Department of English
University of Victoria
PO Box 1700 STN CSC
Victoria BC V8W 2Y2

For God’s sake, let us sit upon the ground
And tell sad stories of the death of kings—
How some have been deposed, some slain in war,
Some haunted by the ghosts they have deposed,
Some poisoned by their wives, some sleeping killed,
All murdered.
(Shakespeare, Richard II)

It is no secret that professional anthropologists have shown little interest in generative anthropology. At our annual meeting in Los Angeles in 2013, we were honoured to hear a plenary address from the distinguished anthropologist and Africanist Jean-Loup Amselle. Though his talk was extremely interesting, Amselle did not at any point discuss the core premises of generative anthropology. In hindsight, the audience should not have been surprised by this reluctance to address the originary hypothesis. In his major anthropological work, Mestizo Logics, Amselle flatly rejects the idea that origins are important: “The analysis. . . of ‘mestizo logics’ allows one to escape the question of origin and to hypothesize an infinite regression. It is no longer a question of asking which came first, the segmentary or the state, paganism or Islam, the oral or the written, but to postulate an originary syncretism, a mixture whose parts remain indissociable.”[1] Evidently, in mainstream anthropological circles, the fear of recreating the naive evolutionism of nineteenth-century anthropology is still very much alive. As Ernest Gellner colourfully puts it, after Malinowski slew the king of anthropology James George Frazer (whose theory of the dying god was once so influential), there has been little talk among anthropologists about the tragic origins of the state in the death of the scapegoat king.[2]

As someone who teaches Shakespeare for a living, I am drawn to theories that explore the relationship between the death of kings and statehood.[3] I was therefore very happy to learn that Simon Simonse’s great book, Kings of Disaster, which was first published in 1992 by E. J. Brill, has been revised and updated in a 2017 edition by Fountain Publishers.[4]

Simonse’s book, which I first read when I was in graduate school, is an account of the Nilotic peoples of southeastern Sudan, in particular, the communities between the upper Nile and Kidepo rivers, homeland of the Bari, Lotuho, Pari, Lokoya, and Lulubo. Simonse argues that these polities provide an especially revealing picture of human social organization. Sharing elements of the acephalous “segmentary” society of the Nuer people to the north (dualism) as well as aspects of more centralized societies such as the Buganda kingdom to the south (centralism), these Nilotic communities represent a transitional stage between segmentary groups and state societies. Not yet full-fledged states, these polities nonetheless have kings or “rainmakers,” who are revered and, in times of disaster, sacrificed.

Simonse’s thesis is that it is possible to see the continuity between acephalous or kingless communities and full-fledged kingdoms or states when one recognizes the central role of sacrifice in both dualist and centralist modes of social organization. Because of the century-long taboo on evolutionary models of human social organization, however, anthropologists have been reticent to make this sort of argument, preferring for the most part to stick to synchronic models, as in the structuralism of Levi-Strauss or the structural-functionalism of Evans-Pritchard, whose landmark study of the Nuer first presented the idea that African societies can be divided into dualistic and centralized modes of social organization (though, as Simonse points out, Evans-Pritchard refused to speculate on the historical relationship between the two modes of social organization). As I suggested above, Amselle’s refusal to address the question of origins is, in this sense, perfectly in line with mainstream anthropology since Malinowski, Evans-Pritchard, and Levi-Strauss.

Simonse does not return to the evolutionism of James Frazer (though Frazer is treated quite sympathetically in the book). Instead, he aligns himself with René Girard, whose theory of sacrifice he believes to be the best theory available to explain the continuity between dualist and centralist modes of social organization. There have not been many anthropologists willing to take on Girard’s theory. At the very least, therefore, Simonse deserves a hearing for going where other anthropologists have feared to tread.

