Cover Page

What Is Genocide?

Second Edition

Martin Shaw



Many people could probably give an answer to the question, ‘What is genocide?’, and name one or two cases – maybe the Holocaust and Rwanda – to which the term applies. They might assume that if a clear definition is required, a dictionary will answer the question in a few lines. The idea that a whole book could be devoted to this question, rather than to the history and politics of the problem, might seem surprising. Yet if we go beyond a few obvious and well-publicized cases, the scope of genocide is not immediately clear: even scholars do not agree on what should count. The popular idea of genocide, which equates it with mass killing, begs a lot of questions, was not what the originator of the idea meant by it and is not how most academics or the United Nations’ Genocide Convention define it. Yet people who study the question disagree profoundly among themselves about the answer, and people who use the idea in political life often choose the meaning that suits their cause, rather than a coherent idea. Genocide is a highly contested concept, politically as well as intellectually. So when the first edition of What is Genocide? was published a decade ago, it found a ready readership.

I have not changed my answer to the question, so it may well be asked why a new edition was necessary. One answer is that the book left some issues underdeveloped, and the rapid growth of the literature has made these more compelling. In particular, recent work has involved much fuller examination of the ideas of Raphael Lemkin, which stand at the heart of this book. The effect is that the discussion of Lemkin in the first edition is insufficient. While I indicated some problems with his approach, I did not explore these in the depth which now seems necessary and (because of new research) possible. I have therefore expanded the treatment of Lemkin to a full chapter. Two other key contributions which are particularly relevant to my argument, Leo Kuper’s idea of genocidal massacres and Tony Barta’s idea of structural genocide, are also more fully treated.

A second reason for revising the book was to make my idea of structural analysis clearer, and in particular to amplify a key distinction between the structure of genocide itself, on the one hand, and the larger structural contexts in which genocide occurs, on the other. I have rewritten what are now chapters 8, 9 and 10 to reflect this. Chapter 8 also highlights the idea of the hybridity of genocide and war, implicit in the first edition, but which I only made explicit in an article in the Journal of Genocide Research after it was published.

Finally, although the book was complimented on its clarity, I felt that its organization and presentation could be improved so as to make it more accessible, especially for new students of genocide. So I have simplified the introduction and conclusion, highlighted illustrative material and presented my new definitions as the conclusion to the volume. I have also removed some discussions, for example about the role of civilians in war and the Nazification of the social sciences in Germany, which were tangential to the argument.

I have several debts apart from the many which will be obvious from the text. In preparing the original edition, my then Sussex colleagues, John Holmwood and William Outhwaite, gave valuable advice on reading in areas where my sociology was rusty. In preparing the new edition, Dirk Moses provided typically pertinent comments on an early draft of chapter 2 (on Lemkin), and Polity’s two anonymous reviewers made very helpful comments on the book as a whole. The United Kingdom’s Economic and Social Research Council granted me the Research Fellowship in 2004 and 2005, which originally enabled me to write this book, as well as to finish The New Western Way of War, which discusses violence against civilians in a different context, and I remain very grateful to it.

I hope this new edition will explain the issues involved in the idea of genocide more effectively to new readers, while still offering considerable interest to those who were acquainted with the original work. As usual, I alone am responsible for the views expressed.

Martin Shaw
Devon, 2014

— 1 —

This book addresses the question: how should we understand the idea of genocide? Genocide has been a central issue of world politics several times in recent decades, especially in Bosnia, Rwanda and Darfur. Its history has also been a topic of controversy, in countries like Germany, Japan and Turkey over murderous violence in the two world wars, and in North America and Australia over earlier violence against indigenous peoples. The spectres of the Holocaust and the Nakba stalk twenty-first-century conflicts between Israelis and Palestinians. Genocide issues continue to arise in all too many current conflicts where populations are targeted with violence. Allegations of genocide are widely made and, invariably, disputed. All too often, ‘genocide’ becomes a tool in political controversy, claimed by one side and denied by the other. Whenever new challenges arise, the same confused debate occurs over whether attacks on civilians constitute ‘genocide’, ‘ethnic cleansing’ or just the excesses of a dirty ‘civil war’, often as though similar arguments had not already raged in earlier cases. Few ideas are as important in public debate, but in few cases are the meaning and scope of a key idea less clearly agreed.

