by Johan M.G. van der Dennen



Plan of the book

Definitions of Genocide, Politicide, and Democide

Controversies: state terror versus genocide
Controversies: genocide versus war crimes
Controversies: intentionality
The salience of the state in genocide
The concept of politicide
The concept of democide
Is there an entity called 'cultural genocide'?
The concept of ethnic cleansing
The concept of massacre
Remaining ambiguities

Typologies of Genocide
Instrumental (pragmatic) vs. transcendental (ideological) genocides
Motivational classification of genocides

Genocide as a Crime of State
Why states victimize groups
Genocide as abuse of state power. . .
. . . versus weakness of state control


Empirical investigations of geno/democide

Theories and explanations of Genocide
Is 'conflict' sufficient explanation?
So what is the explanation - is there a simple explanation anyway?
Lack of theories of genocide
Value-added models
* Activators of genocidal movements:
* Symptoms of genocidal movements:
Continuum of destruction
Redirected hostility and scapegoating

Idiosyncratic (Non-Orthodox) Explanations and 'Theories'
Elimination of 'surplus people'
Fear of death as an instigator
Functionalist explanation
Human destructiveness: aggression theories
Human destructiveness: fear and the 'paranoid potential'

Fein's subterfuge: "killing on state order needs no special explanation"
Criticism of the scapegoat and political opportunity theories

The Social Psychology of Genocide: Normality of the Perpetrator
Ordinary men
Self-serving bias
Political psychology: individuals as strategists
* Genocidal movements
* Psychodynamics and ideological transformation
* Redistribution of hatred
* 'Bribery' into genocide
Obedience and conformism
Ethnocentrism and perceived injustice
The mechanisms of ingroup/outgroup differentiation (experimental evidence)

Social-Psychological Mechanisms and Dynamics of Distancing
Distancing device: assuaging guilt
Distancing device: suspension of responsibility
Distancing device: pluralistic ignorance
Distancing device: anonymity
Distancing device: habituation and learning by participation and doing
Distancing device: compartmentalization
Distancing device: moral drift
Distancing devices: euphemisms and metaphors
Distancing device: psychic numbing
Distancing devices: 'just world' thinking and blaming the victim
Distancing devices: authorization and 'diffuse sanctioning'
Distancing devices: routinization and bureaucratization
Distancing device: normalization of 'evil'

The Ultimate Distancing Device: Dehumanization

Is Ideology Necessary for Genocide to Occur?

Distancing Devices II: Rituals of Degradation

The Genocidal Process

Valentino's (2000) 'Strategic' Perspective on Mass Killing


by Johan M.G. van der Dennen, University of Groningen


Etymologically, 'genocide' means 'group killing': the Greek root 'genos' meaning 'lineage', 'family', 'tribe', 'people', 'ethnie' or 'race', and the Latin root '-cide' (from the verb caedere) meaning killing (as in suicide, homicide, infanticide, etc.) (Diamond, 1992; Cf. Kuper, 1982; Staub, 1989).
The word may be new, Kuper (1982) observed, but the crime is ancient.
The practice of exterminating entire populations is documented in the oldest historical sources. The biblical Hebrews practiced it (under the supervision of their 'God of Wrath'), as did most other ancient peoples. We all know the biblical account of how the walls of Jericho came tumbling down at the sound of Joshua's trumpets. Less often quoted is the sequel: Joshua obeyed the Lord's command to slaughter the inhabitants of Jericho as well as of Ai, Makkedah, Libnah, Hebron, Debir, and many other cities. This was considered so ordinary that the Book of Joshua devoted only one phrase to each slaughter, as if to say: of course he killed all the inhabitants, what else would you expect? (Diamond, 1992).
When we consider early literate civilizations, written records testify to the frequency of genocide. Their wars proceeded to a common end: the slaughter of the defeated irrespective of sex, or else the killing of the men and enslavement and/or massive rape of the women.
One recalls the more horrifying genocidal massacres, such as the terror of Assyrian warfare in the eighth and seventh centuries B.C., when many cities were razed to the ground and whole populations carried off or brutally exterminated, until the Assyrian empire itself became the victim of its own wars of annihilation; or the 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); or the Roman obliteration of Carthage, men, women and children, the site of the devastated city sown with salt, symbolic of desolation (Kuper, 1982).
One can only conclude that the razing of cities and the slaughter of peoples were not isolated episodes in ancient times, and that, contrary to popular belief, genocide is not a 20th-century phenomenon (it may not even be a phenomenon confined to 'civilized' peoples, or even the human species, as will be seen later on [Part 4]).
The names of Genghis Khan and of Timur Lenk (Tamarlane) have become synonyms for the genocides and massacres of a later period. The Mongols in the thirteenth century, though they were eventually absorbed by Chinese culture, it was not before they had carried out the greatest genocide in history: an estimated forty million Chinese were systematically slaughtered by Genghis Khan's soldiers to depopulate the northern areas of the country and free them for nomadic herding. And Iraq, which was visited by the Mongols for only two years, in 1258-60, was so thoroughly devastated that its population did bot recover to the pre-Mongol level until this century (Dyer, 1985: 46).
Also the war practices of 'civilized' peoples in the Middle Ages were often marked by genocidal massacres (see e.g., Walker, 1899; Cohn, 1967).
The Crusades against the Muslim 'infidel' and for the recovery of the holy places almost all started with pogroms against Jews. And the course of the Crusades was marked by the slaughter of Jewish communities. Thus, when the crusading army had finally taken Jerusalem by storm in the year 1099 and had massacred the Muslims, they drove the Jews into a synagogue, set fire to it and burnt all within its walls (Graetz, 1894; Vol.3).

In our own era, there are the genocidal conflicts linked to the march of colonization and to the process of decolonization. Genocide and colonization were always closely linked (Scherrer, 1998a,b). Advanced technology facilitates the obliteration of whole communities in the course of international warfare. But the major arena for contemporary genocidal conflict and massacre is to be found within the sovereign state: it is particularly a phenomenon of the plural society (Kuper, 1981, 1982).
Kuper views plural societies (also called 'multi-ethnic', 'divided' or 'ethnically stratified societies') as the seedbed of genocide, especially in the contemporary period in which contending tribal groups struggle for domination of the state after decolonization. Kuper stresses that genocide is not an inexorable outcome of any social structure but is a result of decisions people make; one needs to study the constraints inhibiting genocide as well as the structural conditions and polarizing processes making it more likely.
Colonization, in its arbitrary delineation of metropolitan domains has been a great creator of plural (i.e., multi-ethnic) societies, and there have been many genocides in the process of decolonization or as an early aftermath of decolonization. The struggles for power between Hutu and Tutsi in Rwanda and in Burundi, and between mainland Africans and Arabs in Zanzibar, became genocidal. Partition, as in India, and repression of secessionary movements, as in Bangladesh, and some would say in Nigeria, have taken a genocidal form (Kuper, 1982).
In colonial times, revolts, or even riots, were often suppressed with great destruction of life and property, or by admonitory massacres. The slaughter of the Herero by the German rulers of South West Africa was among the most exterminatory and horrifying of the reprisals for rebellion. The genocide of the Herero occurred between August 1904 and mid-1905, and was a systematically organized military offensive concentrated into a single effort. By 1905 the Herero population was reduced from 80,000 to 16,000, mostly ill and weak children, women and old people (Palmer, 1998).

Legacies of colonialism led to genocidal atrocities in different parts of the world: e.g., Indonesians against Timorese; Indonesians against Papuans on Irian Jaya; South American regimes against indigenous peoples such as the Aché in Paraguay, Maya in Guatemala and Yanomamö and other low land Indian people in the Brazilian Amazon region (Scherrer, 1998b).

Hunting and gathering peoples have often been the victims of genocidal attacks, as for example the San (Bushmen) in Southern Africa, or many Indian groups in the U.S.A. and Latin America, or the systematic annihilation of Aborigines in Tasmania (Kuper, 1982). The English eliminated the native population of Tasmania in seventy-three years, and the Dutch who settled the Cape of Good Hope treated the native blacks and Khoisan (whom they called 'bobbejaans', i.e., 'baboons') as 'dangerous vermin' on a par with wild animals (Renwick Monroe, 1995; Palmer, 1998). The destruction of the Aborigines of Queensland lasted between approximately 1840 and the later 1880s, was piecemeal and predominantly a frontier activity, claiming between 8,000 and 10,000 lives (Palmer, 1994, 1998). This same characteristic is valid for many contemporary Amazonian Indian societies (Vide infra: section on developmental genocides). Wherever civilization advances, it spells the doom of the 'non-civilized' societies.

However, the role of religious ideas as warrant for, or stimulus to, genocide, has now been taken over by totalitarian political ideologies, of absolute commitment to the remaking of society in conformity with radical specifications, and a rooting out of dissent, as extreme as in the Inquisition. The major examples of the genocidal potentialities of these ideologies in our day are provided by the Nazi regime with its conception of a brave new world of racially tolerated and ordered societies under German hegemony; the Soviet regime, Under Stalin, with the Gulag Archipelago receiving, as a sort of 'rubbish bin of history', the successive blood sacrifices of the communist utopia; and the recent Pol Pot regime in Cambodia, freely and righteously exterminating in total dedication to a starkly elemental blueprint for living (Kuper, 1982).

In international warfare, technological change facilitates genocidal massacre, as in the bombing of Hiroshima and Nagasaki. The development of long-distance lethal weapons is a very obvious 'distancing device': "These make killing easy... Very few aircrews who are willing, indeed, eager, to drop their bombs 'on target' would be willing to strangle, stab, or burn children (or, for that matter, adults) with their own hands" (Tinbergen, 1968). They would very probably stop short of killing in response to the distress, fear and pain signals of their opponents. An American marine serving in the Vietnam war commented that battlefield "ethics seemed to be a matter of distance and technology. You could never go wrong if you killed people at long range with sophisticated weapons" (Caputo, A Rumor of War, 1977). This is an example of 'distancing' (in this case literally creating physical or geographical distance between perpetrator and victim), which plays a major role in the justification of violence, as will be seen later on.

The victims of genocide or genocidal massacre must be selected because they belong to a group, whether or not each victim as an individual has done something to provoke killing. As for the defining group characteristic, it may be racial (white Australians killing black Tasmanians), national (Russians killing fellow white Slavs, the Polish officers at Katyn in 1940), ethnic (the Hutu and Tutsi, two black African groups, killing each other in Rwanda and Burundi in the 1960s, 1970s and 1990s), religious (Moslems and Christians killing each other in Lebanon in recent decade), or political (the Khmer Rouge killing their fellow Cambodians from 1975 to 1979).
Most genocides in this century have been perpetrated by nation-states upon ethnic minorities living within the state's own borders; most of the victims have been children. The people responsible for mass murder have by and large gotten away with what they have done. Most have succeeded in keeping wealth that they looted from their victims; most have never even faced trial (Simpson, 1993).
On first thought, one might expect that no horror could grip public attention as much as the intentional, collective, and savage killing of many people. In reality, genocides rarely grip the public's attention in other countries, and even more rarely are interrupted by foreign intervention.
Is our puzzling nonresponse because we did not know, or could not find out, about ongoing genocides? Certainly not: many genocides of the 1960s and 1970s received detailed publicity at the time, including those in Bangladesh, Brazil, Burundi, Cambodia, East Timor, Equatorial Guinea, Indonesia, Lebanon, Paraguay, Rwanda, Sudan, Uganda, and Zanzibar (Diamond, 1992).
Plan of the book

In this, and the following chapters, we shall attempt to review the literature on genocide, democide, politicide, massacres, war atrocities, and gross human rights violations, especially definitions and concepts of genocide, typologies of genocide, and theories and explanations of genocide. We also intend to answer, at least in part, the following vexing problems and burning questions such as:
* What are the preconditions of genocide?
* What are the psychological motives and mechanisms?
* Are perpetrators disturbed (psychopathological) or normal?
* What is the role and psychology of 'innocent bystanders', etc.?
* What are the distancing devices and facilitating conditions?
* What, especially, is the role of dehumanization and ethnocentrism?
* What is the explanatory power of concepts like obedience, etc.?
* Who are the victims of genocide?
* Does genocide occur in preindustrial societies (primitive peoples)?
* Did genocide occur in prehistory?
* Does genocide occur in other species?
* Why does genocide occur at all?
* Where does cruelty come from?
* What is 'evil'?

After the inventory of the pertinent literature (from all kinds of sources and disciplines), we shall present some 'insights' from evolutionary psychology and sociobiology which might contribute to the explanation and understanding of these most heinous, terrifying, and disgusting human behaviors.

Definitions of Genocide, Politicide, and Democide

As in every definitional debate, the debate on the definition of genocide knows its particulars and moderates, its lumpers and splitters.
One of the well-known lumpers is Charny. Charny (e.g., 1994) fervently argues for a generic definition of genocide that does not exclude or commit to indifference any case of mass murder of any human beings, of whatever racial, national, ethnic, biological, cultural, religious, and political definitions, or of totally mixed groupings of any and all of the above. Most of the scholars studying genocide appear to prefer a more specific definition, however.

The legal definition of genocide is the following:

In the present convention, genocide means any of the following acts committed with intent to destroy, in whole or in part, a national, ethnical, racial, or religious group, as such: (a) Killing members of the group; (b) Causing serious bodily or mental harm to members of the group; (c) Deliberately inflicting on the group conditions of life calculated to bring about its physical destruction in whole or in part; (d) Imposing measures intended to prevent births within the group; (e) Forcibly transferring children of the group to another group (Article 2, United Nations, Yearbook of the United Nations 1947-48. New York: 1949. pp. 595-599; and Yearbook of the United Nations, 1948-49. New York: 1950. pp. 958-959).

Long before the United Nation's definition of genocide, Raphael Lemkin, in his work on Axis Rule in Occupied Europe (1944), gave a succinct account of what genocide is: "Genocide is effected through a synchronized attack on different aspects of life of the captive peoples: in the political... social... cultural... economic... biological field...; in the field of physical existence...; in the religious field...; in the field of morality...".
A key component in the calculus of genocide was, in Lemkin's own words, "the criminal intent to destroy or to cripple permanently a human group. The acts are directed against groups, as such, and individuals are selected for destruction only because they belong to these groups".

Horowitz (1976) amended the UN definition to emphasize that genocide was "a structural and systematic destruction of innocent people by a state bureaucratic apparatus". Since then, Horowitz has concluded that a totalitarian society is a necessary precondition for the genocidal process, but it is not a sufficient one. Horowitz believes that national culture plays a much more important role in genocide than the ideology of the state. A totalitarian ideology may make class, race, or religion lethal sins, he contends, but the decision to eradicate these sins by committing genocide is largely a function of culture (Horowitz, 1984).

In 1984, Fein proposed a general definition of all types of genocide, as

the calculated murder of a segment or all of a group defined outside of the universe of the perpetrator by a government, elite, staff or crowd representing the perpetrator in response to a crisis or opportunity perceived to be caused or impeded by the victim. Crises and opportunities may be a result of war, challenges to the structure of domination, the threat of internal breakdown or social revolution and economic development... Motives may be ideological, economic, and/or political... Genocides, as are other murders, may be premeditated or an ad hoc response to a problem or opportunity (1984, 4-5).

This has been criticized as inadequate as a definition of genocide (Huttenbach 1988): Fein (1990) agrees: Rather than a definition, she contends, it is an explanatory sketch.
Fein (1992) subsequently defines genocide as "sustained purposeful action by a perpetrator to physically destroy a collectivity directly or through interdiction of the biological and social reproduction of group members, sustained regardless of the surrender or lack of threat offered by the victim".
The Oxford English Dictionary defines genocide even more narrow than the legal deinition as "the deliberate and systematic extermination of an ethnic or national group".
Chalk (1990) and Chalk & Jonassohn (1990) define genocide as "a form of one-sided mass killing in which a state or other authority intends to destroy a group, as that group and membership in it are defined by the perpetrator". They tend to regard genocide as purposeful action related both to the interests of the state and leading social classes and elites
Genocide, in Kuper's (1982) terms, is a crime against a collectivity, taking the form of massive slaughter, and carried out with explicit intent.
DuPreez (1994) defines genocide very broadly as "the deliberate and intended killing of a category of persons". The main difference between these last definitions and the legal one presented above is that these do not have a priori restrictions on the types of groups to be included or excluded.
According to Palmer (1994, 1998) genocide is "the intention to destroy physically a whole or a substantial part of a group because they are members of that group and whose membership is defined by the perpetrator, regardless of whether or not the whole group is actually destroyed". This clearly includes cases where unintended genocidal consequences of action are noted by the perpetrator, and are continued or increased and where genocide is attempted but fails. The onus of this definition is thus upon the intention to destroy the group, regardless of any further aims which might be achieved by such destruction or the success of the genocide.

As noted by Melson (1997), the UN definition of genocide is both too broad and too narrow for scholarly comparative purposes. It is too broad because it does not distinguish between massacre and progrom and partial genocide, and it is too narrow becausse it limits its scope only to communal groups and excludes economic classes and political movements.
Melson (1997) suggests the following definition of genocide: actions intended to destroy in small part (massacre), in large part, or in whole a social collectivity or category including its culture and identity. This definition, he claims, makes explicit the distinctions between 'genocidal massacre', 'genocide-in-part', and 'genocide-in-whole'.
One writer plainly argues that only the Holocaust fits his narrow definition of 'genocide'. Katz (1994) defines genocide as "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".
Katz departs from the widely accepted UN definition by excluding the partial destruction of groups (genocide-in-part), while widening the scope of the definition to include other than communal categories. Since earlier he noted that only in the Holocaust did the perpetrator intend physically to murder every Jew in the world, it follows from his definition that only the Holocaust is truly a genocide.

Controversies: state terror versus genocide

Fein (1994) has proposed a paradigm for the detection and tracing of genocide which includes the following conditions:

1.    There was a sustained attack or continuity of attacks by the perpetrator to physically destroy group members;
2.    The perpetrator was a collective or organized actor (usually the state) or commander of organized actors;
3.    The victims were selected because they were members of the collectivity;
4.    The victims were defenseless or were killed regardless of whether they surrendered or resisted; and
5.    The destruction of group members was undertaken with intent to kill and murder was sanctioned by the perpetrator.
On the basis of this paradigm, Fein argues that despite the similarities between state terror and genocide, there is a major difference: "victims of terror are selected because they are believed to have committed 'subversive' acts or they are chosen arbitrarily rather than as members of a group as are victims of genocide".

Controversies: genocide versus war crimes

Another area of disagreement is the connection between war crimes and genocide. Kuper (1994) has argued that the atomic bombing of Hiroshima and Nagasaki, the pattern bombing of Hamburg and Dresden, and the firebombing of Tokyo constituted genocide. As both Fein (1994) and Chalk (1994) argue, however, such a perspective conflates the concept of war crime and genocide.
On the other hand, mass bombings can easily be subsumed under the concept of 'mass killing' as defined by Valentino (2000). Vide infra (§ 3.9.).

Controversies: intentionality

Another area of disagreement is intentionality. For Fein (1994), intent can be demonstrated by "showing a pattern of purposeful action" leading to the destruction of a significant part of the targeted group, regardless of the reasons offered by the perpetrator for his actions. Her critics do not understand, she submits, the difference between intent and motive.
For Charny (1994), the intentionality criterion is irrelevant; he uses the term genocide indiscriminately to refer to the "mass killing of substantial numbers of human beings, when not in the course of military action against the military forces of an avowed enemy, under conditions of the essential defenselessness and helplessness of the victims". According to him the quest for discriminatory definitions cannot but lead "into assigning hierarchical value to different kinds of mass death". While Charny properly cautions against some of the dangers of excessive 'definitionalism', his generic labeling of most mass killings as genocide will raise questions concerning its analytical value. Chalk (1994), for example, reasoned that if perpetrator intent is not crucial to the definition of genocide, then all sorts of unintended lethal consequences of human action would be classified as genocide.

Charny (1994) proposed the following definitional matrix for genocide:

A. Generic Definition of Genocide
    1. Genocidal Massacre
    2. Intentional Genocide
        a. Specific Intentional Genocide
        b. Multiple Intentional Genocide
        c. Omnicide (term proposed by Somerville, 1983)
    3. Genocide in the Course of Colonization or Consolidation of Power
    4. Genocide in the Course of Aggressive ('Unjust') War
    5. War Crimes against Humanity
    6. Genocide as a Result of Ecological Destruction and Abuse
B. Accomplices to Genocide
C. 'Cultural Genocide'
    1. Ethnocide
    2. Linguicide (term proposed by Rudnyckyj, 1984)

According to Wallimann & Dobkowski (1987), it is increasingly difficult to locate the issue of intentionality on the societal level "because of the anonymous and amorphous structural forces that dictate the character of our world". They argue that since we are living in a world in which "individuals are dominated by anonymous forces such as market mechanisms, bureaucracies, and distant decision making by committees and parliaments, the emphasis on intentionality appears anachronistic".
The idea that "only intentional or planned massive destruction of human lives should be called genocide", Walliman & Dobkowski (1987) contend, leads to "the neglect of those processes of destruction which, although massive, are so systematic and systemic, and that therefore appear so 'normal' that most individuals involved at some level of the process of destruction may never see the need to make an ethical decision or even reflect upon the consequences of their action".
It is but a short jump from the arguments of Dobkowski, Walliman, and similar ones of Barta (1987) to those of Seamus Thompson, a sociologist studying Northern Ireland, for whom genocide is "inherently a continuous variable". In Thompson's view, "the search for a clear dichotomous definition of genocide may be a manifestation of... [the] general phenomenon of avoidance". Our definitional attempt to mark the boundary between 'genocide' and 'not genocide' is intellectually unprofitable, he asserts, because "It renders impossible a research finding that incremental steps, involving much normal behavior, and quite conceivable deviations from conventional ideals, can move us to genocide. In short, it cuts us off from the possibility of the understanding that we so earnestly claim to seek through it" (quoted in Chalk, 1994).
Genocide, Chalk (1994) states succinctly, is primarily a crime of state, and an intentional crime at that. There is no evidence that genocide is a "continuous variable". Systemic variables facilitate genocide, but it is people who kill. (italics added).

The salience of the state in genocide

Another major issue which emerges in the recent literature concerns the identity of the perpetrator: are genocides committed when small, relatively isolated aboriginal groups are destroyed by perpetrators acting as individuals _ peasants, ranchers, miners, rubber tappers, lumberjacks and land speculators _ and are 'permitted' (or tacitly condoned) by governments that either cannot or will not stop the process?

Horowitz (1980) reasoned that genocide is not simply a sporadic or random event such as the Katyn Forest Massacre in which 15,000 Polish troops were presumed to have been destroyed by the Red Army during World War II (Zawodny, 1962). In addition to its systematic character, genocide must be conducted with the approval of, if not direct intervention by, the state apparatus. Genocide is mass destruction of a special sort, one that reflects some political support base within a given ruling class or national group. This contrasts sharply with vigilantism, which represents the maintenance of order without law, or some kind of mass participation without corresponding state support (Rosenbaum & Sederberg, 1976). Genocide is also quite distinct from other types of violence, such as tyrannicide in which murder is perceived not as a crime but as a liberating act. These are not absolute categories. There is a slim line, according to Horowitz, between systematic and sporadic destruction.
Others (e.g., Palmer, 1994, 1998), however, point out that in certain circumstances the state may not be overtly involved in genocide but might give tacit consent to the genocidal conduct of local authorities, or the private initiatives of 'frontier-of-civilization peoples'. Consequently, she suggests that colonial genocide is not a convincing category, and that a distinction between state and societal genocides is more useful.

Bartov (1996) introduced the terms 'industrial killing' and 'militarized genocide' to refer to the mechanized, impersonal, and sustained mass destruction of human beings, organized and administered by states, legitimized and set into motion by scientists and jurists, sanctioned and popularized by academics and intellectuals. According to him, the First World War had introduced to the West a whole new concept of (total) war with its mass slaughter in the trenches. Bartov argues that the genocide of the Jews was not only a perfection and extension of the industrial killing of World War I, but also an enterprise that could have been accomplished only by a highly modern, disciplined, bureaucratic society, in which people had respect for law and order, science and technology, that is, a society very much like the Western industrialized states in which we now live.

The concept of politicide

Harff & Gurr (1988, 1995; Cf. Gurr & Harff, 1994) discriminate victims of 'genocide' from those of 'politicide': communal and group victims are distinguished from real and alleged political opponents, including those in rebellion.
The essential quality of all the collective violence episodes accumulated by Harff & Gurr (see the section on 'Crime of State') is that the state or dominant social groups make a concerted, persistent attempt to destroy a communal or political group, in whole or in part. More precisely geno/politicide is defined by Harff & Gurr (1995) as the promotion, execution and/or implied consent of sustained policies by governing elites or their agents - or in the case of civil war either of the contending authorities _ that result in the deaths of a substantial portion of a communal, political, or politicized communal group.
In genocides the victimized groups are defined primarily in terms of their communal characteristics. In politicides, by contrast, groups are defined primarily in terms of their political opposition to the regime and dominant groups.

The concept of democide

As an analogous concept for public murder, that intentionally done by government agents acting authoritatively, Rummel (e.g., 1996) offers the concept of democide (from Greek demos [äçìoò] or people). Unlike the concept of genocide, it is restricted to intentional killing, and does not extend to attempts to eliminate nations, races, or religions by means other than killing members of the group. Moreover democide is not limited to genocide, nor to politicide, mass murder or massacre, or terror. It includes them all and also what they exclude, as long as such killing is a purposive act, policy, process, or institution of government. In short, democide is government murder.
Rummel deviced his concept of 'democide', Fein (1994) criticized, because he does not discrimate victims of terror from victims of genocide.

Is there an entity called 'cultural genocide'?

If we broaden our approach to include an entity called 'cultural genocide', the results might be counterproductive. Horowitz (1980) believes that a restrictive, rather than an omnibus concept of genocide, is the most operationally valid.
Genocide means the physical dimemberment and liquidation of people on large scales; an attempt by those who rule to achieve the total elimination of a subject people. Genocide does not mean simply depriving people of their cultural heritage. This might better be called ethnocide (e.g., DuPreez, 1994).
Some scholars have argued that ethnocide (i.e., the suppression of the culture, religion, and language of the targeted group leading to its forced assimilation with the dominant group, or the attempt to wipe out the ethnic or cultural identity of the targeted group) is a form of genocide. This, however, seems to be a minority view.

The concept of ethnic cleansing

Bell-Fialkoff (1993) has reasoned that ethnic cleansing defies easy definition. At one end it is virtually indistinguishable from forced emigration and population exchange while at the other it merges with deportation and genocide. At the most general level, however, ethnic cleansing can be understood as the expulsion of an 'undesirable' population from a given territory due to religious or ethnic discrimination, political, strategic or ideological considerations, or a combination of these. Under this defnition, then, the slow dispersal and annihilation of North America's indigenous population was indeed ethnic cleansing. Population removal and transfer have occurred in history more often than is generally acknowledged.
The earliest example was cleansing carried out by Tiglath-Pileser III (745-727 BC), the first Assyrian ruler to make forced resettlements a state policy. Under his reign about half the population of a conquered land would be carried off, and its place taken by settlers from another region.
As part of a general process toward greater homogeneity within states that began in the Middle Ages, 'ethnic' cleansing took on medieval notions of religious purity, targeting minorities of 'nonbelievers'. In the Middle Ages, massacre and expulsion were the most common methods of religious cleansing, which tended to target Jews, the onley sizable minority in most countries.
With the profound secularization of the modern world, cleansing later manifested itself in political ideology, namely as part of communism and fascism.
Nationalism, too, as a kind of modern religion, contains quasi-spiritual aspects that lend to its most extreme manifestation a desire to 'purify' the nation of 'alien' groups.
Although still couched in religious terms, the dirst cleansings based on ethnic discrimination were carried out by England. In the 1640s and 1650s, when war and plague swept away half the Irish population, England (Cromwell) seized the opportunity to expel most of the remaining Irish Catholics from Ulster.
It was only in the nineteenth century that the complete destruction of an ethnic group manifested itself as the goal of a state, when Turkey began directing cleansing efforts against Greeks and Armenians. Having come to view those minorities as enemies within, the Turkish sultan Abdul Hamid II encouraged Kurdish depredations on Armenian villages until hostilities grew into a veritable war. By 1894 Turkish regular troops had joined with the Kurds, and about 200,000 Armenians were killed. In the 1915 holocaust, Armenians lost an estimated 1.5 million people - more than half their population - as well as about 90 percent of their ethnic territory.
It was with the Nazi campaigns against Jews that ethnic cleansing reached its height: annihilation. The Nazi campaigns were an ethnic cleansing in the sense that they were intended to remove Jews from territories of the Reich. The German term Judenrein, 'clean of Jews', which was used to designate areas from which all Jews had been deported, testifies to this fact. But the Holocaust was much more. It combined elements of deportation, expulsion, population transfer, massacre and genocide. In that way it was 'complete': truly a final solution. Altogether about six million European Jews were murdered between 1933 and 1945. About 250,000 Gypsies and an equal number of homosexuals were also killed by the Nazis.
After World War II, what was probably the largest and most sweeping ethnic cleansing in history took place: the removal of ten to twelve million Germans from Eastern Europe. About 2.1 million of these died from a combination of war, hunger, cold and disease.
Within its borders, the Soviet Union cleansed about 600,000 people from regions that had proved themselves 'unreliable' in the war, such as the Kalmyk and Checheno-Ingush republic. Twentieth-century communist ideology introduced yet another type of cleansing, that of economic class. The destruction of propertied classes in Stalinist Russia or Maoist China bore all the markings, including vocabulary, of an 'ethnic' cleansing.
Only about fifty years ago - that is within the lifetime of an individual _ Croatian nationalists carried out massacres of Serb civilians in a Nazi puppet state comprising most of today's Croatia and Bosnia-Herzegovina. The Ustashi, as these nationalists were known, regarded Croatia's more than two million Serbs as a threat to national integrity. The Croatian minister of education, for example, speaking at a banquet in June 1941, remarked that "one-third of the Serbs we shall kill, another third we shall deport and the last third we shall force to embrace the Roman Catholic religion and thus meld them into Croats". It is estimated that the Ustashi killed between 300,000 and 340,000 Serbs. The Serbs took revenge after the war by killing 100,000 Croats.