Simonse argues that both dualism and centralism depend upon scapegoating. In dualist societies, the victims are taken from enemy groups. At the village level, communities are organized by age, territory, and lineage. For example, each village is led by the monyomiji, the adult males who are responsible for keeping the peace. They are opposed, on the one hand, by the younger males, who will eventually become monyomiji themselves and, on the other, by the elders or retired monyomiji. Cutting across age sets are the clans or moieties, which are associated with particular places or animals. Each clan has a separate meeting place in the village, and these locations are organized spatially so as to represent alliances between clans. When one clan fights with another, its neighbouring clan is expected to lend support. The system is designed to prevent one clan from dominating others. Like the schoolyard pick, the aim is to create an even distribution of power within the village. Fights between rival clans in the same village are non-lethal. The only weapons allowed are sticks, and the spilling of blood is strictly forbidden. Fights between villages, however, are lethal. And it is in these inter-tribal fights that sacrificial victims are explicitly sought. Victims killed in fights between villages are left for the vultures and hyenas in the bush, the latter representing the disorder associated with untamed nature. The victors, meanwhile, must undergo a purification ritual before they reenter the village, to ensure that they do not introduce the disorder of bloodletting. Simonse argues that the main difference between dualist and centralist modes of sacrifice is that in the former the enemy is external whereas in the latter it is internal. In other words, there is an implied ethical development when societies begin to sacrifice their kings. It shows a recognition that violence comes from within rather than without.

The bulk of Simonse’s book is devoted to an examination of the rainmaker king, a figure Simonse regards as a precursor to the authoritarian kings of more developed states such as the Buganda kingdom to the south of his area of study. The rainmaker’s position is inherited, in the sense that he must come from a family of previous rainmakers. Wives may take the role of rainmaker if the husband dies and there is no son or brother to inherit the title of rain-king. In times of fruitfulness, when the rains are regular, the people show their gratitude by bestowing gifts upon the rainmaker. But in times of drought, antagonism between the two parties breaks out. The people accuse the rainmaker of withholding rain. The rainmaker, in turn, accuses the people of incivility and ingratitude, and he threatens to continue withholding the rain if the people do not show him the respect and gifts he deserves. This usually leads to a flurry of gift giving, as the people seek to restore their good relationship to the rainmaker. If the rains fall, the relationship is renewed and peace restored. If the drought continues, however, the people begin to turn against the rain-king, and, if the crisis gets really bad, they sacrifice him.

Simonse provides tremendous historical and ethnographic detail in his account. Much of this detail is firsthand, based on his observations and interviews with informants. But he has also scoured the historical records of missionaries and government officials, and this enables him to provide a full historical picture of his anthropological field of study, from early colonial contact in the mid-nineteenth century to the period of his field studies in the 1980s. Obviously, it would be impossible for an anthropologist to gather the same quality of field data now.

In terms of the basic argument, nothing has changed significantly in the new edition. Simonse has added, however, a considerable number of new photographs. The bibliography has also been expanded to include recent scholarship. Comparing the two editions, I do not see any substantial changes in the text, though I note that the first chapter, which introduces the theoretical framework, has been expanded to include some additional reflections on the genesis of the state.

In what follows, I will focus on the theory rather than the fieldwork. As far as the latter is concerned, I can provide no further comment except to express my sincere admiration and gratitude. It is wonderful indeed to have a book in which this scrupulous fieldwork has been gathered and synthesized.

*       *       *

The central question is, Why sacrifice a king? Simonse’s answer is taken from Girard. We sacrifice the king because we must sacrifice somebody. Without sacrifice, violence and disorder will ensue. The king is merely the most obvious choice. If the king truly is better than everybody, then it stands to reason that he will also make the best sacrifice. So, when the disaster is really bad, when the rains do not fall and the social fabric is put under intolerable strain, the king must go. Of course, this just begs the question, What if there is no king? Whom do we sacrifice then? Simonse’s answer is that we sacrifice the enemy. In dualist societies, the enemy is the rival village or tribe. Hence the importance of inter-village raids. By channeling violence onto an external enemy, peace is restored at home.

Simonse’s key insight is that the rain-king is a highly convenient substitute for the external enemy of dualist societies. Most people resist becoming victims, especially if the attacker is your equal and you have equipped yourself with a club or spear to protect yourself. But the rainmaker is not your equal. On the contrary, he has the power of life and death over every living thing. Drought is a sign of the rainmaker’s displeasure towards his people, who will search diligently for anything that may have offended him. Internal quarrels and disagreements are mended, gifts are bestowed on the rainmaker, and offenders of the peace are banished or made to pay reparations.