It might seem axiomatic that scholarship should assist in the clarification of ‘genocide’, and thus help all those who feel that the idea assists them to understand terrible episodes of human history. Yet to many, ‘definitional’ discussion over horrendous experiences of violence can seem beside the point. On this subject, normal academic assumptions cannot be taken for granted. The Auschwitz survivor Charlotte Delbo speaks of ‘useless knowledge’, when she refers to experiences that were ‘so dark as to be unforgettable but also so overpowering that the more one encounters their stark realities – even in reading about them, let alone in the flesh or in personal memory – the more likely we are to be disoriented and overwhelmed by them.’1 Genocide has often been seen as involving murderous tendencies so horrible and irrational as to be both utterly exceptional and virtually inexplicable. It can seem devaluing to discuss them within the explanatory frameworks that scholars adopt. This crime of crimes demands more than a normal commitment to scholarship and truth. Study presupposes, John Roth argues, ‘values that are not contained in historical study alone. . . . Any debate . . . is worthwhile just to the extent that it never loses sight of the fact that ethical reasons are the most important ones for studying these dark chapters in history.’2 Here scholars must bear witness, show solidarity with victims and stand unequivocally on one side of the historical process.

Hence, even scholars complain about ‘definitionalism’, excessive attention to the details of the concept. The psychologist Israel Charny warns that extended debate on definition can lead to the point ‘where the reality of the subject under discussion is “lost”, that is, no longer experienced emotionally by the scholars conducting the enquiry’.3 The historian Herbert Hirsch concurs: ‘It is unfortunate that Holocaust and genocide studies are being pressured into a phase of social science rationality . . . only to become bogged down in the elusive variable and definition, as everyday life becomes almost entirely eliminated from their concern.’4 But if we are to do justice to the victims, and help understand the enormities of violence, we cannot but engage with these issues abstractly as well as concretely. The point is certainly to ‘prevent and punish’ genocide; but to do this, we must clearly understand the beast. Issues of definition cannot be avoided in this task, and they will take time and care because simple ideas are often too simple. As the social theorist Max Weber put it: ‘The apparently gratuitous tediousness involved in the elaborate definition of . . . concepts is an example of the fact that we often neglect to think out clearly what seems to be “obvious”, because it is intuitively familiar.’5

Lemkin and the necessity of classification

In any case, definition is part of the subject matter of genocide. The Nazi genocide was a crime of social classification, a sociological crime in which pseudoscience defined and classified people according to their ‘race’. The journalist William L. Shirer described the Nazis, whom he observed at first hand, as ‘sociologists’ because of how they were obsessed with these classifications.6 Not all genocide is systematically pseudoscientific, but classifying populations and individuals in racial and other hostile terms is one of its essential components. The danger of classification is always, Nigel Eltringham suggests, that ‘we “misplace concreteness” and set out to “prove” that our abstract concepts . . . really do correspond to reality, rather than being contingent approximations.’7 Genocidists try to enforce their classifications through physical violence, which backs up the conceptual violence of arbitrary representations.

This is, however, an abuse of definition and classification, which are inescapable parts of human cognition and social life. It is important to note that the targets of genocide also classify, and implicitly or explicitly advance, their own definitions. They assert their understandings of groups to which they belong, their versions of identity, rather than simply accepting their attackers’ classifications. They assert their status as civilians, refuting genocidists’ beliefs that unarmed people can be treated as combatants. They define themselves not only as victims, the passive recipients of genocidal violence, but also as resisters, civilian or armed. In the struggles over genocide, people who are targeted with violence also try to impose counter-classifications on those who would classify them, often describing them as tyrannical, cruel and criminal.