The concept of massacre and mass killing

Carlton (1994) defined 'massacre' as "the indiscrimate killing of unresisting or defenceless people". It can be subsumed, he asserts, under the more general term 'atrocity', which normally refers to some horrible or cruel act. Massacres cannot be classified in terms of any single factor.
Carlton distinguishes and describes in some detail the following categories:
* Massacre as a military norm: Mzilikazi and the Matabele;
* Massacre by 'consent': Mass killing as a sacrificial offering;
* The Greek experience I: Massacre as a military postscript;
* The Greek experience II: Massacre as a political expedient;
* Roman slavery I: Massacre as social retribution;
* Roman slavery II: Massacre as condign punishment;
* Massacre as a spectator sport: The Roman games;
* Massacre as a succession device: Achaemenid Persia;
* Massacre and counter-massacre: Turks, Greeks and Armenians;
* Massacre and economic efficiency: The Stalin era;
* Massacre as fratricide: The Spanish Civil War;
* Massacre as tactical exigency: The choice-of-evils problem;
* Massacre as a matter of policy: The SS at Lidice and Oradour;
* Massacre by proxy: Vichy and the Jews;
* Massacre as a strategic extremity: The Allied bombing offensive 1942-45;
* Massacre and liberation I: The colonial situation;
* Massacre and liberation II: The colonial aftermath;
* and, finally, Massacre and the occult (on witch hunts).

Valentino (2000) defines 'mass killing' as the intentional killing of a significant number (somewhat arbitrarily defined as at least 50,000 intentional deaths over the course of five years) of the members of any group of noncombatants (as the group and its membership are defined by the perpetrator). Victims of mass killing may be members of any kind of group (ethnic, political, religious, etc.) as long as they are noncombatants and as long as their deaths were caused intentionally. This definition does not only include 'direct' killings such as executions and gassings, but also deaths caused in more indirect ways, such as the forced deportation of individuals to regions which lack the resources to support them.
From this perspective, genocide is best defined as a specific subset of mass killing in which the victim group is defined on the basis of its ethnicity or nationality. This definition is similar in some ways to Rummel's concept of 'democide'. Rummel's definition, however, includes the killing of any number of civilians, no matter how small. In addition, Rummel specifies that democide must be carried out by government groups, while the perpetrators of mass killing can belong to any kind of group (Valentino, 2000: 4).
Limiting mass killing to the total annihilation of a group, as Valentino (2000: 5) observes, would create a highly restrictive definition. In fact, by this definition genocide may never have been committed, only attempted (and even the attempt to destroy an entire group has been rare; by this definition, the Holocaust itself would qualify as attempted genocide only if the victim group is defined as 'European Jews', since there is little evidence that the Nazis attempted to kill. for example, Jews living in the United States).

Remaining ambiguities

While collective killing is thus the essence of genocide, one can argue over how narrow a definition to adopt. The word 'genocide' is often used so broadly that it loses meaning altogether. Even if it is to be restricted to large-scale cases of collective killing, ambiguities remain. Here is a sample of the ambiguities (based on Diamond, 1992):

* How many deaths are needed for a killing to count as genocide rather than mere murder? This is a totally arbitrary question. Australians killed all five thousand Tasmanians, and American settlers killed the last twenty Susquehanna Indians in 1763. Does the small number of available victims disqualify these killings as genocidal, despite the completeness of extermination?

* Must genocide be carried out by governments, or do private acts also count? Horowitz (1976 et seq.) distinguished private acts as 'assassination', and defined genocide as "a structural and systematic destruction of innocent people by a state bureaucratic apparatus". However, there is a complete continuum from 'purely' governmental killings (Stalin's purges of his opponents) to 'purely' private killings (Brazilian land-development companies hiring professional Indian killers). American Indians were killed by private citizens and the U.S. army alike, while the Ibos in Northern Nigeria were killed both by street mobs and by soldiers. In 1835 the Moriori of the Chatham Islands were exterminated by a Maori tribe of New Zealand. By Horowitz's definition, this and many other equally well-planned exterminations of one tribal group by another do not constitute genocide, because the tribes lacked a state bureaucratic apparatus (see also section on 'Genocide as a crime of state').

* If people die en masse as a result of callous actions not specifically designed to kill them, does that count as genocide?

* Related to the above question: is 'intent' (as evidenced by a masterplan, purposeful preparation, propanganda and admonition to butcher as by 'Radio Milles Collines') a necessary component of a generic definition of genocide?

* When a provocation is followed by massive retaliation all out of proportion to the provocation (e.g., the French in Algeria; the Dutch in Indonesia), how do we decide when 'mere' retaliation becomes genocide?

* If it is stipulated that genocide involves sustained killing or systematic destruction, a one-sweep genocidal massacre does not, strictly speaking, constitute genocide.

Typologies of Genocide

For the explanation and understanding of genocides it is of utmost importance to draw some basic distinctions and build a sophisticated classification or typology of genocides.

Clay (1989), among many others, distinguished two types of genocide. The first type is "official genocide, carried out by states and directed at groups that, although distinct, had long been part of the society". The second type is "the elimination of small relatively isolated groups on the frontiers of expanding political, social and economic systems". It is the second type, Clay argues, that is far more common in the twentieth century.
Savon (1972) distinguished substitution genocide ("le génocide de la faim"), devastation genocide ("le génocide de la peur"), and elimination genocide ("le génocide de l'humiliation").
Kuper (1981, 1982) distinguished genocides against hostage or scapegoat groups; genocides following decolonization of a two-tier system of domination; genocides against indigenous peoples; and genocidal massacres (such as Hiroshima). Later on, Kuper (1994) draws a basic distinction between 'domestic' genocides and genocides committed in the course of international war. The main elements in the scapegoat type of genocide are, first, an identifiable and differentiated group within a society. Where there are problems in recognizing members of this group, a solution may be found by the use of devices to ensure identifiability. Usually, perhaps invariably, hostile stereotypes are projected onto the victims, and vilifying propaganda directed against them. These often take the form of 'dehumanizing' the target group. Vulnerability seems to be an essential element: the group is an easy prey. The ability of a group to defend itself, and to exact reprisals, is of course some guarantee against genocide. Commonly, there are material advantages to be gained - funds for war or other projects, elimination of competition in trade, distribution of spoils to one's followers, loot for the mob. But the group serves also as scapegoat, as diversion, as surrogate (Kuper, 1982).
Dadrian (1974, 1975) lists the following forms: (1) cultural; (2) violent-latent (that is, genocide as a by-product of other operations); (3) retributive (either punitive or admonitory); (4) utilitarian (Kuper thinks this overlaps with others of his categories); and (5) optimal (massive, relatively indiscriminate, sustained and aiming at total obliteration).
Fein (1984; 1990) distinguished ideological, retributive, developmental, and despotic genocides.
R. Smith (1987) distinguished ideological, retributive monopolistic, utilitarian, and institutional genocides.
Chalk & Jonassohn (1990) distinguished genocides to implement a belief, ideology or theory; genocides to eliminate a real or potential threat; genocides to acquire economic wealth; and genocides to spread terror among real or potential enemies.
The first type of genocide in history, according to Chalk & Jonassohn, was used to build and maintain empires. These are illustrated in Carthage (by Rome), by the Mongols, by Shaka (creating the Zulu Empire) and by European settlers in Tasmania, the Amazon, and South West Africa. Characteristically, the victims are defeated peoples or enemies - alien to the perpetrator or viewed as alien (indigenous peoples). These are to be discriminated from genocides to acquire wealth, land or resources and those which serve primarily to spread terror among real or potential enemies. Lastly, Chalk & Jonassohn focus on genocides committed to implement a belief, ideology or theory which they view as preeminently modern, including imaginary groups as well as real groups in studying this type and its precursors. The victims include the Cathars (13th century), Knights of the Temple (14th century), the victims of the Great Witch Hunt (17th century), Armenians (1915), Jews and Gypsies (1939-1945) and Cambodians (1975-1979).
Chalk & Jonassohn thus conclude that the first genocides in history arose from attempts in antiquity of imperial powers to destroy their recurrent enemies; best known of these is the destruction of Carthage. After the Romans had won the Punic Wars, which lasted from 264-146 BC, they massacred the Carthaginians and destroyed their city completely.
It does not seem very likely, according to Chalk & Jonassohn, to assume that early hunter-gatherers practiced genocide. Because people lived in small kin- or extended family groups and overall population densities were very low, there was little reason for exterminating other peoples at that time.
After the agricultural revolution, however, the situation changed dramatically: it divided the world into two groups of peoples; the settlers, living in villages, and the nomads, requiring vast spaces for their roaming herds. In times of hardship the nomads quickly learned to attack the stores of the settlers in order to obtain food. They were not interested in exterminating the settlers because they had the intention to repeat their raids in the future. The settlers, on the other hand, had better reasons to exterminate the nomads, but they had neither the means nor the skills to do so.
After the rise of the (city)state and the first empires, when the agriculturists produced surpluses and were able to support cities, armies, priests and rulers, armed conflicts arose over territory, wealth, trade and trade-routes. In this situation, the only way to assure a stable life for the victorious group was to eliminate the defeated party immediately. Genocide for this reason, elimination of a potential future enemy, appears to be the first reason for genocide in history.
The terror-spreading and deterrence type of genocide distinguished by Chalk & Jonassohn, which concerns terrorizing real or potential enemies, was probably invented later. The Assyrians must most likely be credited for discovering that the spreading of terror is as effective and efficient an instrument, or even more so, than expensive wars themselves.
Surveying such wars in antiquity, Lerner (1986) infers slavery arose from the successful separation of conquered peoples, slaughtering the males and incorporating the females into the nation of the conqueror. Changes in sex roles and the patriarchal organization of society now makes women more vulnerable to genocide than ever before, R. Smith (1989) observes. In premodern times, women belonging to enemies defeated in war were enslaved and raped as they were valued for their reproductive power and could be incorporated in a new society, isolated from social participation and power. But women in the twentieth century have been both perpetrators and victims of genocide (Fein, 1990; vide infra).

Harff & Gurr (1987) distinguished genocides with victims defined communally; politicides against politically-active communal groups; and episodes with mixed communal and political victims.
Harff & Gurr (1988, 1995) further distinguish between two types of genocides, hegemonical genocide and xenophobic genocide. In the former the primary motive of the ruling group is to subordinate a communal group by killing enough of its members that the survivors have no will or capacity to resist, whereas in the latter, elite ideology calls for the elimination of the 'offending' communal group. (They do not consider 'ethnocide').
Harff & Gurr furthermore distinguish four variants of politicide:
(1) A common variant is repressive politicide, in which ruling groups retaliate against adherents of political parties, factions or movements because of their support for oppositional activities. Common tactics are secret operations in which Communist sympathizers are executed or murdered, for example in Argentina, El Salvador, Indonesia and India. Other cases have taken place in African states in which newly-empowered leaders use extreme and deadly repression against any and all groups suspected of opposition.
(2) Repressive/hegemonical politicides are similar except that the political opposition often coincides with the victimized group's communal identity.
(3) Revolutionary mass murder is another common type of politicide. Here new regimes committed to bringing about fundamental social, economic and political change kill those perceived as standing in the way to achieve such ends. Marxist-Leninist regimes which came to power through protracted armed struggles provide most of the post-1945 examples (e.g., Campuchean killing fields under the Pol Pot regime).
(4) The least common type of politicides is retributive mass murder. In these cases subordinate or oppositional groups seize power and kill their former masters/oppressors in an act of vengeance. Examples are the Hutu rulers killing their former Tutsi masters in Rwanda (1963-64) and the Pinochet regime retaliating against leftist supporters of Allende. The latter is an example of counter-revolutionary politicide.
According to Melson (1997) and Scherrer (1998a,b) genocide means actions carried out by a state or ruler with the intent to systematically kill a particular community of people or social collectivity, resulting in destroying the targeted group in whole or in part. Full-scale or total genocide means that the perpetrators were aiming at the complete extermination and total destruction of a particular community of peoples. Partial genocide means that the perpetrators aimed at destruction-in-part of a particular community or group of people in order to dominate the group.
By combining the dimensions total-partial and domestic-foreign, Melson (1997) arrives at a simple four-fold typology:

1. total-domestic-genocide  

2. total-foreign-genocide  

3. partial-domestic-genocide  

4. partial-foreign-genocide  

Melson (1997) combined the distinction of scale (total/partial) with the equally obvious distinction of place (domestic/foreign), resulting in a 2x2 matrix. An obvious third distinction and sort of third dimension would, according to Scherrer (1998), be the type of perpetrator (state/nonstate actor). Scherrer additionally includes other types of mass murder (such as progroms and massacres) to be distinguished from total or partial genocide, resulting in a 3x4 matrix.

     Genocide and mass murder  

Genocide and mass murder  

Full-scale genocide  

Partial genocide  

Mass murder (Pogroms, massacres)  

Domestic/state actors  

extreme worst case: total extermination of a minority planned and executed by the state  

destruction of a minority perpetrated by the state  

massacres against a minority orgenzied by state agents/agencies  

Domestic/non-state actors  

European adventurerers and invaders against indigenous groups in settler colonies of the Americas and Australia  

genocidal atrocities directed against a minority committed by extremists/interest groups  

genocidal atrocities directed against a minority committed by extremists/interest groups  

Foreign/state actors  

colonial genocides committed by European powers against indigenous peoples (mainly Spain, Britain, Portugal, Germany)  

colonial genocides and slavery committed by European powers against indigenous and rebellious groups  

massacres committed by invaders against civilian populations and rebellious or resisting groups  

Foreign/non-state actors  

no evidence found/probably invalid  

colonial genocides and slavery committed by European settlers and colonial companies against vulnerable groups  

massacres committed by invading settlers against local populations and rebellious or resisting groups  

A Typology of Mass Killing

The history of mass killing in the twentieth century indicates, according to Valentino (2000), that such violence occurs when powerful leaders come to believe it is the most 'practical' way to accomplish certain radical goals, counter powerful threats of solve difficult military problems. Broader elements of society need not share these goals or the conviction that mass killing is the best way to achieve them. A tiny minority, well armed and well organized, can generate an appalling amount of bloodshed when unleashed upon unarmed and unorganized victims.
These conclusions indicate that the causes of mass killing will be best understood when the phenomenon is studied from a 'strategic' perspective. This approach suggests that mass killing should be recognized as a goal-oriented policy calculated to achieve leaders' most important political and military objectives with respect to other groups - a bloody solution to leaders' most urgent problems (Valentino, 2000: 29).
The strategic approach provides a common perspective for understanding mass killing, but it does not posit a single explanation for this diverse phenomenon. Rather, the strategic approach identifies several unique motives and scenarios which generate strong incentives for leaders to consider mass killing. These scenarios may be grouped into two general categories. First, when leaders' plans result in the near complete material or political disenfranchisement of large groups of people, leaders are likely to believe that mass killing is necessary to overcome resistance by these groups or, more radically, that mass killing is the only practical way to remove these groups or their influence from society. Valentino refers to this general class of mass killing as 'dispossessive' mass killing. Second, mass killing can become an attractive solution when groups find themselves engaged in military conflicts in which conventional military options are ineffective or inefficient. When governments encounter great difficulties defeating their enemies' military forces directly, they face incentives to target the civilians whom they suspect of providing support to those forces. Valentino refers to this class of mass killing as 'coercive' mass killings. He arrives at the following typology (p. 20-33):









Ideological-political mass killings comprise communist transformations as well as chauvinist (ethnic, national or religious) mass killings. The territorial category comprises colonial enlargement as well as expansionist wars. The terroristic category comprises strategic bombing, starvation blockades/siege warfare, and mass terrorism.
Dispossessive mass killings are the result of policies which, by design or by consequence, have the effect of stripping large groups of people of their possessions, their homes or their way of life. Historically, such policies have emerged in two general scenarios. First, ideologies such as hard-line variants of communism and radical racism or chauvinism have driven leaders to seek a revolutionary transformation of society at the expense of certain groups. Second, the territorial ambitions of colonial or expansionist powers have often stripped preexisting populations of their land and means of subsistence (Valentino, 2000: 30-31).

Instrumental (pragmatic) vs. transcendental (ideological) genocides

DuPreez (1994), in what is probably the most basic dimension underlying the phenomenon, distinguishes instrumental (pragmatic or limited) versus transcendental (ideological) genocide (very broadly defined as the deliberate and intended killing of a category of persons).
If genocide has a clear economic or political purpose, then it will continue until that purpose is achieved. If genocide has a transcendental purpose, then it will continue until all members of a category have been eliminated or until the perpetrators have been arrested.
An example may clarify the difference between limited (instrumental, pragmatic) and transcendental (or ideological) genocide. If a particular group occupy land which settlers desire they may be killed. We may think of the German killing of the Herero in South West Africa. These kinds of killings shade into 'normal' warfare.
Transcendental genocide, on the other hand, is based on theories of the absolute need to eliminate all members of a category, because of their intractable vileness, wickedness, dangerousness, or opposition. Religious theory of a certain kind might impose the duty of eliminating all heretics, and Nazi theory demanded the elimination of all Jews and many other 'impurities' from the nation. What is transcendent about these forms of genocide is that they are based on absolutes and do not cease until these absolutes have changed or the perpetrators have been defeated. They are not mere means-ends murders. The elimination of the victim is the end which is sought, because the victim embodies the evil which must be removed from the world.

DuPreez (1994) illustrates his basic distinction by comparing the typologies of Fein (1984, 1990), Kuper (1981, 1985) and Harff & Gurr (1988), introduced above.
The first thing to observe is that it is not difficult to recognize the split between ideological and pragmatic genocides. All the authors identify an ideological form of genocide. Harff & Gurr write that revolutionary genocide is "mass murder of class or political enemies in the service of new revolutionary ideologies" (1988, p. 363).
Revolutionary genocide in our time often takes the specific form of politicide _ or the killing of members of political groupings, movements and parties rather than of ethnic groups or nationalities. Thus, the killing of the kulaks in the Soviet Union was politicide - the extermination of a faction or class construed as deliberately opposing the policy of collectivization. No matter if the kulaks did not construe themselves as a political movement. They were so construed by Lenin and Stalin and relentlessly persecuted as an obstacle, a political anachronism and a counter-revolutionary force. It is the perpetrators of genocide (or politicide) who do the defining. This is the prerogative of power.

One other item that stands out immediately is retributive genocide. It is listed by all the authors. A retributive genocide is very much a practical affair. Someone has injured you in the past. It is therefore only sensible to do them in before they have a second chance. This often occurs in the wake of decolonization or at the end of a war (in the Netherlands after liberation this episode of general revenge was aptly called 'bijltjesdag'), because some distinct group has been used by the colonizers or the invaders against the rest of the population. Hence the Harkis in Algeria, the Hutu in Burundi and the Tamil in Sri Lanka were singled out for advancement by their colonial masters and subsequent massacre by the revolutionary movements.

The next form of genocide which stands out is what Fein calls developmental genocide and Kuper calls genocide against indigenous peoples. Harff & Gurr subsume developmental genocides under hegemonial genocides. What are developmental genocides? The usual pattern is for colonizers to arrive to occupy a land and exploit it. The people they encounter are technologically unsophisticated and have a different economy and way of life. The colonized are organized in tribal systems and are politically weaker than the invaders, who have a well-developed sense of their own superiority. They often believe that development and progress make the displacement of the indigenous people essential. Those who can use the land more efficiently should own the land. Among the innumerable straightforward cases are the extermination of the Australian Aborigines and Tasmanians by the English; the Araucanian Indians by Argentinians; the Bushmen and Hottentots by the Boer settlers of South Africa; the North American Indians; the Herero and Dama in German South West Africa; the Amazonian Indians at the moment; the Papuans of Irian Jaya, the Bushmen in Botswana, etc. (The latter case is sometimes called 'ethnocide', or the killing off of a culture. In southern Africa at the present time, ethnocide is a consequence of a failure of policy rather than as a deliberate plot of extermination).
Developmental genocide is easily recognized. It is a fate that has threatened technologically backward peoples since the beginning of history, and hunter-gatherers in particular.
Like the nineteenth-century United States in relation to its Indian Minorities, the twentieth-century Brazilian state has adopted entirely genocidal standards in relation to its Indian minorities. Lewis (1974) mentions the Mundurucus, Guaranis, Carajas, Cintas Largas, Kadiweus, Chavantes, Bororos, and Tapaiunas. The latter have disappeared altogether. It is not difficult to add a number of other societies at present (1998) such as Aché, Pataxo and Yanomamö.

There remain only the forms of hegemonic genocide to consider. Hegemonic genocide is the mass murder of various groups of people in order to force the survivors to submit to the authority of the state. This takes place as part of an attempt to establish, expand or maintain the state. Usually, there has been a crisis: a colonial power has withdrawn, leaving the contenders for power to fight it out among themselves; or the balance of power has changed in such a way that previously submissive groups begin to challenge authority and claim the right to secede. These are the sorts of struggle we see in the disintegration of the federation that used to be Yugoslavia. Mass murders are committed as people are forced to submit to what is left of central authority. How to explain it? The difficulty is that we have layer upon layer of history as group after group has seized power. Each group remembers this history in a time of crisis, and each has reason to fear every other group when the rules of the political game have been destabilized.
In the former Yugoslavia, for example, each national group was identified with a different period of domination. Each was 'dirty' to the others in a different way. The Serbians were associated with communist rule; the Croatians with Nazi collaboration during the war; the Muslims with wealth and privileges dating from the Ottoman Empire.

What are 'despotic' genocides? They are based on personality cults. Sometimes, in 'new' states, artificially held together by personalities rather than principles, unity is achieved by identifying the state with a despot. Naturally, strong measures have to be taken against those who dissent. Without such measures, the state will disintegrate.

DuPreez believes that the classification system which enables us to achieve the maximum resolution with the maximum clarity is the following:









Motivational classification of genocides

A motivational classification of genocides - which is fully compatible with DuPreez' basic distinction - has been presented by Diamond (1992). Genocides prove as hard to pigeonhole in their motivation as in their definition. While several motives may operate simultaneously, it is convenient to divide them into four types. In the first two types there is a real conflict of interest over land or power, whether or not the conflict is also disguised in ideology. In the other two types such conflict is minimal, and the motivation is more purely ideological or psychological.

1. Perhaps the commonest motive for genocide arises when a militarily stronger people (mostly technologically 'advanced') attempts to occupy the land of a weaker people (mostly 'primitive' or 'preliterate' or 'preindustrial'), who resist. This is 'developmental' genocide.

2. Another common motive involves a lengthy power struggle within a pluralistic (multi-etnic) society, leading to one group's seeking a final solution by killing the other. Cases involving two different ethnic groups are the killing of Tutsi in Rwanda by Hutu in 1962-63 and the 1990s, of Hutu in Burundi by Tutsi in 1972-73, of Serbs by Croats in Yugoslavia during World War II, of Croats by Serbs at the end of that war, and of Arabs in Zanzibar by blacks in 1964. However, the killer and killed may belong to the same ethnic group and may differ only in political views. Such was the case in history's largest known genocide, claiming an estimated twenty million victims in the decade 1929-39 and sixty-six million between 1917 and 1959: that committed by the U.S.S.R. government against political opponents among its own citizens. Political killings lagging far behind this record are the Khmer Rouge purge of several million fellow Cambodians during the 1970s, and Indonesia's killing of hundreds of thousands of Communists in 1965-67.
In these two just-described motives for genocide, the victims could be viewed as a significant obstacle to the killers' control of land or power.

3. At the opposite extreme are scapegoat killings of a helpless minority blamed for frustrations of their killers. Jews were killed by fourteenth-century Christians as scapegoats for the bubonic plague, by early twentieth-century Russians as scapegoats for Russia's political problems, by Ukrainians after Word War I as scapegoats for the Bolshevist threat, and by Nazis during World War II as scapegoats for Germany's defeat in World War I. When the U.S. Seventh Cavalry slaughtered several hundred Sioux Indians at Wounded Knee in 1890, the soldiers were taking belated revenge for the Sioux's annihilating counterattack on Custer's Seventh Cavalry force at the Battle of the Little Big Horn fourteen years previously. In 1943-44, at the height of Russia's suffering from the Nazi invasion, Stalin ordered the killing or deportation of six ethnic minorities who served as scapegoats: the Balkars, Chechens, Crimean Tatars, Ingush, Kalmyks, and Karachai.

4. Racial and religious persecutions have served as the remaining class of motives. Diamond does not claim to understand the Nazi mentality, but the Nazis' extermination of Gypsies may have stemmed from relatively 'pure' racial motivation, while scapegoating joined religious and racial motives in the extermination of Jews. The list of religious massacres is almost infinitely long. Is includes the First Crusaders' massacre of all Moslems and Jews in Jerusalem when that city was finally captured in 1099, and the St. Bartholomew's Day massacre of French Protestants by Catholics in 1572. Of course, racial and religious motives have contributed heavily to genocides provoked by land struggles, power struggles, and scapegoating.

Genocide as a Crime of State

As we have seen, Horowitz (1976 et seq.) heavily implicated "a state bureaucratic apparatus" in the commitment of genocides.
Horowitz (1982), like Rubenstein (1983), views genocide as an act of state but an act intended as the ultimate means of social control, an act which can only be perpetrated by a totalitarian state. (This book has gone through several editions but this has not changed - as Fein [1990] observed). Horowitz distinguishes states along a continuum of means of social control and ranges of tolerance of deviance: permissive societies, tolerant societies, guilt societies, traditional shame societies, harassment societies, torture societies, deportation or incarceration societies, and genocidal societies.

To Chalk & Jonassohn (1990) also, genocide is preeminently a crime of state. They assume that most people are reluctant "to carry out a mass slaughter of defenseless victims" and therefore "genocide has always required a high degree of centralized authority and quasi-bureaucratic organization" excepting when the victims are few (small indigenous groups).
The assumption that most people are reluctant to kill defenseless people is "contradicted" (as Fein [1990] put it quite inappropriately) by their previous observations on the many cases of spontaneous killings by settlers which were not state-organized and were "often opposed by governments" ineffectually (1988). Fein continues: "The extensive research on pogroms, lynchings and massacres does not usually probe the extent of mass participation but gives us much evidence that people are not reluctant to kill defenseless victims of another group. But the victimizers in the past have stopped when they were sated, drunk, or eager to enjoy their booty; genocides on state order press perpetrators to continue". (Fein probably needs this universal-non-reluctance-to-kill because she could not possibly explain why some people do and some people don't - her psychology is totally inadequate. We shall come back to this important issue).

Harff (1986, 1987) considers genocide to be an instance of state terrorism, and she compares its use in the Ottoman Empire, East Timor, Burundi and Kampuchea. She proposes that genocide occurs only under special circumstances in the internal development of the state and the international system. The former result from national upheavals resulting from defeat in war or from post-revolutionary regimes. "National upheaval is an abrupt change in the political community, caused, for example, by the formation of a state through violent conflict, when national boundaries are reformed, or after a war is lost. Thus, lost wars and the resultant battered national pride sometimes lead to genocide against groups perceived as enemies" (Harff, 1987: 43).
Harff's conception of 'national upheaval' comes very close to Melson's (1992 et seq.) notion of 'revolution' as a structural precondition for genocide (or at least for a small set of what he called 'total domestic genocides' [see § 4]), Mazian's (1990) 'internal strife', and Staub's (1989) 'difficult life conditions' (which can include economic problems; wars; structural crises or stresses of state and society; and rapid changes in technology, social institutions, values, and ways of life).

Insulation from foreign influence and the indifference of foreign states and international and regional organizations enables and reinforces the use of genocide by the perpetrator. This corresponds to the propositions and observations of Fein (1979b, 1990) and Mazian (1990).
Looking at the internal social structure of genocidal states, Harff's propositions do not differ substantially from those of Kuper (1981 et seq.), Fein (1979 et seq.), and Chalk & Jonassohn (1988, 1990): in states with pre-existing internal cleavages and real opposition, the dominant elite is more likely to target an excluded and oppositional racial, religious, or ethnic group. But, Harff also notes, genocide is not restricted to such states (as shown in Cambodia). She sometimes does not distinguish between conflicts stemming from real inter-group opposition and symbolic conflict in which the perpetrator unilaterally decrees the victim a threat (Fein, 1990).

Why states victimize groups

Harff & Gurr (1995) have presented an insightful summary of 'Why States Victimize Groups'.
Discrimination against communal and political groups has many historical origins. Some are residues of historical conflicts among groups, others are the result of immigration by alien minorities, some are incidental consequences of the political expansion of a nation-state's authority over previously autonomous groups.
What is it about the state, and the victims, that make some kinds of groups targets of the most severe kinds of repression and violence? In some instances an inactive communal group is singled out for attack and dispersal because it stands in the way of national expansion. Historical episodes include the treatment of many native peoples of the Americas and the aborigines of Tasmania. In other instances governments respond tit for tat with violent tactics in an attempt to quell politically organized groups that actively seek to alter power relations within an internationally recognized state, as is the case in Kurdistan. When state repression is met with further resistance, leaders may be provoked to return violence disproportionately. This is the strategy of politicide, in which governing authorities choose to respond to challenges by killing as many members of the group as is necessary to shatter their capacity to persist and act as a collective. In some such cases many individual members of the group survive.
The worst of all possibilities is that in which a state systematically seeks to destroy, as a matter of policy, all members of a communal group irrespective of their actions. 'Guilt' is established not by action or association, but is assigned to all those who share the defining ascriptive characteristics. This was the Nazis' intent with respect to Jews and Roma (gypsies).
The state's involvement in genocides and politicides may be more or less direct. Not all are carried out by uniformed agents of the government. In others, leaders assist or knowingly acquiesce in the killing of undesirable groups by vigilantes, 'deathsquads', or private militias. And in some instances governments simply neglect their obligations to protect vulnerable minorities who are attacked by murderous mobs or profiteers. The Tasmanian aborigines were exterminated when settlers hunted them down like rabbits, with the acquiescence of colonial authorities.

