules: to shoot a woman was a source of shame (sramota) for the entire clan.(60)
That these trends are not historical relics was attested to by an Agence France-Presse dispatch from northern Albania in June 1999. Mihaela Rodina described vendetta killings in the town of Shkoder. When a bar-owner refused  to pay protection money to local mafiosi, and instead killed five gang members, a local villager observed: "Isufu's family knows it will not escape kanun, which requires bloodshed to be avenged by bloodshed. The males, even the youngest, will be cowering in their homes, hoping to escape the vendetta that will be mounted by the relatives of the five dead." Rodina added: "According to non-governmental groups, the men of some 25,000 families in northern Albania live thus, never going out of the house for fear of being victims of similar feuding. The women, who are unaffected by the kanun, are left alone to provide for the family's needs."(61) In the Middle East, the Caucasus, Colombia, and certain urban zones of the United States, such acts of execution and mass murder similarly display a selective, ritualized, and intricately-coded gendering. This framing could be extended as well to children and women, who are overwhelmingly the victims of domestic mass murder,(62) the perpetrators of which are largely (though not exclusively) male. Such acts of extermination possess their own coded and ritualized character (including, frequently, the suicide of the perpetrator), although the assailant acts outside the centrally-directed power structure that characterizes standard politico-military genocides and gendercides.
Retributive and "Pre-Emptive" Gendercides
The most useful single concept for the study of "gendercide" that I have found in the genocide literature is Vahakn Dadrian's idea of retributive genocides. Dadrian depicts "this form of genocide [as] limited in scope insofar as its objective is confined to localized atrocities as a form of meting out punishment to a segment of the minority, challenging or threatening the dominant group." The strategy "possesses the concomitant function of warning and/or intim[i]dating potential challenges and of deterring a recurrence of trouble."(63) This captures quite well the operation of the gender variable in gendercides against men, where the wider collectivity is "culled" and "sifted" to isolate a minority considered threatening, according to the blanket application of diverse variables (usually gender and age). Furthermore, the "challenge" and "threat" to "the dominant group" captures something of the competitive and belligerent character of intra-male politics, the principal challenge of which has always been to suppress perceived male rivals or competitors.
The retributive strategy might also be a "pre-emptive" one. Indeed, one line of investigation that offers real promise is the notion of gendercide as a tripwire or harbinger of fuller-scale "root-and-branch" genocides. A gendered understanding of the dynamics of genocide throws important new light on key cases of mass killing throughout modern history. The Armenian genocide of 1915-16, for example, is remembered primarily -- and rightly -- as a fullscale assault on the ethnic-Armenian population of Turkey. But the dynamics and development of the genocidal attack exhibited a little-noticed gendering that may be predictable, and thus theoretically useful:
The first step in the genocidal process was the emasculation of the Armenian population.  It was initiated by the disarming of the many soldiers serving in the Turkish army, followed by the disarming of the civilian population. ... In the early part of 1915, the Armenian soldiers, mostly combatants, were stripped of their arms and transformed into road labourers, and into pack animals, stumbling under the burden of their loads, and driven by the whips and bayonets of the Turks into the mountains of the Caucasus. They were given only scraps of food; if they fell sick, they were left where they had fallen. In many cases, they were dealt with in even more summary fashion, "for it now became almost the general practice to shoot them in cold blood."
Once the "battle-age" Armenian males swept up in military conscription had been exterminated, a similar cull of remaining community males was carried out before the wider program of deportation was effected. Drawing on the work of Arnold Toynbee, Kuper describes "a common pattern of deportation" as starting
with a call from the public crier that male Armenians forthwith present themselves at the Government Building. This was the usual procedure, though in some cases the warning was given by the soldiers or gendarmes slaughtering every male Armenian they encountered in the streets. When the men arrived, "they were thrown without explanation into prison, kept there a day or two, and then marched out of the town in batches, roped man to man ... They had not long to ponder over their plight, for they were halted and massacred at the first lonely place on the road ... The women and children were not disposed of by straightforward massacre like the men. Their destiny under the Government scheme was not massacre but slavery or deportation" [Toynbee]. Usually after a few days, the women and children, and the remnant of men who, through sickness, infirmity or age, had escaped the general fate of their sex, were ordered to prepare themselves for deportation. For the women, the alternative of conversion to Islam (if available) could only be ratified by immediate marriage to a Muslim and the surrender of children to be brought up as true Muslims. "Deportation was the alternative adopted by, or imposed upon, the great majority."
