Few terms carry the weight of genocide. It names what is widely regarded as the gravest crime in international law, and its application to a set of events changes how those events are understood, remembered and acted upon. That weight is also the source of persistent difficulty: a term that determines legal liability, shapes political mobilisation and confers a particular moral standing on victims and perpetrators alike is unlikely to be defined quietly.

This article examines the concept rather than any single case: where it came from, how it was fixed in law, how scholars have contested that fixing and what happens when it is applied. The cases discussed illustrate conceptual problems rather than surveying atrocity.

Lemkin and the Coining of the Term

The word was invented by Raphael Lemkin, a jurist of Polish-Jewish origin, in Axis Rule in Occupied Europe, published in 1944, combining the Greek genos, meaning a people or race, with the Latin suffix denoting killing. He had pursued the idea for over a decade, having proposed in the 1930s that the destruction of collectivities be made an international crime.

His original conception was broader than the one later adopted in law. Lemkin understood the destruction of a group as a coordinated process operating through many means, attacking political institutions, culture, language, religion, economic existence and physical security. Mass killing was one method among several, not the definition itself. This breadth matters because much of the subsequent scholarly argument has consisted of attempts to recover parts of Lemkin’s conception that the legal definition left out.

The Legal Definition and Its Constraints

The Convention on the Prevention and Punishment of the Crime of Genocide was adopted by the General Assembly of the United Nations (hereinafter: UN) on 9 December 1948 and entered into force on 12 January 1951. Article II defines the crime as any of five enumerated acts, including killing members of a group and imposing measures intended to prevent births within it, committed with intent to destroy, in whole or in part, a national, racial or religious group, as such.

Three features of this text have generated most of the subsequent debate. The first is the requirement of specific intent. It is not sufficient that a group suffers destruction, nor that the perpetrator knew destruction would follow. The prosecution must establish an intent to destroy the group as such. This is a demanding standard, and it is the reason a great many episodes of mass death fall outside the legal category, however severe they were.

The second is the restriction to four protected groups. Political and social groups were considered during the drafting and excluded, in part through the objections of states that anticipated scrutiny of their own conduct. The consequence is that the mass killing of people identified by their politics or class is not genocide in law, whatever its scale.

The third is the exclusion of cultural destruction. Earlier drafts covered the destruction of a group’s language, religion and culture; the final text retains only the forcible transfer of children. Destruction of a group’s collective existence without physical killing, therefore, sits outside the legal definition, though the term cultural genocide remains in scholarly use.

The definition was not applied by an international court for half a century. The first conviction came at the International Criminal Tribunal for Rwanda in the case of Jean-Paul Akayesu, decided on 2 September 1998, which also held that sexual violence could constitute a genocidal act. The International Criminal Tribunal for the former Yugoslavia subsequently found genocide at Srebrenica, and the International Court of Justice (hereinafter: ICJ) addressed the question in proceedings decided on 26 February 2007.

Definitions of Genocide Beyond the Law

Scholars have rarely accepted the legal text as an adequate analytical tool, and a field of comparative genocide studies grew up around attempts to improve on it. The figures below are not incidental commentators: most built that field’s institutions, and their formulations are the ones later researchers argue with.

Helen Fein, founding president of the International Association of Genocide Scholars, defined genocide as sustained purposeful action to destroy a collectivity, including by preventing its biological and social reproduction. Frank Chalk and Kurt Jonassohn, whose 1990 volume was among the first systematic comparative treatments, centred their formulation on one-sided destruction of a group as the perpetrator defines it; a move that matters because targeted groups are frequently constructed by those attacking them.

Barbara Harff, whose risk models have been used in governmental early-warning work, introduced politicide for groups defined by political position. Rudolph Rummel’s wider notion of democide is often cited alongside it, though his mortality estimates have been criticised by other quantitative researchers as inflated: his concept has worn better than his figures.