The idea of ‘genocide’ fits into this pattern. Its originator, Raphael Lemkin, wanted to impose, through international law and historical inquiry, a new kind of classification on the perpetrators of violence. In his definition – which I discuss fully in the next chapter – all kinds of destructive anti-group acts, committed by any actors, are seen as belonging to the same class, and are thereby criminalized. The strength of the idea is its breadth: for Lemkin acts like killing, deportation, dispossession and cultural destruction were not simply distinct crimes but manifestations of the overarching crime of group destruction. He aimed to entrench this idea both in international law and in historical inquiry so that a general problem of violence against population groups would be widely recognized.

Lemkin’s is a powerful legal and sociological classification, imbued with universal and humanistic values, containing huge moral and political significance. It would not be so powerful if it was vague or imprecisely defined. We do not have to adopt Lemkin’s terminology or definitions; indeed we cannot avoid modifying them. Yet if it was important that he defined ‘genocide’, it is also important that we are aware of how we change its meaning. If we use it in new ways, or introduce new terms to describe some of the phenomena it originally designated, we need to explain why.

The changing problem of definition

Lemkin coined the word in 1944, and initially it meant what he said. However, ‘genocide’ quickly became widely used and – as is normal when a word escapes its inventor – its meaning began to change, in subtle and not so subtle ways. In particular, it was redefined for a crucial international legal document, the 1948 Convention on the Prevention and Punishment of the Crime of Genocide (the Genocide Convention) in the drafting of which Lemkin played an important role. After that, it entered popular political discourse in most languages and – somewhat belatedly – became a concept of academic social science and history from the 1980s onwards. After the end of the Cold War, in 1989–91, it gained new leases of life in all these fields. World politics opened up more to human rights concerns, legal institutions actually began to try genocide cases and academic study deepened.

A large part of the problem of this book is that public, legal and scholarly discourse has changed genocide’s meaning in questionable ways, often without bothering to justify it, or has simply used the term loosely. In the expanding field of genocide studies, amidst the array of often impressive case and comparative studies, the debate on what genocide means has hardly advanced since the early 1990s. Many scholars (not just lawyers, who are bound to acknowledge its centrality to genocide law) uncritically use the Convention as their benchmark, despite its generally admitted inadequacies. This situation means that, despite many insights, much scholarship gives inadequate answers to the vexed question of the meaning of genocide.

Partly because of the powerful emotional, moral and political interests at stake in all these discourses, ‘genocide is an essentially contested concept par excellence’, Christopher Powell notes.8 This is not just because of these interests, however: it is also because it is an inherently complex matter that can be described in a variety of ways.9 Lemkin deliberately proposed a concept that covered a wide range of acts and diverse historical events, so complexity was unavoidable from the start. In the current debate, one of the few things that everyone agrees on is that genocides are large-scale, violent episodes. It is in the nature of such events that they are individually complex and collectively varied. Adding complexity and variation to the deep interests that arise in discussing such matters, it is not surprising that ‘genocide’ is contested.

It follows that there is no single correct answer to this book’s question, ‘What is genocide?’ Genocide is not a simple reality, ‘out there’, which I can just get hold of and of which I can give the reader the ‘correct’ definition. So the question ‘What is genocide?’ really boils down to ‘What should genocide mean?’ and to this many answers have been given. Since the word has to bear the pressure of many different moral, legal and political as well as academic demands, it is difficult to devise a definition which will satisfy them all. Still this does not mean – as too many scholars as well as others assume – that we can simply define genocide anew each time in whichever way appears most convenient for the particular moral, political, legal or academic project that we embark on. On the contrary, usage of ‘genocide’ must be respectful to the history of thinking about the word. All serious concepts must be used consistently – with internal coherence of meaning as well as valid reference – and must be capable of extended justification. We need a concept whose parameters are clear and logical, which makes the most sense of a range of cases.