Simonse sees that the figure of the rain-king represents an ethical advance over the dualist solution to violence, which projects violence onto an external enemy. In the case of the rainmaker, violence is initially deferred by offering reparations and gifts. Typically, the rainmaker is himself consulted. If he is wise, he will use the occasion to solidify the bonds between himself and the people as well as within the community. Of course, he has no power to make it rain, but he does have power to criticize the community, which may benefit ethically from his advice. As Simonse suggests, what we see taking place here is a form of gift exchange. The rainmaker is the king of big men. His gift is rain. As long as the rain falls, the people are in his debt, and they shower him with gifts in return for a plentiful harvest. But if it does not rain, if the harvest fails, then the king is in the people’s debt because he has not repaid the gifts they have given him.

As Simonse realizes, gift giving can only take place nonviolently. If you bring gifts to the rain-king, then he cannot be a scapegoat. This leads Simonse to suggest that gift giving is a form of displaced or deferred violence. The term he uses is suspense:

We should imagine a scene where a group of individuals caught up in an uneasy, conflictual, relationship is edging to the brink of violence. One individual sets himself apart from the wrangling and confronts the others. The pent-up negativity now directs itself at him. He becomes the focus of the hostility of all. At that very moment the scene changes. The disgruntled lot turns into a group because of a common focus. The impasse is broken.
The longer the stand-off lasts, the stronger will be the bonding of the group. The emerging sense of togetherness does not go unnoticed nor does its association with the figure who triggered it. The group may welcome its new state of being and realize that it is in its interest to make the suspense triggered by the exceptional figure last. The group—or the exceptional figure—may also realize that the new cohesiveness gives it an edge over similar, less cohesive, competing groups. This configuration may be the springboard from which the earliest forms of kingship were launched. What is important here is that the bonding was not the result of a deliberate agreement between individuals or the outcome of mutual compatibility, nor the side-effect of the pursuit of a common interest, let alone an expression of group solidarity. The unity is the product of the suspense of the stand-off between the group and its antagonist. (39)

Simonse modifies slightly the Girardian scenario. Unlike Girard, who emphasizes the violent passage from disorder to order in the frenzied immolation of the victim, Simonse stresses the moment of suspended action or deferral prior to the immolation. All that is necessary to turn Simonse’s primal scene of scapegoating into a genuinely originary account of human difference—that is, into an account that does not beg the question of the difference between humans, who ritually sacrifice their fellows, and their hominid ancestors, who, like nonhuman primates today, occasionally kill their conspecifics but do not sacrifice them—is to add that this suspension of violence is a ritual or symbolic representation of the central victim. The difference between humans and other primates is not that humans are unique in killing their own kind, but that they do so out of a specifically religious or ethical motivation, which is to say, they represent the killing as a sacrifice. In other words, the violence inflicted on the victim has a symbolic meaning that cannot be reduced to what is perceivable independently of the representation. The difference between a dead body and a sacrificial victim is, precisely, the difference between an index and a symbol.[5] The bleeding body of the immolated victim may, of course, function as an index of the violence that precedes it. But for the body to be interpreted as a sacrifice requires more than simply the indexical perception of violence. It requires the shared intentionality of the group, what Simonse elsewhere calls the group’s consensus—the experience that each individual is participating in the collective representation of the victim as an antagonist or enemy. And this representation can only take place, as Simonse implies, in a nonviolent moment of contemplation or suspended action in which each individual exchanges his index-cum-symbol with the others, whose own similarly aborted gestures now take on the status of symbolic signs. A symbol is not an index of its object. In the latter case, the sign is a part of the action that follows. A violent gesture indicates violence because it is part of the overall pattern of violence. A dog will shrink from a raised hand or stick, just as a human will shrink from a dog’s snarl. These are not symbols but indexes of well-established patterns of behaviour. The aborted gesture of appropriation is, in contrast, the negation of violent behaviour. More precisely, it is the deferral of violence through the negation of the previously existing indexical relationship between the sign and its object.

