The last military dictatorship in Argentina was responsible for the violent, systematic and clandestine murder of thousands of citizens between 1976 and 1983, a unique event in the country’s history.footnote These facts have created a social need to better comprehend what really happened during those years, to search for the causes and conditions that made this horror possible. The Argentine case is a recent example of a more general phenomenon, the historical massacre, which has troubled historians since classical antiquity, because of the difficulty inherent in its narration. In what follows, we attempt to identify what we term the ‘representational formulae’ that have been used to portray historical massacres. First, we define what a historical massacre is and give the theoretical grounds for our arguments. Secondly, we follow the long-term evolution of three of these formulae—hunt, martyrdom, hell—up to the twentieth century. Thirdly, we provide a number of examples that point to a possible new formula: the multiplication of silhouettes and Doppelgänger. Finally, we discuss some of the risks of an endeavour of this kind.

There are several uses of the term ‘massacre’ in everyday language, applied to situations ranging from a murderous attack by a deranged individual to a crushing victory in sport. The term ‘historical massacre’ relates to a different phenomenon: the mass murder of a large group of people, who are usually unarmed and have a limited capacity to defend themselves. The origin of the word dates back to sixteenth-century France. In the 1556 pamphlet Histoire mémorable de la persécution et saccagement du peuple de Mérindol et Cabrières, the term massacre was used to describe a campaign of religious cleansing against the local Waldensians. It soon became the word of choice for Protestants recounting the worst episodes of the French Wars of Religion, including the most spectacular of such occurrences, the Saint Bartholomew’s Day Massacre, perpetrated in August 1572. Generally, in a historical massacre a group rather than an individual is responsible, and exceptionally cruel methods are used.footnote1 Moreover, the victims, dead or alive, are treated with utter contempt, while the perpetrators face no great physical risks in their endeavours. Although historical massacres are difficult to explain and the chains of cause and effect appear broken as they are taking place, this type of slaughter occurs in clearly delimited spaces and time-frames, and the people responsible for them are identifiable.

During the twentieth century a new type of mass murder emerged, related to historical massacres but different in several respects: the crime of genocide. The un Convention on the Prevention and Punishment of the Crime of Genocide, approved in 1948, defines it as:

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.footnote2

Of course, the concept of genocide is closely associated with the Shoah, but there is strong evidence that Raphael Lemkin had begun to develop the notion earlier, taking account of the Armenian Genocide.footnote3 There are obvious similarities between a historical massacre and a genocide, but also significant differences. Among them is the fact that in a genocide, there is usually a criminal state with which collective responsibility rests. Moreover, penal responsibility falls upon the individual perpetrators and those who ordered the slaughters. Several historical massacres may take place within a genocide, but the verification of a historical massacre does not mean that a genocide is taking place.

More than almost any other kind of event, historical massacres and genocides test to the utmost the relationship between fact, truth and narration. Whenever deliberate and systematic attempts to destroy a human group as such have occurred, they have been considered limit experiences that defy all available ethical, rhetorical and analytical categories. The perpetrators have attempted to hide their actions, while the survivors and defenders of the victims have stubbornly sought to bear witness to the facts. In doing so, they have faced what Saul Friedländer called ‘the limits of representation’.footnote4 In 1923, Aby Warburg suggested that magic, art and religion provide a Denkraum, a space for reflection, which allows us to approach objects that force us to confront our deepest fears and anxieties, principally our fear of death.footnote5 Our contention here is that representations of massacres, and the rhetorical and aesthetic devices associated with them, similarly allow for the creation of a certain distance or Denkraum. These devices in turn open up the possibility of addressing events that would otherwise be unbearable.

We introduced the term ‘representational formulae’ above, but it demands a more precise definition. The concept owes a certain debt to Aby Warburg’s Pathosformel—never defined explicitly, but present in most of his crucial works.footnote6Pathosformeln are not ahistorical, nor are they anthropological invariables. On the contrary, specific historical periods and different cultures are characterized by coherent clusters of perceptions and feelings. The expression of these demands a certain consistency of formal approach—principles of configuration that appear in different artistic domains and embody a wide variety of cultural concerns. In turn, these Pathosformeln are capable of change and adaptation, and are subject to appropriation and subsequent modification. Representational formulae are slightly different from Pathosformeln, in that they need not refer only to humanity’s historically determined primal experiences; they are thus more numerous and nimbler than Pathosformeln. In our understanding, a representational formula is a set of cultural devices—words, images, performances—that have been historically shaped and are, at the same time, relatively stable, so that they are readily recognizable by the reader or observer. Its aim is to allude to and show objects, figures, facts, and the relationships between then, which a given society considers fundamental elements of its common life and experience. Representational formulae are also subject to change, in that they can be modified to represent new—different—phenomena and convey novel meanings, although generally these are related to the old ones. It is precisely when links between new facts and existing formulae are difficult to trace or become highly problematic that we may find ourselves facing the limits of representation.