Martin Shaw argues that the concept belongs to sociology rather than criminal law. Others contend that the legal concept has shielded some mass violence by treating only racial and ethnic destruction as the supreme crime while violence conducted for security escapes it, a deliberately unsettling position that has drawn substantial disagreement.

Scholars from outside Europe and Neo-European states have pressed the definitional question further. Mahmood Mamdani, who directed the Makerere Institute of Social Research, argues that the framework inherited from Nuremberg misreads violence bound up with the formation of nation states, and that treating it as criminal rather than political obstructs its resolution. Abdelwahab El-Affendi, at the Doha Institute for Graduate Studies, shifts attention from perpetrator ideology to the narratives of existential insecurity through which populations come to understand mass violence as self-defence.

Gregory Stanton’s model of the stages of genocide, developed for prediction and prevention, runs from classification and symbolisation through dehumanisation, organisation and preparation to extermination and denial. It is a warning framework rather than a definition, valuable for identifying processes early rather than classifying events afterwards.

The Central Disputes

Three disagreements recur across this literature. The first concerns intent. Intentionalist readings require a demonstrable purpose to destroy, located in the decisions of identifiable actors. Structuralist readings hold that destruction can result from the cumulative logic of a system, such as settler expansion or forced labour extraction, without a single directing decision. The legal standard is firmly intentionalist, which is why episodes of enormous colonial mortality often fail to qualify.

The second concerns colonialism. A substantial body of work argues that genocide has been theorised around the European experience of the twentieth century, and that this obscures the destruction of indigenous peoples over several centuries. Moses and others contend that settler colonialism tends structurally towards the elimination of native societies, whether by killing, removal, assimilation or destruction of the means of subsistence.

The third concerns the risks of expansion. Robert Hayden, writing in a volume on the historiography of genocide edited by Dan Stone, argues that invoking the term is frequently a political process that creates images of the past to serve present purposes, and that its rhetorical force can forestall inquiry into whether the label is accurate. Examining the reasoning of the tribunals in the Bosnian cases, he contends that broadening the definition to encompass expulsion from strategically or symbolically important territory relativises the concept to the point where it no longer distinguishes attempts to exterminate a people from other forms of mass killing. Critics of this position reply that a definition narrow enough to satisfy such objections would exclude most historical destruction of peoples and would make the legal category nearly unusable. The disagreement is unresolved, and it is the central tension in the field.

Earlier Cases and What They Show

The colonisation of the Americas illustrates the intent problem in its sharpest form. The demographic collapse of indigenous populations after 1492 was of enormous magnitude, and its largest single cause was epidemic disease to which those populations had no immunity, which no legal definition treats as genocidal. Yet alongside this ran deliberate massacre, enslavement, forced removal, destruction of subsistence and the compulsory transfer of children to institutions intended to eliminate their languages and cultures. Benjamin Madley’s study of California between 1846 and 1873 documents killing organised and funded by state authorities, meeting the intent requirement directly, while Roxanne Dunbar-Ortiz reads the wider process as eliminationist in structure. Whether the whole is best described as genocide, or as a long process containing discrete genocides, remains disputed.

The destruction of the Herero and Nama peoples in the German colony of South West Africa between 1904 and 1908 is the case on which the agreement is broadest. The extermination order issued by Lothar von Trotha on 2 October 1904, the driving of survivors into waterless terrain and the subsequent camp system produced mortality that scholars estimate at a large majority of the Herero population and around half of the Nama. Explicit orders make the intent question comparatively tractable. Germany formally acknowledged the events as genocide in 2021, though the accompanying arrangements were criticised by Herero and Nama representatives who considered themselves inadequately consulted.

The Congo Free State under Leopold the Second, between 1885 and 1908, points the other way. Mortality from forced rubber extraction, mutilation, killing and disruption of subsistence was immense, with widely cited estimates of population loss reaching several million, though the figures are debated. Some scholars, mostly European, nonetheless decline to classify it as genocide, because the object of the system was extraction of labour and rubber rather than destruction of a group as such. Georges Nzongola-Ntalaja and other historians of the Congo treat it as a system of atrocity and exploitation. The case, again, demonstrates the difficulty of tracing intent in a case that is rooted in a deep conviction of racial hierarchy between peoples.

