Review of Jacques Sémelin, Purify and Destroy, Journal of Genocide Research, 2009

Posted: December 12, 2009 in genocide reviews

Genocidal Massacres and the ‘Rarity’ of Genocide

Martin Shaw

Contribution to the review forum on Jacques Sémelin, Purify and Destroy (London: Hurst, 2007), Journal of Genocide Research, 11, 3, March 2009, pp. 149-63.

The paradox of genocide studies is that while an enormous growth in the literature is producing ever richer case and comparative studies, there remains little agreement on essentials. Although it would be wrong to speak of ‘crisis’, there is (as Scott Straus has argued) no consensus among major authors on the definition of genocide, the universe of cases, or the type of explanation that is appropriate.1 When a scholar of Jacques Sémelin’s standing produces an ambitious, wide-ranging work like Purify and Destroy, we should ask, first, how does the book work in its own terms, but also second, how well does it address these underlying problems of the field?

Sémelin has written a hugely knowledgeable, wise and stimulating book. Its primary strength is not just his mastery of the three cases (Nazi Germany, Rwanda, Bosnia-Herzegovina) around which he constructs his argument. His supreme talent is to present new ideas through constant engagement with the historical realities and theories of genocide, so that at every turn in the discussion we have the sense of the author working through his analysis, evaluating the ideas of fellow scholars, and coming to mature and reflected judgements. Sémelin’s translator, Cynthia Schoch, has succeeded admirably in rendering his fluent, attractive writing into English, although he has been less well served by his British publisher who has allowed too many small errors to appear.

In contrast to writers who have examined a wider range of cases, Sémelin’s more selective approach means that the cases illustrate his general argument rather than facilitating an analysis of genocide in general even in a given period. But the smaller number of cases means that he is able to organise his material thematically while still providing a detailed discussion of each issue for each case. This will make his book very useful for students. Sémelin is superb at calibrating general arguments with specific accounts, and all readers will find much food for thought in his finely balanced judgements.

Sémelin’s initial focus is ‘massacre’, rather than ‘genocide’, and he invites his reader to follow him ‘along the tortuous paths that lead human beings from peace to barbarity.’ (p.7) His starting-point is that ‘massacres are mainly born out of a mental process, a way of seeing some “Other” being, of stigmatising him, debasing him, and obliterating him before actually killing him.’ It is a ‘type of discourse’ that ‘can set the stage for mass violence’. (p.13) These discourses, in turn, work in circumstances of a ‘collective trauma’ for a nation or community, where ‘the very “soul “ of a disoriented and paralysed population … appears to be at stake’. Then the social and political actors who own these discourses ‘take charge of the collective emotions associated with this mass trauma’ (p.15),  ‘eliciting in society the desire to destroy what has been designated as the source of fear’. They develop ‘an imaginaire of death’. (p.17)2 Sémelin then develops an account of three core themes in the ideologies of massacre: identity, purity and security.

Sémelin justifies his approach through Frank Chalk and Kurt Jonassohn’s notion that the perpetrator defines the target group: ‘Good methodological sense thus dictates that we give priority to analysing this imaginary construct, its structure and its basic themes, in order to understand how the process of mass violence is triggered.’ (p.21) We do indeed need to grasp how particular civilian populations come to be constituted as ‘enemies’, and how this idea is legitimated among other civilians, the constituency of the state, party or armed group that is responsible. Sémelin’s answer is that the ‘regressive desire for perfect “unity”’ and ‘the narcissistic fantasy of complete peace characteristic of the infant expelling the bad object … is not far off.’ (p.33) The discourse offered to the public is more important than the individual psychology of a paranoiac leader, and what is most important is that the world view of a ‘sect’, like the Nazis originally were, ‘succeeds in irrigating a broader social body to the extent of seeming commonly shared by all.’ (p.45) The imaginaire ‘spurs men on inexorably’ towards rational action to ‘eliminate’ the enemy, towards ‘the threshold of war’. (p.48)