Like Girard, Simonse seems at times to regard scapegoating as an animal instinct or biological imperative:

Girard’s victimary model is silent about the social organization of the proto-humans before the first eruptions of the mimetic crisis. It figures a single group that suffers endemic conflict and that by arbitrarily picking on one of its members (an individual or sub-group) and driving him or her out, finds peace. The proto-humans, however, were not clean slates. Like other primates, they must have lived in territorial groups that were in contact with other groups. Recent primatological research shows that demonstrations of hostility are part of inter-group relations of primates. Chimpanzees even engage in warfare including acts of killing and cannibalism.
It is plausible that the proto-humans who acquired the capacity to use symbols [my italics] incorporated existing pre-human behavioural patterns when they started reproducing the salutary victimary event. It is likely that the first mimetic crises after the primeval event were resolved by discharging the urge of salutary scapegoating on groups competing for the same territory and resources. (17)

What is most remarkable about this passage is that the origin of language is assumed but not explained. How did these proto-humans acquire language? And why was language necessary? Simonse assumes that Girard’s scenario explains the origin of language. The object of the group’s attention—the scapegoat—is also a signifier, because it represents the miraculous passage from violence to peace: “The victim, its substitute (name, emblems, totem) or what remains of it (the body, the tomb) functions as a signifier for the power that saved the community from destruction” (15). But to speak of the victim as a signifier already assumes the capacity for language. The question then becomes, When did this capacity originate? And here Girard has no answer. As Simonse admits, Girard reduces the model to a mechanism that elides the question of representation.

Because of the self-contained, inescapable way in which the process unfolds, Girard speaks of a mechanism. A similar sequence of interactions was, according to René Girard, the course of events that led to the emergence of modern humans. Exactly how the mechanism developed over time is an open question. It is worth emphasizing that Girard never envisaged the collective murder as a single event—as Freud did and as some of Girard’s interpreters do—but as a frequent occurrence during the stages of our development as a species. (15)

I think there are two basic errors here. First, Simonse, like Girard before him, assumes that scapegoating preexists language. That is why Girard describes scapegoating as a mechanism. Second, this leads to the further assumption that human social organization is founded on an unconscious impulse or, to put it more emphatically, a biological instinct for scapegoating. Scapegoating is part of our biological makeup, built into all forms of social organization, from the segmentary and egalitarian tribe (dualism) to the hierarchical state (centralism). Ultimately, Girard refuses to countenance the minimal human difference, the shift from index to symbol described by the originary hypothesis. Girard assumes this difference but he has no hypothesis to explain its emergence.

As Gans has shown, the really crucial element in the Girardian scenario is not the empirical presence of a human victim but, rather, the process of sacralization.[6] This process must be symbolic in structure. That is, there must be a shift from perceiving an object of appetitive interest to representing it as sacred, as collectively significant to the group. The sacred is not reducible to what can be perceived independently of the collective scene in which it appears. Simonse understands that this shift from index to symbol lies at the heart of human culture, but he provides no hypothesis to explain its origin. Scapegoating may well be a universal feature of human culture, but the scapegoat scenario is no substitute for a minimal originary hypothesis.

Language is our most minimal cultural form. But it too partakes of the basic structure of the originary scene. Simonse quotes Gellner on how the charismatic leaders in the Moroccan Atlas are selected:

Agurram-hood is in the eye of the beholder. But that isn’t quite right: agurramhood is in the eyes of the beholders—all of them in a sense squint to see what is in the eyes of other beholders, and if they can see it there, then they see it also. Collectively this characteristic is an ascription, but for any one man, it is an objective fact, an inherent characteristic: if all others can see it in a man, then, for any single beholder, that man truly has it. (Gellner, quoted in Simonse 35)

What Gellner says of agurramhood can be applied to all culture, including language. The same process of collectively “squinting” lies behind the arbitrary signs of language. Words have meaning not because they are reducible to their objects but because language-users acquire them by “squinting” to see how others use them.

If the crucial element in the origin of human culture is the capacity to interpret the world symbolically, why does Girard insist that scapegoating precedes language? One can answer this question by distinguishing the origin of language, on the one hand, from the origin of what Gans calls “cultural universality,” on the other.[7] Scapegoating exists as a universal cultural phenomenon only after the sacred centre of the originary scene has been fully humanized. The first language-users were incapable of interpreting the centre as occupiable by a human being (hence Gans argues that the focus of attention is more plausibly imagined to be a prey animal). Here desire is maximally alienated from the individual. The centre stands opposed to the periphery as the god stands opposed to its human worshipers. In contrast, the ethic implied by the kind of humans capable of not only elevating one of their own to the centre but periodically sacrificing him too represents, not an originary historical moment, but on the contrary an advanced stage of human culture, presumably emerging only with modern Homo sapiens after a period of some several million years of co-evolution between language and the brain. Cultural universality means, precisely, the takeover of culture as the principal means of historical change, genetic change being no longer able to keep pace with cultural change. The fact that all humans are born with the same innate capacity for language illustrates that the significant difference between humans is cultural rather than genetic. We are born with same genetic capacity, but we acquire very different languages and cultures.