Genocide as abuse of state power. . .

Given the state's monopoly of political power, Hedrick-Wong (1998) reasoned recently, whether by individuals or impersonal institutions, its abuse is inevitable. Indeed, recorded history is abundant with episodes of such abuse, the most extreme being genocide - premeditated large-scale murder of certain groups of people, whose existence is perceived as dangerous to the society by those in control of the state machinery of violence and coercion. It is by no means a rare occurrence. Gurr & Harff (1988), for example, estimated that racially and ethnically motivated genocide has claimed between 7 and 16 million lives since World War II. Even more disturbing are claims made by social scientists who have examined genocide in depth that no society is immune from genocidal tendencies. It is somehow built into the institution of the sovereign state. Kuper (1981) put the matter precisely and bluntly: "the sovereign territorial state claims, as an integral part of its sovereignty, the right to commit genocide, or engage in genocidal massacres, against peoples under its rule, and that the United Nations, for all practical purposes, defends this right".

. . . versus weakness of state control

Keeley (1996), on the other hand, reasoned that the weakness or lack of state control is conducive to massacres and exterminations.
In general, nonstate groups preserved the lives of captives only when some material benefit would accrue; this approach generally limited the persons spared to women and children. States, by contrast, often have a strong material interest in preserving the lives of defeated enemies - even adult males - because they can become tax- and tribute-paying subjects, serfs, or slaves. The life-preserving rituals of formal surrender and widespread official distaste for killing noncombatants are expressions of this interest. Economically, the state is usually best served by the submission of its enemies, not by genocide. The atrocities that do occur in civilized warfare usually happen when commanders lose control of their soldiers, whose primary motive may be the primitive one of avenging combat losses or previous real or fictive enemy atrocities. And slaughters of noncombatants can occur as a matter of policy, when the policymakers themselves are consumed by ethnic hatred or when they make a calculated attempt to use state terrorism to cow a conquered populace.
The reaction of the German government during the Herero-Nama uprising in Southwest Africa in 1904 is an example of the self-interested mercy of states and of the conditions under which it fails. The local military governor, General von Trotha, issued an extermination order against the Hereros. The imperial chancellor and the German colonial office successfully demanded that this order be countermanded by the Kaiser; it was inhumane, was bad for public relations, and (perhaps most important) would "undermine the potential for development" by eliminating native labor. The governor, his troops, and the German colonists paid little heed to the Kaiser's order, however. When the fighting ended several years later, only one-half of the Nama and one-sixth of the Herero had survived (Edgerton, 1988; Pakenham, 1991). Precisely this weakness of state control over frontier 'militias' made massacres of native peoples more common by such agents than by the 'regular' forces of the state. Indeed, the most notorious massacres of North American Indians, such as those at Sand Creek and Camp Grant, and the only actual genocides (that is, complete extinction of a tribe primarily by homicide) during the European conquest were all inflicted by local militias (e.g., Utley & Washburn, 1977).


Who are the (preferential) victims of massacres and geno/politicides? Some general observations can be offered about the victims. According to Harff & Gurr (1995), the victims of politicide typically have either long-standing aspirations of independent nationhood (secessionist movements) or are members of groups actively opposing existing regimes (political opposition parties). The chosen victims of genocides are often traditional enemies. Genocidal victims are most often minorities whose cultures are sharply distinct from the dominant group.

Sacrifice groups

DuPreez (1994) observes that all genocides are intended to 'purify' the nation, but in some the purpose of purification seems to be more obvious than in others. In such cases we often find groups which are a permanent affront to the majority, because their very existence challenges the universal validity of the beliefs of that majority. This was the misfortune of the Armenians in Turkey, of the Jews in Europe, and of heretics everywhere, particularly during periods of national insecurity or looming catastrophe. When these groups are vulnerable and enduring minorities, they are sometimes referred to as 'hostage groups'. But hostages are usually released when blackmail has succeeded. These groups seldom are. They are more appropriately called 'sacrifice groups'. They are, from time to time, sacrificed to the purity of the majority. In times of crisis and defeat there is usually a sense of betrayal and the question is asked: Who has betrayed the people? Sacrifice groups have valuable characteristics at such times. They are outsiders who have not only a different but an opposing faith, the very existence of which may easily be construed as a threat to the majority; they often have outside links with the enemy; and they have wealth and positions which leaders of the dominant group would like to redistribute to their most loyal followers. In fact, the success of the sacrifice group often 'explains' the failure of the insiders. Quite obviously, the majority are poor because the outsiders are rich; the majority do not have jobs because the outsiders have taken them all, especially the best ones; and catastrophe is looming because the outsiders have betrayed the country. A government which sacrifices the outsiders shows that it is on the side of 'the people', or whatever the current term is for the sacred majority. In this way, old quarrels are used to distract people from the causes of present failure. Firm action shows that the state (embodying the people) is strong and not weak. People begin to believe that they can face up to dangers which previously made them feel helpless. By smiting the weak, who have become in fantasy a threatening monster, they gain confidence (DuPreez, 1994).

That the victims played a part in their own destruction has been suggested, mainly by Hannah Arendt (1963) but also by Hilberg (1961), Dawidowicz (1975), Bettelheim (1979), and others.
The behavior of victims affects the perpetrators' resolve. It can make the devaluation of victims, the evolution of a genocidal ideology, and its expression in action easier or more difficult. But it is not the origin of the motivations that lead to mass killing or the cause of victimization (Staub, 1989).


Van den Berghe (1981) observed that no other group is so vulnerable to pogroms, expulsions, and genocide as middlemen-minorities [MM's] _ powerless alien minorities who migrate to other societies to fill niches (usually in trade) not filled by native classes.

"In the catalogue of human bestiality to man, MMs almost invariably appear as victims: Jews in Europe, Armenians in Turkey, Indians in Uganda, the 'boat people' of Vietnam (who are mostly ethnic Chinese), the Chinese in Indonesia and many other such MMs have been repeatedly victimized throughout human history" (1981, 140).

But many unobserved MM's do not become victims (Fein 1987b, 1990) and other groups which have been victims - particularly indigenous peoples - are not MM's. Zenner (1987), reviewing research on MM's, concludes that "There is no necessary connection between MM status and victimization".
Fein (1990) seems to regard this as a refutation of - or at least a devastating blow to - Van den Berghe's observation. But Van den Berghe nowhere claims that MM's are the only (ethnic) victims of genocide, or exclusively so, or necessarily so. He states: "No other type of ethnic group is in such a perilous, vulnerable and defenseless position as the MM. No other groups are as frequently the victims of pogroms, expulsions, confiscation of property, discriminatory taxation and even wholesale genocide". One can only read this to mean that no other group is as frequently subjected to a whole range of oppressive and repressive measures as the MM's.

According to Fein (1990) there are three commonalities among the classes from which victims have repeatedly been drawn, including some MM's and indigenous people. The first
(1) is not the roles they play but that they are alien or perceived as alien _ the reason why MM's are apt to migrate and ready to play scorned roles _ and they are outside the universe of obligation of the dominant group (Fein 1977, 1979b, 1990).
(2) Secondly, they are seen as unassimilable (either for religious or racial reasons) and may themselves reject assimilation. As Van den Berghe (1981) notes, "Even MM groups that do achieve a large measure of assimilation are not safe from persecution, as shown by the tragic history of German Jewry".
(3) Lastly, their elimination either removes a threat (real or symbolic) or opens up opportunities (or both). While both MM's and indigenous people are viewed as aliens, MM's may also be seen as loyal to some other state from which they came. They are in double jeopardy both because they can not be protected by their state of origin (e.g. overseas Chinese and Indians) and are apt to be seen as threats to their state of residence if the latter is involved in hostilities with their state of origin (e.g., Chinese in Vietnam 1978-79).

Kelman (1973) deals with sanctioned massacres, which (1) occur in the context of an overall policy that is genocidal in character, being designed to destroy all or part of a category of people defined in ethnic, national, racial, religious or other terms, and (2) are directed at groups that have not themselves threatened or engaged in hostile actions towards the perpetrators of the violence. He adds that there are, of course, historical and situational reasons why a particular group becomes a suitable target for massacres, but that it cannot be said "in any objectively meaningful sense, that they provoke the violence against them by what they have done. They are not being murdered because they have harmed, oppressed, or threatened their attackers. Rather, their selection as targets for massacre at a particular time can ultimately be traced to their relationship to the pursuit of larger policies. They may be targeted because their elimination is seen as a useful tool or because their continued existence is seen as an irritating obstacle in the execution of policy".

Empirical investigations of geno/democide

Brzezinski (1994) labels the twentieth "the century of megadeath". He present a "concise statistical accounting of the extraordinary toll of politically motivated killings" in this century, an unparalleled and heinous record of mass murder.
According to Brzezinski (1994; as summarized by White, 1997), no less than 87,000,000 lives were lost during the wars of the century, including the two World Wars. These include casualties inflicted on the civilian populations during the course of the war, but not committed as a deliberate policy of extermination. Brzezinski adds another 80,000,000 (80 million) lives or more as belonging to people in groups being deliberately murdered, and not as a result of combat. Thus, he asserts, more than 167,000,000 people were murdered in the course of the twentieth century. If Brzezinski is even half right, the title 'The Century of Megadeath' is unfortunately all too apt. Alas, he is almost certainly more than half right.
These mass murders are distinguished into intra-ethnic and inter-ethnic. Intra-ethnic include mass murder under Lenin and Stalin (up to 30 or more million), mass murder under Mao Zedong (up to 30 million), and mass murder under Pol Pot (1 million). Inter-ethnic mass murders include mass murder under Hitler, including the Holocaust (17 million), Chinese civilian deaths during the Sino-Japanese war (15 million), Hindu-Moslem killings in India (3 million), and the Armenian massacre by Turks (1 million).
White notes that of the 80,000,000 noncombat mass murders, fully 60,000,000 or three-fourths were basically intra-ethnic. Insofar as they were based on socio-economic class and (alleged) resistance to the Communist regime, these murders were, strictly speaking, not genocidal, since the victims were not targeted merely because of their ethnic group membership.
It is clear that one can go beyond Brzezinski's account. In particular, the inter-ethnic strife in Africa between Hutus and Tutsis and between Ibos and other tribes in Nigeria comes to mind, with each costing perhaps up to a million or more lives. During the nineteenth century up to 24 million lives were lost in China as a result of the internecine Taiping Rebellion. Contemporary (and mostly 'hidden') examples of what may turn out to be large-scale mass murders include Tibetans, Kurds, Timorese, and the ongoing 'ethnic cleansings' in former Yugoslavia (besides the numerous examples already mentioned above).
Other recent cases of large-scale genocide since 1945 were the Khmer Rouge in Cambodia (1975-79) and Sudan's Arab or Arabized regimes (1954-1972; 1984 until today) against the Nuba in Central Sudan and against Dinka, Nuer and other nation peoples in Southern Sudan (Scherrer, 1998b).

According to Leitenberg (pers. comm. January 23, 2000), Brzezinski's figure of 87 million mentioned above is quite wrong. From his personal archival material it appears that the number is closer to 127 million. Leitenberg also notes that "The figure used by Hobsbawm in his 1995 book The Age of Extremes: A History of the World, 1914-1991, is 187 million 'people killed or allowed to die by human decision' for the 'short century' that he examines. Hobsbawm notes that this accounts for about 10 percent of the global population at the year 1900. Hobsbawm's category - 'by human decision' - includes such non-wartime politically caused deaths as those in the Soviet Union (1930s Ukrainian starvation and the 'Gulag') and in China between 1949 and 1975 (the massive starvation of the 'Great Leap Forward,' and various 'repression campaigns'). However, likely deaths in those two countries for political, government-decided reasons are on the order of 35 million and 45 million respectively, or 80 million, for a total of around 205 million (125 plus 80), rather than Hobsbawm's figure of 187 million" (Leitenberg, p.c.).

The most extensive empirical investigation of casualties of geno/democide has been conducted by Rummel and has been reported in a great number of publications. The following account is based on Rummel (1996).
Near 39,000,000 people have been killed in genocide, or near 23% of the total toll. This itself is more than all the war-dead of all this century's international and civil wars. Surprisingly, genocide is a pattern of democide independent of other empirical democide patterns. That is, genocide is largely uncorrelated with other kinds of democide. The immediate causes and conditions of genocide are different from those for other types of democide or democide overall. At a more basic level there still may be causes and conditions that encompass genocide and other patterns of democide (and war).
The more totalitarian and less democratic a regime the more democide, the more genocide, and the greater the annual rate of democide that it commits. That is, although the independent patterns of domestic democide, foreign democide, genocide, and the others, are not correlated, together they are accounted for by a regime's totalitarian power. Power is a means through which a regime can accomplish its goals or whims. When a regime's power is magnified through its forceful intervention in all aspects of society, including its control over religion, the economy, and even the family, then when conjoined with an absolutist ideology or religion, mass killing becomes a practical means of achieving its ends. When the regimes finds for whatever reason that the continued existence of a social group is incompatible with its beliefs or goals, totalitarian power enables it to destroy that group. Genocide follows. On the other hand, democratic elites generally lack the power to, and democratic culture anyway opposes, the outright extermination of people or social groups for whatever reason.
The more power a regime has, the more it is likely to commit foreign violence and to have rebellions against it. The empirical evidence on this is overwhelming. The least warlike regimes are democratic, while the most warlike are totalitarian. Clearly, war or rebellion provide an excuse and cover in the fog of war for a regime to eliminate those social groups it finds objectionable.
The social diversity of a nation - the nation's ethnic, religious, racial, linguistic, or national divisions - the nature of a regime's culture, or the level of education or economic development of a nation, are all uncorrelated with the overall democide of a regime (Recall that Germany and Japan were both highly developed and the best educated on the brink of the Second World War). It is not diversity that predicts to genocide, but a regime's power.
Rummel would probably modify the adage "power corrupts; absolute power corrupts absolutely" to read: "power kills; absolute power kills absolutely".

Rummel and others have marshaled strong evidence to demonstrate that democratic forms of government are associated with lower levels of mass killling than other governmental systems, especially totalitarian and communist regimes. The fact that democracies engage in less mass killing of their own citizens than other forms of government is one of the most careful documented findings of the theoretical literature on mass killing. Nevertheless, two problems limit the relevance of this finding.
First, as Rummel acknowledges, although democracies almost never kill their own citizens, they do engage in mass killing during foreign wars and in their own colonies abroad.
Democracies have also been known to use mass killing against domestic groups not defined as citizens. The United States, for example, waged a relentless series of wars on the indigenous people of North America, a process that eventually resulted in their near total extermination (Stannard, 1992).
Democratic states have also provided economic and military assistance to foreign regimes engaged in mass killing.
A second and perhaps more important limitation of Rummel's power principle is that it provides relatively little ability to predict when mass killing will occur. As Rummel acknowledges, power is a necessary but not sufficient condition for mass killing (Valentino, 2000: 19-20).

Theories and Explanations of Genocide

What causes genocide? Explanations of genocide generally fall into one of two categories: the top-down approach (starting from social structure, intergroup relations, and collective behavior) and the bottom-up approach (starting from the psychology of the individual or from social psychology) (Fein, 1990).
Besides the top-down and bottom-up approaches to genocide, two 'orthogonal' explanations are frequently offered. The first stresses group disparities in political-economic situations and the desire of a dominant group to use its power to obtain better living conditions, more land, and the material wealth held by an ethnic minority. Such explanations offer reasons the mind can comprehend; they suggest a glimmer of rationality, a kernel of sense underlying the pathology of a renegotiation of the political-economic balance gone tragically awry. In contrast to such rational explanations, we find genocide explained through ancient hatreds festering in the body politic, hatreds that remain inherently unresolvable through moderate forms of political negotiation because of their primordial force and passion.

Is 'conflict' sufficient explanation?

Both explanations, however, stumble on one striking fact: similar disputes, conflicts and political-economic disparities exist elsewhere and are resolved without resort to genocidal brutality.
The use of the category of 'conflict' as an explanation of genocide is not sufficient and often tautological, as noted also by Fein (1990). "Conflict between groups sometimes becomes so severe that physical destruction of one by the other becomes an accepted goal" assert Simpson & Yinger (1985, 20-21; see also Berry & Tischler [1978, Ch. 16]), a landmark text in intergroup relations, without telling us how or why. Since most conflicts, including conflicts over resources, are settled by other means, this is not a sufficient explanation.

So what is the explanation - is there a simple explanation anyway?

An examination of memoirs by the architects of genocide and of legal testimony by men who actually committed genocidal massacres suggests that genocide occurs when existing ethnic differences are overlain by specific situational factors (primarily economic distress and political instability) and then further combined with a critical cognitive perception of one's neighbors as 'the other'. This initial psychological distancing from former friends and fellow citizens facilitates an eventual process of dehumanization of 'the other' - a cognitive perception that is the match that lights the tinder of genocidal violence. Only when friends and fellow citizens are dehumanized does the unimaginable become possible (Renwick Monroe, 1995).
The concepts of distancing and dehumanization will be discussed in more detail later on.

Lack of theories of genocide

Unfortunately, there are no 'real' theories of genocide. What we have is a number of ingredients, determinants, preconditions, prerequisites and precipitating factors, facilitating or inhibiting factors, etc. which have been identified or hypothesized by researchers from a variety of disciplines. What comes closest to 'real' theory are either 'value-added' (cumulative) hierarchical models (e.g., Mazian, 1990; Hovannisian, 1994) or 'steps-along-a-continuum-of-destruction' models (e.g., Staub, 1989), which specify and order the explanatory elements from general preconditions to specific situational idiosyncracies. An excellent example, combining both models, would be Els de Temmerman's (1995) account (in Dutch) of the recent Hutu-Tutsi genocide in Rwanda. All these will be presented in more detail below.

The basic elements of a proto-theory of geno/politicides are spelled out by Harff (1988) and Harff & Gurr (1987, 1988, 1990):

The likelihood that geno/politicides will occur is increased among states of lesser international status [after 1945], in states with new elites and regimes, in states with a history of repressive policies against groups, in societies with persisting internal cleavages, and among those states in which elites use their power to differentially reward groups for perceived loyalties. Revolutionary takeovers are particularly likely to lead to geno/politicides when ideological considerations target groups outright as expendable. Groups not fully incorporated into the social structure are likelier targets. Revolutionary one-party states are the likeliest offenders. Genocides occur with alarming frequency during or shortly after the revolutionary takeovers. Especially dangerous are situations in which long-standing ethnic rivalries erupt and radicalized groups armed with a revolutionary ideology gain the upper hand. Communist ideologues tend to be most aggressive in their dealings with potential or past opposition groups. Interestingly enough, the length of democratic experience is inversely related to the occurrence of geno/politicides (Harff 1988, 23-24).

Value-added models

Mazian (1990) sketches a theory based on a revised version of Smelser's Theory of Collective Behavior (1963), positing a value-added or cumulative interacting set of determinants of genocide:

1.    Creation of 'outsiders'
2.    Internal strife [or turmoil]
3.    Destructive uses of communication [composed of aggressive ideology and propaganda war]
4.    Powerful leadership with territorial ambitions forming a monolithic and exclusionary party
5.    Organization of destruction
6.    The failure of multidimensional levels of social control [powerlessness and weakness of the victim and lack of checks by the state committing genocide, other states and religious institutions] (1990, ix-x).

Some of these determinants are actually aspects of genocide rather than preconditions.

What were the factors contributing to the unleashing of the Armenian genocidal process? According to Hovannisian (1994) they include:

1.    The existence of a plural (multi-etnic) society (Ottoman Turkey) with clearly defined racial, religious, and cultural differences;
2.    A sense of deprivation/danger felt by the perpetrator groups (Turks) [threat perception];
3.    The espousal and propagation by the perpetrators of an ideology/belief system (Pan-Turkism) emphasizing the mobility and distinctiveness of its own group as opposed to the exploitative nature of the intended victims (the need to subhumanize the intended victim); and
4.    The determination to establish a new regional order and in that process eliminate elements posing real, potential, or perceived threats [The felt need to eliminate the threat or danger].

Porter (1982) attempts to synthesize the contributions of Dadrian, Fein, Horowitz, Hilberg, and Porter, recapitulating the conditions favoring genocide:

a)    Minority groups have previously been and are presently defined outside the universe of moral obligation by the dominant group...
b)    Pervasive racialistic ideologies and propaganda are found in the nation-state's society.
c)    There is a strong dependence on military security.
d)    Powerful, monolithic exclusionary political parties are present.
e)    The leadership has strong territorial ambitions.
f)    The power of the state has been reduced by defeat in war and/or internal strife.
g)    The possibility of retaliation for genocidal acts by kin of the victims of or interference by neutral nations is at a minimum (Porter 1982, 17-18).

Related to the models presented above is the conception of 'genocidal potential' (DuPreez, 1994): A situation has genocidal potential if there is a genocidal movement and a set of genocidal activators. It is the interaction of the two which is critical. Let us now list the set of activators and it will be immediately evident that they are cumulative rather than all-or-nothing. This set is inspired by Melson's (1989) comparison of the Armenian genocide and the Holocaust.

* Activators of genocidal movements:

* Long history of conflict between perpetrators and victims
* Violent precedents
* Catastrophe and hard times
* Collapse of the centre
* Polarization
* Police unable or unwilling to control paramilitary street violence
* The existing government is regarded as illegitimate
* War

The work of Littel (1988) on an early warning system to recognize genocidal movements is a good basis from which to start:

* Symptoms of genocidal movements:

* Dehumanizing propaganda about target group
* Conspirational history. The people have been betrayed.
* Utopian ideology. The target group stands between the people and 'the good life'.
* Politics of polarization pursued and middle ground scorned.
* Paramilitary wing; private armies
* Quasi-religious ceremonies of initiation and archaic, secret symbols
* Closed cell organization
* Infiltration and subversion of public institutions
* Young and very young mobilized against the older generation and their parents.
* Total loyalty to the movement and its leader
* Violent purges of suspected dissidents

The reason why we cannot deliver a predictive theory of society and societal events such as genocides and massacres is quite simple. We can summarize it as the strategic dilemma: the effect of any move made by any participant in a social/political event will depend on what everyone else does (DuPreez, 1994).

Continuum of destruction

Staub (1989) believes that - tragically - human beings have the capacity to come to experience killing other people as nothing extraordinary. Some perpetrators may feel sick and disgusted when killing large numbers of people, as they might feel in slaughtering animals, but even they will proceed to kill for a 'good' reason, for a 'higher' cause. The perpetrators change, as individuals and as a group, as they progress along a continuum of destruction that ends in genocide.
People learn and change by doing, by participation, as a consequence of their own actions. Small, seemingly insignificant acts can involve a person with a destructive system; for example, accepting benefits provided by the system or even using a required greeting, such as 'Heil Hitler'. Initial acts that cause limited harm result in psychological changes that make further destructive actions possible. As Kressel (1996: 199) observed "Once people commit their first evil act, often without much thought, a new logic pushes them on toward more heinous atrocities".
Deeply ingrained, socially developed feelings of responsibility for others' welfare and inhibitions against killing are gradually lost. Often the leaders assume responsibility, and accountability is further diminished by compartmentalization of functions and the denial of reality. Often there is a reversal of morality, and killing the victims comes to be seen as good, right, and desirable.
Some people become perpetrators as a result of their personality; they are 'self-selected' or selected by their society for the role. But even they evolve along the continuum of destruction. Others who were initially bystanders become involved with the destructive system and become perpetrators. Even bystanders who do not become perpetrators, if they passively observe as innocent people are victimized, will come to devalue the victims and justify their own passivity (Staub, 1989). People often prefer to remain bystanders and not get involved. This collective indifference may be the most fundamental prerequisite for genocides and gross human rights violations to take place.
Hollander (1997; as summarized in Beck, 1999: 292) discusses four personality types of specialists in coercion and political violence, including those involved in both the Nazi and Soviet apparatuses. The first group is represented by the ideologically driven, supposedly incorruptible, puritanical executioners exemplified by Heinrich Himmler in the Nazi case. The second group embodies the 'banality of evil', of which Adolf Eichmann, described by Hannah Arendt in Eichmann in Israel: A Report on the Banality of Evil, is the prototype. This group is composed of supposedly very ordinary human beings who simply follow orders without being driven by strong convictions. Often money and privilege are factors. The third category comprises well-educated careerists who find satisfactory employment and mobility opportunities in the organization. The fourth group is composed of individuals who gravitate toward organizations of violence and coercion. Many of the notable torturers belong to this group. Their personalities are congenial to sadistic and repressive actions.

Redirected hostility and scapegoating

Threats and frustrations (what Staub generally - and rather indefinitely _ calls "difficult life conditions") give rise to hostility and the desire to harm others. The appropriate targets of this hostility are, of course, the people who caused the problems, but usually they cannot be identified. Often no one is to blame; the causes are complex and impersonal. At other times those responsible are too powerful, or they are leaders with whom people identify too much to focus their hostility on them. The hostility is therefore displaced and directed toward substitute targets. Hostility is especially likely to arise if people regard their suffering as unjust, as they often do, and especially if some others are not similarly affected.
Some of the internal processes to cope with life conditions are basic psychological tendencies common to all human beings: differentiation of ingroup and outgroup, 'us' and 'them'; devaluation and dehumanization of those defined as members of an outgroup; just-world thinking (Lerner, 1980; also known as 'blaming the victim'), which is the tendency to believe that people who suffer, especially those already devalued, must deserve their suffering as a result of their deeds or their characters; and scapegoating, or blaming others for one's problems. Sometimes having a scapegoat is the glue in the formation of the group. But even if the ideology does not begin by identifying an enemy, one is likely to appear when fulfillment of the ideological program proves difficult.
Societal values can embody a positive or negative evaluation of human beings and human well-being. But even in societies that do value human welfare, an outgroup may be excluded from the moral domain (Cf. Fein, 1990; see section on dehumanization).
'Us' - 'them' differentiation is a basic human potential for which we even carry 'genetic building blocks'. It is one source of cultural devaluation. Negative stereotypes and negative images of a group can become deeply ingrained in a culture (Staub, 1989; see Ch. on 'Evil' for elaboration).

De Temmerman (1995) presents a perceptive and illuminating synopsis of the causes of the Rwanda genocide as follows:
De verscheurdheid zit niet alleen in de kerk maar ook in de vertegenwoordigers van die kerk en in elke mens die bij dit drama betrokken was, denk ik later als ik de ingrediënten beschouw van de explosieve cocktail die geleid heeft tot de genocide in Rwanda.
    De basis voor die cocktail vormen armoede en overbevolking, noteer ik. Het ineenstorten van de prijzen op de wereldmarkt van koffie en thee in de jaren tachtig en het westerse protectionisme voor andere producten uit het Zuiden hebben Rwanda in de spiraal van schulden, IMF-programma's en de devaluatie gebracht waar weinig landen in Afrika aan ontsnappen. Maar ook droogte, ongelijke interne verdeling, grondschaarste en de snelle bevolkingsgroei dragen bij aan die basis. Voeg daarbij een totalitaire staat die via de samenlevingsstructuren, de (voormalige) eenheidspartij, het militair apparaat (leger, politie, gendarmerie, presidentiële garde), de media, de steun van de kerk en van het buitenland een totale controle uitoefende op het volk.
Dat mengsel begon te borrelen toen na de Koude Oorlog de democratiseringswind ook over Afrika ging waaien en het Westen aandrong op meerpartijenstelsels en verkiezingen. Met de democratisering dreigde de kliek rond de macht haar privileges te verliezen. Het ondermijnen van de oppositie en het etniseren van de politieke strijd behoort tot de typische middelen van Afrikaanse leiders om de democratie te doen mislukken en de macht te behouden, maar hoeft niet noodzakelijkerwijs tot genocide te leiden.
De cocktail begon gevaarlijk te sissen met de inval van het RPF in 1990 en de oorlog met al zijn menselijke ellende (vluchtelingen) en economische gevolgen (hongersnood). Volgde een militaire opbouw: een snel groeiend en weinig gedisciplineerd leger en de vorming van de Interahamwe. Bandieten en straatjongens konden door het vermoorden van de vijand kans maken op sociale promotie en materieel gewin. Maar de vijand was niet alleen RPF - en dat maakt de Rwandese oorlog verschillend van alle andere - de vijand bestond vanaf het begin ook uit onschuldige burgers, oppositiemensen en Tutsi's, vermeende bondgenoten van het RPF.
Het mengsel werd aan het koken gebracht door de propaganda van de extremistische Hutu's die de kern uitmaakten van de macht, de radio-omroepen met hun simplistische, drieledige boodschap: de Tutsi's komen om de macht te grijpen, ze gaan alle Hutu's uitroeien, we moeten ze vóór zijn! De propaganda speelde in op een fundamentele angst voor overheersing, op de overlevingsreflex van een volk. Allusies op gebeurtenissen in Burundi - en dan vooral de moord op Hutu-president Ndadaye in oktober 1993 - en de vaak mondeling overgeleverde verhalen over de Tutsi-monarchie van vóór 1959 konden die angst alleen maar versterken. De radio's appelleerden bovendien aan een diepgeworteld minderwaardigheidscomplex van de Hutu's. De Belgische kolonialen die door allerlei metingen hadden aangetoond dat de Hutu's de aangeboren slaven waren, hebben daar een flink deel aan bijgedragen.
Het deksel van de Arusha-akkoorden en het daaropvolgende toezicht van de VN-macht wist het bruisende mengsel even in bedwang te houden maar niet te neutraliseren. Met de aanslag op president Habyarimana kwam het met een verschrikkelijke kracht tot ontploffing: het bracht een gedrocht voort dat de wereld van walging vervulde. Het vertrek van de VN-macht nam de laatste hindernis voor de genocide weg. De moord op tien Belgische Blauwhelmen, waarop Willy Claes eenzijdig besloot de troepen terug te trekken, de lobby van Frankrijk, van de Rwandese ambassadeur in de Veiligheidsraad en van de VN-gezant Booh-Booh dragen een grote verantwoordelijkheid voor een beslissing waar in de Westerse wereld straf op staat: het weigeren van bijstand aan mensen in doodsgevaar. Het Somalië-effect deed ook de belangrijkste supermacht, de Verenigde Staten, terugdeinzen. De soldaten en milities kregen vrij spel in de uitvoering van het goed voorbereide monsterplan: de uitroeiing van de oppositie en van de Tutsi's. Waar de bevolking zich verzette tegen het moorden, gingen de president en de eerste minister zelfs persoonlijk op bezoek om de burgers aan te zetten 'het werk te beginnen' (p. 237-38)... 'De mensen doodden met zoveel bloedlust,' vertelt Leonard. 'Ze hakten en hakten alsof ze over speciale krachten beschikten'" (p.216).