Contemporary Cases and Contested Determinations

Rwanda in 1994 is the reference case for legal certainty. Between April and July, several hundred thousand people were killed, overwhelmingly Tutsi, with estimates commonly placed between five hundred thousand and eight hundred thousand. Organisation, propaganda and explicit targeting were extensively documented, and the tribunal convicted on genocide charges. Mahmood Mamdani’s analysis is conceptually important: he argues that Hutu and Tutsi identities were substantially constructed as political categories under colonial administration, which bears on how the Convention’s protected groups should be understood, since the groups destroyed are often those the perpetrators have defined.

Gaza and the wider Palestinian territories are the subject of an active and unresolved legal process. South Africa instituted proceedings against Israel at the ICJ under the Genocide Convention in December 2023, and in January 2024 the Court ordered provisional measures, finding some asserted rights plausible; this is a procedural threshold, not a determination on the merits, which remains pending. In September 2025, the independent commission of inquiry established by the Human Rights Council concluded that genocide had been committed, and the International Association of Genocide Scholars adopted a resolution to similar effect. Israel rejects this situation, however, without substantively countering the intent argument. Scholarly opinion is divided, with scholars including the Jewish Omer Bartov, Raz Segal and Amos Goldberg arguing the threshold has been met, and others disputing that specific intent is satisfied. Due to the proliferation of digital media, various statements of Israeli politicians and officials, defending a racial hierarchy between Israelis and Palestinians, have led to the classification of the situation as genocide.

The situation of the Uyghur Turks in East Turkestan is contested mainly along political lines. Mass internment from 2017, pervasive surveillance, restrictions on religious practice and language and reported measures affecting birth rates are documented in scholarship by Sean Roberts, Darren Byler and others. The assessment published by the Office of the High Commissioner for Human Rights in August 2022 found serious violations that may constitute crimes against humanity, and made no finding of genocide. The government of the United States of America made a genocide determination in January 2021, and parliaments in several states have adopted non-binding motions to the same effect. A non-governmental tribunal convened in London in 2021 concluded that genocide had occurred on the basis of birth-prevention measures. The Chinese government rejects this situation. The scholarly disagreement turns largely on whether the measures at issue satisfy Article II, with some analysts arguing for cultural destruction or crimes against humanity as the more accurate classification. Given the history between Turks and Chinese, the systematic targeting of culture, identity and race and the purely political basis for denial of the genocide on the Chinese side, make this another clear instance of a genocide, where the disputed parts arise from political interests.

Conclusion

The concept of genocide was created to name something for which no word existed. This alone shows an alarming decline in the normative structure of the global society. Ancient wars were never directed at the full extermination of a people, but at full victory and conquest. What needs to be stressed is that the victors, back in the day, might have oppressed people in conquered lands, but seldom was the extermination of the oppressed the ultimate goal. At some point, probably driven by increased globalisation, the policy idea of extermination of other peoples emerged and produced the atrocities discussed above.

Concluding, what is the most deterministic characteristic of a genocide is the aggressor’s intent to remove a racial group completely. As psychological, biological and health-related, societal and economic warfare advanced and became increasingly more utilised, the definition of genocide expanded to these realms as they are very representative of the intent of the aggressor. If an aggressor is targeting the reproductive capacities of a people, it needs to be counted as genocide just as equally as the direct slaughter of a people.

In turn, the most effective countermeasure against genocide, next to great defensive capabilities, is to increase the reproductive capacity of a people. This can be done through advanced public health policies, countermeasures to psychological manipulation, sophisticated education infrastructures and policies and strengthening of socio-political identity. At a broader level, states will need to adopt devletist structures to realign their state goals with the purpose of our existence, genuine knowledge production. Only that way, we can exist cooperatively to advance global society towards a valuable goal, and do so while keeping the enriching characteristics of each people.