This is a ‘psychological approach’ to a ‘delusional rationality’, but Sémelin recognises that this ‘cannot replace historical or sociological enquiry.’ This core approach has been categorised by Straus as ‘idealism’, because of its explanatory priority for ideas3, but it could also be called social-psychological materialism. Sémelin emphases group dynamics – the roles of intellectuals, political forces and states in fostering destructive ideologies, and of individual and social receptivity to them, the combination of which ‘determine[s] the risk of a group sliding into acts of mass violence.’(p.53) And he acknowledges that ‘Everything will depend in fact on how the political situation develops, on whether or not the economy is in decline, to what extent there is social instability, the state of the international environment, etc.’ (p.62). The ‘international’ element is emphasised by a whole chapter which contains many useful insights.

So from an ‘idealist’ starting-point Sémelin ends up with a more balanced account of the causes of violence, which chimes quite well with other recent accounts like that of Michael Mann.4 Indeed he acknowledges that ‘The dynamics by which mass murder develops are extremely varied: they stem from a larger number of factors …’ (p.168) – a welcome, even too open, explanatory pluralism but one that leaves problematic the status of some stronger theoretical positions he adopts. However the social-psychological returns with his pseudo-religious account, derived from Réne Girard, of the violence itself: ‘murder is no longer a taboo to be observed but becomes, on the contrary, a fundamental practice – not of a new religion, but of different conception of collective transcendence … violence … that regenerates the group through the sacrifice of those designated as being responsible for the crisis.’ (p.90)

There are three problems with Sémelin’s core approach. First, it gives too much importance to direct three-way dynamics between genocidal elites, the nations they mobilise, and ‘enemy’ groups. The ‘political rise’ of the former, he writes, ‘occurred above all through the stigmatisation of a minority group.’ Yet these relationships were not always so direct, as Sémelin immediately acknowledges when he writes that ‘it is by no means certain’ that anti-Semitism ‘played a decisive role in [Hitler’s] electoral success.’ (p. 69) This qualification could be deepened: the more extreme policies the Nazis pursued against  the Jews in occupied eastern Europe, leading to the Final Solution, may have assumed their earlier public stigmatisation, but they were not achieved through public mobilisation of Germans. Likewise there was a considerable disconnect between Milosevic’s political mobilisation in Serbia and Serbian violence against non-Serbs in Croatia, Bosnia and Kosovo, as Sémelin points out (p.213).

Second, the argument overemphasises the emotional meaning of the violence to national publics, as well as to the perpetrators: far from filling an openly ‘sacrificial’ role, in the Nazi and Serbian cases it was mostly hidden from them and denied. The argument could fit better with Rwanda, where popular mobilisation and genocide were simultaneous and violence was manifest to the ‘nation’ that the génocidaires mobilised. But even here the ‘sacrificial’ element is overstated: most studies stress more mundane factors.

Third, this leads Sémelin to depoliticise genocide. He claims that unlike in war, which as Clausewitz noted is politically constrained, in massacre-violence ‘there is no longer any barrier or restraint that would seem capable of reining in violence as it unfurls.’ (p.224) This seriously misrepresents both war (Clausewitz’s point was surely that escalation can take violence beyond initial political objectives) and genocide, which is generally politically driven, and in which murderous escalation always remains in tension with the political objectives of the perpetrator organisations.

Another difficulty is opened when Sémelin comes to pinpointing the ‘decision’ for massacre. His discussion now assumes competing nationalisms, so that ‘The joint refusal of two factions to compromise precipitates the transition into action.’ (p.177) One can see the relevance to Bosnia and Rwanda: but to the Holocaust?5 Interestingly Sémelin has got himself into the same position, of advancing general theses about mass violence that don’t fit the archetypal Nazi case, in which Mann found himself. Is this common dilemma an unfortunate accident, or does it tell us something more general about studying genocide? I came to genocide studies from the study of war. In that field, one would not assume that just because the First World War, the Iraq War and the Congolese Civil War are all ‘wars’, one should look for common explanations. So why should we assume that Bosnia, Rwanda and the Holocaust must have fundamentally similar explanations?