Idiosyncratic (Non-Orthodox) Explanations and 'Theories'

Elimination of 'surplus people'

Rubenstein offers us a single-factor explanation of genocide in The Age of Triage (1983). This assigns the principal trigger of genocide to population growth assuming, as did Malthus, that population growth leads to an unsupportable surplus: such problems can be resolved (according to Rubenstein) by emigration, famine, war, and genocide.
"The lack of any independent definition of surplus people enables his argument to unfold without the reader's noting its circularity. For surplus people are simply, Rubenstein acknowledges obliquely, any people the authorities want to eliminate (1983: 9, 18, 26). Thus, the original thrust of the argument is contradicted without note. And genocide is explained as a program to eliminate people which the state wants to get rid of for various causes. Surely, this is of interest chiefly in the history of recurring fallacies in social theory" (Fein, 1990).
Rubenstein's theory is, in fact, far less 'single factor' than Fein tries to convey (she is somewhat prone to caricature her opponents), but it would take us too far away from our main theme to try to present an accurate picture of Rubenstein's work.

Fear of death as an instigator

Charny (1982) draws on a psychodynamic theory of aggression based on a dialectic between life-affirming forces and life-denying forces, building on the early dichotomy of Freud (1938) between Eros and Thanatos. The fear of death and non-being (see Becker 1973) propels the 'genocider' to sacrifice others to still his fear (Charny 1982, 192-193; cf. Safranski, 1998).
Charny asserts that societies are organized around life-promoting and life-destroying forces, the latter providing the precursors for genocide. However, he does not specify the role of the state, elites, social structure and intergroup conflict in promoting genocide, stressing the basic likeness of human nature and ambivalent potentialities of channeling behavior violently and nonviolently. What he does is to try to trace the sequence of events leading to genocide using Smelser's (1963) theory of collective behavior, based on a 'value-added' process as a model (Fein, 1990; see above).
Lifton (1986) propounds a theory of genocide based on his work on Nazi-doctors, based on medicalized etiology. "The model I propose includes a perception of collective illness, a vision of cure, and a series of motivations, experiences and requirements of perpetrators in their quest for that cure" (467).
Lifton, like Charny, stresses fear of death as the underlying motif of Nazi culture and takes seriously the Nazi perpetrators' fear - as well as their dehumanization - of their Jewish victims who were seen as "the bearer of death and therefore the embodiment of evil" (476-479). This leads him to propose that "Genocide is a response to collective fear of pollution and defilement. It depends upon an impulse toward purification resembling that given collective expression in primitive cultures" (Lifton, 1986, 481).


In much contemporary writing on colonization, especially radical writing, there is a tendency to equate colonization with genocide. This is a conception expressed especially by Sartre (1968).
"Since victory, easily achieved by overwhelming fire-power, provokes the hatred of the civilian population and since civilians are potential rebels and soldiers, the colonial troops maintain their authority by the terror of perpetual massacre, genocidal in character" (Sartre, 1968; quoted in Kuper, 1981).
It is a theoretically untenable position.

Functionalist explanation

In a specifically structural-functional analysis, Dadrian (1974) advances the proposition (in the peculiar language of functionalism) that "even though in conception, design and execution, genocide may be regarded as a phenomenon sui generis, in terms of underlying structural contingencies and projective goals, it is functional; it subserves the ultimate end of equilibrium of a system beset by disarray through acute group conflict" (Dadrian, 1974: 5). (Small wonder that structural-functionalism went bankrupt).

Human destructiveness: aggression theories

To these theories focused on specific contexts or strata, one might add any of the more general theories of aggression and/or violence such as frustration-aggression theory (e.g. Dadrian, 1974). Then too, given the fact that genocide is almost invariably a crime of governments, or of organized groups, it is to be expected that élite theory should be invoked. The following proposition seems plausible enough: "When the ruling élites decide that their continuation in power transcends all other economic and social values, at that point does the possibility, if not the necessity, for genocide increase qualitatively. For this reason, genocide is a unique strategy for totalitarian regimes" (Horowitz, 1976).
Many commentators place the responsibility on élites, mobilizing ethnic support in a struggle for power. Kuper (1982) would add his own view that whatever the responsibility of élites, they are working with social forces present within the society, and not creating a genocidal situation out of a vacuum or transforming a harmonious equilibrium into a genocidal conflict.

In addition to these sociological theories, there are theories of human destructiveness which would derive genocide from the 'biological' nature of man (e.g., Lorenz, 1977) or his personality structure as moulded by the advance of civilization (e.g., Fromm, 1975).
Lorenz, as well as Tinbergen (1968) and Koestler (1978), argue that Homo sapiens may be an aberrant biological species, an evolutionary misfit, afflicted by an endemic disorder, the source of which lies in the conflict between the new and the old structures of the brain. The neo-cortex, seat of our intellect, has failed to establish proper control over our ancient brain, seat of archaic emotion-based beliefs, the reptilian brain which we carry inside our skulls (Kuper, 1982). This theory is based on MacLean's (e.g., 1990) Triune Brain theory; see Ch. on Evil).
By contrast, Fromm (1975) places the emphasis, in his analysis of human destructiveness, on the interaction of various social conditions with man's existential needs. He distinguishes two different types of aggression, defensive aggression and malignant aggression. Defensive aggression is part of human nature and is a benign form of aggression.
"However, man differs from the animal by the fact that he is a killer; he is the only primate that kills and tortures members of his own species without any reason, either biological or economic, and who feels satisfaction in doing so. It is this biologically nonadaptive and nonphylogenetically programmed 'malignant' aggression, that constitutes the real problem and the danger to man's existence as a species" (Fromm, 1975). The malignant forms of aggression arise out of man's reactions to conditions which are unfavorable to the development of his genuine needs and capacities, his human self-actualization and creative power as ends in themselves.

Human destructiveness: fear and the 'paranoid potential'

Fear, man's 'hypertrophied threat perception' (Fromm, 1975), and 'paranoid potential' (Storr, 1991) as potent (causative) factors in all kinds and forms of human (and animal) violence has long been recognized by many researchers (for a thorough and extensive review see Van der Dennen, 1985). It has not been a common observation in genocide research, however (in contrast to its eminent role in the etiology of war atrocities, more fully dealt with in the next chapter).
Fear-inspired violence is commonly more bloody, destructive, disproportionately savage and cruel, extreme, quixotic, bizarre and disgusting. At the same time, man's threat perception is hypertrophied, as Fromm (1975) explained in great detail. We are able to detect dangers and threats to our possessions, world view, group-identity, or self-esteem in even the most powerless persons, if not now then at least in the near future. This is what Storr (1991) called the 'paranoid potential': "When threatened, we look for saviors and devils: saviors [strong leaders] who will rescue us, devils [scapegoats] to blame for our predicament" (p. 122).
In her lucid observations of the recent Hutu-Tutsi genocide, De Temmerman (1995) clearly recognized the role of fear as a precipitant of the massacres, as well as the close connection between fear and hatred in the violence-provoking propaganda of 'Radio milles collines':

De angst- en haatzaaiende propagandamachine heeft niet alleen ingespeeld op de armoede en uitzichtloosheid van dit volk, ze heeft ook geappelleerd aan een fundamenteel minderwaardigheidsgevoel bij de Hutu, die volgens de stereotypen kleiner en gezetter is. Het is een complex dat als een kankergezwel is gegroeid gedurende eeuwen van feodale Tutsi-overheersing, aangewakkerd is in de koloniale periode en dat door een samenloop van omstandigheden - de staatsgreep in oktober in Burundi - op een verschrikkelijke manier is opengebarsten. Uit frustratie en gebrek aan morele weerbaarheid hebben ze de rijzige, slanke Tutsi's letterlijk een kopje kleiner gemaakt, hun vingers verkort, hun benen afgesneden. De angst dat de heersers van weleer weer de heersers van morgen worden - een angst die gevoed is door al dan niet correcte verhalen uit het verleden - heeft het licht van de rede uitgedoofd. (p. 227).

She also noticed the role of fear in the silent collaboration and the behavior of 'innocent bystanders':

Veel Hutu's wilden niet meewerken aan de slachtingen maar hadden ook de moed niet om de Tutsi's te beschermen. De Hutu's van het zuiden wendden dus onverschilligheid voor opdat men hen niet voor bondgenoten van het RPF zou aanzien. Ze applaudisseerden voor de moordenaars, uit angst. Of ze deden mee om te kunnen plunderen. De meeste Hutu's van het noorden hebben daarentegen wel van harte meegedaan. De vrouwen waren het ergst. Jaloezie speelt daarbij een rol. Er werd altijd gezegd dat de Tutsi-vrouwen de mooiste vrouwen van het land zijn (p. 116).

The Hutu, vasty superior in number, remembered past years of oppressive Tutsi rule, and many of them not only resented but feared the minority.
But this genocide was not an uncontrollable outburst of rage by a people consumed by "ancient tribal hatreds". Nor was it the preordained result of the impersonal forces of poverty and over-population. This genocide resulted from the deliberate choice of a modern elite to foster hatred and fear to keep itself in power (Human Rights Watch Rwanda Report, 1999).
Like the organizers, the killers who executed the genocide were not demons nor automatons responding to ineluctable forces. They were people who chose to do evil. Tens of thousands, swayed by fear, hatred, or hope of profit, made the choice quickly and easily. They were the first to kill, rape, rob and destroy. They attacked Tutsi frequently and until the very end, without doubt or remorse. Many made their victims suffer horribly and enjoyed doing so.
Hundreds of thousands of others chose to participate in the genocide reluctantly, some only under duress or in fear of their own lives (Human Rights Watch Rwanda Report, 1999).
The genocide was not a killing machine that rolled inexorably forward but rather a campaign to which participants were recruited over time by the use of threat and incentives. As authorities played on popular fears and greed, some people picked up their machetes and came readily. Others came more slowly and some refused to come, even at the risk of their lives (Human Rights Watch Rwanda Report, 1999).

We may imagine that so-called normal people could never believe in anything so ludicrous as the delusional systems of the insane. Yet, historical evidence suggests the opposite. Whole societies have been persuaded without much difficulty to accept the most absurd calumnies about minority groups (e.g., witches, heretics, Jews, 'enemies of the people') portrayed as enemies of the majority. Such accusations, Storr (1991) claims, originate from a particular type of fantasy which is comparable with, indeed equivalent to, paranoid delusions of the kind found in psychotic subjects. As Cohn (1975) explains: "The essence of the fantasy was that there existed, somewhere in the midst of the great society, another society, small and clandestine, which not only threatened the existence of the great society but was also addicted to practices which were felt to be wholly abominable, in the literal sense of anti-human" (p. xi). The practices of which this clandestine society was accused included ritual slaughter of babies or young children, often accompanied by drinking their blood; cannibal feasts at which their remains were eaten; erotic orgies featuring every variety of sexual behavior, including homosexuality and incest; and a caricature of religion ('black masses').
The fantasied practices of the clandestine society sound like a list of all the worst sins the conformist citizen can think of, and so combine to constitute an image of absolute evil (Storr, 1991). In his seminal book Warrant for Genocide, Cohn (1967) has demonstrated that medieval paranoid fantasies about a Jewish conspiracy were revived in modern dress and provided at least part of the motive force behind the Holocaust. "Exterminatory antisemitism appears where Jews are imagined as a collective embodiment of evil, a conspiratorial body dedicated to ruining and then dominating the rest of mankind. This kind of antisemitism can exist almost regardless of the real situation of Jews in society" (Cohn, 1967: 252).
"Over a period of some eight centuries, and over a wide range of countries, the myth of a Jewish world-conspiracy has enabled organised groups to kill Jews, by providing them with an ideology and by bewildering the rest of the population" (Cohn, 1967).

This ludicrous fantasy became a keystone of Nazi propaganda and was extensively employed by Rosenberg, Streicher (editor of the rabidly antisemitic periodical Der Stürmer), and Goebbels to foment antisemitism. When Hitler came to power, he exploited the paranoid potential latent in the minds of men more successfully than any other leader has ever done. He was building on a long tradition of quasireligious mystical writings claiming that there had once been a golden age in which an uncorrupted race of Aryans had ruled supreme (but now degenerated by interbreeding with inferior 'races'). Goodricke-Clarke (1985) has recorded the extraordinary but influential racist and nationalist fantasies of writers like Guido von List and Lanz von Liebenfels. If anyone still questions the power which myth, as opposed to reason, exercises over the human mind, he should read The Occult Roots of Nazism.
The more insecure and helpless people feel, the more will they look for a savior to rescue them and a scapegoat to blame for the crisis. The pattern of social collapse followed by the emergence of a leader who both makes promises of a millennium to come and identifies the enemies of the society is a familiar one which has repeated itself throughout history (Storr, 1991).
"The appeal of Nazism was based on powerful fantasies designed to relieve acute feelings of anxiety, defeat, and demoralization. An anti-German conspiracy of Jews and their minions was supposed to be threatening the very survival of the German nation... Only the total destruction of the Jews could thus save the Germans and enable them to enter the promised land" (Goodricke-Clarke, 1985, p. 203).

Kuper (1982) presents some general observations based on the preceding exposition of theoretical positions:

(1) Though animals do engage in intra-specific killing, Kuper (1982) holds, genocide is essentially a human crime. But this does not mean that it is rooted in human nature. There are convincing arguments to the contrary. Conflict of a potentially genocidal character is not the normal pattern of interaction between social groups. Even in our contemporary world, ravaged as it is by genocidal conflicts, most societies develop and relate to each other without interruption by group annihilating destruction (see also Ch. on Evil)..

(2) Since genocide is a crime against a collectivity, it implies an identifiable group as victim. The more specifically focused theories we have been discussing deal with hunting and gathering peoples; hostage groups; the colonized; other nations; the racial and ethnic sections of a society. Maximum identifiability is present where there are marked racial differences: but cultural differences may be equally divisive.
The plural society, discussed by Kuper, is by definition characterized by the presence of identifiable groups, He uses the term to describe societies with persistent cleavages between racial, ethnic or religious groups. They are considered to be a major arena for genocidal conflict.


Staub (1989) views the origin of genocide in 'difficult life conditions' (depression, war, social disorganization, etc.) which, reinforced by cultural and personal preconditions (involving orientation to hierarchical authorities, devaluation of out-groups, aggressiveness) lead to motivational readiness to defend the self (physically and symbolically) and to harm others. These preconditions make people more likely to devalue, scapegoat and hurt other groups, and to follow an authoritarian leader idealizing their group over others and justifying punishing or eliminating the others.

This is a probabilistic conception. The combination of difficult life conditions and certain cultural preconditions makes it probable that motives will arise that turn a group against another. This combination makes it probable that initial acts of harm-doing will be followed by further steps along the continuum of destruction. The behavior of bystanders can facilitate or inhibit this progression. Genocide arises from a pattern, or gestalt, rather than from any single source. The outcome of this evolution and the immediate cause of the genocide is that perpetrators come to believe either that the victims have something they want or (more likely) stand in the way of something they want (Staub 1989, p. 23).

Fein's (1990) major criticism is that Staub has overgeneralized the preconditions - hard times (given the variety of his indicators) have been prevalent in many non-genocidal societies and not sufficiently specified the role of historical situations and structural preconditions that render the state vulnerable to dissolution, revolution, or seizure by elites apt to authorize genocide.

Some specific internal factors which lead toward genocide and politicide can be identified, based on the observations of Gurr (1988 et seq.), Harff (1986 et seq.), Gurr & Harff (1994), Harff & Gurr (1995) and others (e.g., Kuper, 1981; Wallimann & Dobkowski, 1987; Fein, 1990, 1993). Plural (multi-ethnic) societies, especially those in which one communal group has dominated others, are especially prone to political mass murder. The violent history of a state, i.e., its recurrent abusive treatment of opponents and minorities, also figures prominently in the genesis of future geno/politicides. A violent history helps establish, and is reinforced by, a cultural disposition to accept violence as a means to maintain power and to settle disputes between peoples. This is fertile ground for the emergence of exclusionary, racist doctrines of national protection or social purification. Such doctrines help justify the destruction of victim groups by blaming and dehumanizing them.

Staub further mentions as preconditions and predisposing characteristics:

* Strong respect for authority and strong inclination to obedience are other predisposing characteristics for mass killing and genocide.
* A monolithic, in contrast to a pluralistic, culture or society is another important precondition (This contrasts inexplicably with Kuper's and most other researchers' view).
* Another important factor is the (legitimating) role of bystanders.
* Genocide is usually organized and executed by those in power, by a government or ruling elite.

Social-psychological features such as external threat, opposing beliefs, competition for scarce resources, and frustration-aggression-scapegoating syndromes are often present in social systems without genocide taking place. These characteristic are often extant during wartime conditions (Fein, 1990).
Gross (1978) claims that genocide is a function of political tensions between nation-states.
Thompson and Quets (1987), employing an 'unbounded' definition of genocide, which includes phenomena from a single lynching or murder to a massacre or pogrom to full-scale organized protracted extermination - relate genocide to the normative climate. But they do not go beyond this to explain how and why the normative order changes and genocide is instigated.

Based on a number of authorities (Lerner, 1992; DuPreez, 1994; Renwick Monroe, 1995; Harff & Gurr, 1995, etc.), certain background conditions appear as virtual prerequisites of genocide.

1. Genocide most commonly occurs in a pluralist (multi-ethnic) society in which diverse racial, ethnic, and/or religious groups experience persistent and pervasive cleavages. These cleavages may exist in many different spheres, from institutionalized inequality in political incorporation to more informal socioeconomic inequalities. Genocide is greatly facilitated when such long-standing inequalities in political participation overlap accentuated economic and social cleavages and when there is a history of conflict between the groups.

2. Unstable political conditions that threaten the social order compound the effects of these preconditions. Wars and revolutions are particularly powerful triggers for genocide; this is especially true when they bring geographic and psychological dislocations. A government that is losing battles grasps greedily for scapegoats. Beyond this, a modernizing minority may be politically linked to an enemy of the host state. This was the case for the Armenians, who had close ties to Armenian communities in Russia. Economic downturn also prompts genocide as may the disintegration of communism in the former Yugoslavia.

3. What critical psychological factors trigger genocide? There must be an intentional identification of a particular group for destruction, and these individuals must differ in some way from the majority population. Most frequently these differences are in ethnicity, race, or religion, and are reflected in distinct communal dissimilarities... Allegiances and goals are thus ascriptive and flow inevitable from birth. This sets genocide apart from other conflicts in which allegiances can shift and opponents can be converted.

4. Psychologically, both a scapegoat and a victim are needed to explain the disintegration of the old economic, political, and social order and to justify the beginning of the new. The stereotypes of wealthy, cosmopolitan Jews and affluent Armenian merchants provided handy scapegoats; the good German people and the noble Turkish peasants filled the useful role of victims. This may explain the power of genocide at both the local level and at the level of more sophisticated political leadership; both elite and masses respond to threats to their economic situation, political power, and way of life. The political elite may do so in a more cynical and calculating way while the followers - the ones who actually perform the genocidal acts - are moved out of their own personal frustration and hostility as much as in response to orders.

5. Finally, secrecy helps (e.g., Taliban in present-day Afghanistan).

Fein's subterfuge: "killing on state order needs no special explanation"

The most puzzling and enigmatic paragraph in Fein's (1990) monograph is the introduction to the section on Social-Psychological Theories (p. 43). It reads like this:

What is common among these theories is their starting point; they ask how people can kill others because they are members of another group or class or hold the wrong political views, beginning with individuals as the basic unit of analysis.
    To begin with, I must confess a bias. It seems to me that this has already been answered and that inferring psychological motives is basically a complementary level of explanation to a macrosociological perspective. What is needed is a theory of how structural, situational and cultural forces lead potential perpetrators of genocide to define the situation so as to mandate and justify the killing of the victims. One might even refocus on what is taken for granted - the right of the state to authorize killing - and assert that killing on state order needs no special explanation, given the long-term record of people killing at the call of tribal and national leaders.

Rereading this passage for the 100th time, I still do not understand what she means - and I am still at a loss whether I have to doubt my own faculties of judgment or whether it is just plain gibberish. What exactly has already been answered? The question how people can kill others without compunction? Then what is the answer? That killing on state orders needs no special explanation? Is that an answer? If that is an answer, then why bother to explain genocides, or anything else for that matter. The question why some human beings - that is, not all human beings - are ready, willing and able to kill on state orders, or anybody's orders, is the most central, crucial and important question to be answered yet, as well as its counterpart: why other humans are not susceptible, and resist and refuse to do so. Research on combat motivation during the major wars since World War II has revealed that the majority of soldiers are not ready to kill on state orders, despite special training and indoctrination. If people were so prone to killing without compunction as Fein suggests, there would be little need to develop Posttraumatic Stress Disorders, as so many Vietnam veterans did. To pretend that these questions have already been answered is at least premature. Not even the most cynical sociobiologist would contend that killing on state orders is in any way 'normal' and not in need of explanation. On the contrary.

Criticism of the Plural Society, Scapegoat and Political Opportunity Theories

As Valentino (2000: 14-17) points out, national crises do not inevitably or even usually lead to mass killing. While national crises are relatively common events, mass killing is rare.
More importantly, historical evidence casts doubt on the relevance of the basic psychological mechanisms described by many proponents of the scapegoat theory. The scapegoat theory suggests that mass killing serves to alleviate the psychological frustration and fear generated by national crises in society at large. If this claim is correct, we should find that mass killing is usually a popular undertaking among members of dominant social groups. Members of these groups should demonstrate spontaneous support for the killings.
History provides little confirming evidence for these propositions. Mass killing is often carried out with little public support, even from members of dominant social groups. It is difficult to imagine, for example, how the Great Terror in the Soviet Union could be characterized as a reaction to the desires of Soviet society.
Other cases of mass killing in countries such as Cambodia, El Salvador, Uganda, Guatemala, Mozambique, Algeria (since 1992) and Afghanistan (since 1989), to name only a few, also took place with very little public support. These killings were not driven by 'bottom-up' public discontent or a popular desire to blame others, but rather by powerful political and military interests working from the top down.
Even the Holocaust offers considerable confounding evidence for the scapegoat theory. This evidence is particularly relevant since proponents of the scapegoat theory often cite the Holocaust as the paradigmatic example of this explanation. Although a small group of authors, most notably Daniel Goldhagen, assert that most Germans wished to see the Jews exterminated, the weight of scholarly opinion firmly opposes this interpretation (e.g., Steinert, 1977: 132-47; Kershaw, 1983: 274-77; Gordon, 1984; Broszat, 1986; Marrus, 1987; Kulka, 1989; Bankier, 1992; Friedländer, 1997).
The Nazi government ultimately felt compelled to try to keep the final solution secret from the German people.

Another family of explanations which focuses on the effects of national crises suggests that these events encourage genocide not because they prompt societies to seek scapegoats, but because these events provide the incentives, opportunity and cover for revolutionary elites seeking to consolidate political power or implement genocidal ideologies.
Although the political opportunity theory offers a convincing explanation of the correlation between national crises and mass killing, it ultimately offers little leverage for understanding or predicting genocide and mass killing. After all, as Valentino (2000: 18) notes, the majority of national crises, be they wars (won or lost), civil wars, revolutions or severe economic depressions, do not lead to mass killing.

A closer look at the history of mass killing indicates that unusually deep, preexisting social cleavages are neither necessary not sufficient conditions for mass killing.
Valentino (2000: 9) notes that proponents of the plural society theory offer surprisingly little evidence that social cleavages are more intense in societies that have experienced mass killing than in those that have not. Recent quantitative research on ethnic conflict and genocide has found little correlation between the severity of ethnic, social, economic and cultural diferences and the likelihood of large-scale violence between groups (e.g., Singer, 1994; Rummel, 1995; Fearon & Laitin, 1997; Krain, 1997). As some proponents of the plural society theory acknowledge, even societies torn by exceptionally deep social cleavages can exist for long periods without erupting in mass killling (e.g., Kuper, 1981: 189-209).
Conversely, history also provides numerous examples of mass killing between members of similar social and economic groups with little history of unususally intense discrimination or dehumanization. Mass killings carried out by communist governments in the Soviet Union, China and Cambodia - three of the bloodiest episodes of mass killing in history - are especially illustrative. Although specific ethnic and national groups wre singled out in each of these cases, vast numbers of victims were also drawn from dominant social groups. Ethnic and national minorities appear to have been targeted primarily because of suspected resistance to communist rule, not because of their ethnic or national affiliations per se. Nor can deep cleavages between economic classes explain the violence in these countries.
Social cleavages and dehumanizing attitudes often seem more like an effect than an independent cause or precondition of mass killing. It is easy to overemphasize the importance of cleavages between groups after violent conflict has broken out. Widespread violence tends to exacerbate the negative aspects of group relations, making it harder to appreciate long periods of relative harmony and cooperation between groups. Severe discrimination, moral exclusion and dehumanizing attitudes, however, are often the result of a consciouis strategy designed to facilitate violence against victim groups, not preexisting features of society (Valentino, 2000: 11-12; Freeman, 1991; Human Rights Watch, 1995).

The Social Psychology of Genocide: Normality of the Perpetrator

Psychopathology and bloodthirstiness of leaders

Frank (1994) investigated the etiology of the bloodthirstiness syndrome in five contemporary and recent national leaders (Saddam Hussein of Iraq, Idi Amin of Uganda, Macias Guema of Equatorial Africa, Joseph Stalin and Adolf Hitler.
As to character traits, all five leaders fit the description of paranoid political actors (Robins & Post, 1987). Their suspicion of others' hostile intentions was coupled with ostensible supreme self-confidence and grandiosity (delusions of grandeur or megalomania). In moderation such attitudes are adaptive. To climb the ladder of power successfully, all aspiring leaders must have confidence in their own abilities, and must also be alert to detect clandestine efforts of subordinates, rivals and superiors to block their rise or bring them down. Bloodthirsty leaders, in addition, would have every reason to suspect that numerous persons whose relatives they have tortured and killed were secretly plotting revenge. As a result they are hypervigilant (Crum, 1993), interpreting about everything as justifying their suspicions.
The longer bloodthirsty leaders remain in power, moreover, the more numerous the clandestine enemies they acquire.
Knowledge of the childhoods and (psychopathological) personalities of leaders and followers can inform us about their susceptibility to fanaticism but it cannot, by itself, explain mass killing and genocide (Staub, 1989).

Perhaps the most extraordinary fact about the psychology of genocide is that there is no need to search for abnormality. Genocide is the work of perfectly normal and ordinary people. We cannot explain genocide (and other kinds of political violence) in terms of psychological abnormality. It has never been necessary, as DuPreez (1994) put it somewhat sardonically, to empty the lunatic asylums in order to recruit people for pogroms or genocides or holy wars.
Also Staub (1989) states that earlier characterizations of perpetrators as psychopathic seem to be unfounded on the limited evidence we have from assessments of the personality of major Nazi war criminals (based on Rorschach data); when proper scientific controls were employed in analysis, investigators concluded there was no reason to distinguish them from other men (Borofsky & Brand 1980). The investigating psychiatrist at Nuremberg (1946), Douglas Kelley reported: "From our findings, we must conclude not only that such personalities are not unique or insane, but also that they could be duplicated in any country of the world today" (Borofsky & Brand 1980, 362).

Perpetrators of genocide may be self-selected on the basis of readiness to commit violence or they may learn what is expected of them (Staub, 1989).
A mosaic of information suggests that perpetrators often have one or both of two constellations of characteristics: Staub (1989) calls these potentially antisocial and authority oriented.
Perpetrators can be ordinary people who have long filled certain roles _ prison guards, combat soldiers - in which the devaluation of some other people is inherent. If the definition of their role comes to include acts of cruelty, many will adapt (cf. the Sanford simulated prison experiment by Zimbardo et al., 1974; see section on Moral Drift).