That it was no alternative at all became progressively clear, however, as the horrors of the march mounted. Toynbee wrote of the deportations that
Women who lagged behind were bayoneted on the road, or pushed over precipices, or over bridges. The passage of rivers, and especially of the Euphrates, was always an occasion of wholesale murder ... The lust and covetousness of their tormentors had no limit. The last survivors often staggered into Aleppo naked; every shred of their clothing had been torn from them on the way. Witnesses who saw their arrival remark that there was not one young or pretty face to be seen among them, and there was assuredly none surviving that was truly old ...(64)
Thus, although the element of "gendercide" in the Armenian holocaust is important to an understanding of the Turks' genocidal strategy, it is far less significant in describing or explaining the broader exterminationist impulse towards ethnic Armenians. The destruction of males was a necessary, but not a sufficient, condition for the expression of that impulse. The generalized nature of the severe atrocities inflicted on the entire Armenian population places such genocides in a different category than, say, the Balkans wars of the 1990s -- though there are superficial similarities, both in the genocidal massacres of males and the forced deportation of women, the elderly, and children. 
The Jewish holocaust under Nazi rule and occupation similarly represented an attempt to eliminate an entire people. Again, though, the "gendercide" framing sheds new light on Nazi procedures. Daniel Goldhagen has examined Einsatzgruppen killing operations on the eastern front, which accounted for some two million Jewish lives before the main apparatus of death camps and "work" camps was fully operational. He points out in Hitler's Willing Executioners how this "up-close," intimate killing of manifestly defenseless civilians was incrementally managed according to gender:
The Einsatzgruppen officers ... could habituate their men into their new vocation as genocidal executioners through a stepwise escalation of the killing. First, by shooting primarily teenage and adult Jewish males, they would be able to acclimate themselves to mass executions without the shock of killing women, young children, and the infirm. According to Alfred Filbert, the commander of Einsatzkommando 9, the [execution] order from Heydrich "quite clearly" "included also women and children." Yet, "in the first instance, without a doubt, the executions were limited generally to Jewish males." By generally keeping units' initial massacres to smallish numbers (by German standards) of a few hundred or even a thousand or so, instead of many thousands, the perpetrators would be less likely to become overwhelmed by the enormity of the gargantuan bloodbaths that were to follow. They also could believe that they were selectively killing the most dangerous Jews, which was a measure that they could conceive to be reasonable for this apocalyptic war. Once the men became used to slaughtering Jews on this sex-selective and smaller scale, the officers could more easily expand the scope and size of the killing operations.
In the early weeks of these murder campaigns, the Einsatzkommandos, again according to Goldhagen, "were the equivalent of genocidal scouting parties, developing the methods of killing, habituating the perpetrators to their new vocation and, generally speaking, working out the feasibility of the overall enterprise."(65) Acts of gendercide can be seen in such cases as a vanguard for the genocide as a whole, an initial barrier to be surmounted and "threat" to be removed, before the remainder of the community is consigned to violent death.(66)
The development of the mass killing of Jews and others on the Eastern Front by poison gas, in specially-designed vans, was also apparently a response to the reluctance some executioners felt to killing women. Christopher Browning is emphatic on this point: "Faced with the complaints ... about the psychological burden on the men of killing women and children, Himmler ordered the search for alternative killing methods that led to the development of the gas van."(67) But though they would subsequently be disproportionately targeted for this less "stressful" form of slaughter -- as far as the perpetrators were concerned, anyway -- women were not the first to be gassed by the Nazis, either in vans or in gas chambers. The victim generally selected for early tests of poison gas was the male Soviet prisoner-of-war. This seems to have been true both on the Eastern Front(68) and at the first gas chamber in Auschwitz.
The link between the gendercide against Soviet POWs and the Jewish holocaust may extend further still. In passages that are nothing short of revelatory, Christian Streit has pointed out that the very infrastructure and  techniques of the death camps were originally developed to enslave and exterminate Soviet POWs, not Jews:
Two large groups of Soviet prisoners were involved. The first comprised those prisoners who were selected and executed as "politically intolerable." Before the end of December 1941 at least 33,000 such prisoners had been executed in the concentration camps of the Reich and the General Government [in occupied Poland]. The second group consisted of those Soviet POWs who had been allotted to Himmler as slave labourers in the SS enterprises. The decision to turn these POWs into Himmler's slaves also resulted from the basic decision to brush aside international law in the war against the Soviet Union. ... Repeatedly during the summer of 1941, and starting with the a convoy of several hundred in July, groups of Soviet prisoners of war, who had been selected as "intolerable," had been taken to the Auschwitz concentration camp to be executed there. To ease the mental strain of the shooting squads and to save costs and energies the executors soon started looking for a simpler method. It was probably deputy commander Karl Fritzsch who experimented in early September with a pesticide, Zyklon B, to murder some 600 such prisoners and another 250 camp inmates who had been selected as "unfit for work." After more such "test gassings" -- there were at least two more convoys of Soviet prisoners among the victims, one numbering 900 men -- the gassings of Jewish victimsCategory: Yugoslav genocide