We should recognise that the explanatory commonalities of all genocides are of a fairly general order (for example, connections with war and militarism6), and so can only indicate the direction in which specific hypotheses may be sought. Much comparative genocide study, as in this volume, often assumes stronger affinities than are possible between episodes in different times and places. Genocides, like wars, are large-scale, open-ended historical episodes, each of which is ‘unique’ in an important sense. They are not as comparable as regular, narrowly constrained social events like elections and football matches. Comparison is crucial to understanding but it must be informed by a much stronger sense of historical specificity and periodisation. Otherwise good scholars will continue to put forward over-general theses and then be obliged to acknowledge so many exceptions and secondary differences that their theses will not fully stand up.

Sémelin relies on the concept of ‘massacre’, which he defines as ‘a form of action that is most often collective and aimed at destroying non-combatants.’ (p.4) He uses ‘massacre’ because he is cautious about the concept of genocide, for good reasons and bad. He worries about ‘genocide’s’ legal entanglement, and rightly argues that ‘it is crucial for genocide research to disengage itself from the legal approach’. (p.310) However he also believes that by starting with ‘massacre’ he is adopting a ‘non-normative vocabulary’- yet ‘massacre’ surely also carries normative connotations. He believes that before defining genocide, ‘research must first concentrate on massacres; not all massacres can be considered as genocide, and genocide necessarily involves one or several massacres.’ He appeals to ‘simple good methodological sense’ to justify ‘privileging massacre as an object of study, the point being one of identifying at what point and in what circumstances a massacre becomes genocide.’ (p.323) In order to do this, the study of massacres ‘must necessarily fit within the framework of a destruction process, of which it is the most spectacular expression.’ (p.324)

Here Sémelin begins to suggest a relationship between massacre and genocide, but his account is inadequate. First, ‘massacre’ usually refers to a specific event in a particular time and place: his use of it to describe large-scale historical episodes is confusing. Second, although genocides include massacres, it is not as obvious as Sémelin suggests that we should start with massacres in order to define genocide. Massacre is only one of the manifestations of genocide: why start with massacre rather than with expulsion, rape or execution, each of which can also signal a destructive process? Third, surely we need to begin with the idea of what is being destroyed? Sémelin suggests that we are talking about ‘an organised process of destruction of civilians directed at persons and their property.’ (p.325) However this does not accurately capture the common feature of genocide: the destruction of persons and property are ways of attacking the social networks, identities and ways of life of particular groups.

Sémelin now proposes a further distinction: between ‘destroying to subjugate’ and ‘destroying to eradicate’. In the first case, the aim is ‘to annihilate a group partly in order to force the rest into total submission. The destruction process is, by definition, partial but is intended to have an overall impact.’ (p.327) Massacres, therefore, are instruments of subjugation.  ‘The second dynamic’, Sémelin asserts, ‘is altogether different. It aims not so much to subjugate individuals to a given political power as to eliminate a community from a more or less extensive territory controlled or coveted by a state.’ (p. 335)

Some such distinction is important: a subjugated society, which remains in situ, may survive repression and the massacre of some of its members. In this case partial destruction does not lead to total destruction. There may be what has been widely called a ‘genocidal massacre’, in which some members a particular civilian group are killed because of their group membership, but there is no large-scale destruction of the group as a whole. However one of the real oddities of Sémelin’s book is that in hundreds of pages discussing ‘massacre’ he never mentions the idea of genocidal massacre – presumably because it doesn’t fit his restricted idea of genocide, which I discuss shortly.