A sociological and historical concept

Although Lemkin first proposed ‘genocide’ in order to establish it as a legal category, it is clear that he always regarded it as a sociological and history category too. Thus, if it was in legal and political contexts that it first gained currency, it was always accepted that it refers to a certain class of historical social phenomena. Lemkin quickly followed his initial legal and political ventures with the beginnings of a comprehensive historical study. He continued to offer his own essentially sociological definitions despite the legal and political primacy of the Convention’s. It follows that the main task of definition belongs to those who professionally study historical social phenomena, namely social scientists and historians.

Of course, legal authorities have grappled with our essentially sociological question, as have scholars in other disciplines. Unusually, this sociological concept was defined in international law before social scientists could offer their own definitions, and this created a situation in which legal authorities continued to address the meaning of genocide autonomously from wider academic debate. Social scientists and historians are well advised to address legal discussions, as I shall in this book, but we should not be unduly deferential to them. The main task is to develop a sociological concept that will be useful in historical inquiry. Here sociology stands for the concept-producing role of the social sciences in general, and history for the empirical study of past, present and future phenomena that is undertaken by social scientists and other humanities scholars as well as by professional historians. The historical and sociological projects of comprehensively understanding and explaining classes and episodes of phenomena are more fundamental than the legal project, which is ultimately directed towards ascertaining the culpability of individual and collective actors for specific acts. Although legal cases (especially against state and military leaders) are sometimes claimed to provide historical truth, they function according to rules designed to assess the culpability of individuals and concepts that result from political compromise. Therefore, they cannot substitute for sociological concept-making or historical research.

This book approaches the question of definition, then, from the standpoint of sociological theory and method. It does so, however, based on a commitment to empirical historical and social scientific research, in the belief that conceptual inquiry is only a preliminary undertaking. This inquiry needs to lead to both the individual study of particular historical episodes and the general study of classes of episodes. More precisely, my approach belongs to historical sociology, a sub-discipline of sociology that links sociological analysis closely to broad trends of historical development, as well as applying social theory to particular historical cases. Historical sociology in general can be considered a kind of theoretical history, or the application of sociology to historical problems. Most historical sociology involves middle-range analysis, and elsewhere I have carried analysis of this kind on genocide, including a discussion of its methodological issues.10 The present book, while based on the same approach, is instead preoccupied with the conceptual and theoretical questions. In this sense, it might be considered a work of historical social theory rather than historical sociology in the most common sense. Yet while the book focuses unapologetically on sociological conceptualization, at all stages I link the argument to particular cases and historical themes.

Structure of the book

The book addresses two general problems in the conceptualization of genocide. First, the discourses surrounding definition are often incoherent, with widely divergent proposals which often take little account of each other’s rationales and are poorly reconciled – as I have suggested, it often seems as though scholars believe they can simply propose their own definition and follow it in their research. Second, the discourses of genocide studies, although informed by sociological as well as legal conceptualizations, are often poorly reconciled with social theory – not only does legal discourse often address sociological problems without recourse to sociological knowledge, but empirical students of genocide often find it convenient to leave conceptual and theoretical issues at the door.