Ordinary men

Browning (1991), in his fascinating book Ordinary Men: Reserve Police Battalion 101 and the Final Solution in Poland, describes how on July 13, 1942 a unit of German reserve police arrived in the village of Jozefow in eastern Poland. Methodically, they rounded up 1500 Jews, marched them into the woods, ordered them to lie face down, and shot them in the backs of their heads.
Browning, an American historian, points out what is so unusual in this "numbingly familiar tale of mass murder". The German police who carried out this 'execution' were not young recruits raised on Nazi propaganda but middle-aged family men from Hamburg of working-class and lower middle class background. Their formative years had been spent in the pre-Nazi era. They had been considered unfit for regular army duty and had been drafted into this Reserve Police Battalion. Most had not seen military action before arriving in Jozefow. But what was particularly unusual and is of much interest to us is the fact that these 500 men were given an opportunity to not take part in the massacre: the battalion commander, a major Wilhelm Trapp, apparently disconcerted by the orders he had received, announced to his men that morning that anyone who did not feel up to the task could exempt himself by stepping forward. About a dozen of the men did so. Once the executions began, a few others balked at shooting women and children; some became sickened by the sight of blood and death. They too were allowed to drop out without any sanctions or penalties. At least 400 of the 500 men voluntarily took part in the killings, then, and later went on to other assignments.
How did the men themselves explain their actions when queried about them in the 1960s? The first surprise, as Browning (1993) notes, is that antisemitism was never offered as an explanation. "What is clear is that the men's concern for their standing in the eyes of their comrades was not matched by any sense of human ties with their victims. The Jews stood outside their circle of human obligation and responsibility" (as will be seen later on, indifference and dehumanization are constant psychological components of genocide).
When the Hamburg state prosecutor investigated the unit in the 1960s, many of the men explained their behavior as resulting from a fear of 'looking cowardly' before their comrades; or not wanting to 'lose face' or admit they were 'too weak'. Some presented the argument that even is they had refused to participate, this would not have saved the Jews. Another explanation was lack of choice. We also find a kind of twisted compassion, the idea that the men were putting people out of their misery. Finally, although ethical or political principles were seldom offered as a reason for asking to be excused, something that may border on this category does occur: sheer physical revulsion and extreme distaste for the messiness of the task. One of the men testified: "Truthfully I must say that we did not at the time reflect about it all." (Browning , 1991, 1993; Moses, 1993; Renwick Monroe, 1995).
In his conclusion, Browning offers another explanation: dehumanization and distancing (emphasis added). Browning dismisses the literature that attributes wartime atrocities to the frenzy of battle since this was not battle. It is very clear from Browning's account that for the men of Battalion 101 the killing was neither impersonal nor bureaucratic. Hannah Arendt's explanation of the banality of evil thus cannot apply, for these men killed face to face. They "were quite literally saturated in the blood of victims shot at point-blank range. No one confronted the reality of mass murder more directly than the men in the woods at Jozefow. Segmentation and routinization, the depersonalizing aspects of bureaucratized killing, cannot explain the battalion's initial behavior there").
Browning also considers and rejects, some of the traditional explanations found in the literature, such as geographic origin, social background, party membership, and Steiner's (1980) argument of a sleeper effect. Browning seems to agree more with Zygmunt Baumann (1991) who argues that most of us are capable of great cruelty and fall into the roles society assigns us. Browning also considers Theodor Adorno's (1950) argument on authoritarianism and Stanley Milgram's (1974) work on obedience to authority - not obedience out of fear of immediate reprisal but rather obedience at a more general level of deference and arising out of long-term socialization. Browning also discounts this explanation for the men of Battalion 101 (Renwick Monroe, 1995)
Browning does not believe that the incessant propaganda of Germanic racial superiority was sufficient to prepare the men for the killing at Jozefow. He argues that something more had to occur. While Browning does not label them as such, his critical factors are what Renwick Monroe (1995) would call identity and group ties. By refusing to shoot, a man left the dirty work to his comrades. Refusal therefore constituted an asocial act toward one's fellows. It also left the defecting policeman open to ostracism and rejection, a significant factor for men living in a hostile occupied country, with only their comrades to provide support and social contact. Refusing to shoot also constituted a moral statement that what the others were doing was wrong; hence, Browning argues, men who refused to shoot couched their refusal in terms of being too weak, not too humane, to kill. Such a stance implied no criticism of the others, but rather reinforced the dominant ethos of toughness that the unit so carefully cultivated. Browning concludes: "Pervasive racism and the resulting exclusion of the Jewish victims from any common ground with the perpetrators made it all the easier for the majority of the policemen to conform to the norms of their immediate community (the battalion) and their society at large (Nazi Germany)".
Renwick Monroe (1995) regards as the key psychological factor in the distancing that is critical for genocide: the idea that psychological or emotional connection to a victim affects one's willingness to participate in inhumane behavior toward that person. The final critical pschological factor is the dehumanization of the 'other', which contributes immeasurably to the psychological distancing that facilitates the killing (emphasis added).

Though psychopaths and sadists in Nazi Germany and elsewhere frequently made the most of opportunities afforded by genocidal programs, the preponderance of crimes of mass hatred can be traced to those whom psychologists would regard as 'normal' (Kressel, 1996: 259).
On the other hand, one should reject the theory that mass murderers are everyday, nice people who carry out their orders dutifully, and with the greatest of guilt and regret. Few participants in crimes of mass hatred comply out of the mechanical, slavish obedience found in the Milgram experiments. And very few kill because they fear death if they do not obey. Situational pressures push many to join a genocidal program, but these pressures are more subtle and drawn out than those in Milgram's experiment. Most perpetrators of atrocities understand exactly what they are doing, yet they persist. Few feel much troubled by conscience.
Still, there are exceptions to each of the above generalizations. To understand why a particular individual participates in mass murder, one must consider many possibilities. Individuals may participate because: (1) they fear punishment by superiors if they do not; (2) they are deeply committed to an ideology of hate; (3) they lack awareness of the consequences of their actions; (4) they possess very weakly developed consciences; (5) they seethe with hatred for the target group; (6) they see murder as a means of obtaining material reward; (7) they want revenge for real or imagined offenses; (8) they fear retaliation of punishment for crimes already committed by their group; (9) they encounter situations where they cannot control their aggressive or sexual drives; (10) they cannot muster personal resources to disagree with their peers; (11) they perceive no justification for disobeying orders from a legitimate auhtority; or (12) their sense of identity depends on continued association with the killing group.
Individuals may take part in genocidal or terrorist acts for any of these reasons. But it is possible to classify crimes of mass hatred into two broad categories: crimes of submission and crimes of initiative. These categories reflect the psychological activity or passivity of the killer (Kressel, 1996: 259-60).

Self-serving bias

Psychologists have shown that people naturally have a self-serving bias that they use when they explain life's ups and downs. While people usually take credit for accomplishment, they frequently assign blame for failure to others. In every spot where hatred has flourished, a popular ideology has capitalized on this psychological tendency by exaggerating real injustice and by teaching that all, or most, of a group's suffering arises because of some evil 'other'
Ideologies of hate do more than identify the evil enemy, they excite a lust for revenge. Nearly always, destructive doctrines couch the call to arms in a vocabulary of self-defense: "We must attack them now in order to prevent them from attacking us later". Often, no real threat exists (Kressel, 1996: 249-50).
The risk of murderousness increases greatly when ideologies of hate take steps to dehumanize victims. To carry out the destructive agenda set forth in an ideology of hate, leaders often create special organizations to hone supporters and recruits into sharp instruments of mass murder. These units or 'elite' troops undergo harsh training, designed to eliminate the sensitivities of civilian life. The toughness of the discipline also serves to build committment (Kressel, 1996: 250-51).

Political psychology: individuals as strategists

The schema for interpreting every action is:

who / does what / with whom (or to whom) / where and when / for what purpose / with what consequences / according to what (cultural) rules / in what societal context

The only social scientists who have a slight chance of understanding social situations are those that realize that people are strategists (though frequently very bad strategists) and that to understand them we have to understand their strategies in a societal context. To attempt to understand strategies is much more useful than to attempt to grasp the 'laws' of psychology (DuPreez, 1994).
What are the elements of a social psychology of politics - one which would enable us to understand events such as genocide?
According to DuPreez (1994), on which the following paragraphs are largely based, the first is culture. People act within a set of assumptions or rules.
The second important basic ingredient is the concept of a movement. People join movements, or create movements, or support movements. A movement is a group of people who act together (some of the time - no collection of people is in complete harmony) to achieve a political aim. They want to take over the government or influence the government. Movements are the social units of action in politics. Movements are essential to secure political power and achieve ideological transformation of the social world. The difference between a harmless crackpot and a deleterious societal force is a political movement.
The third important ingredient of political psychology is the psychodynamics and cognitive functioning of leaders and followers. In particular, we are interested in the process by which their private ambitions, miseries and pleasures are politicized and transformed into public concerns.

* Genocidal movements

Every society has within it a number of movements which differ from each other in many ways. One of these ways is their use of violence to achieve their aims. Every society has movements which range from genocidal to pacifist, with others somewhere in between.
This proposition may be greeted with incredulity by those who live in tranquil societies. 'Where are our genocidal movements?' they will ask. The answer is that the would-be perpetrators of genocides belong to fringe movements of cranks while a society is stable and prosperous. Nevertheless, they exist. They emerge rapidly during troubled times, or when the government appears to be betraying its followers, or when there is a power vacuum. They also emerge when a country is conquered by like-minded people and they are offered an opportunity to put their theories into practice. The Nazis found genocidal collaborators wherever they went. The Nazis were a fringe movement of 'cranks' who would have been totally insignificant in a prosperous society.
(Also Melson (1992: 9-10) has noted that "Without state power Hitler would have remained a harmless crank, and without the opportune context provided by the German revolution and the Second World War, the institutional spheres of the German state and society would not have been organized into a machinery of destruction against the Jews").
If this hypothesis is correct, it explains why the people who commit genocide are not abnormal and can very rarely be classified as mentally ill. They are the normal members of a political movement acting in an unstable and threatening social environment. They are often evil and repulsive, but they are sane.

It is not difficult to understand why parties of violence should emerge during times of catastrophe (shortages, unemployment and economic decline; military defeat; war). Genocidal parties have a particular appeal since they clearly identify the enemy and attribute responsibility for the anxiety and sufferings which people are experiencing. They are seldom inventive. They do not need to stretch the imagination since the chosen victims are those who have always been the enemies. What gives the genocidal movement its freshness is ritual and some new theories to justify an old sentiment.

Psychological processes in groups may have different meanings from those of inviduals. If an individual blames members of a minority group for his problems and his beliefs are not shared, he will be seen as paranoid rather than visionary. Individual solutions to frustration, threat, or incomprehension may include individual violence, psychotherapy, or a new religious faith. Only shared problems, motives, and 'solutions' will lead a group to turn against another. Hoffer (1951) has suggested that "a rising mass movement attracts and holds a following not by its doctrines and promises, but by the refuge it offers from the anxieties, barrenness and meaninglessness of individual existence". Staub (1989) agrees that joining mass movements fulfills important personal needs, but in part it does so by providing doctrines and promises that offer hope, a vision, and a sense of significance.
Membership in a group changes people. Intense group beliefs are intensely defended by denial, selective perception, selective exposure to information, and other methods.
Psychoanalytic thinking suggests that groups, like individuals, project unacceptable aspects of themselves onto others. Pinderhughes (1979) posits an urge to join or come together in groups and an urge to differentiate from the self and repudiate. Those who are repudiated become 'bad'; they possess the rejected and renounced parts of the self, which remains pure and good.

* Psychodynamics and ideological transformation

What interests us here is the process by which private miseries, conflicts and deprivations come to be seen as collective miseries, conflicts and deprivations. The sufferer joins a group and what was private becomes public when experience is collectivized and ideologized. People create theories and movements to achieve this. This process is best called ideological transformation.

We now come to the second part of the process, which is ideological acceleration. People take a series of steps as they join groups, say things which commit them more and more completely from others outside the group. In the movement, ideological acceleration takes the form of a sequence of actions which take you into faster and faster lanes. Each acceleration is the result of an action which removes you further from people at the centre. The sequence consists of acts of increasingly violent contempt for outsiders. It may start with words and uniforms and end in killing. In between, there have been oaths of loyalty, hate slogans, acts of damage to property, humiliations inflicted on enemies and grand processions. A combination of rewards and punishments, of conviviality and terror, is used to define the movement and attach members to it. But the pressure is always to take the next step of commitment. For those who have gone a certain way, there is the pressure not to retreat. The pressure to take the next step frequently takes the form of rewards. Just as frequently the pressure not to retreat takes the form of punishment and terror.
The right and the assumed capacity to define utopia and separate the good from the bad are the crucial properties of the complete ideological accelerator. The most complete separation is always moral, as is the most perfected system of terror. The ideological accelerator is always a moral accelerator, finally splitting the world into two hostile camps. Now we can see where the ideological accelerator leads: it leads to utopia. In more ambitious times, it leads to heaven.
The psychological consequences of this moral process are not difficult to conjecture. The world is split in such a way that those who are victimized deserve what they get (and get what they deserve) [Manichaean dualism; split moral universe].

* Redistribution of hatred

Political parties exist not only to redistribute goods, but to redistribute rage, anger and hatred. Those who have been filled with rage because of defeat, catastrophe, and frustrated expectations can either hate themselves or someone else. It turns out that people prefer to hate and blame someone else (as might be expected on the basis of neodarwinian psychology; see last chapter).
The engine of ideology is hatred. Fascism, Nazism, nationalism, communism, and religious fundamentalism are about the punishment of those who are to blame. They are about smiting the sinners. Of course, the elect are promised good things, but the engine of ideology is hatred. This is why good things for the elect have to be obtained by so much killing. This is why the struggle for peace has to continue until no house is left standing.
The point here is that there is a lot of misery waiting to be converted into anger and directed at enemies. These need not necessarily be the richest nations. Hatred may be directed against hostage groups or neighbours who appear to have blighted the hopes of the people. Violence is usually directed at first against the weakest targets. In revolutionary situations, members of the oppressed group first attack each other ("Die Revolution frißt ihre Kinder").
Finally, people may acquiesce in what a movement does because its ideas ('ideologemes' or 'memes') can be attached to existing ideas. This 'host susceptibility' paralyzes resistance (DuPreez, 1994). Furthermore, religious and ideological 'memes' tend to be intolerant.

* 'Bribery' into genocide

But the ideology of these authoritarian parties and even their seizure of state power are not necessarily enough to trigger a genocide. The leading perpetrators need mass mobilizations to actually implement their agenda.
The Nazi genocide probably would not have been possible without the active or tacit cooperation of many collaborators who did not consider themselves Nazis and, in some cases, even opposed aspects of Hitler's policies, yet nonetheless cooperated in mass murder. Put bluntly, the Nazis succeeded in genocide in part through offering bystanders money, property, status, and other rewards for their active or tacit complicity in the crime (Simpson, 1993).
In the Rwanda genocide, authorities offered tangible incentives to participants. They delivered food, drink, and other intoxicants, parts of military uniforms and small payments in cash to hungry, jobless young men. They encouraged cultivators to pillage farm animals, crops, and such building materials as doors, windows and roofs. Even more important in this land-hungry society, they promised cultivators the fields left vacant by Tutsi victims. To entrepreneurs and members of the local elite, they granted houses, vehicles, control of a small business, or such rare goods as television sets or computers. Many poor young men responded readily to the promise of reward (Human Rights Watch Rwanda Report, 1999).
In some regions, particularly those where Habyarimana's supporters were strongest, authorities needed to do little more than give the signal for Hutu to begin attacking Tutsi. In other areas, such as central and southern Rwanda, where Tutsi were numerous and well integrated and where Habyarimana's party had little standing, many Hutu initially refused to attack Tutsi and joined with them in fighting off assailants.
In some places, authorities apparently deliberately drew hesitant Hutu into increasingly more violent behavior, first encouraging them to pillage, then to destroy homes, then to kill the occupants of the homes (Human Rights Watch Rwanda Report, 1999).
Just as communities were readier to kill some Tutsi than others, so individual Hutu would agree to attack one person and not another, or, in an extension of the same logic, would attack one person and save another (Human Rights Watch Rwanda Report, 1999).

It has been argued that in the case of Nazi Germany material rewards were rather immaterial in comparison with the ideological factor of 'exterminatory antisemitism':

De holocaust vond in Duitsland plaats omdat, en alleen omdat drie factoren zich tegelijkertijd voordeden. Ten eerste grepen de meest fanatieke, rabiate antisemieten uit de geschiedenis van de mensheid er de macht en besloten zij hun moordlustige fantasieën tot een kernonderdeel van het staatsbeleid te maken. Ten tweede deden ze dat in een maatschappij waar hun opvattingen over de joden in hoofdlijnen door de meerderheid van de bevolking werden gedeeld. De derde factor is dat alleen Duitsland in de geo-militaire positie was om een volkerenmoord van een dergelijke omvang te begaan.
Als een van de eerste twee factoren (of de derde) zich niet had voorgedaan, zou de holocaust absoluut niet hebben plaatsgehad. De meest kwaadaardige vormen van haat, of deze nu voortkomt uit antisemitisme of een andere vorm van racisme of welk vooroordeel dan ook, monden niet uit in systematische slachtingen, tenzij de politieke leiding degenen die haten doelbewust tot moorden aanzet door die haat te mobiliseren en in banen te leiden (Goldhagen, 1996b).

In contrast to Goldhagen's concept of 'exterminatory antisemitism' (derived from Cohn [1967]), Fein (1990) states: "... it was not the growth of antisemitism and Jew-hatred but the growth of moral indifference and exclusion of the Jews from the political community that allowed Germans to overlook and deny the deportation and destruction of the Jews (Kershaw, 1987)".

Essentially, Goldhagen's (1997) thesis is that the Holocaust was the culmination of 'exterminationist anti-Semitism', and that the perpetrators participated because they thought the Jews ought to die, that the annihilation of the Jews was socially desirable, that the Jews were a particularly inferior form of subhumans.
According to Goldhagen, Milgram fails to understand the nature of the problem: "Milgram's own experiment undermines the notion that his findings are relevant to an explanation of the perpetrator's claims - though Milgram does not draw this conclusion. By varying the conditions of this experiment, he discovered that the more the people who administered the shocks confronted the apparent pain of the person being shocked, the more frequently they are willing to defy the authority of the Yale University experimenter" (p. 592).
There was no evidence that such recognition had any role in compliance by Holocaust perpetrators. Likewise, Goldhagen dismisses Kelman & Hamilton's Crimes of Obedience (1989) as "flawed, especially with regard to the perpetrators of the Holocaust, because it depends upon the notion that the actors recognize their actions to be crimes" (p.592).
Brannigan (1998: 267) correctly points out that Goldhagen overplays his hand. "Although Hitler's cynical plans for the murder of European Jewry are covered well by Goldhagen, he pays little attention to how the cover of war also permitted Hitler to plan the virtual annihiliation and subjugation of the Slavs, particularly the Russians... Before Operation Barbarossa was finally over, 5,750,000 Russian soldiers fell under German control, but barely a million survived the war (Shirer 1990, p. 1241). Two million were starved to death or died from exposure in open camps during the bitterly cold Russian winter of 1942-1943, whereas another million were never accounted for and may have died from similar circumstances and from mass executions at the hands of the SS. In some camps, all the Asian POWs were screened and shot (Shirer 1990, p. 1242)".
Furthermore, Nazi strategists estimated that "many millions of persons will be starved to death if we take out of the country the things necessary for us" (Shirer 1990, p. 1093). In addition of this deliberate starvation of millions of civilians, the Führer had decided to have Leningrad wiped off the face of the earth.
Brannigan (1998: 268-70) continues: "In the longer term Nazi plans, the eastern territories were to become Germany's breadbasket and a zone of Aryan expansion. The displacement of the existing population was inevitable. In this regard, the actions of the Germans were calculated to bring about massive annihilation, whether of soldiers or of civilians. How are these plans to be explained? Do they reflect 'exterminationist racism' or 'exterminationist anti-Bolshevism'? And in what respect were these different from 'exterminationist anti-Semitism'?....[or exterminationist anti-homosexuality, exterminationist anti-gypsy-ism?]
If we shift our perspective for a moment, it does not appear to me that there was much concern in the British circles into which I was born about how many Germans or Japanese civilians were being killed during the war in Allied bombing raids. I would venture to guess that my clan probably thought that the more, the better. They thought the German civilians ought to die. And it did not take a huge propaganda machine to inculcate racial or ethnic animosities... Goldhagen's case for exterminationist anti-Semitism ultimately falls short of the mark".

Obedience and conformism

People will often alter their reported judgments in order to conform to a majority, but sometimes a minority can swing the majority around by sticking to its guns.
Most people will obey even outrageous instructions once they have accepted an authority. Some people are more given to identify with authority and be hostile to outsiders than others. People most apt to accept authority appear to have three measurable characteristics: submission to authority, conventionalism (or conformism) and aggression towards outsiders.
Everyone who has been a member of an army unit will know how strong the culture of obedience is and how difficult it is to disobey even clearly illegitimate orders. Young soldiers can become torturers if they are commanded to torture by superior officers and are surrounded by young men who have already complied. The strongly reinforced culture of obedience, the close interdependence of men together in danger, the fact that killing is normal in war, and ever-present fear make it easier to obey than to disobey _ often leading to further and even grosser acts (DuPreez, 1994; Staub, 1989).
In military settings, the soldier who obeys orders is described as being loyal and 'doing his duty'., so that disobedience appears morally reprehensible as well as a dangerous infringement of military discipline. Many who took part in the horrors of the concentration camps disowned responsibility for what they did. Even those who were proven to have been grossly cruel defended their actions in terms of obedience. When the notorious Nazi torturer Irma Grese was taxed with cruelty, she said defiantly: "It was our duty to exterminate antisocial elements, so that Germany's future should be assured" (quoted in Storr, 1991, p. 107). Similarly, as Hannah Arendt (1963) observed, Eichmann was not a monster but an obedient civil servant. The subtitle of her book on Eichmann's trial is A Report on the Banality of Evil. (This theme will be considered in more detail in the next chapter).
Storr also relates the methods (mainly Kadaverdisziplin to create an 'elite' and gradual introduction and gradual habituation to torture and cruelty) used in Greece to train torturers when the junta of the colonels was in power.
Milgram's (1974) famous (or notorious) experiments go some way toward demonstrating how it comes about that average citizens can be persuaded to participate in horror. Milgram himself was surprised and dismayed by how far ordinary individuals would go in complying with the experimenter's orders. Subjects interviewed after the experiments repeatedly said: "I wouldn't have done it myself. I was just doing what I was told". Milgram comments:

It is the old story of 'just doing one's duty' that was heard time and time again in the defense statements of those accused at Nuremberg. But it would be wrong to think of it as a thin alibi concocted for the occasion. Rather, it is a fundamental mode of thinking for a great many people once they are locked into a subordinate position in a structure of authority. The disappearance of a sense of responsibility is the most far-reaching consequence of submission to authority (Milgram, 1974, p. 8).

Milgram calls this mode of thinking and acting the 'agentive mode', meaning that the individual acts as the agent or instrument of an authority regarded as legitimate, and consequently 'suspends' his own responsibility and common sense (cf. Sabini & Silver, 1982).
Nevertheless, there usually is a definable point when the individual makes a choice which proves irrevocable. According to Bettelheim (1979) this moment of choice came for Eichmann when he for the very first time visited the extermination camps and saw what happened to the Jews:

He nearly fainted. But instead of heeding his emotional reaction, he pushed it down to go on with the task that he had been assigned and that he embraced as his own obligation. This was Eichmann's point of no return. Then and there he abdicated from reacting as a human being and made himself a mere tool of the state... My thesis is that if one does not stand up to one's experience in accordance with one's values, if one takes the first step in cooperating with the totalitarian system at the expense of one's convictions and sentiments, one is caught in a web that tightens with each step of cooperation until it becomes impossible to break free (Bettelheim, 1979, p. 270-71).

This latter mechanism is called an involvement trap. For many ordinary people, the Holocaust involved a process of gradual and escalating commitments: the subject is trapped by his gradual involvement, whether in a Milgram experiment, or in bureaucratic genocide (cf. Sabini & Silver, 1982).
Milgram postulates that man's capacity to lose his humanity when subordinating his individuality to an institutional structure actually threatens the survival of our species: "It is ironic that the virtues of loyalty, discipline and self-sacrifice that we value so highly in the individual are the very properties that create destructive organizational engines of war and bind men to malevolent systems of authority" (p. 188).

Distance from the victim (whether physical or psychological), Milgram (1974) found, has a profound effect on the level of obedience and on the stress experienced in obeying. It has been noted time and time again by psychologists, ethologists (notably Tinbergen, Lorenz and Eibl-Eibesfeldt), and psychoanalysts that the invention of weapons which kill at a distance has overridden any lingering inhibitions against killing other human beings. "It has also facilitated the commission of acts of cruelty. A pilot who drops napalm upon people he cannot see may do so with no feeling of guilt. If he were ordered to pour gasoline over a child and then ignite it, he might well recoil and disobey. Yet the injuries inflicted would be closely similar" (Storr, 1991, p. 113).

The Milgram paradigm produced considerable anguish in many. To shock an innocent stranger at such high voltage levels obviously was an act of extreme violence against another human being. Most subjects complained and protested. The majority of the subjects dissented but they did not disobey.
Personality tests did not reveal any differences between those people who refused to comply and the majority who showed total, blind obedience to authority.
In these studies the situational forces are identified as: the presence of a 'legitimate' authority who assumes responsibility for the consequences of one's actions; a victim who is physically remote; acceptance of a subordinate role with functions governed by rules; and, finally, allowing oneself to become part of a social system where public etiquette and protocol are more important to maintain than one's personal values and private beliefs (Zimbardo, 1978).
Snow (1961; Cf. Koestler, 1967) reminds us: "When you think of the long and gloomy history of man, you will find more hideous crimes have been committed in the name of obedience than have been committed in the name of rebellion". The role of authority is also stressed by Fromm (1965, 1973) and Alice Miller (1983).
While obedience is an important force, Staub (1989) observes, it is not the true motive for mass killing or genocide. The motivation to obey often comes from a desire to follow a leader, to be a good member of a group, to show respect for authority. Those who willingly accept the authority of leaders are likely to have also accepted their views and ideology.

Also Kressel (1996) discussed the relevance of the Milgram obedience studies.
Insofar as Milgram's experimental situation differs in so many essential aspects of circumstances surrounding mass atrocities in the real world, the study proves little about how such crimes could occur. However, the obedience research does help to explain several elements of mass murder in Nazi Germany, Bosnia, and Rwanda. First, if people lacked a powerful tendency to obey the orders of authorities, the architects of mass murder would not be able to coordinate effective policies; pogroms and other manifestations of mass hatred would occur, but they would not reach the levels observed in Rwanda, Bosnia, Nazi Germany, or elsewhere. Second, perpetrators nearly always find it easier to commit atrocities when they can surrender moral and legal responsibility to an authority whom they perceive as legitimate. Third, some percentage of murderers in every genocide act without any hatred toward their victims, or at least without much hatred, and some experience considerable guilt; these obedient killers are one type among several, not the majority, but they do kill for much the same reason that Milgram's subjects pull the lever. They simply lack the presence of mind or moral courage to challenge an authority figure. Fourth, group pressure apparently drives many people to suppress any inclinations they might have to question or disobey authority.
The 'obedience' model also detracts attention from the overarching importance of an ideology of hatred - the Nazi racial doctrines, the vision of a Greater Serbia, Hutuism, Islamic extremism, and the like. Over the course of years, opportunistic or genuinely hateful leaders use control of the state, the media, and other institutions to flood people's minds with messages teaching them to devalue their victims.
Many atrocities are crimes of obedience, but many more are crimes of agreement and even initiative (Kressel, 1996: 195-97).

Ethnocentrism and perceived injustice

Ethnocentrism, the belief in the superiority of one's own cultural group or society (ingroup) combined with derogation of other groups (outgroups), is probably universal. It has been observed in virtually all preindustrial as well as industrial societies; it can be experimentally induced very easily by mere categorization; and its nonhuman equivalent has been observed in other socially-living animal species. Ethnocentrism, and its concomitant xenophobia, results in the splitting of the moral universe in a manichaean dualism (We [the ingroup] are good and superior; they [the outgroups] are bad and inferior).

"It is not just the seeming universality of ethnocentrism that makes us think it uneradicable but rather that it has been traced to its source in individual psychology, and the source is the individual effort to achieve and maintain positive self-esteem. That is an urge so deeply human that we can hardly imagine its absence" (R. Brown, 1965, p. 534; cf. Horowitz, 1985).

Ethnocentrism alone, however, is not in itself sufficient to bring about hostile action between groups. In addition it is necessary that groups be similar enough so that they compare their outcomes, the distribution of rewards and costs, and find that it is unfair or inequitable. In considering hostility between groups we are concerned not with the state of fairness or equity but with perceived inequity. A state of inequity exists when a person fails to receive a reward for which he is qualified or incurs some cost he has not deserved or, generally, when rewards or costs are not in the expected ratio with investments.
Unfair or unjust disadvantage or frustration is the sovereign cause of anger and aggression. Disadvantage or frustration will not alone do it; perceived injustice is critical. Since good things and bad things are never distributed equally, the perception of injustice is pretty well guaranteed for any groups that compare their outcomes (R. Brown, 1986). Like ethnocentrism, stereotypes are not the cause of group hostilities.