A society that is eliminated from its home territory, on the other hand, largely loses its social networks, identity and ways of life. By distinguishing ‘destroying to eradicate’ from subjugation, Sémelin recognises the importance of space and territory: people and their social relations are territorially located, and destroying the spatial forms of these relations is a fundamental way in which, in the fixed agrarian and urban societies of the civilised era, a society is destroyed. Here small massacres, often going hand in hand with expulsion, rape, etc., may be designed to cause large numbers of people to flee in fear and so lead to the destruction of the society. Here too a massacre would conventionally be called ‘genocidal’, but the larger destruction of the society would also be ‘genocide’ in terms of Raphael Lemkin’s original definition and, arguably, of the Genocide Convention too.

Yet (and here is another similarity with Mann) Sémelin wants to make yet another distinction before ‘the notion of “genocide” can be reintroduced, this time as a concept of the social sciences.’ (p.340) It is only in the ‘final stage of total eradication’, he says, that we should talk about genocide. This final stage can be seen in the Holocaust and Rwanda, but not, he argues, in Bosnia. In effect, genocide is reduced to total mass murder, and Lemkin’s idea that we need a single, overarching concept for all processes designed to lead to the destruction of a people or group has been cast aside yet again.

To justify this conclusion Sémelin makes some curious digressions. He makes a lot of the fundamental difference between fascism and communism, because seeing them as similar would ‘equate totalitarianism with genocide’ (p.342). This is a strange argument: surely we can distinguish ‘subjugation’ and ‘eradication’ tendencies within both fascism and communism? He also insists on the difference between the destruction of racial enemies by the Nazis and of class enemies by Communist regimes (because ‘what exactly is a kulak?’) as though the category of race was more ‘real’ than that of class. Thus he loses the Chalk-Jonassohn thread that he earlier endorsed, which would lead to seeing both these ideas as social constructs of the perpetrators. He also seems to believe that a broader concept must open the door to seeing the US atomic bombings of Japan as genocide, although it seems fairly easy to distinguish these acts as forms of (highly degenerate) war.

Sémelin’s bottom line, though, remains ‘discerning the difference in destructive intensity and extremity between what has been called here “ethnic cleansing” on the one hand and “genocide” on the other. … We have seen that the ethnic violence in Bosnia had certain limits, whereas it apparently had none in Nazi Europe or Rwanda.’ (p.345) But it would be more accurate to say that it had different limits. Sémelin asks that we consider genocide as process, but then defines it by the end of the process in particular cases: this is a kind of teleological approach. He is therefore wrong to say that a process-approach ‘means to consider genocide as a specific dynamic of violence.’ (p.347) It would be more accurate to say that genocide involves variable dynamics of violence.

By trying to categorically separate Bosnia from the Holocaust, Sémelin has fallen into a common analytical trap. On the one hand, despite his emphasis on imaginaires, his formulations make it impossible to recognise the genocidal ideas and policies of the Serbian and Croatian leaderships, merely because their policies didn’t undergo the ‘final’ radicalisation to total murder that happened elsewhere. On the other hand, despite his emphasis on process, they make it difficult for us to see the process of the Nazi genocide as a whole, since they must lead us to categorically separate the ‘non-genocidal’ expulsions of Jews and others (pre-1941) from the ‘genocidal’ mass murders (1941-45). This is despite the fact that Nazi ‘ethnic cleansing’ and ghettoisation policies were informed by the same genocidal imaginaire that later led to the extermination camps.

Even more seriously, making ‘the ultimate fate of the victims’ the criterion for genocide (p.346) confines genocide to a few, truly exceptional episodes: it is ‘a rare event’, as Straus also contends.7 This is a dangerously complacent perspective: for all the smaller-scale massacres, mass rapes and brutal expulsions are just as informed by genocidal imaginaires as the very few cases that conclude with total murder. For example, the mega-genocide in Rwanda was preceded, surrounded and followed by smaller but often quite extensive genocidal massacres – in Rwanda itself, in Burundi, and in the Congo. We need to see genocide genuinely as historical process, not rare outcome.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s