The fact that one can identify hidden victims in the many different myths of origin from around the world does not mean that the originary event must have produced a human victim. As Gans argues, one need not deny that the myths refer to actual historical victims. But the mere presence of the myths demonstrates an awareness that ritual sacrifice stands in need of narrative explanation. The myth is a theory of ritual sacrifice, an attempt to explain and justify the need for victimization.

In The End of Culture, Gans argues that social hierarchy emerges with the figure of the big man who takes over the distributive function of the ritual centre, a function formerly reserved for the god or the god’s representative. The rainmaker is not, in this sense, a big man. His role is not primarily economic but sacrificial. As Simonse points out, however, the rainmaker has advantages over the big man, whose power is limited by his ability to attract clients. In contrast, the rainmaker’s clientele includes the entire village. The rainmaker is thus a hybrid between the ritual specialist of egalitarian hunter-gatherer societies and the big man of sedentary societies. The gifts he acquires from his clients make him a de facto big man; he has the largest house, the most wives, and the most wealth. But his power over the rain also makes him the people’s antagonist. If the rains do not come, he will be subject to the people’s wrath.

What Simonse’s analysis suggests is that the rainmaker king is caught between the opposing forces of centre and periphery. Dualism organizes itself around the “fearful symmetry” of the periphery. No individual may occupy the centre, which remains the reserve of the god. Centralism, which only becomes possible with the creation of a surplus among agricultural or sedentary societies, lifts the taboo on the centre. But as the example of the rainmaker shows, the centre remains a locus of suspicion and resentment. Occupation of the centre is not for the faint of heart. The periphery may decide that the rainmaker possesses agurramhood. But if his agurramhood should desert him prematurely, which in the case of the rainmaker is more or less inevitable, then his occupation of the centre will be tragically brief.

Though Simonse is guilty of making the same theoretical mistake as Girard when it comes to understanding the motivation for the originary hypothesis, it would be churlish to suggest that this mistake somehow undermines his specific analyses of the rainmaker kings of southeastern Sudan. What I admire most about this study is the wealth of ethnographic detail Simonse provides to illustrate the constant tug-of-war between dualism and centralism, the people and their king. In the end, this is not a work of philosophy. It is a theoretically sophisticated ethnographic study. As such it is one of the best of its kind.


[1] Jean-Loup Amselle, Mestizo Logics: Anthropology and Identity in Africa and Elsewhere (Stanford: Stanford University Press, 1998), 161.

[2] Ernest Gellner, “James Frazer and Cambridge Anthropology,” in Anthropology and Politics: Revolutions in the Sacred Grove (Oxford: Blackwell, 1995), 102-17.

[3] I have explored this question in Shakespeare’s Big Men: Tragedy and the Problem of Resentment (Toronto: University of Toronto Press, 2016).

[4] Simon Simonse, Kings of Disaster: Dualism, Centralism and the Scapegoat King in Southeastern Sudan, rev. ed. (Kampala, Uganda: Fountain Publishers, 2017); henceforth cited parenthetically in the text. The new edition is also available as an e-book from Michigan State University Press.

[5] See Terrence Deacon, The Symbolic Species: The Co-Evolution of Language and the Brain (New York: Norton, 1997). I apply Deacon’s distinction between index and symbol to the originary hypothesis in “Cognitive Science and the Problem of Representation,” Poetics Today 24 (2003): 237-95, reprinted in The End of Culture: Essays in Anthropological Aesthetics (Aurora, CO: Davies Group Publishers, 2009).

[6] See, in particular, The End of Culture: Toward a Generative Anthropology (Berkeley: University of California Press, 1985).

[7] Gans, End of Culture, 144.