The structure of the book follows from this diagnosis. Part I focuses on key authors and themes in the genocide literature, in order to address their inadequacies and incorporate their achievements in a more rounded conceptual framework. This part starts with the history of the genocide idea: chapter 2 provides an extensive discussion of Lemkin’s founding approach, and chapter 3 shows how this has been modified in the Genocide Convention and the later genocide literature. This part then continues by addressing three issues which, in my view, confuse the conceptualization of genocide: chapter 4 discusses how the Holocaust has been used as a standard of genocide in general; chapter 5 outlines the issues involved in the substitution of ‘ethnic cleansing’ for genocide; and chapter 6 discusses the wider proliferation of ‘-cide’ concepts that has affected the study of genocide. Building on the arguments in these chapters, Part II proposes a new sociological conceptualization of the key issues, focusing on the relationship of agency and structure. Chapter 7 deals with the idea of intentionality and the alternative idea of structure in the analysis of genocide. Chapter 8 moves on to issues in the analysis of the structure of genocide conflict and its relationship to war. Chapter 9 discusses the actors of genocide, including the key idea of ‘groups’, and issues of genocidal process. Chapter 10 moves to issues of explanation, examining the structural contexts of modernity from which the causes of genocide arise: culture, economy, politics, war and international relations. Chapter 11 concludes by summarizing the arguments of the book and presenting my new definitions.

The argument

This book therefore presents a comprehensive outline of the idea of genocide and guides the reader in weighing how it can be best understood. However, it is important to emphasize that it also puts forward a particular point of view. I contend that Lemkin’s original proposal, that we need a general concept to describe all targeted destruction of population groups, is seminal and needs to be restored to its central place in the understanding of genocide. At the same time, I address the inadequacies of the intellectual underpinning that Lemkin provided for this idea. He was not a trained social theoretician, and his framework was an insufficiently developed reflection of some common ideas of his time. This led to difficulties in his conceptualizations of groups, biology and culture, and even in his terminology of ‘genos’ and ‘genocide’.

Lemkin’s was, I argue, appropriately a broad concept. In contrast, many later definitions, from the United Nations (UN) onwards, have followed a remorseless trend of narrowing the scope of genocide. In the end, genocide has become for many writers little more than mass murder, so that cases in which the majority of a population are not killed are excluded from its scope. In the course of this narrowing, while the Convention created an important legal standard and academic writers propose some important advances on Lemkin’s thinking, the literature has undermined the core of Lemkin’s approach. This undermining is complemented on the one hand by a distorted role for the Holocaust in the general understanding of genocide, and on the other by a proliferation of problematic new concepts like ‘ethnic cleansing’, ‘politicide’ and other ‘-cides’ that describe aspects of what Lemkin, appropriately, saw as genocide.

At the same time, the literature has had enormous difficulty with key concepts introduced by Lemkin and embodied in the Convention. Misinterpretations of these ideas have been cemented, especially through international law, as obstacles to understanding. Here, I give three examples, but there are more in the body of the text. First, the idea that genocide is ‘intentional’ action has been reified into a sociologically unrealistic concept of a fixed ‘special’ and ‘ulterior’ intention, which is given excessive weight in understanding genocide. This idea of intentionality has blocked the normal social scientific and historical concern with the interplay of what is generally called ‘agency’ – action that determines the course of events – with ‘structure’ – recurring patterns of social relations that shape the practice of genocide. Second, the idea that targeted populations must be stable ‘groups’ has involved the sociologically unviable idea that ethnic, national and similar collectivities are more fixed than they really are, leading to the exclusion of much targeted destruction of populations from the scope of genocide. Third, the idea that genocide can be practised in ‘peacetime’ as well as war has led to a failure to understand the deep connections of genocide with war.

All these failures arise, I argue, from the lack of a coherent sociological understanding of genocide, not only as social action but also as a structure of conflict involving social relationships between different collective actors. These actors cannot be described, moreover, in the conventional trinity of ‘perpetrators’, ‘victims’ and ‘bystanders’ because this terminology ascribes agency only to genocidists, excluding both the resistance of targeted populations and the impact of third parties’ actions. Moreover, understanding the structure of genocide as conflict must be complemented by understanding the larger structural contexts in which this kind of conflict arises. Explanations of genocide, dealing with objective ‘causes’ as well as the subjective orientations of the actors, must necessarily attend to these contexts. Reviewing the main possibilities, I argue that the principal direct causal context concerns the nexus of political and military power, understood in international and global as well as national and local terms.


The Genocide Idea