The mechanisms of ingroup/outgroup differentiation (experimental evidence)

Social identity research indicates that the stereotypic behavior and attitudes of the ingroup are positively valued, while outgroup behavior and attitudes are negatively valued. Furthermore, people very easily adopt negative stereotypes about outgroups and these stereotypes possess a great deal of inertia (i.e., they are slow to change and are resistant to countervailing examples).
Social identity theorists propose that it is the need for high self-esteem which drives the entire process (MacDonald, 1994). Also Horowitz (1985) posits the quest for the affirmation of 'personal worth' as a central motive of human behavior: "[S]elf-esteem is in large measure a function of the esteem accorded to groups of which one is a member". Hence, "the sources of ethnic conflict reside, above all, in the struggle for relative group worth".
The contest for group legitimacy, for political inclusion and exclusion, merges with the quest for group worth to form "a politics of ethnic entitlement". For Horowitz, this is the engine of mass ethnic conflict. He also understands ethnicity to be a form of "greatly extended kinship". The particular fierceness, bitterness and cruelty of ethnic conflicts can be understood, at least in part, through the relatedness of familial and ethnic consciousness: "If group members are potential kinsmen, a threat to any member of the group may be seen in somewhat the same light as a threat to the family" (Horowitz, 1985).
The result of these categorization processes is group behavior which involves discrimination against the outgroup; beliefs in the superiority of the ingroup and inferiority of the outgroup; and positive affective preference for the ingroup and negative affect directed toward the outgroup. Although groups may be originally dichotomized on only one dimension, there is a tendency to expand the number of dimensions in which the individuals in the groups are categorized and do so in an evaluative manner.
Besides categorization on specific negatively evaluated traits, social identity theorists have noted that the stereotyping process can also result in scapegoating (i.e., the explanation of complex events as resulting from the behavior of the outgroup). Another common correlate is the process of dehumanizing the outgroup. These tendencies toward ingroup cohesiveness and devaluations of the outgroup are exacerbated by real conflicts of interest (MacDonald, 1994; van der Dennen, 1995).

Tajfel (1970 et seq.), and many other social psychologists, provided experimental support for the hypothesis that an individual will discriminate against a member of an out-group even when (a) there is no conflict of interest; (b) there is no past history of intergroup hostility; and (c) the individual does not benefit personally from this behavior. Mere (random) categorization is sufficient to produce intergroup discrimination and prejudice (see MacDonald, 1992, 1996; and Van der Dennen, 1995 for reviews).
Sometimes the ingroup devalues most strongly another group that is highly similar: this serves to protect the identity, integrity, and purity of the ingroup. The communist hatred of 'revisionist' social democrates was often greater than their hatred for capitalist enemies. Small differences in dogma often resulted in the persecution of religious heretics (DuPreez, 1994).

Long before Tajfel begun his work, Sherif et al. (1961) in the Robbers Cave experiment, and other following his lead, had shown that it is possible to produce ethnocentrism and all the phenomena of group conflict by assigning unacquainted strangers to groups and putting those groups into competition.
One of the few attempts to replicate the Sherif experiments was that of Diab (1970). This experiment had some frightening consequences for the subjects as well as for the researcher who had to be hospitalized for exhaustion after the experiment was abruptly terminated. He had been too successful in arousing intergroup hostility. The conflict got completely out of hand; some boys knifed each other and the police had to evacuate the camp to prevent further violence (Rabbie, 1982).

Anthropological evidence indicates the universality of the tendency to view one's own group as superior (e.g., Davie, 1929; Vine, 1987; Shaw & Wong, 1989; Van der Dennen, 1995; MacDonald, 1996), and the empirical results of social identity research are highly compatible with an evolutionary basis for ethnocentric group behavior. A. Flohr (1994) similarly concludes her extensive review that there is a biological disposition toward ethnocentrism. Lopreato (1984), Irwin (1987, 1990), Shaw & Wong (1989), and Wuketits (1993) provide some compelling arguments why humans are genetically predisposed to ethnocentrism: in the environment in which it evolved, ethnocentrism and xenophobia enhanced individual reproductive success and survival. It can thus be considered to be a (bio)rational disposition.
In addition to the suggestion of universality, an evolutionary interpretation of these findings is supported by results indicating that these social identity processes also occur among 'advanced' animal species such as chimpanzees (e.g., Goodall, 1986).
Moreover, as MacDonald (1992, 1996) points out, the powerful affective component of social identity processes is very difficult to explain except as an aspect of the evolved machinery of the human mind. As Hogg & Abrams (1987) note, this result cannot be explained in terms of purely cognitive processes, and a learning theory seems hopelessly ad hoc and gratuitous. The tendency for humans to place themselves in social categories and for these categories to assume immense affective, evaluative overtones is the best candidate for the biological underpinning of ethnocentrism (MacDonald, 1992, 1996; Shaw & Wong, 1989).
Within the framework of social identity theory, there is clearly no requirement that the beliefs regarding the ingroup or the outgroup be true. Bigelow (1969) notes that "each group requires something intimate, unique to itself, around which its members can cohere. Irrational beliefs serve this purpose far better than rational ones; they are not only easier to produce, but also less likely to be confused with enemy beliefs. Irrational fantasies produce a continuous supply of 'group uniforms', promoting and maintaining internal cohesion within each group, and segregation between groups".
Also Tiger (1969) suggested that "males bond in terms of either a pre-existent object of aggression or a concocted one".

Xenophobia is a widespread trait throughout the animal kingdom, according to Southwick et al. (1974), but it is by no means universal. Among vertebrates, xenophobic aggression has been demonstrated experimentally in a great number of species, especially those with prominent territorial and/or relatively closed social groups, which are organized on a hierarchical basis (e.g., Holloway, 1974; Southwick et al., 1974; E.O. Wilson, 1971, 1975; see Van der Dennen [1987] for a review). The introduction of unfamiliar conspecifics to such groups (e.g., rodents, many primate species) may release massive attacks and even killing from the resident animals.
When it occurs in natural settings, xenophobia may be considered to be a functional and adaptive trait in that it maintains the integrity of the social group. It ensures that group members will be socially familiar. It limits the flow of individuals between groups, and can therefore affect patterns of both social and genetic evolution. Xenophobia has apparently evolved in those species where discrete, bounded social groups are adaptively favored (Southwick et al., 1974).
Also Hebb & Thompson (1968) cite the evidence in favor of the mammal's xenophobia; the fear of and hostility towards strangers, even when no injury has ever been received from a stranger. The enmity aroused by conspecifics which are different (in anatomy, in coloration, in behavior, in language use) or by strangers, may easily lead toward discrimination, ostracism and cruelty in animals as well as man.
Markl (1976) deduced the following general rule from observations such as these: species with highly cooperative social behavior within the group are particularly apt to be very aggressive towards conspecifics that are not members of their group.
Several authors have suggested that mistrust and fear of the foreigner or the stranger may have biological origins. McGuire (1969) discussed the possible genetic transmission of xenophobia: "[I]t appears possible for specific attitudes of hostility to be transmitted genetically in such a way that hostility is directed toward strangers of one's own species to a greater extent than toward familiars of one's own species or toward members of other species. It would not be impossible for xenophobia to be a partially innate attitude in the human". Vine (1987) argued for a genetically primed, generalized, weakly xenophobic and suspicious tendency as a defense against being deceived. Holloway (1974) would submit that at least for adult humans, xenophobic responses are normative unless there has been strong cultural training and conditioning against it. Clannishness, or strong intragroup affiliation coupled with distantiation of other ethnic, religious, racial, or political groups, is an enforcing mechanism for continued xenophobias. The demagogue, he observes, knows this fact only too well.
Trivers (1971) speculated that 'moralistic aggression' - an urge to attack someone who is acting unjustly or unfairly - evolved in humans as indispensable protection against excessive failure to reciprocate altruistic acts. 'Moralistic aggression' seems to be readily mobilized against individuals believed to be deviating from basic group norms and symbolic allegiances; that is, it can help enforce collective intolerance.
On the other hand, Hebb & Thompson (1968) argued that fear or dislike of the stranger is not innate, since it depends on certain prior experiences, yet it still does not have to be taught. "If, therefore, man is not born with a dislike for those who differ from him in habits or appearance, he can still pick up the dislike with no help or encouragement" (Hebb & Thompson, 1968).
Also Hamilton (1975) and Alexander (1979) argue that social interactions of an individual with his close relatives can provide all of the experiential background necessary to produce xenophobia. We tend to react negatively to countenances which are uncommunicative, and which convey contradictory or paradoxical messages.
It is not clear whether the transient phenomenon of the fear of strangers in infants - which predictably develops between 6 and 9 months of age - has any impact on adult xenophobia (See e.g., A. Flohr, 1994). This infantile fear of strangers is also reported in other social species (e.g., canids), and its development does not depend upon aversive experience with strangers. Furthermore, it also develops in congenitally deaf and blind children (e.g., Eibl-Eibesfeldt, 1982).
Although the expression of these predispositions varies, Emmert (1984) and Shaw & Wong (1989) conclude, it seems that initial distrust of social strangers is universal among humans and nonhuman primates. Also H. Flohr (1987) concludes that xenophobia seems to be universal, i.e., it seems to occur in all cultures. This is no proof, he states, but strong evidence in favor of a biological basis of xenophobia (cf. Markl, 1982; A. Flohr, 1994). The biological basis concerns, of course, the tendency towards xenophobic prejudices, not their specific content. Peck (1990) has shown through formal models that mechanisms of outsider exclusion can be favored by evolution.

The role of fear

For months before the outbreak of genocidal violence in April, folk songs, slogans, speeches, and inaccurate news reports (transmitted by Radio Televion Libre des Milles Collines, Rwanda's 'killer radio') demonized the Tutsis and warned against Tutsi plans to retake the land, kill the Hutu, and reclaim their preindependence role of dominance. Broadcasts depicted Tutsis as snakes, animals, and most often, cockroaches. Government leaders would urge the killers on with slightly more cautious euphemisms, telling them to 'clear the bush', 'get to work', or 'clean around their houses'. Peasants understood well these encouragements to slaughter their neighbors.
After Habyarimana's death, Radio milles Collines became a mobile genocidal headquarters, traveling in an armored vehicle. It told the Hutus, "The enemy is out there - go get him" and reminded them that, "The graves are only half full".
Many Hutus in the militias and the general public believed that Tutsis were killing Hutus indiscriminately, that they would continue to do so if they were victorious, and that the only safe route for Hutus was to kill Tutsis first.
As in the case of Bosnia, many outsiders interpret the bloodshed as a consequence of 'ancient' animosities between ethnic groups. As we have seen, the Hutu and the Tutsi are not simply ethnic groups and, at least before the late 1950s, the groups seldom fought. Moreover, many moderate Hutus were numbered among the victims of the extremists. Historian René Lemarchand correctly attributes the violence to "the extent to which collective identities have been reactivated, mythologized and manipulated for political advantage" (1994: 31). But ruthless politicians did not manufacture these identities and historical 'myths' ex nihilo. The historical reality of Hutu-Tutsi relations in Rwanda and Burundi made the manipulations possible.
The precolonial Tutsi domination of Hutus in Rwanda had certain benevolent aspects, but it was clearly an exploitative system. When the system fell apart and the exploited realized what had happened, their anger intensified. The persistence of denigrating stereotypes added to their fury. ("They think they were born to rule; we will show them.") Moreover, many Hutu remained psychologically fearful or returning to their former predicament. When the realistic threat from the RPF appeared, it was a relatively simple matter to activate a nightmare scenario. The genocide, by this logic, had its deepest root in a precolonial social system based on a 'premise of inequality'...
Thus, Serb and Hutu militants created a public atmosphere of fear, where a strategy of mass murder became widely perceived as the only effective defense against attack. In both cases, a willingness to murder dissidents from their own group strengthened the hand of the extremist leaders and shielded them from internal criticism. Finally, trained and well-indoctrinated militias committed the bulk of the atrocities in both Rwanda and Bosnia (Kressel, 1996: 110-18).

Yet it would be wrong to interpret genocide a ethnic cleansing solely as the product of self-interested rational choice. Moral and social inhibitions, under normal conditions, would keep most people from acting on these interests. A few sociopaths might partake, but not nearly so many as formed the backbone of Serbian ethnic cleansing. A process of psychological preparation for genocide had to happen first. One critical step in this process was the reinvigoration of nationalist sentiments. Three more steps completed the creation of a murderous mindset: (1) the intensification of the culture of toughness; (2) the reactivation of historical fears and resentments stemming from the Second World War; and (3) the stimulation of a desire for revenge (Kressel, 1996).
Among many Serbs, before the current conflict, the personal or inherited memories of brutal mistreatment remained vivid. Psychiatrist John Mack has commented that: "[E]thnonational groups that have been traumatized by repeated suffering at the hands of other groups seem to have little capacity to grieve for the hurts of other people, or to take responsibility for the new victims created by their own warlike actions. Victims kill victims through unendingly repeated cycles that are transmitted from one generation to another, bolstered by stories and myths of atrocities committed by the other people..." (Mack, 1990: 125).

Social-Psychological Mechanisms and Dynamics of Distancing

In the following paragraphs we shall discuss a number of 'distancing devices': psychological mechanisms which serve the purpose of creating distance between perpetrator and victim, and rationalizing, assuaging, denying or avoiding feelings of responsibility and guilt about the perpetrated acts of violence.
Bandura (1976 et seq.) mentions as 'mechanisms of moral disengagement' or 'mental distancing devices': (1) reconstructing aggression by palliative comparison; (2) justification of aggression by euphemistic labeling; (3) displacement of responsibility; (4) diffusion of responsibility; (5) dehumanization of victims; (6) attribution of blame to victims; (7) misrepresentation of consequences; and (8) graduated desensitization.

Distancing device: assuaging guilt

There are ways people rationalize guilt that seem so preposterous that one wonders how anyone could have believed them. Their very implausibility shows how hard people have to try to bring themselves to believe them.
The explanation for these bizarre rationalizations is probably twofold. First, the perpetrator wants very strongly to believe them. Second, they are superficially plausible enough that the person can accept them as long as he or she doesn't think about them very carefully. The combination of some kernels of truth and an intense will to believe enable some perpetrators to get by with dubious justifications (Baumeister, 1997).
One strategy for avoiding guilt feelings is based on the interpersonal nature of guilt. People feel guilty when they hurt people with whom they share some kind of social bond. Therefore, to hurt people without guilt, make sure that your victims do not share any bond with you. The lower the fellow-feeling, the less guilt.
The strategy of distancing oneself from one's victims is implicit in the common device of labeling them as subhuman creatures. To kill a human being is wrong; to kill a fly or a rat is acceptable and perhaps even praiseworthy. Perpetrators often denounce their victims as vermin, as slime or filth, as a disease or the carriers of a disease, and so forth. Such remarks are insulting, of course, but they are more than mere insults. They are important devices for securing legitimacy. They justify killing and reduce guilt because they rule out any human connection between perpetrator and victim.
Another version of this strategy is to see the victims as utterly foreign enemies who deserve their fate. It is not surprising that so much bitter violence arises between groups that regard each other as different. Many racial, ethnic, and religious groups see themselves as close, important communities - and see anyone who is different as outside that community.
One of the most popular justifications of evil is the appeal to necessity: "I couldn't help it, I had no choice". Responsibility is typically understood as involving freely chosen behavior. But if there were no options to choose among, then there was no choice, and hence no responsibility.
The most famous and familiar version of the "I can't help it" defense involves following orders. The "I was only following orders" justification became a cliché among Nazis and other Germans who were implicated in killing Jews (Baumeister, 1997).
If guilt arises from harming others, another possible way to ward off guilt would be to claim that the victims actually benefited from what one did to them. A Rwandese woman who had participated in the recent genocide, testified that she regarded killing the children as merciful acts because the youngsters would not be able to survive on their own, now that their parents had been killed (Baumeister, 1997).
"Juliana Mukankwaya van de Interahamwe vertelde in een gesprek met het Amerikaanse persbureau AP in Kigali hoe zij met een andere buurvrouw twee buurkinderen had doodgeslagen. 'Ze gaven geen kik omdat ze ons kenden. Ze zetten allen grote ogen op. Ach, al met al hebben we er zoveel gedood, ik ben de tel kwijtgeraakt.' Volgens de vijfendertigjarige vrouw stond de wezen toch alleen maar een hard leven te wachten nadat hun vaders waren afgeslacht en hun moeders waren verkracht en vermoord" (de Temmerman, 1995, p. 57).
A rather novel and frightening phenomenon in the recent Rwandan Hutu-Tutsi genocide was the active participation of women, even nuns, in the killings. Other Hutu women encouraged the killing and rape of Tutsi women (African Rights Report, 1995; Human Rights Watch, 1996, 1999).
"Het moorden duurde meer dan twee uren. Daarna kwamen de vrouwen en kinderen om te plunderen" (de Temmerman, 1995, p. 110).
"'Aanhangers van alle partijen hebben gedood,' zegt Bernadette Mujawimana die voor een Frans ontwikkelingsproject werkte. 'Ook de militairen, ook de vrouwen - zij zweepten de moordenaars op. 'Jullie zijn Inkotanya,' riepen ze. 'Jullie hebben de dood verdiend.' Maar zij vermoordden ook rijke Hutu-handelaren, voor materieel gewin." (de Temmerman, 1995, p. 111).
When almost an entire hamlet was massacred in Sri Lanka, it was not only the images of children chopped to death that shocked people but the knowledge that the butchery was led by women (the Tamil Tiger guerrillas founded their women's wing 'Freedom Birds' in 1986). According to the press release (Times of India, Sept. 22, 1999 on the internet), United Nations 1998 Human Rights Prize winner Sunila Abeyesekera said the massacre showed that women employed by a fighting machine were no less fierce than males. "It is fatal to have this idea that women will show kindness. Once they have been indoctrinated by a killing machine, be it with the rebels of government forces, there is no difference between men and women... These women may even have a sense of satisfaction that they have done their job well." A witness of the carnage testified "I heard the Tamil women laughing as if they had gone mad."

During the 1994 genocide, Rwandan women were subjected to brutal forms of sexual violence. Rape was widespread. Women were individually raped, gang-raped, raped with objects such as sharpened sticks or gun barrels, held in sexual slavery (collectively or individually) or sexually mutilated. In almost every case, these crimes were inflicted upon women after they had witnessed the torture and killings of their relatives, and the destruction and looting of their homes. Some women were forced to kill their own children before or after being raped. Survivors report that during the genocide, militia even raped the corpses of women they had just killed or women who had been left for dead. Both Tutsi and Hutu women were raped, but there was a difference both in the numbers assaulted and in the reasons for the rape. Most of the women raped were Tutsi and they were attacked as one more means of terrorizing and destroying the Tutsi ethnic group. Hutu women, fewer in number, were targeted ordinarily because they were close to Tutsi: either wives of Tutsi men, supporters of political groups associated with Tutsi, or protectors of Tutsi. Some women were simply caught in the general increase in violence (Human Rights Watch, 1996).
How especially young, poor males may be bribed into genocide by presenting 'a license to kill and rape' as well as presenting a prospect of material rewards can be glanced from the following account:

In Kigali, the Interahamwe and Impuzamugambi ['Those who have a common purpose'] had tended to recruit mostly among the poor. As soon as they went into action, they drew around them a cloud of even poorer people, a lumpenproletariat of street boys, rag-pickers, car-washers and homeless unemployed. For these people, the genocide was the best thing that could have happened to them. They had the blessings of a form of authority to take revenge on socially powerful people as long as these were on the wrong side of the political fence. They could steal, they could kill with minimum justification, they could rape... This was wonderful. The political aims pursued by the masters of this dark carnival were quite beyond their scope. They just went along, knowing it would not last (Prunier, 1995: 231-2).

The government and military authorities gave the militias full license to commit egregrious human rights abuses, including rape, with impunity (Human Rights Watch, 1996). Atrocities can only (?) be committed in a carefully created moral vacuum.

Distancing device: suspension of responsibility

People predisposed to harm-doing may find membership in certain groups highly satisfying. Hostility toward outgroups becomes desirable; the authoritarian structure is familiar and comfortable; the camaraderie provides a haven in a hostile world.
Belonging to a group makes it easier for people to act in ways that are out of the ordinary. They no longer need to take individual responsibility for their actions: no one is responsible, or the group is responsible, or the group's leader (Staub, 1989).
This mechanism of responsibility suspension may well be a human universal.
The Avatip, who live on the Sepik River in New Guinea, see themselves, according to Watson (1995), as essentially peaceful. A description belied by their long history of inter-communal conflict. They fight as often as their neighbors, and as hard, but insist that this behavior has nothing to do with aggression. They do not feel anger or hostility towards their opponents, they claim, but talk instead of putting on a different or 'bad' face. An alter ego made visible in the form of a chest ornament in the shape of a face, or in a terrifying mask. And they change their personalities in the process, becoming 'deaf' to their enemies, putting themselves beyond the reach of sympathy, remorse or shame. They manage this all by magic, by becoming possessed with the spirit of an ancestor or a guardian who relieves them of responsibility for killing. It is a familiar device.

"Feelings of responsibility are subverted by excluding certain people from the realm of humanity or defining them as dangers to oneself and one's way of life and values. At the extreme, a complete reversal of morality may occur, so that murder becomes a service to humanity" (Staub, 1989; emphasis added).

Distancing device: pluralistic ignorance

Research on helping in emergencies has shown that, when a number of people are present, responsibility is diffused, and each person is less likely to help (e.g., Darley & Latané, 1968; Latané & Darley, 1970).
Another consequence is what Latané & Darley (1970) call pluralistic ignorance. People tend to inhibit expressions of feeling in public. In an emergency, the fact that all bystanders are hiding their feelings may lead them all to believe that there is no need for concern and nothing need be done. Hiding reactions is also common when suffering is inflicted by agents of society on members of a minority.
Why then are bystanders so passive and silent? Sometimes silence results from fear, but that is not the whole explanation. Everywhere people tend to accept a definition of reality provided by 'experts', their government, or their culture. Lack of divergent views, just-world thinking, and their own participation or passivity change bystanders' perceptions of self and reality so as to allow and justify cruelty.

Distancing device: anonymity

Anonymity can lead to the loss of a well-defined separate identity that embodies inhibitions limiting antisocial behavior. Psychological research has shown, for example, that wearing a hood facilitated violence by Ku Klux Klan members (e.g., Zimbardo, 1969). Wearing masks, facial tatoos, etc. has a similar effect on the cruelty of warfare in preliterate societies.
Zimbardo's and others' studies have shown that anonymity increases the likelihood of stealing and cheating as well as of aggression and vandalism. Treating other people as if they were impersonal objects and making them feel anonymous sets the occasion for a wide range of antisocial behavior in normally law-abiding people.

Distancing device: habituation and learning by participation and doing

Hannah Arendt (1963) describes a turning point for Eichmann. When he was first exposed to the bodies of massacred Jews, he reacted with revulsion. But 'higher ideals' (that is, powerful motives) such as Nazi ideology and loyalty to the Führer, as well as a desire to advance his career, led him to ignore his distress and continue with his 'work'. The distress eventually disappeared (Staub, 1989).
Bruno Bettelheim (1979) described the inner struggle of a man who was against the Nazis but had to use the obligatory greeting 'Heil Hitler'. Even such a limited participation can result in substantial psychological reorganization.
The Greek torturers also learned by participation. First they stood guard outside interrogation and torture cells. Then they witnessed torture and provided help in beating up prisoners. They had to perform these duties satisfactorily before they were given a role as torturers (e.g., Gibson & Haritos-Fatouros, 1986; Haritos-Fatouros, 1988; Gibson, 1991).
Bystanders also learn and change through passive or semiactive participation. Another very important phenomenon is self-persuasion, especially among leaders and decision makers (Staub, 1989).

Once people have become accomplices, either by habituation or by learning by participation, there is no way back. This is also a favorite strategy in criminal gangs to commit individuals to the gang, e.g., the Manson massacre; They can be (morally) blackmailed for example. Another example is presented by De Temmerman in her account of the recent Hutu-Tutsi genocide: "De Interahamwe wilden het aantal moordenaars opvoeren om zich in te dekken, om iedereen schuldig te maken" (de Temmerman, 1995, p. 225).
One psychological consequence of harm-doing is further devaluation of victims (Lerner, 1980).

Distancing device: compartmentalization

Compartmentalization is a psychological mechanism which enables people to focus and act on goals that conflict with important values. When the discrepancy persists, a splitting of the self can occur that enables people to live with it (Lifton [1986] calls this 'doubling').
Dedicated or fanatical perpetrators may come to value killing; there is no inconsistency or need for splitting. However, less fully committed perpetrators must be able to compartmentalize. They may concentrate on the immediate task, ignoring ethics and long-term consequences. Many Nazi doctors focused on medical 'achievement' in their cruel experiments (Lifton, 1986). Camp commanders focused on efficiency. Bureaucrats prepared regulations and train schedules for transporting victims. Over time, internal changes will increasingly diminish the need to compartmentalize. Two psychological developments are of great importance: a reversal of morality and relinquishing a feeling of responsibility for the welfare of the victims (Staub, 1989).

Distancing device: moral drift

Zimbardo (1971) planned a two-week simulation of prison experience to examine the reactions of normal subjects who would play the parts of guards and prisoners. Although the study was designed to last for two weeks, Zimbardo was forced to stop it after six days due to the level of brutality the guards had reached and the signs of severe psychological distress on the part of several prisoners.
"The guards could be characterized as falling into one of three groupings. There were the tough but fair guards whose orders were always within the prescribed rules of prison operation. Then there were several guards who were the good guys according to the prisoners, who felt genuinely sorry for the prisoners, who did little favors for them and never punished them. And finally, about a third of the guards were extremely hostile, arbitrary, inventive in their forms of degradation and humiliation, and appeared to thoroughly enjoy the power they wielded when they put on the guard uniform and stepped out into the yard, big stick in hand" (Zimbardo, 1971).
Matza (1968) pointed out a similar drift into brutality in some youth gangs he extensively interviewed. He claimed that individual gang members reported that they privately disavowed the delinquent behavior the gang engaged in but that each was afraid to say this out loud for fear of ridicule. Behavior can deteriorate to the level of the least-restrained members of the group, not because the other members fully endorse that behavior, but because they are unwilling or unable to make their view known. Sabini & Silver (1982) call this phenomenon 'moral drift'.
Once brutality becomes standard procedure within an institution, it takes on an added legitimacy. As Berger & Luckmann (1966) argue, institutions by their very existence are taken by people, at least prima facie, to be legitimate. This legitimacy is conferred and reinforced by the process of socialization into the institution. The fact that everyone else seems to accept, or at least no one opposes, what would appear to naive first glance to be brutality suggests prima facie that in the judgment of other people the behavior is acceptable, and this suggests that the behavior is acceptable. Tillion's (1975) account of the 'socialization' of a new Aufseherin (a female guard) at Ravensbruck, suggests such a process at work.

Distancing devices: euphemisms and metaphors

Another, and very effective, way to assuage guilt is to present the shocking and horrific in mundane and ordinary terms: 'ethnic cleansing', 'final solution' (Endlösung) for the extermination of the Jewish people, 'special handling', Einsatzgruppen for extermination squads, and similar euphemisms.
Two sets of metaphors for killing deserve special mention because many different groups have found them appealing. The first involves extermination. To say that the enemy must be exterminated is to reduce the enemy to the status of subhuman vermin and to present killing as a matter of getting rid of worthless, troublesome pests.
The second common metaphor is medical one: amputation. By cutting off a diseased limb, amputation saves the rest of the body, and this metaphor has proved appealing to executioners everywhere.
There is one additional reason that perpetrators use more innocuous words than victims or dispassionate bystanders would use: the magnitude gap (Baumeister, 1997). Of course it is shocking to the victims that the people that are killing them one after another would speak of them as 'units' to be 'processed'. But that is because so incredibly much is at stake for the victim. For the perpetrator, in contrast, very little is at stake. Whether he kills 35 or 40 people on a particular day will make little difference in his own life and feelings and what he will experience in the coming weeks.
Diamond (1992) mentions another distancing device: the use of military language (e.g., the Pequod War, the Battle of Wounded Knee), which implies declared warfare waged by adult male combatants, to refer to what actually were sneak attacks (often by civilians) on villages or encampments to kill Indians of any age and either sex.

Dehumanization might be conceived as the relegation of the victims to the level of animals or of objects or to a purely instrumental role. The denial of a common humanity would seem to be an important component of any definition, since it emphasizes the element of exclusion (Kuper, 1982).
Kelman (1973) identifies dehumanization as one of the processes by which "the usual moral inhibitions against violence become weakened". He comments that inhibitions against murdering fellow human beings are generally so strong that the victims must be deprived of their human status if systematic killing is to proceed in a smooth and orderly fashion, and that to the extent that the victims are dehumanized, principles of morality no longer apply to them and moral restraints against killing are more readily overcome.
Kelman defines dehumanization as the denial of identity and community. The animal world has been a particularly fertile source of metaphors of dehumanization.
Hunters and gatherers have been a frequent repository of images borrowed from the bestiary. Described as animals, they have often been hunted down like animals, which is not to say that describing others as animals necessarily implies any inclination to massacre them (Kuper, 1982).

Metaphors sometimes serve a function subsidiary to, and analogous with, ideologies of dehumanization. Diseases which strike horror or repel, and which call for radical surgery, are specially favored. Susan Sontag (1978), in an analysis of 'Disease as Political Metaphor', writes that modern totalitarian movements, whether of the right or the left, have been peculiarly and revealingly inclined to use disease imagery, the Nazis repeatedly analogizing European Jewry to syphilis and to a cancer that must be excised, while disease metaphors were a staple of Bolshevik polemics. She continues that: "To describe a phenomenon as a cancer is an incitement to violence".
In a similar vein, there was the Nazi metaphor of the Jew as a dangerous parasite or bacillus, to be eradicated at all costs.

Hannah Arendt (1963) and Raul Hilberg (1961) both emphasized the use of euphemistic language that veiled reality not only from outsiders but also from the perpetrators themselves. Euphemistic language was used even by the victims.
Bureaucratic compartmentalization and euphemistic language serve to deny reality and distance the self from violent actions and their victims. However, bureaucratization and euphemistic language are not the source of, or the motivation for, genocide or mass killing. Nor are they crucial. In Cambodia and Turkey, for example, there was little bureaucratic organization (Staub, 1989).

Distancing device: psychic numbing

Scott Peck's (1983) account of the My Lai massacre shows how stress and distress, which are among the usual consequences of difficult conditions, affect human behavior. Peck suggested that humans regress under prolonged stress or discomfort; they become more primitive, childish.
Another response to stress is the mechanism of defense that Robert Jay Lifton (1971) called 'psychic numbing'. When our emotions are overwhelmingly unpleasant or painful, we anesthesize ourselves; soldiers become able to tolerate mangled bodies, and the capacity for horror becomes blunted. While this diminishes suffering, it also makes us insensitive to the suffering of others, especially when the other is defined as different, the member of an outgroup, or an enemy bent on our destruction.
While moral rules arise to serve human welfare, rules can be reified or held as absolutes, and at times the group rather than the individual is made the focus of their concern. This makes it easier to exclude specific individuals or subgroups from the universe of moral concern. In addition, given the widespread belief in a just world, victims will often be seen as deserving their fate (Staub, 1989).

Distancing devices: 'just world' thinking and blaming the victim

Genocidal societies also show a marked tendency toward what psychologists call 'just world' thinking: Victims are believed to have brought their suffering upon themselves (by their evil actions or by their intrinsically bad character) and, thus, to deserve what they get, and get what they deserve (Lerner, 1980; Staub, 1989; Simpson, 1993; DuPreez, 1994).
We are all vulnerable to blaming the victim because of our cognitive liability to stereotypy and 'just-world' thinking which assumes the innocent are not punished arbitrarily (Staub, 1978 I, 151). We are also liable to hurt others even without blaming the victim when we conceive of ourselves as simply agents because of the norm of obedience to authority (Milgram 1974). Both vulnerabilities are reinforced by authoritarian social structures which create organizations idealizing hardness, force, obedience, loyalty to the organization, and hatred of the enemy. Such learning is made possible by role expectations, group influences to conform, and subcultures which resocialize members; it is learning by doing building our capacity for compartmentalization of norms and obligations (Staub 1978, 1989, Sabini and Silver 1980; Milgram 1974).

Distancing devices: authorization and 'diffuse sanctioning'

Genocide and sanctioned massacres by definition occur in the context of an authority situation, a situation in which, at least for many of the participants, the moral principles that generally govern human relationships do not apply. Thus, when acts of violence are explicitly ordered, implicitly encouraged, tacitly approved, or at least permitted by legitimate authorities, people's readiness to commit or condone them is enhanced, That such acts are authorized seems to carry automatic justification for them. Behaviorally, authorization obviates the necessity of making judgments or choices. Not only do normal moral principles become inoperative, but - particularly when the actions are explicitly ordered - a different kind of morality, linked to the duty to obey superior orders, tends to take over (Kelman & Hamilton, 1989).
In an authority situation, individuals characteristically feel obligated to obey the orders of the authorities, whether or not these correspond with their personal preferences. They see themselves as having no choice as long as they accept the legitimacy of the orders and of the authorities who give them (though there are great individual differences in this obedience). Often people obey without question even though the behavior they engage in may entail great personal sacrifice or great harm to others.
An important corollary of the basic structure of the authority situation is that actors often do not see themselves as personally reponsible for the consequences of their actions (see above). Again, there are great individual differences.
Thus, when their actions cause harm to others, they can feel relatively free of guilt. A similar mechanism operates when a person engages in antisocial behavior that was not ordered by the authorities but was tacitly encouraged and approved by them - even if only by making it clear that such behavior will not be punished. In this situation, behavior that was formerly illegitimate is legitimized by the authorities' acquiescence. This has been called 'diffuse sanctioning' by Sanford & Comstock (1971).

Distancing devices: routinization and bureaucratization

Authorization processes create a situation in which people become involved in an action without considering its implications and without really making a decision. Once they have taken the initial step, they are in a new psychological and social situation in which the pressures to continue are powerful (this is the involvement or commitment trap treated before). For example, concern about the criminal nature of an action, which might originally have inhabited a person from becoming involved, may now lead to deeper involvement in efforts to justify the action and to avoid negative consequences.
Despite these forces, however, given the nature of the actions involved in genocide and sanctioned massacres, one migh still expect moral scruples to intervene; but the likelihood of moral resistance is greatly reduced by transforming the action into routine, mechanical, highly programmed operations. Routinization fulfills two functions. First, it reduces the necessity of making decisions, thus minimizing the occasions in which moral questions may arise. Second, it makes it easier to avoid the implications of the action, since the actor focuses on the details of the job rather than on its meaning. The latter effect is more readily achieved among those who participate in genocide and sanctioned massacres from a distance - from their desks or even from the cockpits of their bombers.
Routinization operates both at the level of the individual actor and at the organizational level. Individual job performance is broken down into a series of discrete steps, most of them carried out in automatic, regularized fashion. It becomes easy to forget the nature of the product that emerges from this process. When Lieutenant Calley said of My Lai that is was "no great deal", he probably implied that it was all in a day's work. Organizationally, the task is divided among different offices, each of which has responsibility for a small portion of it. This arrangement diffuses responsibility and limits the amount and scope of decision making that is necessary. There is no expectation that the moral implications will be considered at any of these points, nor is there any opportunity to do so. The organizational processes also help further legitimize the actions of each participant. By proceeding in routine fashion - processing papers, exchanging memos, diligently carrying out assigned tasks - the different units mutually reinforce each other in the view that what is going on must be perfectly normal, correct, and legitimate. The shared illusion that they are engaged in a a legitimate enterprise helps the participants assimilate their activities to other purposes, such as the efficiency of their performance, the productivity of their unit, or the cohesiveness of their group (Kelman & Hamilton, 1989; see also Janis, 1972).

Sabini & Silver (1982) explained the necessity of bureaucratization in the Holocaust. A pogrom, an instrument of terror, is typical of the long-standing tradition of European antisemitism, not the new Nazi order, not the systematic extermination of European Jewry. Mob violence is a primitive, ineffective technique of extermination. It is an effective method of terrorizing a population, keeping people in their place, perhaps even of forcing some to abandon their religious or political convictions.
Mob violence rests on the wrong psychological basis, on violent emotion. People can be manipulated into fury, but fury cannot be maintained for two hundred years. Emotions have a natural time course; lust, even blood lust is eventually sated. Further, emotions are fickle, can be turned. A lynch mob is unreliable; it can sometimes be moved to sympathy, say, by a child's suffering. To eradicate a 'race' it is essential to kill the children. Comprehensive, exhaustive murder required the replacement of the mob with a bureaucracy, the replacement of shared rage with obedience to authority. The requisite bureaucracy would be effective whether staffed by extreme or tepid antisemites, considerably broadening the pool of recruits; it would govern the actions of its members not by arousing passions, but by organizing routines. It was this bureaucratization of evil, the institutionalization of murder, that marked the Third Reich.
Participating in a 'legitimate' enterprise allowed subjects and bureaucrats to ignore the immoral implications of their actions in two ways. First, the issue of moral responsibility for the goals of the organization just does not come up in legitimate institutions; to do their jobs, people do not have to think about such matters. Second, even if subjects or bureaucrats had addressed the question of moral responsibility, bureaucratic structure would have helped them answer the question incorrectly; the relation between an individual's action and the rules and commands of an organization obscures personal responsibility (Sabini & Silver, 1982).
Because our feelings of responsibility are grounded in our intentions, and bureaucracies arrange that everyone need only intend to follow the rules, the result is that bureaucracies have a genius for organizing evil. The covering excuse of superior orders allows the individual to pursue personal goals - wealth, status, power - as well as more altruistic goals - providing for his family (Sabini & Silver, 1982; Baumeister, 1997).

Distancing device: normalization of 'evil'

As the destruction process evolves, harming victims can become 'normal' behavior. Inhibitions against harming or killing diminish, and extraneous motives can enter: greed, the enjoyment of power, the desire for sex or excitement. This is helped along by the belief that the victims do not matter and deserve to suffer, and even that any form of their suffering furthers the cause the perpetrators serve. 'Vicarious' rather than direct participation can also contribute to this evolution (Staub, 1989).
Selection for mass murder may be based not on cultural devaluation, but on a newly evolving or speedily adopted ideology. Usually, the ideology still draws on existing divisions in society. The identification and elimination of 'class enemies' has often been part of established communist practice.
In genocides and mass killings that follow decolonization, as in Burundi and Biafra, deep-seated historical conflicts can come to the fore in the context of profound social-political change. A history of conflict and antagonism (resentment and revenge, Chosen Trauma) fuels a power struggle that ends in genocide.

The ultimate distancing device: dehumanization

Chalk & Jonassohn, Charny, Dadrian, Fein, Hilberg, Horowitz, Kelman, Kuper, Porter, Renwick Monroe, Staub, Storr, and virtually all other writers on genocide accept that a precondition of genocide is the devaluation of the victims, their definition "outside the web of mutual obligations" (or "the exclusion of the victim from the universe of obligation", as Porter, Staub and Fein call it), defined as the other, unequal and threatening or subhuman. "The greater the perceived gap between 'the people' [a group's name for itself] and the out-group, the less were the values and the standards of 'the people' applicable to the out-group" (Chalk & Jonassohn, 1990, 28).
Dehumanization of the victims or defining them as a member of another species has been widely recognized as a precondition both for genocide (Erikson 1964; Kelman 1973; Charny 1982) and 'guilt-free massacres' (Duster 1971: "The most general condition for guilt-free massacre is the denial of humanity to the victim").
The following news item appeared in the San Francisco Bulletin during the 1860s. It is a telling example of dehumanization in action: "Some citizens of this city, while hunting in Marin County yesterday, came upon a large group of miserable Digger Indians. They managed to dispatch 30 of the creatures before the others ran away" (cited in Barash, 1991: 172).

Together with the concept of cultural pseudospeciation (to be elaborated in the chapter on 'Evil'), dehumanization is probably the most important proximate concept for understanding (mass)violence phenomena, including warfare, 'ethnic cleansing', massacres and genocide, in humans (and probably as 'dechimpization' [Goodall, 1986] in chimpanzees too). There is a profound paradox involved in the process of dehumanization in the sense that one can only dehumanize what is recognized and acknowledged to be human in the first place (Van der Dennen, 1996).
Another curious paradox is that both superhuman and debased characteristics may be ascribed simultaneously to certain groups in order to justify discrimination or violence against them. The foreigner, for instance, is seen at once as 'wicked, untrustworthy, dirty', and 'uncanny, powerful, and cunning'. Similarly, according to the canons of race prejudice, contradictory qualities of exceptional prowess and extraordinary defect, together make them a menace toward whom customary restraints on behavior do not obtain.
In its more complete form, however, dehumanization entails a perception of other people as nonhumans - as statistics, commodities, or interchangeable pieces in a vast 'numbers game'. Its predominant emotional tone is that of indifference and callousness (Bernard, Ottenberg & Redl, 1971; also see Sanford, 1971; Sanford & Comstock, 1971; Kelman, 1973; and Volkan, 1988, 1991).
De Temmerman (1995) relates how she perceived total indifference in the eyes of a Hutu boy who was about to horribly kill a Tutsi woman:

Ik denk aan het tengere, trillende lichaam van de Tutsi-vrouw van gisteren, aan de angstplekken op haar gezicht en haar verre, vreemde blik, een blik die al dood was. Maar het zijn de ogen van de moordenaar, de zestienjarige jongen met het hakmes, die me bijblijven. Er was geen wreedheid in die ogen, ook geen haat of genot maar totale onverschilligheid. Niet één keer keek hij haar aan. Niet één seconde ook nam hij aanstoot aan ons, buitenlandse toeschouwers. Geen moment kwam het in hem op dat wij dit misschien niet goed zouden vinden, dat wij dit zouden aanklagen. Het was alsof hij, lachend en keuvelend met zijn vrienden, een dier naar de slachtplaats sleepte (p. 27)

Dehumanization is more than calling the enemy 'cockroaches'. It refers to a combination of malignant psychological processes. Dehumanization is a response to the group's need to keep alive the principle of not being like the enemy. It represents an attempt to establish firmer boundaries between the two groups. Hostility and fear maintain dehumanization; in turn, dehumanization eliminates feelings of guilt, since it is acceptable to kill what is not human. Indeed, under these conditions the act of killing can be accompanied by feelings of pleasure and triumph, as it represents another step towards absolute control over the group's psychological distance from the enemy (Volkan, 1991).
Volkan identifies two elements in the group dynamics toward violence and war: the 'Chosen trauma' and the 'Chosen glory' of the group. Similarly, Galtung (1994) identifies Chosenness, Trauma and Myths of a Glorious Past, which toegether form a syndrome: the Chosenness-Myth-Trauma (CMT) complex or, more evocatively, the collective megalo-paranoia syndrome. Chosenness means the idea of being a people chosen by transcendental forces, above all others, endowed, even anointed, to be a light unto others, with the right and even the duty to govern them. Trauma means the idea of being a people hit and hurt by others, possibly out of their envy, by enemies lurking anywhere, intent on hitting again. Chosenness induces collective sentiments of grandeur relative to all others. This is then built into the Myths of a Glorious Past to be recreated, the present being suspended between the glorious past and the glorious future. But the traumas can also be used to validate the idea of chosenness; "we have suffered so much, there must be a deeper meaning to that suffering". New traumas are then expected for the future, with a mixture of fear and the lustful anticipation of self-fulfilling prophecies coming true. The three parts of the syndrome reinforce each other socially, not only as ideas, in a vicious circle (Galtung, 1994).
The group incorporates the mental representation of the traumatic event(s) into its identity, thus leading to the intergenerational transmission of historical enmity. Once a trauma becomes a chosen trauma, the historical truth about it does not really matter. In war or war-like situations, the leader evokes the memory of the chosen trauma, as well as that of the chosen glory, to galvanize his people and make his group more cohesive (Volkan, 1991).

"Offenses against persons outside the universe of obligation will not be socially recognized and labelled as crime... Collective violence is an offense against a class whose members are outside the universe of obligation" (Fein 1977, 18-19).
Fein's (1977, 1979, 1990; cf. Staub, Kuper, and many others) proposition that the exclusion of the victim from the universe of obligation is a necessary but not sufficient condition for genocide has often been equated with previous notions of dehumanization. However, the exclusion of the victim from the universe of obligation need not, according to her, entail dehumanization. She does not explain how it is psychologically possible, however, to exclude someone from the universe (or rather an individual's 'idioverse') of obligation without at least a modicum of dehumanization.
This is, I think, a fine example of what Freud called the 'narcissism of minor differences' in the conceptual realm. Fein's insistence on the distinction is more pedantic and pretentious than rooted in empirical reality. Another possibility is that 'exclusion of the victim from the universe of obligation' is a convoluted equivalent of 'indifference', in which case it is superfluous. In the remainder of this chapter we shall treat the two concepts ('dehumanization' and 'exclusion of the victim from the universe of obligation') as equivalent to all practical purposes.

Dehumanization seems to be a universal phenomenon. It also operates in preindustrial societies. Watson (1995), for example, states about a cannibalistic New Guinea society: "Even for a cannibal, killing isn't easy. So it has to be hedged about with ritual and restraint, and given social and religious dispensation. It has to be justified, and one of the easiest way to do so is to find some way of making the enemy unworthy, giving him an inferior or unclean status. Turning him from one of us, the human beings, to one of them, those less than human. Those who clearly deserve to die" (Watson, 1995).

In the context of the Holocaust, Horowitz (1980) states that "the precondition for mass extermination was engineered dehumanization: the conversion of citizens into aliens, first by executive decree, then by legislative enactment, and finally by judicial consent. These legal and sociological events represent the precondition for the technical performance of genocidal policies by totalitarian states" (italics added).

Is ideology necessary for genocide to occur?

"We are thus driven to the unfashionable conclusion that the trouble with our species is not an excess of aggression, but an excess capacity for fanatical devotion" (Koestler, 1978).

"There must have been periods in the history of human societies when groups committed genocide without benefit of ideologies to incite or justify commission of the crime. In the genocides of our own era, they seem to be invariably present" Kuper (1982) states, although even in biblical times genocides were 'justified' by explicit commands from the 'God of Wrath'.

The cumulative work of Norman Cohn (1967, 1970, 1977), ably synthesized by R. Smith (1989), and in its turn summarized by Fein (1990), is an impressive trans-historical line of research showing how ideas promising messianic deliverance regularly lead to murder.
Cohn views genocide not as the result of calculated action but as the acting out of messianic, apocalyptic fantasies promising salvation.
In short, only through massacre can the world be purged of evil and oppression: extermination is the price of virtue and happiness.
These fantasies stem from specific apocalyptic traditions which, adapted to change to new conditions, are revived in times of social crises. Cohn traces them from the Romans to the Nazis, showing how the mythic Jewish world-conspiracy depicted in the Protocols of the Elders of Zion proved to be a 'warrant for genocide' (1967). Such beliefs postulate a Manichean universe, stigmatizing the potential victims or enemy as the embodiment of evil: non-human, diseased, unclean, diabolical. They are not only outside the universe of obligation, they are outside humanity.
It is not hardship as such that leads people into the quest for total salvation, but rather the collapse of the normative order, the fact that society no longer seems to make any sense.
From this there emerges a party or group to enact this fantasy on earth, one usually led by a charismatic leader. The group and myth gives meaning to members' lives, ventilates and displaces their passion and legitimates ruthlessness. The myth becomes a warrant for genocide.
Others join the movement and participate in vandalism and killing from other motives: loot, and pleasure in others' pain. But more important than these is the indifference of the multitude who learn to blame the victim, the indifference which allows them to overlook and fail to deter discrimination, persecution and murder. Thus, Cohn incorporates the range of motives Staub (1989) and others have shown lead to the bystanders distancing themselves from the victims.

In his book The Pursuit of the Millennium, Cohn (1957) linked 'revolutionary chiliasm' to the psychopathological syndrome of paranoia. Where revolutionary chiliasm thrives best is where history is imagined as having an inherent purpose which is preordained to be realized on this earth in a single, final consummation. It is such a view of history, at once teleological and cataclysmic, that has been presupposed and invoked alike by the medieval movements described in Cohn's study and by the great totalitarian movements of our own day.
Chiliastically minded movements are ruthless not simply in order to safeguard or further specific interests but also - and above all - in an effort to clear the way for the Millennium. What else, he argues, can have induced the Nazis, in the middle of a desperate war, to allot manpower, money and materials to the wholly irrelevant enterprise of exterminating millions of Jews?
In his book Warrant for Genocide, Cohn (1967) distinguishes exterminatory antisemitism from other forms (as did Goldhagen [1997] later on).
This deadliest kind of antisemitism, he argues, has little to do with real conflicts of interest between living people. At its heart lies the belief in a Jewish world conspiratorial body, employed in medieval times by Satan for the spiritual and physical ruination of Christendom, and in modern times, banded together for the ruin and domination of the rest of mankind.
"Pogroms as spontaneous outbreaks of popular fury seem to be a myth, and there is in fact no established case where the inhabitants of a town or village have simply fallen upon their Jewish neighbours and slaughtered them".
Probably some such theory of 'demonization' is required to explain the prevalence and virulence of the genocidal massacres of Jews in Christian societies over so many centuries, and the rapid spread of Nazi doctrines, as envisaged by Hitler (Kuper, 1982).

Buried within Cohn's title and Kuper's (1982) chapter, 'Warrant for Genocide', is a liberal assumption as to the nature of human nature or of man in society. The assumption is that massive slaughter of members of one's own species is repugnant to man, and that ideological legitimation is a necessary precondition for genocide.
Thus Cohn (1967), in his discussion of the theory of a Jewish world conspiracy as a legitimating ideology for organized killings of Jews over a period of some eight centuries, writes that it seems "certain that however narrow, materialistic, or downright criminal their own motives may be, such men cannot operate without an ideology behind them. At least, when operating collectively, they need an ideology to legitimate their behavior, for without it they would have to see themselves and one another as what they really are - common thieves and murderers. And that apparently is something which even they cannot bear" (Cohn, 1967).

It would seem, Kuper (1982) contends, that the ideology may emerge after the massacres, as a posteriori rationalization or as justification, rather than as a warrant, but generally there would be ideas circulating which encourage the commission of genocide and the mobilizing of murderous mobs and of organized killers.
The most widely held theory is that these ideologies act by shaping a dehumanized image of the victims in the minds of their persecutors. We are, of course, again in the presence of the liberal assumption of powerful moral inhibitions against the slaughter of one's own kind. Since the victims are not human, the inhibitions against their slaughter cease to be operative (Kuper, 1982).
There are many examples of the application of this theory to the interpretation of genocide. Colin Legum (1966), in an account of 'The Massacre of the Proud Ibos' in Northern Nigeria, concludes that "in such conditions, political leaders can exploit actual or imaginary grievances against the alien minority. The danger signal is when there is official sanction for talking about a minority group in non-human terms. This process seems essential to provide some kind of justification for dealing with other human beings as one would treat dangerous animals - exterminate them".
Or again, in an analysis of the sociology of evil, Coser (1969) refers to concentration camp and prison guards who had brutally murdered Jewish, Ukrainian and other prisoners during the Second World War, but who had staunchly maintained that they had been justified in acting as they did because their victims were not 'people like you and me'. "These concentration camp guards, though they lived in the same physical environment as their victims, nevertheless managed to build so immense a social distance between 'them' and 'us' that they perceive their victims as not belonging to the same human race".

The elimination of hunters and gatherers is said to be an inevitable consequence of progress, a regrettable cost. Colonization was linked to evolution, the conquered peoples being conceived as lower in the scale of evolution with rights and capacities by no means comparable to those of their conquerors. In the march of progress, the unfolding of evolution provided the justification that many peoples are destined for 'the rubbish bins (the trash cans) of history'. In the Nazi phrase, these are 'Mistmenschen'.

Ideologies of dehumanization abound in the sacred writings of universal religions, yielding sanction for destructive conflicts, between Christians and Jews, Christians and Muslims, Hindus and Muslims, the Faithful and the Infidels, the Believers and the Heretics. They were waged with great atrocity and cruelty, and genocidal massacre (Kuper, 1982).
It is striking that most cases of genocide are marked by religious differences between the killers and their victims. It is quite possible, as Kuper suggests, that we underestimate the contemporary significance of religion in genocidal conflict (The role of religion is further elaborated in the chapter on cruelty).

Because it is state-sanctioned, if not state-induced, genocide also requires a legitimizing principle or ideology to justify the scale of human destruction. For the unfortunate Armenians, the ideology was pan-Turkism; for the Jews, anti-Semitism and the myth of Aryan superiority. Ironically, then, genocide is related to the persecuting groups' search for group meaning, identity, and power. Collective violence on such an extensive scale must make reference to a myth or theory in order to justify killing people not because of their own acts but simply because they fit into a particular category by virtue of their birth.
Ironically, the doctrine of biological determinism serves as a justification of genocide and genocide is frequently equated with a holy crusade to free the body politic of diseased tissue. Thus genocide becomes a scientific prevention of contamination by agents of 'racial pollution' who are viewed as parasites and bacteria causing sickness, deterioration, and death in the host peoples they supposedly infect. The mass murder of a people is thus justified through a twisted logic in which it becomes necessary to prevent the members of a biologically degenerate group from destroying a biologically superior one. What is at stake is the very existence of the in-group; in the Alice-in-Wonderland world of genocide, mass murder becomes an act of survival, in which even the most draconian measures are justified. This bizarre double-think exists at all levels of society (Lerner, 1992).

The dynamics of the genocidal or massacre process itself further increase the participants' tendency to dehumanize their victims. Those who participate as part of the breuacratic apparatus increasingly come to see their victims as bodies to be counted and entered into their reports, as faceless figures that will determine their productivity rates and promotions. Those who participate in the massacre directly - in the field, as it were - are reinforced in their perception of the victims as less than human by observing their very victimization. The only way they can justify what is being done to these people - both by others and by themselves - and the only way they can extract some degree of meaning out of the absurd events in which they find themselves participating (see Lifton, 1971, 1973) is by coming to believe that the victims are subhuman and deserve to be rooted out. And thus the process of dehumanization feeds on itself (Kelman & Hamilton, 1989).

Although Kuper (1982) observes the role of dehumanization in facilitating genocide, he notes "there may be dehumanization without massacre - this is surely the general case - and presumably massacre without dehumanization" (1981, p. 92). His challenge to the assumption that we can take normative constraint against genocide for granted is perhaps Kuper's most original observation, according to Fein (1990), and one often overlooked.

A broader question is whether democide in general involves such dehumanization and demonization. Rummel (1996) believes the answer is no. Much of the nongenocidal killing took place because the victims opposed a regime, criticised it, were killed as examples to deter others from opposition or sabotage (as in hanging ten subjects selected at random in retaliation), were of the wrong class (as of a landowner), did not work hard enough, violated a minor law or rule, were disrepectful (as in hanging one's coat on a bust of Lenin), or just happened to come under suspicion of being an enemy of the state or people. Tens of million of people were killed simply as expendable bricks and lumber in the building of a utopia as in Cambodia under the Khmer Rouge, Vietnam, or communist China. Near 40,000,000 people died or were killed in the Soviet slave-labour system alone, a number that exceeds all the genocides of this century. And although once within the system political prisoners were systematically dehumanized as 'enemies of the people', they may have been before imprisonment highly respected members of society. Many, in fact, were former communist party members themselves. Therefore, Rummel (1996) argues, demonization and seeing or treating the other as a threat is not a necessary preliminary to democide in general. It is, however, an intrinsic part of the process of genocide. Dehumanization is a handmaiden of power. At the extremes of power, totalitarian governments have slaughtered their people by the tens of millions, while many democracies can barely bring themselves to execute even serial murderers.

At least part of the answer to the problem why ideological conflicts are so malevolent has been suggested by Berger & Luckmann (1966): "The confrontation of alternative symbolic universes implies a problem of power which of the conflicting definitions of reality will be 'made to stick' in the society". He who has the bigger stick has the better chance of imposing his definition of reality (Berger & Luckmann, 1966).

According to Melson (1992: 7) "No doubt ideology played an indispensable role in the Holocaust. It provided the conceptual framework, the cognitive map, the 'warrant for genocide' that gave the Nazis the illusion that they had located the sources of defeat, crisis, and disintegration bestting Germany. It also provided part of the élan that the Nazis needed to fight the enemy they had discovered or, better yet, whom they had constructed...
Relying on ideology alone as the main explanation for historical events, however, raises at least three important questions: (1) What is the source of the ideology itself? Where did it come from? (2) Why was it believed? And this is most important: (3) What if anything does ideology have to do with action and behavior? What was the connection between antisemitism and genocide?".

Distancing Devices II: Rituals of Degradation

Ideological notions of the depravity of these groups remain abstract, compared to the physical sight, smell, and expressions of the victims themselves. The captives must somehow not only be labeled as inferior, but must be made to appear that way. This problem was not insolvable. A constant theme of both the literature on the German concentration camps and Solzhenitsyn's (1973) accounts of the Russian camps is the perpetual degradation of the inmates.
Starvation and constant hunger are extremely effective tools of degradation. Perhaps the most important technique of degradation is to make the individual filthy, to make him stink.
Prisoners in the concentration camps inadequately clothed, malnourished, and forced to perform hard labor. It was impossibe to stay clean, and many prisoners soon lost any concern with personal hygiene. Dysentery was rife, toilet facilities were grossly inadequate, and permission to use latrines during the working day was often refused. Prisoners were frequently covered with their own excrement and smelled repulsive as well as looking like filthy scarecrows (Storr, 1991).
'We do not claim that all the camp officials recognized the 'desirable' consequences of keeping the prisoners perpetually degraded and planned the camps with that in mind, but at least one official noticed the utility of bestiality, after the fact. In an interview with Franz Stangl, commandant of the extermination camp Treblinka, the issue of why the guards degraded the prisoners became explicit:
"Why", I asked Stangl, "if they were going to kill them [the Jews] anyway, what was the point of all the humiliation, why the cruelty?".
"To condition those who actually had to carry out the policies", he said. "To make it possible for them to do what they did" (Sereny, 1974; cited in Des Pres, 1976, and Storr, 1991).

German bureaucratic norms of procedure, industrial norms of efficiency, Prussian military norms of honor and discipline had converged with Nazi ideology in an ultimate dehumanization. The reality of this dehumanization was the antithesis of the dispassionate killer. It found expression in the most sadistic cruelty - in the brutality of the deportations, the running amok in the ghettoes of Poland, the stripping of social identity in the death camps to a number and an object, the gratuitous torments, the terror, humiliation, degradation and torture designed to destroy in the spirit before annihilating in the flesh. The absolute power of the killers, and the total vulnerability of the victims, encouraged the commission of every conceivable atrocity (Kuper, 1982).
Steiner (1971) sees in the death camps the realization of "the millenary pornography of fear and vengeance cultivated in the Western mind by Christian doctrines of damnation". He describes them as a complete, coherent world. "They had their own measure of time, which is pain. The unbearable was parcelled out with pedantic nicety. The obscenities and abjections practiced in them were accompanied by prescribed rituals of derision and false promise. There were regulated gradations of horror within the total, concentric sphere. 'L'univers concentrationnaire' has no true counterpart in the secular mode. Its analogue is Hell. The camp embodies, often down to minutiae, the images and chronicles of Hell in European art and thought from the twelfth to the eighteenth centuries... It is in the fantasies of the infernal, as they literally haunt Western sensibility that we find the technology of pain without meaning, of bestiality without end, of gratuitous terror. For six hundred years the imagination dwelt on the flaying, the racking, the mockery of the damned, in a place of whips and hell-hounds, of ovens and stinking air" (Steiner, 1971).

The Genocidal Process

Kuper (1982) summarized the genocidal process as follows
*    Some genocides have appreciable spontaneity, others are highly organized. There is probably always a measure of organization.
*    Genocides may explode with sudden violence or take a quite protracted course.
*    In rare cases, there may be an equal capacity to annihilate, following for example the changing fortunes of war. In India, partition created a situation in which Hindus and Muslims each constituted majorities in different parts of the country, with the capacity to engage freely in reciprocal genocidal massacre.
*    Over and above the atrocity of mass slaughter, there seems to be almost invariably the additional gratuitous atrocity of torture, perpetrated with incredible brutality, and, as appears from many accounts, with hilarious and joyful abandon. Sexual torments and mutilations are common enough. It would seem that the élite are by no means immune to the fascination of these forms of torture.
*    Ideological dehumanization of the victims is a constant feature, the mass slaughter itself being the denial of a common humanity. It is expressed too in the handling of the victims, in the disposal of their bodies, and in the obscene mutilation of corpses. There are often 'rituals of degradation' which deliberately reject, with brutal contempt, the most deeply held human values, and the deepest sentiments of human attachment. Thus men are tortured before their wives and children, women are repeatedly raped in the presence of their families, children are killed in the arms of their mothers, and prospective victims are forced to slaughter their fellow victims by the most fearful means.
*    Euphemisms are commonly used by those in authority to describe the genocidal process. Solzhenitsyn (1978) refers to the Soviet phrase 'special settlers' for the exiled nations; elsewhere he mentions the versatile category of 'social prophylaxis'. In Nazi Germany, we had such euphemisms as 'the final solution' [Endlösung]. For exterminations in warfare, there is the convenient euphemism of military necessity.

For the most part, Kuper (1982) does not think it is possible to write in general terms about the genocidal process. The only valid approach would be to set up a typology of genocides (see § Typologies). According to DuPreez (1994), genocides have the following stages: 1. Preparation; 2. Justification; 3. Precipitation; and 4. Massacre.
Stanton (1999) distinguishes seven stages of genocide:
1. Classification: All cultures have categories to distinguish people into 'us' and 'them' by ethnicity, race, religion, or nationality. Bipolar societies that lack mixed categories, such as Rwanda and Burundi, are the most likely to have genocide.
2. Symbolization: We give names or other symbols to the classifications: We name people 'Jews' or 'Gypsies', or distinguish them by colors or dress; and apply them to members of groups. Classification and symbolization are universally human and do not necessarily result in genocide unless they lead to the next stage, dehumanization.
3. Dehumanization: One group denies the humanity of the other group. Members of it are equated with animals, vermin, insects, or diseases. Dehumanization overcomes the normal human revulsion against murder. At this stage, hate propaganda in print and on hate radios is used to vilify the victim group.
4. Organization: Genocide is always organized, usually by the state, though sometimes informally (Hindu mobs led by local RSS militants) or by terrorist groups. Special army units or militias are often trained and armed. Plans are made for genocidal killings.
5. Polarization: Extremists drive the groups apart. Hate groups broadcast polarizing propaganda. Law may forbid intermarriage or social interaction. Extremist terrorism targets moderates, intimidating and silencing the center.
6. Identification: Victims are identified and separated out because of their ethnic or religious identity. Death lists are drawn up. Members of victim groups are forced to wear identifying symbols. They are often segregated into ghettoes, forced into concentration camps, or confined to a famine-struck region and starved.
7. Extermination begins, and quickly becomes the mass killing legally called 'genocide'. It is 'extermination' to the killers because they do not believe their victims to be fully human. When it is sponsored by the state, the armed forces often work with militias to do the killing. Sometimes the genocide results in revenge killings by groups against each other, creating the downward whirlpool-like cycle of bilateral genocide (as in Burundi).
8. Denial is an eighth stage that always follows a genocide. The perpetrators of genocide dig up the mass graves, burn the bodies, try to cover up the evidence and intimidate the witnesses. They deny that they committed any crimes, and often blame what happened on the victims. They block investigations of the crimes, and continue to govern until driven from power by force, when they flee into exile.

Scherrer (1998) presents a 95-item list applicable to any situation of contemporary genocide, divided into ten broad categories covering the main aspects of genocidal processes: perpetrators, victims, motivations, the central role of the state and its transformation, process characteristics, processual dynamics, the framework/context and the reaction/aftermath.
Harff (1987), Melson (1992, 1997), and Scherrer (1998a,b) have identified a number of common elements and patterns in the four total modern genocides, which happened in the wake of crushing or crumbling states and empires:
*    The modern nation state is the "predominant culprit in genocides" (Harff). Evidently, the likelihood to realize such aims are much higher in a totalitarian system than in a democratic one. However, periods of democratization may be conducive to genocide. The elimination of so-called "'foreign' elements from within" (Harff) is one of the common denominators of total modern genocide.
*    Since 1945 there were more than 250 violent conflicts occurring world-wide - until the late 1980s nearly exclusively in the Third World. Ethnic or ethnicisized violence became the single most dangerous source of violence. In 2 out of 3 cases of mass violence we find an 'ethnic' component. That means the ethnic factor (ethnicity) is a dominant or influential component.
*    Awareness is growing that failed states are the most dangerous states (Scherrer, 1998: 6).
*    One frequently overlooked characteristic separates societies where mass hate flourishes from those where it does not: an established democratic political culture. Genocidal policies in nearly every instance have been launched by dictators in the midst of military conflict (Kressel, 1996: 251-52).
*    The context of the crime is characterized by rapid political, social and structural changes which were described by the context of 'national upheaval' (Harff), separatist conflict, internal strife, rebellion, or 'revolution' (Melson): abrupt historic changes following an extended period of crisis. The aim of genocide is part of a larger project of nation state formation or its revision. This includes all the different processes of changing regimes, moving of boundaries or loss of territory, warfare (especially lost wars) or security threats resulting from (or perceived as) challenges to the dominant groups identity and to the identity of the 'national' political community.
*    The redefinition or mystification of that 'national' identity by the power-elites is a central point. The perceived "struggle for 'national' survival" against internal and external enemies has to be somewhat plausible for a majority group. The 'foreign' minorities shall function as scapegoats. The 'nation' needs to be purified. The elites need to "reconstruct society, (and) revitalize support for the state by way of a new system of legitimation" (Melson, 1997). In reality all total domestic genocides were preceded not so much by real challenges to national identity rather than by challenges to the dominant power strata, having won or maintained its power in an outright unstable situation.
*    Of utmost importance is a subservient state bureaucracy and obedient or extremist sections of armed forces or special troops. Conducive to the genocidal aim is the great fear and confusion of the 'national population' and at least lukewarm support among larger sections of the grassroots. Support for the genocide aim can only be won if the victimized groups are presented as 'racially', ethnically, religiously or morally different from the dominant group. Most successful proved the construction of close links between domestic 'enemies' and external 'aggressors' by the genocidal elites. Individual victims may often not be easily identifiable, and there is usually a large 'grey area'; however, the target group is visible and easily identifiable as an imagined collective entity.
*    The intentions of the killers are expressed in their ideology. This ideology always takes up older stereotypes. The aim of the power elite is to single out and exclude a group as "enemy of state and society". Extremist regimes are essentially combining militarism, xenophobia/ethnochauvinism and ultra-nationalism (militant narrow nationalism) with promises for the majority population's "bright future".
*    Fierce and bellicose world-views are often reinforced by ideologies that hold the life or welfare of individuals to be of no consequence if they impede progress toward over-arching goals. Such goals include a new social order, the glory of God, or even the leaders' own deification. Leaders and their followers dominated by such ideologies have no compunction about torturing and slaughtering numberless individuals, whom they regard as faceless interchangeable members of an evil, hated, feared or despised class. The victims are infidels, germs destroying the body politic, capitalist swine and the like. The perpetrators, similarly, view themselves as anonymous instruments serving such highly valued goals as defending the faith (e.g., Torquemada), protecting the country, or avenging injustice (Frank, 1994).

Both large-scale genocides of this century (the Aghet - Turkish genocide of the Armenians - and the Holocaust) were committed under the cover and during periods of warfare (Scherrer, 1998b). Why are revolutions and wartime circumstances conducive to genocide? Melson (1997) provided the following answer. When revolutionary vanguards come to power, in a situation where most institutions have been undermined and the identity of the political community is in question, they need to reconstruct society, revitalize support for the state by way of a new system of legitimation, and forge new identities. In a revolutionary context they will redefine the identity of a subset of the political community as 'the people', 'the nation', the race', 'the religion', or 'the class'. These are the group or groups that are celebrated by the ideology of the revolutionaries and from whom they hope to draw their support. Groups that are not included and are singled out as racial, national, religious, or class enemies, run the danger of being defined as 'the enemies of the revolution and the people' And it is such groups that may become the victims of repression or genocide.
There are three ways in which revolutionary war is closely linked to genocide: First, wartime aggravates feelings of vulnerability and paranoid fears which link supposed domestic 'enemies' to external aggressors. The victims of all the major genocides were said to be in league in a nefarious plot with the enemies of the revolutionary state. Wartime may also intensify feelings of invincibility. Second, war increases the autonomy of the state from internal social forces, including public opinion, public opposition and its moral constraints. Third, war closes off other policy options of dealing with 'internal enemies', leaving genocide as a strong choice for an already radicalized regime (Melson, 1992, 1997).

It should always be borne in mind that, in the final analysis, genocides result from individual human decisions. As Kuper (1981: 332) asserted: "The involvement of governments and elites in many genocides is a reminder that human actors make choices and decisions, and carry out actions which constitute, or lead to, genocide. Genocide is not an inevitable consequence of certain social conditions within a society".

Valentino's (2000) 'Strategic' Perspective on Mass Killing

Valentino's (2000: 21) research suggests that the impetus for mass killing usually originates from high political or military leadership, not from factors present in society at large.
The broader public sometimes approves of mass killing, but often it does not. Whatever its position, the active support of a large portion of the public is usually not required to carry out mass killing. The killing itself is almost always performed by military or paramilitary organizations, often with little more than the passive acceptance of the rest of society - including members of the perpetrators' own social groups.
In light of these unsettling truths, Valentino argues that the causes of mass killing are best understood when the phenomenon is studied from a 'strategic' perspective. The strategic approach suggests that mass killing is most accurately viewed as a goal-oriented policy - a brutal strategy designed to accomplish leaders' most important objectives, counter their most dangerous threats and solve their most difficult problems. Like war, mass killing can be a powerful political and military tool. Unfortunately, states and other powerful groups throughout history have proved all too ready to use this tool when it seemed to serve their purposes (Valentino, 2000: 3).

Public support for mass killing

Perhaps one reason that so many scholars have searched for the causes of mass killing in the structure or psychology of society at large is due to an implicit assumption that a very large segment of society must be involved in, or at least activively support the implementation of mass killing (This belief has even led some authors to suggest that societies involved in mass killing should be considered collectively mentally ill [e.g., Charny, 1984; Aronson, 1987]).
This assumption is unfounded. Although the general public sometimes supports mass killing, often it does not.
Leaders who possess the political and/or military support necessary to gain control of a state usually also possess sufficient support to carry out mass killing.
Large segments of the public may support violent regimes while remaining indifferent or even opposed to mass killing itself. Individuals may support specific policies without approving of the violent means used to implement them. Public support for mass killing may also be based on misinformation and propaganda.
Regardless of whether the general public supports mass killing or opposes it, however, civilians almost never play a major role in the killing itself. Rather, in nearly all instances of mass killing, the majority of the actual violence has been carried out by a relatively small segment of society. These killers are almost always young men, typically members of an organized military group, militia or police organization.
Widespread participation of civilians in killing operations, as Valentino (2000: 23) observes, has occured in only four instances: the partition of India in 1947, the anticommunist massacres in Indonesia in 1965, the Rwandan genocide of 1994 and during some episodes of mass killing in Communist China. In each of these cases, however, much of the killing was organized and coordinated by (national or local) political and military leaders. Military, paramilitary or militia units still played a significant role in the killings, often resorting to coercion to enforce civilian participation.
More significant segments of the public frequently, although by no means always, lend their assistance to mass killing in indirect ways. Indirect cooperation may involve activities such as producing weapons, providing logistical and administrative support to organizations directly involved in the killing, or informing on fellow citizens. Many cases of mass killing would not have been possible without widespread participation of this kind. Nevertheless, indirect collaboration should not automatically be equated with approval of mass killing. Individuals may participate in these kinds of activities for a variety of other reasons ranging from an abstract sense of patriotic duty, material or careerist ambition, to outright fear and coercion. Human beings seem capable of psychologically separating their participation in such activities from their personal approval of or sense of responsibility for the killing itself (Sabini & Silver, 1980). Even these more indirect forms of cooperation, however, seldom seem to occupy the majority of the public. In most cases, all that is required of the great mass of citizens is their powerlessness, passivity or indifference to the suffering of others (Valentino, 2000: 24-25).

The perpetrators of mass killing

Although mass killing typically requires the participation of only a relatively small group of people in acts of violence, it remains important to understand how these individuals are motivated to engage in such brutal behavior. At least part of the answer to this question can be found in the substantial body of psychological research demonstrating that ordinary people are capable of doing considerable violence to others when the situation seems to require it and/or when they are ordered to do so by those in positions of authority. Undoubtedly, the most famous of these experiments was performed by Stanley Milgram in the early 1960s.
Milgram's findings are all the more powerul because his experiment could not accurately reproduce many of the conditions and incentives that make compliance more likely in the real world, In particular, peer pressure, group conformity and other situational factors common to the military and paramilitary organizations usually tasked with carrying out mass killing provide powerful added inducements to participate, pressure that could not be realistically simulated in Milgrams's study. Christopher Browning's study of German military units involved in mass killing operations finds that these pressures were central to individual compliance with orders to kill.
Very little evidence is available that might shed light on the personal beliefs of the individuals who carry out mass killing. To make a broad generalization, however, it is probably fair to say that most perpetrators believe that killing defenseless civilians is a distasteful but ultimately necessary task. They are convinced that what they are doing is basically right. These justifications are often encouraged by exposure to propaganda specifically designed to portray the victims as a serious threat and to promote acts of violence against them (though Browning argues that indoctrination played a much smaller role than group loyalty and obedience in inducing the soldiers he studied to kill Jews). Nevertheless, for many individual killers - as opposed to the leaders who direct them - the understanding of the ideologies that justify such brutality often appears surprisingly 'shallow'. Although some participants in mass killing are committed fanatics, drawn to their task by strong beliefs, most killers do not seem so enthusiastic about the killing that, if given the opportunity, they would have ordered it themselves. They are willing, sometimes even eager, to go along with the killing but it probably would not have been their idea.
Individual perpetrators of mass killing often display only the most vague and simplistic sense of why they participate in these horrific acts. As Gérard Prunier describes the state of mind of many perpetrators of the 1994 genocide in Rwanda, "If the notion of guilt presupposes a clear understanding of what one is doing at the time of the crime, then there was at that time... a lot of innocent murderers" (Prunier, 1995: 253). Similarly, it seems doubtful that, despite attempts at 'political education', the young and largely illiterate soldiers who carried out the mass killings following the communist revolution in Cambodia could have grasped the abstruse political and economic ideology which motivated Pol Pot.
If the willingness of human beings to kill others in the service of goals they scarcely understand seems inconceivable, one need only reflect upon the willingness of soldiers throughout history to offer up their own lives in pursuit of such goals. History provides numerous examples of soldiers' willingness to fight for political or military goals they barely comprehend, sometimes in far away places they have never even heard of before (Valentino, 2000: 25-28).

Ideological-political mass killing

The first major scenario of dispossessive mass killing results when powerful groups attempt to implement an ideology calling for the rapid and radical economic or social transformation of their society. Few scholars who have studies genocide and mass killing have failed to comment on the central role that ideology has played in many of the twentieth century's most bloody mass killings. Various authors have suggested that the most dangerous ideologies are those which seek to purify the nation, dehumanize victims, place national security above all other goals, or expound a political formula that excludes victim groups from the larger community or nation (e.g., Lopez, 1986; Chalk & Jonassohn, 1990; Fein, 1993; DuPreez, 1994). From a strategic perspective, however, what ideologies that lead to mass killing share is not their specific content, but the magnitude, scope and speed of the changes they seek to bring about at the expense of certain groups.
Ideologies which call for the rapid and radical transformation of society often result in the near total material or political disenfranchisement of existing groups. Leaders with such radical agenda soon discover - or sometimes anticipate - that members of disenfranchised groups will not cooperate with the implementation of a new social order in which they stand to lose their livelihood, their homes or their very way of life. Leaders fear that the existence of large, highly dissatisfied groups within society will pose a serious threat to their most important goals, or even their survival. Regardless of their original intentions, such considerations force leaders to contemplate the most brutal options available to them. Even if the victim group is not totally annihilated, leaders may believe that mass killing will prevent or deter surviving members from mounting further resistance. Despite their deadly consequences, it is important to recognize that the ideologies which motivate this kind of mass killing need not aim at the destruction of specific groups as an end in itself (Valentino, 2000: 31).
Ideological mass killings in the twentieth century can be divided into two main groups - mass killings by communist states and the ideologically or politically motivated killing of ethnic, national or religious groups.

Communist mass killings

The most brutal mass killings in history have resulted from the effort to transform society according to communist doctrine (Rummel estimates that communist regimes have killed approximately 110,000,000 civilians in this century; see also Culbertson, 1978; Brzezinski, 1989; Courtois et al., 1999). In their study on the use of violence by communist governments, Alexander Dallin and George Breslauer argue that the leadership of communist regimes,

    ...committed as it is to substantial transformations - whether for ideological or power related reasons - is bound to encounter and generate resistance and alienation, since the chnages it is determined to carry out will necessarily clash with the values and perceived interests of some significant sectors of society. Anticipating such hostility, the authorities, in line with their preconceptions and images of class or group loyalties and grievances, may identify certain strata as requiring preemptive, or prophylactic, suppression, intimidation or removal (Dallin & Breslauer, 1970: 6; see also Adelman, 1991).

Mass killings associated with the collectivization of agriculture provide the most striking examples of this process. In the Soviet Union, millions were excecuted or perished in famines which appear to have been intentionally exacerabted by Soviet authorities as part of the effort to crush peasant resistance.
Violence, coercion and devastating famines also accompanied the implementation of radical communist agrarian policies in China.
Cambodia's experiment with collectivization followed an exceptionally violent path.
In addition to mass killings associated with radical agricultural policies, communist mass killings have also taken the form of brutal intraparty purges and attacks on social and cultural elites, intellectuals, and members of opposition political parties. The Great Terror in the Societ Union and the Cultural Revolution in China represent the most violent examples of this type of communist mass killing.
The belief that counterrevolutionaries were guided by their class and social origins also tended to promote the conclusion that suspect groups must be suppressed prophylactically and as a whole, even in the absence of any evidence of individual involvement in specific opposition acticities (Valentino, 2000: 34-37)

Ethnic, national, and religious mass killings

While communist leaders have sought to reshape the economic fabric of society, groups seeking the radical ethnic, national or religious transformation of society have also killed millions in places like Turkey, Nazi Germany, Burundi and Rwanda.
Radical racist or chauvinist groups have relied on two primary methods to accomplish the transormation of their societies. Although neither method calls for killing for its own sake, both have often resulted in mass killing. The first method, commonly known as ethnic cleansing, entails physically removing large numbers of the victim group from the country. Ethnic cleansing often results in mass killing because perpetrators typically find it necessary to use extreme violence to force people to abandon their homes and belongings. Even after victims have been coerced into flight, the process and aftermath of large population movements itself can be deadly. One of the most recent examples of ethnic cleansing occurred in the former Yugoslavia as Serbs (and later Croats) attempted to rid certain territories of their ethnic-political enemies. According to Mark Danner, Serbian officers

    were making rational, systematic use of terror as a method of war. Rather than being a regrettable but unavoidable concomitant of combat, rapes and mass executions and mutilations here served as an essential part of it. The Serbs fought not only to conquer territory but to 'clear' it of all traces of their Muslim or Croat enemies; or, as a notorious Serb phrase has it, to 'ethnically cleanse' what they believed was 'their' land. Of course making use of terror in such a way is probably as old - and as widespread - as warfare itself (Danner, 1997: 59).

Ethnic cleansing, however, is not always a 'practical' way for leaders to eliminate the influence of their ethnic-political opponents. When expelling victims to neighboring countries is judged too difficult, too risky, or when neighbors refuse to accept refugees, the killers may seek instead to deprive their victims of the ability to organize politically or militarily by killing its elites, intellectuals or sometimes all males of military age. In extreme cases, leaders may attempt to annihilate the victim group altogether.
The Nazi genocide of between five and six million Jews during the Second World War stands as the most deadly example of racial-ideological mass killing in history.
The decision to launch the Final Solution appears to have occurred only after the Nazis concluded that deportation was impractical or impossible in view of wartime exigencies, particularly the problems raised by the millions of Jews in conquered Eastern territories.
The racial-ideological ideology that motivated the Nazi Holocaust is unique in the history of mass killing. No other ethnic mass killing has been motivated by such elaborate 'scientific' principles. Ethnic, national or religious groups are usually targeted not because their killers believe that these groups will poison the gene pool, but because they are perceived (rightly or wrongly) to represent a threat of a more political or military nature. As one suspected Hutu organizer of the 1994 Rwndan genocide of of Tutsis argued, "the Tutsis were not killed as Tutsis, only as sympathizers of the RPF [the main Tutsi political-military organization]... ninety-nine percent of Tutsis were pro-RPF. There was no difference between the ethnic and the political". Underlining the political motives behind the genocide, between 10,000 and 60,000 Hutu suspected of cooperating or sympathizing with the RPF were slaughtered alongside more than 500,000 Tutsis (Prunier, 1995: 265; Human Rights Watch, 1999: 15).
Political conflict escalates to mass killing when powerful groups decide that only a fundamental restructuring of society along ethnic lines will permanently end the conflict, secure their political domination or achieve other important goals.
René Lemarchand's conclusions about the motivations which prompted the Burundi 1972 mass killing (in which at least 1000,000 civilians, nearly all Hutus, were killed) are worthy of quotation at length:
    Acting on the principle that exceptional problems call for exceptional solutions, the more radical elements [of the Tutsi leadership]... saw the annihilation of all educated Hutu elements as the most sensible course to make Burundi safe for the Tutsi minority. By striking at all Hutu elites, students and schoolchildren indiscriminately, even at the least suspicion of subversive intentions, they aimed not only to decapitate a potential counterelite but to spread terror throughout the entire Hutu community and thus create an enduring sense of fear and submission among the living and the unborn - in short, to teach a lesson that would be remembered by generations to come... From this macabre ethnic 'cleansing' emerged a maimed society in which the only elites were Tutsi elites... For the next fifteen years, only Tutsi were qualified to gain access to power, influence, and wealth (Lemarchand, 1994: 101-3).

The Turkish genocide of Armenians in 1915 arose from similar motives but ultimately proved even more violent, leaving over 1,000,000 Armenians dead (Dadrian, 1997: xviii) (Valentino, 2000: 34-42).

Territorial mass killing

The second general scenario of dispossessive mass killings occurs when powerful groups attempt to resettle territories already populated by native people. The desire for land is one of the oldest motives for mass killing. In the ideological mass killings described above, political conflicts and ideological goals may lead states to opt for ethnic cleansing of certain territories. In territorial mass killings, it is the desire for the land itself that drives the killing. Killers do not seek to rid the territory of its inhabitants because of political or ideological conflicts with them, but rather because the killers want to cultivate or populate the land with their own people. In such cases, leaders may simply decide to annihilate preexisting groups or force them to relocate to other areas. In either event, the result is often mass killing.
Territorial motives for mass killing have emerged in two closely related scenarios. First, territorial motives have provoked the extermination of indigenous people in many colonial societies. Second, territorial mass killing has often occurred during expansionist wars.
Mass killing has frequently resulted as as settler colonies have sought to expand their territories into regions already populated by indigenous peoples. This scenario has occurred primarily in colonial settings, most notably in the European colonies of North and South America and to a lesser extent in Africa.
Violence is frequently required to force indigenous people to abandon their homes. The process of moving large populations often proves deadly, as the relocation of the Cherokee people in 1838 on the infamous 'trail of trears' demonstrated. Between 4,000 and 8,000 out if the 17,000 Cherokee died during the 1,200 mile forced march to Oklahoma (Stannard, 1992: 122-5).
To make matters worse, starvation, disease and depredation by other tribes often await those who survive relocation since colonists have seldom chosen to deport indigenous populations to hospitable regions.
This deadly competition for land played a major role in the destruction of many indigenous tribes of America. In many instances, however, the ravages of European diseases so decimated native populations that settlers were capable of expropriating native lands without resorting to mass killing.
Two French colonies in North America provide telling examples of how the economic relationship between colonists and indigenous peoples can influence the likelihood of mass killing. In what is now Canada, the Huron people became an integral part of France's fur trade, serving as guides and skilled trackers for the French. French trappers relied on Huron villages for supplies and protection from other Indians. The French and the Hurons maintained a relatively peaceful relationship until the Huron were decimated in a war with the Iroquois in 1649. The relationship between the French and the Natchez people of the lower Mississippi, on the other hand, ended in the annihilation of the Natchez by the French colonists. In the lower Mississippi the French planned a large colony based on agriculture. They imported slaves from Africa for their servants and laborers. The Natchez people simply stood in the way of expanding French plantations. When the Natchez would not abandon their land peacefully, the French decided to remove them by force. By 1731 the Natchez had ceased to exist (Valentino, 2000: 43-45).

Mass killing and wars of territorial expansion

In addition to the mass killing of indigenous people, territorial mass killing has also resulted when groups engaged in expansionist wars have sought to resttle areas already occupied and developed by others. Perhaps the most horrific example of this kind of mass killing occurred during the Second World War a Germany attempted to expand its territory eastward into regions inhabited by Poles, Russians and other Eastern European peoples. This effort to acquire Lebensraum (living space) for Germany's growing population was one of Hitler's primary obsessions, rivaled in importance only by his interrelated campaign to free Europe from the Jews.
On the eve of the German invasion Hitler ordered his forces to put to death "without pity or mercy all men, women and children of Polish descent or language. Only in this way can we obtain the living space we need" (quoted in Lukas, 1986: 3). By the end of the war more than 22 percent of the prewar Polish population was dead (Valentino, 2000: 45-46).

Coercive mass killings

Sometimes mass killing is simply war by other means. Coercive killings results when groups engaged in conflict lack the capabilities to defeat their opponents' military forces with conventional military methods. The decision to target civilians intentionally has emerged in three closely related scenarios - counterguerrilla warfare, terror warfare and imperial warfare.

Counterguerrilla warfare

Mass killing can be an attractive strategy for governments engaged in counterguerrilla campaigns. Examples of this bloody brand of counterguerrilla fighting abound in the history of the twentieth century. Counterguerrilla tactics were applied in an especially ruthless manner in Ethiopia, Afghanistan, Iraq, and East Timor. Counterguerrilla motives were also prominent in the Guatemalan mass killings of the 1980s. More than 100,000 Guatemalan civilians were killed as the government sought to crush a guerrilla rebellion (Jonas, 1991: 49).
In many cases, regimes engaged in counterguerrilla wars have attempted to deprive guerrillas of local support by relocating vast numbers of civilians into areas that can be more easily policed. While in theory more 'humane', in practice this strategy can also result in mass killing. Large populations are still often relocated to overcrowded camps or inhospitable regions where disease and starvation take their toll (Valentino, 2000: 47-50).

Mass killing as mass terror

A related scenario of coercive mass killing occurs when combatants engaged in protracted wars of attrition search for means to bring the war to speedy end. As in counterguerrilla killings, leaders may choose to target enemy civilians in the hopes of coercing surrender without having to defeat the enemy's military forces directly. The ultimate goal of this type of mass killing is straightforward - to speed up the end of the war.
The advent of strategic air power has rendered this strategy an especially attractive and deadly weapon. During the Second World War, the United States and especially Britain intentionally bombed German cities in an effort to weaken German public support for the war and force an early surrender. By the end of the war, American and British bombing probably killed between 300,000 and 600,000 civilians in Europe.
The strategy of terror warfare existed long before the advent of long-range bombers. Military forces throughout history have relied on the practice of siege warfare and the use of starvation blockades to achieve exactly the same effect.
Finally, powerful sub-state groups have sometimes used coercive mass killing to terrorize their enemies, typically a colonial government and its supporters among the native population. By inflicting high casualties on specifically targeted groups of civilians, these terrorists hope to achieve their political goals without directly engaging the superior military forces of their enemies. Algerian resistance groups relied heavily on this strategy during their war for independence from France, eventually killing nearly 70,000 people - nearly all native Algerians (Horne, 1977: 538) (Valentino, 2000: 51-53).

Imperial mass killing

The final scenario for mass killing has been closely associated with empire. History provides numerous example of mass killings committed by imperial powers. Imperial mass killings are motivated by the desire to diminish the costs of building and maintaining empires. Large empires, after all, would be prohibitively expensive if each conquered city, state or province had to be defeated by force and then policed to a man. Imperial leaders, therefore, have strong incentives to adopt a strategy of mass killing as a method of deterring rebellions and resistance within their empire and as a method of intimidating future conquests into submission. When individual regions, cities or states resist annexation or attempt to rebel against exploitation by the empire, imperial powers often respond with mass killing, sometimes in an exceptionally violent and gruesome manner. The annihilation of rebellious subjects is intended to demonstrate to other provinces considering rebellion what fate awaits them if they resist.
The Mongol Empire ruled by Genghis Khan and his progeny was one of the first recorded and most efficient practitioners of this kind of mass killing. According to Paul Ratchnevsky, "Genghis Khan used terror as a strategic weapon in his military plans... terrible destruction was threatened in the event of resistance; bloody examples were designed to spread fear and reduce the populace's will to resist" (Ratchnevsky, 1991: 160, 173; cf. Saunders, 1971: 66-67).
Like territorial mass killing, however, imperialism has also been a declining cause of mass killing in the twentieth century as the great European colonial empires have steadily dissolved (Valentino, 2000: 53-55).

To the literature on genocide and mass killings
To the next part of The 'Evil' Mind